CN119689871A - Evaporation energy-saving self-adaptive regulation and control method and system under multi-parameter analysis - Google Patents
Evaporation energy-saving self-adaptive regulation and control method and system under multi-parameter analysis Download PDFInfo
- Publication number
- CN119689871A CN119689871A CN202411923679.1A CN202411923679A CN119689871A CN 119689871 A CN119689871 A CN 119689871A CN 202411923679 A CN202411923679 A CN 202411923679A CN 119689871 A CN119689871 A CN 119689871A
- Authority
- CN
- China
- Prior art keywords
- evaporation
- solution
- fitness
- initial
- control parameter
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Feedback Control In General (AREA)
Abstract
本申请提供了多参数分析下的蒸发节能自适应调控方法及系统，涉及数据处理技术领域，通过获取单效蒸发器的蒸发控制参数空间，并随机选取满足均匀分布约束的多个初始蒸发控制参数；基于待蒸发盐水的溶液属性、环境参数和期望浓缩指标，执行多个初始蒸发控制参数在预定时区的多个节点下的蒸发效率预测，获取多个预测蒸发效率集合，评价得到多个蒸发适应度；以蒸发控制参数空间为约束，根据多个蒸发适应度进行蒸发控制参数寻优，输出最优蒸发控制参数，控制单效蒸发器进行蒸发作业。本申请解决了现有技术由于缺乏多参数协同优化方法，导致盐水蒸发过程中能耗较高且对环境变化适应性较差的技术问题，达到了提升蒸发效率和降低蒸发能耗的技术效果。
The present application provides an evaporation energy-saving adaptive control method and system under multi-parameter analysis, which relates to the field of data processing technology. The evaporation control parameter space of a single-effect evaporator is obtained, and a plurality of initial evaporation control parameters that satisfy uniform distribution constraints are randomly selected; based on the solution properties, environmental parameters and expected concentration index of the brine to be evaporated, the evaporation efficiency prediction of the multiple initial evaporation control parameters at multiple nodes in a predetermined time zone is performed, and a plurality of predicted evaporation efficiency sets are obtained, and a plurality of evaporation fitness degrees are obtained by evaluation; with the evaporation control parameter space as a constraint, the evaporation control parameters are optimized according to the multiple evaporation fitness degrees, and the optimal evaporation control parameters are output to control the single-effect evaporator to perform evaporation operations. The present application solves the technical problem that the prior art lacks a multi-parameter collaborative optimization method, resulting in high energy consumption and poor adaptability to environmental changes during the evaporation of brine, and achieves the technical effect of improving evaporation efficiency and reducing evaporation energy consumption.
Description
Technical Field
The application relates to the technical field of data processing, in particular to an evaporation energy-saving self-adaptive regulation and control method and system under multi-parameter analysis.
Background
The salt water evaporation is a key link in chlor-alkali chemical industry, and chloride ions and hydroxide ions in the salt water are subjected to ion migration and combination through a single-effect evaporator to generate chlorine, hydrogen and sodium hydroxide.
Currently, energy-saving regulation and control methods for brine evaporation are generally based on simple analysis of the operating conditions of the evaporator, and perform empirical parameter adjustment or single variable optimization, such as control of fixed temperature, pressure or flow. However, there are significant limitations to this single parameter or empirical rule based regulation approach. On one hand, the cooperative effect among a plurality of evaporation control parameters cannot be effectively excavated, so that the overall optimization capability is insufficient, on the other hand, the adaptability to environmental condition changes (such as initial concentration of solution, external temperature and humidity changes and the like) is poor, and the evaporation process cannot be accurately adjusted according to the actual working condition, so that the evaporation efficiency is difficult to reach the optimum, the energy consumption is higher, the stability of the production process is also difficult to ensure, and the sustainable development of the chlor-alkali chemical industry is not facilitated.
Disclosure of Invention
The application provides an evaporation energy-saving self-adaptive regulation and control method and system under multi-parameter analysis, which solve the technical problems of higher energy consumption and poorer adaptability to environmental change in the brine evaporation process due to the lack of a multi-parameter collaborative optimization method in the prior art, and achieve the technical effects of improving the evaporation efficiency and reducing the evaporation energy consumption.
In view of the above problems, on the one hand, the application provides an evaporation energy-saving self-adaptive regulation and control method under multi-parameter analysis, which comprises the steps of obtaining an evaporation control parameter space of a single-effect evaporator, randomly selecting a plurality of initial evaporation control parameters meeting uniform distribution constraint, based on solution properties, environment parameters and expected concentration indexes of brine to be evaporated, executing evaporation efficiency prediction of the plurality of initial evaporation control parameters under a plurality of nodes in a preset time zone, obtaining a plurality of prediction evaporation efficiency sets, evaluating to obtain a plurality of evaporation fitness, carrying out evaporation control parameter optimization according to the plurality of evaporation fitness by taking the evaporation control parameter space as constraint, and outputting optimal evaporation control parameters, wherein parameter optimization is carried out according to a dynamic optimization direction and an adaptation optimization step length, the dynamic optimization direction is a trend adjustment or a degradation avoidance adjustment, and controlling the single-effect evaporator to carry out evaporation operation according to the optimal evaporation control parameters in the preset time zone.
On the other hand, the application also provides an evaporation energy-saving self-adaptive regulation and control system under multi-parameter analysis, which comprises an initial evaporation control parameter selection module, an evaporation efficiency prediction evaluation module and an evaporation control parameter optimizing module, wherein the initial evaporation control parameter selection module is used for acquiring an evaporation control parameter space of a single-effect evaporator, randomly selecting a plurality of initial evaporation control parameters meeting uniform distribution constraint, the evaporation efficiency prediction evaluation module is used for executing evaporation efficiency prediction of the plurality of initial evaporation control parameters under a plurality of nodes in a preset time zone based on solution properties, environment parameters and expected concentration indexes of salt water to be evaporated, acquiring a plurality of prediction evaporation efficiency sets, evaluating and obtaining a plurality of evaporation fitness, and the evaporation control parameter optimizing module is used for optimizing the evaporation control parameters according to the plurality of evaporation fitness to output optimal evaporation control parameters by taking the evaporation control parameter space as constraint, wherein the dynamic optimizing direction and the adapting optimizing step length are used for optimizing parameters, and the evaporation control module is used for controlling the single-effect evaporator to perform evaporation operation in the preset time zone according to the optimal evaporation control parameters.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
The evaporation control parameter space of the single-effect evaporator is obtained, the subsequent optimization range is determined, and possible parameter combinations can be comprehensively covered by a random selection mode meeting uniform distribution constraint, so that the one-sided performance of parameter selection is avoided, and more possibilities are provided for finding the global optimal solution. By means of the evaporation efficiency prediction by combining the solution properties, the environment parameters and the expected concentration indexes of the brine to be evaporated, the complex situation in the actual evaporation process can be reflected more truly. And the predicted evaporation efficiency is evaluated to obtain a plurality of evaporation fitness, so that the quality degree of different parameter combinations in an actual evaporation scene is quantized, and a guide is provided for subsequent parameter optimization. Based on the evaporation fitness obtained in the previous step, parameter optimization is performed under the constraint of a parameter space so as to find the optimal evaporation control parameter. The introduction of the dynamic optimizing direction and the adaptive optimizing step length can more flexibly cope with different environment and condition changes, and the optimal evaporation control parameters can be found more efficiently and accurately. And in a preset time zone, controlling the single-effect evaporator to perform evaporation operation according to the optimal evaporation control parameters, and applying the obtained optimal evaporation control parameters to actual evaporation operation to ensure that the evaporator operates according to the optimal parameter combination, thereby realizing an energy-saving and efficient evaporation process.
In summary, the application can comprehensively consider the evaporation control parameters, synthesize multiple factors to perform accurate efficiency prediction and fitness evaluation, and find the optimal parameters through dynamic optimization, thereby improving the evaporation efficiency of the single-effect evaporator, realizing the aim of energy saving and consumption reduction, being better suitable for various changes in the chemical production process, improving the stability of the production process and guaranteeing the stability of the product quality.
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.
Drawings
Fig. 1 is a schematic flow chart of an evaporation energy-saving adaptive regulation method under multi-parameter analysis according to an embodiment of the present application.
Fig. 2 is a schematic flow chart of obtaining multiple prediction evaporation efficiency sets in the evaporation energy-saving adaptive regulation and control method under multi-parameter analysis provided by the embodiment of the application.
Fig. 3 is a schematic flow chart of obtaining optimal evaporation control parameters in the evaporation energy-saving adaptive regulation and control method under multi-parameter analysis provided by the embodiment of the application.
Fig. 4 is a schematic structural diagram of an evaporation energy-saving adaptive regulation and control system under multi-parameter analysis according to an embodiment of the present application.
Reference numerals indicate the initial evaporation control parameter selection module 10, the evaporation efficiency prediction evaluation module 20, the evaporation control parameter optimizing module 30, and the evaporation control module 40.
Detailed Description
By providing the evaporation energy-saving self-adaptive regulation and control method and the system under multi-parameter analysis, the embodiment of the application solves the technical problems of higher energy consumption and poorer adaptability to environmental change in the brine evaporation process due to the lack of a multi-parameter collaborative optimization method in the prior art, and achieves the technical effects of improving the evaporation efficiency and reducing the evaporation energy consumption.
In a first embodiment, as shown in fig. 1, an embodiment of the present application provides an evaporation energy-saving adaptive regulation method under multi-parameter analysis, where the method includes:
step S1, acquiring an evaporation control parameter space of the single-effect evaporator, and randomly selecting a plurality of initial evaporation control parameters meeting the uniform distribution constraint.
In particular, the evaporation control parameter space refers to a set of value ranges of all possible control parameters affecting the evaporation process of the single effect evaporator. The uniform distribution constraint is a constraint when randomly selecting initial evaporation control parameters, meaning that when selecting initial parameters, these parameters are uniformly distributed throughout the parameter space to ensure coverage of all possible parameter combinations.
The control system of the interactive single-effect evaporation controller collects historical control records, and extracts all control parameters which possibly influence the evaporation process, such as steam supply quantity, liquid inlet flow, liquid outlet flow, steam temperature, steam pressure and the like. And determining the value range of each parameter according to the historical data and the equipment manual, thereby obtaining the evaporation control parameter space. Next, a random number generator is used to randomly select a plurality of initial parameter combinations within this parameter space, ensuring that these parameter combinations are evenly distributed throughout the space, so that the effect of different parameters on evaporation efficiency can be fully evaluated.
By randomly selecting initial evaporation control parameters meeting uniform distribution constraint in the evaporation control parameter space, possible parameter combinations are fully covered, parameter selection limitation caused by artificial experience or a traditional fixed value mode is avoided, and abundant basic data are provided for subsequently finding globally optimal evaporation control parameters.
And S2, based on the solution attribute, the environmental parameter and the expected concentration index of the brine to be evaporated, executing evaporation efficiency prediction of the initial evaporation control parameters under a plurality of nodes in a preset time zone, obtaining a plurality of predicted evaporation efficiency sets, and evaluating to obtain a plurality of evaporation fitness.
In particular, the solution properties refer to chemical and physical properties of the brine to be evaporated, such as salt concentration, viscosity, etc. Environmental parameters include temperature, humidity, barometric pressure, etc. of the operating environment. The desired concentration index refers to the desired degree of brine concentration achieved during evaporation. Using a computational model or simulation software, evaporation efficiency predictions are made for each initial evaporation control parameter combination based on the solution properties of the brine to be evaporated, the environmental parameters, and the desired concentration index. Evaporation efficiency prediction is performed under a plurality of nodes (e.g., 24 nodes are taken every hour in a day) in a predetermined time zone to simulate dynamic changes of the evaporation process. And then, evaluating the performance of each control parameter combination in the actual evaporation process according to the prediction result, and obtaining the evaporation fitness corresponding to a plurality of initial evaporation control parameters. The evaporation fitness is a comprehensive evaluation index of different initial evaporation control parameters after considering solution properties, environment parameters and expected concentration indexes, and reflects the quality degree of each parameter combination in the actual evaporation process.
By comprehensively predicting and evaluating the evaporation efficiency of multiple factors, the advantages and disadvantages of different initial evaporation control parameters can be evaluated more accurately, the evaluation result which meets the requirements of the actual evaporation process is obtained, a reliable basis is provided for subsequent parameter optimization, and the accuracy and the practicability of the whole evaporation control scheme are improved.
And S3, carrying out evaporation control parameter optimization according to the evaporation fitness by taking the evaporation control parameter space as constraint, and outputting optimal evaporation control parameters, wherein parameter optimization is carried out according to a dynamic optimization direction and an adaptive optimization step length, and the dynamic optimization direction is a trend optimization adjustment or a degradation avoidance adjustment.
Specifically, the dynamic optimizing direction refers to flexibly adjusting the optimizing direction according to the current evaporation adaptability in the optimizing process, and can be optimizing adjustment or degradation adjustment. The optimal adjustment is to adjust the parameters in a direction to make the evaporation fitness better (such as higher evaporation efficiency, closer to the desired concentration index, etc.), and the inferior adjustment is to avoid the deterioration of the evaporation fitness. The adaptive optimizing step length is to adjust the amplitude of the parameter each time in the optimizing process. This step is not fixed but adapted according to specific conditions (e.g. the position of the current parameter, the trend of the evaporation adaptation, etc.). For example, in the region close to the optimal solution, a smaller optimizing step may be required to find the optimal solution more accurately, and in the region far from the optimal solution, a larger optimizing step may be used to increase the optimizing speed.
And (3) taking the evaporation control parameter space obtained in the step (S1) as a constraint condition, namely ensuring that parameters in the optimizing process are always in the legal value range. And (3) optimizing by adopting an optimization algorithm according to the plurality of evaporation fitness obtained in the step (S2), and determining the optimal evaporation control parameters. The optimal evaporation control parameter is an optimal solution obtained through an optimization process, and can realize parameter setting with highest efficiency and lowest energy consumption under specific working conditions. The optimizing process may use genetic algorithm, particle swarm optimization algorithm, etc. In the optimizing process, parameters are adjusted according to the dynamic optimizing direction and the adaptive optimizing step length. For example, the dynamic optimizing direction may be determined according to the comparison result of the current fitness and the previous fitness, and the adapting optimizing step length may be determined according to the estimated distance between the current parameter and the optimal parameter, the change rate of the fitness, and other factors.
The combination of the dynamic optimizing direction and the adaptive optimizing step length ensures that the optimizing process can be quickly converged to the vicinity of the optimal solution, and parameters can be more finely adjusted when the optimizing process is close to the optimal solution, thereby improving optimizing efficiency and precision, and finding optimal evaporation control parameters in a parameter space more efficiently and accurately, thereby providing optimal parameter combination for realizing efficient and energy-saving evaporation operation.
And S4, controlling the single-effect evaporator to perform evaporation operation according to the optimal evaporation control parameter in the preset time zone.
Specifically, the predetermined time zone refers to a time range set in advance in the evaporation work. The optimal evaporation control parameters output in the step S3 are input into a control system of the single-effect evaporator, which can be an automatic control system based on a PLC (programmable logic controller) or a DCS (distributed control system), and the automatic control system controls the operation of the evaporator according to the optimal evaporation control parameters in a preset time zone, so that the evaporator is ensured to run in an optimal working state, thereby realizing an efficient evaporation process, improving the evaporation efficiency, reducing the energy consumption and stably reaching the expected concentration index.
Further, step S1 includes:
And S11, acquiring an evaporation control parameter space of the single-effect evaporator, wherein the evaporation control parameters comprise steam supply quantity, liquid inlet flow, liquid outlet flow, steam temperature and steam pressure.
And S12, randomly selecting a first evaporation control parameter from the evaporation control parameter space to be set as a first initial evaporation control parameter, randomly selecting a second evaporation control parameter again, and setting the second evaporation control parameter as a second initial evaporation control parameter if the Euclidean distance between the second evaporation control parameter and the first initial evaporation control parameter is larger than a preset distance threshold.
And S13, continuing to randomly select a third evaporation control parameter, if the Euclidean distance between the third evaporation control parameter and the first initial evaporation control parameter is larger than the preset distance threshold value and the Euclidean distance between the third evaporation control parameter and the second initial evaporation control parameter is larger than the preset distance threshold value, setting the third evaporation control parameter as the third initial evaporation control parameter, and iteratively selecting until the preset number is met, and outputting the plurality of initial evaporation control parameters.
Specifically, the steam supply amount refers to the amount of steam supplied to the single effect evaporator for heating the brine to cause evaporation thereof. The liquid inlet flow is the flow of the saline solution entering the single-effect evaporator, and influences the liquid level and the residence time of the solution in the evaporator. The liquid outlet flow is the flow of the solution flowing out of the single-effect evaporator, and is related to factors such as liquid inlet flow, evaporation capacity and the like. The steam temperature is the temperature of the supplied steam, and different steam temperatures affect the heat transfer efficiency within the evaporator. The vapor pressure is the pressure of the vapor and affects the energy transfer efficiency of the vapor within the evaporator.
Referring to an equipment manual of the single-effect evaporator, an evaporation control parameter space is obtained, wherein the evaporation control parameter space comprises the range of values of control parameters such as steam supply quantity, liquid inlet flow, liquid outlet flow, steam temperature, steam pressure and the like. For example, for a single-effect evaporator of a certain model, the steam supply is 5-20 kg/h, the liquid inlet flow is 3-10 cubic meters/h, the liquid outlet flow is changed in a certain range according to factors such as liquid inlet flow and evaporation amount, the steam temperature is 100-150 ℃, and the steam pressure is 100-500 kPa. The value ranges together form an evaporation control parameter space, a complete range frame is provided for the follow-up selection of initial evaporation control parameters, various possible parameter combinations are comprehensively considered, and important parameter value omission is avoided.
And randomly selecting a first evaporation control parameter from the evaporation control parameter space to be set as a first initial evaporation control parameter, randomly selecting a second evaporation control parameter again, and calculating Euclidean distance between the first initial evaporation control parameter and the second evaporation control parameter. For the evaporation control parameters, each parameter may be regarded as one dimension of a vector, the euclidean distance of the vector corresponding to the first initial evaporation control parameter and the second evaporation control parameter is calculated using the euclidean distance calculation formula, and compared with a predetermined distance threshold. The preset distance threshold is a preset Euclidean distance value and is used for judging whether the randomly selected evaporation control parameters are sufficiently dispersed or not. If the Euclidean distance between the second evaporation control parameter and the first initial evaporation control parameter is larger than a preset distance threshold, the second evaporation control parameter is set as the second initial evaporation control parameter, otherwise, if the Euclidean distance between the second evaporation control parameter and the first initial evaporation control parameter is smaller than or equal to the preset distance threshold, the second evaporation control parameter is abandoned, and then a new evaporation control parameter is selected again for judgment. By setting the judgment conditions of the Euclidean distance, the selected initial evaporation control parameters have certain dispersibility, and the too close parameter combination is avoided, so that the evaporation control parameter space can be more comprehensively covered, and the possibility of finding the global optimal solution is increased.
After the first initial evaporation control parameter and the second initial evaporation control parameter are determined, continuing to randomly select a third evaporation control parameter, and then setting the third evaporation control parameter as the third initial evaporation control parameter if the Euclidean distance between the third evaporation control parameter and the first initial evaporation control parameter and the Euclidean distance between the third evaporation control parameter and the second initial evaporation control parameter are larger than a preset distance threshold value. Otherwise, discarding the third evaporation control parameter, and re-selecting a new evaporation control parameter from the evaporation control parameter space to perform distance calculation and judgment. And continuously and randomly selecting the evaporation control parameters according to the iterative mode, and judging the distance to determine new initial evaporation control parameters until the preset number is met. By means of the iterative selection mode, good dispersibility of the selected multiple initial evaporation control parameters in the evaporation control parameter space is ensured, the parameter space can be explored more widely, so that more diversified and representative initial parameters are provided for subsequent evaporation efficiency prediction, optimizing and other operations, and the probability of finding the optimal evaporation control parameters in the whole technical scheme is improved.
Further, as shown in fig. 2, step S2 includes:
And S21, collecting solution properties and environment parameters of the brine to be evaporated, wherein the solution properties comprise a solution type, an initial concentration and an initial temperature, and the environment parameters comprise an environment temperature and an environment humidity.
And S22, taking the solution property, the environment parameter and the expected concentration index as condition constraint, taking the property characteristic of the single-effect evaporator as equipment constraint, taking evaporation control as guide, retrieving big data to obtain a sample evaporation control parameter set and a sample evaporation time set, and marking evaporation efficiency under different sample evaporation control parameters and different sample evaporation times to obtain a sample evaporation efficiency set.
And S23, training an integrated learning operator based on the sample evaporation control parameter set, the sample evaporation time set and the sample evaporation efficiency set until convergence to obtain an evaporation efficiency prediction plug-in unit, wherein the integrated learning operator at least comprises a random forest, a BP neural network and a support vector machine.
And S24, determining a plurality of evaporation times according to a plurality of nodes in the preset time zone, and executing evaporation efficiency prediction of the plurality of initial evaporation control parameters under the plurality of evaporation times by using the evaporation efficiency prediction plug-in unit to output a plurality of predicted evaporation efficiency sets, wherein the evaporation time is the time interval between the node and the starting time.
Specifically, the control system or production task is interacted with to obtain the solution properties and environmental parameters of the brine to be evaporated. Wherein the solution properties include the solution type, initial concentration, and initial temperature, and the environmental parameters include the ambient temperature and the ambient humidity. The solution type refers to the specific component type of the brine to be evaporated, such as different salt solution types of sodium chloride brine, potassium chloride brine and the like, and the evaporation efficiency is affected due to different physicochemical properties of different solution types in the evaporation process. The initial concentration is the concentration of the solute of the brine to be evaporated before the evaporation starts, and the concentration can influence the mass transfer driving force in the evaporation process. The initial temperature is the temperature at which the brine begins to evaporate, and the higher the temperature, the greater the kinetic energy of the water molecules, which is beneficial to the evaporation of water to some extent. The ambient temperature is the temperature of the environment surrounding the evaporator and is closely related to heat transfer during the evaporation of brine, for example, in a high temperature environment, the rate of heat dissipation to the environment during the evaporation of brine is relatively slow, which may be advantageous for evaporation. Ambient humidity is the amount of moisture in ambient air, generally expressed as relative humidity, and the higher the relative humidity, the closer the moisture in ambient air is to saturation, which is detrimental to the evaporation of moisture.
A sample evaporation control parameter set refers to a set of evaporation control parameters collected in historical data. The sample evaporation time set refers to a set of evaporation times corresponding to the sample evaporation control parameters. The sample evaporation efficiency set is a set of evaporation efficiencies obtained through labeling under different sample evaporation control parameters and sample evaporation time. The attribute characteristics of the single-effect evaporator comprise the characteristics of the volume, the heating mode, the material quality and the like of the evaporator. The method takes the acquired solution attribute, environmental parameter and expected concentration index as condition constraint, the attribute characteristic of the single-effect evaporator is equipment constraint, the evaporation control is guided, and the retrieval is carried out in a large data platform (such as a database in the chemical industry, a large amount of evaporation process data accumulated in an enterprise, and the like). For example, from evaporation process data accumulated in chemical enterprises for many years, evaporation process data meeting the conditions that the solution type is sodium chloride brine, the initial concentration is in a certain range, the ambient temperature and the humidity are in a specific interval and the like are screened, and evaporation control parameters and evaporation time in the data are extracted to form a sample evaporation control parameter set and a sample evaporation time set. And then labeling the evaporation efficiency of the samples according to actual production records or experimental data to obtain a sample evaporation efficiency set. Related data are obtained from big data by integrating constraint conditions in multiple aspects, a sample evaporation control parameter set, a sample evaporation time set and a sample evaporation efficiency set are constructed, a rich data source is provided for training an evaporation efficiency prediction model, the model can learn evaporation rules under different conditions, and accuracy and reliability of model prediction are improved.
The evaporation efficiency prediction plug-in is a tool which is obtained through training of an integrated learning operator and can be used for predicting the evaporation efficiency, and the evaporation efficiency can be output according to the input evaporation control parameters, evaporation time and other information. The ensemble learning is combined by a plurality of machine learning algorithms to improve the prediction performance, wherein each machine learning algorithm is an ensemble learning operator, and in the embodiment of the application, the ensemble learning operator at least comprises a random forest, a BP neural network and a support vector machine. And carrying out data preprocessing, such as normalization processing, on the sample evaporation control parameter set, the sample evaporation time set and the sample evaporation efficiency set so as to improve the training effect of the model. And then, a machine learning framework (such as a Scikit-learn library in Python) is used for respectively inputting the preprocessed sample evaporation control parameter set, the sample evaporation time set and the sample evaporation efficiency set into an integrated learning operator such as a random forest, a BP neural network, a support vector machine and the like for training. In the training process, parameters of the model (such as the number of decision trees in a random forest, weight and bias in a BP neural network and the like) are continuously adjusted, and whether the model converges or not is judged by calculating errors (such as mean square errors) of the model on a verification set. When the model converges, that is, the parameters of the model are gradually stabilized in the iterative process, the prediction result of the model does not have obvious change any more, and the integrated learning operators are combined to obtain the evaporation efficiency prediction plug-in. The process of combining the ensemble learners first needs to set a trusted weight of each ensemble learner, for example, a mean square error may be used as an evaluation index, the training accuracy of each ensemble learner on the verification set may be evaluated, weight calculation is performed according to the inverse of the mean square error, and the ratio of the inverse of the mean square error of each ensemble learner to the sum of the mean square errors of all ensemble learners is used as the trusted weight of each ensemble learner. When the evaporation efficiency prediction plug-in is used for prediction, the evaporation control parameters and the evaporation time are respectively input into a plurality of integrated learning operators to obtain a plurality of corresponding predicted evaporation efficiencies, and then the evaporation efficiencies of the predictions are weighted and summed according to the credible weights to obtain the final predicted evaporation efficiency. The evaporation efficiency prediction plug-in integrates the advantages of various machine learning algorithms, and has higher prediction accuracy and stability. Compared with a single machine learning model, the method can be better suitable for different data distribution and complex evaporation process relations, and therefore powerful guarantee is provided for accurately predicting the evaporation efficiency.
The evaporation time refers to a time interval from a starting time to a certain node in a predetermined time zone, which is an important factor affecting evaporation efficiency, and the evaporation efficiency varies due to the progress of the evaporation process at different evaporation times. The plurality of nodes are determined according to the predetermined time zone, for example, the predetermined time zone is 24 hours a day, and the nodes are taken every one hour, so that 24 nodes are obtained. For each node, the time interval from the start time is calculated as the evaporation time. And then inputting a plurality of initial evaporation control parameters and evaporation times into an evaporation efficiency prediction plug-in unit, and predicting the evaporation efficiency by utilizing an integrated learning algorithm in the plug-in unit to obtain the evaporation efficiency of each initial evaporation control parameter under each evaporation time, and finally outputting a plurality of prediction evaporation efficiency sets, wherein each prediction evaporation efficiency set corresponds to one initial evaporation control parameter and comprises the evaporation efficiency of the initial evaporation control parameter under different evaporation times. By predicting the evaporation efficiency of the initial evaporation control parameters in different evaporation time in a preset time zone, the evaporation efficiency condition of different parameter combinations at different time points can be comprehensively known, a rich data basis is provided for subsequent evaporation adaptability evaluation and evaporation control parameter optimization, and the optimal evaporation control parameters can be found.
Further, in step S2, a plurality of evaporation fitness values are evaluated, including:
And S25, randomly selecting a first predicted evaporation efficiency set, carrying out mean value calculation and variance calculation on the first predicted evaporation efficiency set, and determining a first predicted evaporation efficiency mean value and a first predicted evaporation efficiency fluctuation coefficient.
And S26, obtaining a first evaporation fitness according to the first predicted evaporation efficiency average value and the first predicted evaporation efficiency fluctuation coefficient through weighted calculation, and adding the first evaporation fitness to the plurality of evaporation fitness, wherein the evaporation fitness and the predicted evaporation efficiency average value are positively correlated, and the predicted evaporation efficiency fluctuation coefficient is negatively correlated.
Specifically, one predicted evaporation efficiency set is randomly selected from a plurality of predicted evaporation efficiency sets output from the evaporation efficiency prediction plug-in, and is denoted as a first predicted evaporation efficiency set. And calculating the arithmetic mean value and variance of all the predicted evaporation efficiencies in the first predicted evaporation efficiency set to obtain a first predicted evaporation efficiency mean value and a first predicted evaporation efficiency fluctuation coefficient. The predicted average evaporation efficiency reflects the average level of evaporation efficiency under a set of evaporation control parameters. The variance is a statistic that measures the degree of dispersion of a set of data, reflecting the degree of fluctuation of the predicted evaporation efficiency. For the set of predicted evaporation efficiencies, the larger the variance, the larger the fluctuation in evaporation efficiency at different time nodes or conditions. By calculating the predicted evaporation efficiency mean and the coefficient of fluctuation, the overall level and stability of evaporation efficiency under a set of evaporation control parameters can be comprehensively understood. The average value gives an average reference value, the fluctuation coefficient reflects the change condition of the evaporation efficiency, and the two indexes provide important basis for the subsequent evaluation of the evaporation fitness.
The purpose of evaporation energy-saving regulation and control is to improve evaporation efficiency and stability of an evaporation process, a formula for weighting calculation is set according to the target, and a mean value of predicted evaporation efficiency and a fluctuation coefficient of predicted evaporation efficiency are weighted calculated to obtain evaporation fitness. In the weighted calculation formula, the evaporation fitness and the average value of the predicted evaporation efficiency are positively correlated, and the fluctuation coefficient of the predicted evaporation efficiency is negatively correlated. For example, the calculation formula may be set to f=a×e-b×v, where F is the evaporation fitness, E is the average value of the predicted evaporation efficiency, V is the fluctuation coefficient of the predicted evaporation efficiency, a and b are weight coefficients, and the values thereof may be set in a customized manner according to the actual situation. Substituting the calculated first predicted evaporation efficiency mean value and the first predicted evaporation efficiency fluctuation coefficient into the formula to obtain a first evaporation fitness, and adding the first evaporation fitness to a plurality of evaporation fitness.
According to the calculation process, the evaporation fitness corresponding to each of the plurality of predicted evaporation efficiency sets is determined, and a plurality of evaporation fitness are obtained. The method for calculating the evaporation fitness by weighting can comprehensively consider the average level and stability of the evaporation efficiency, avoid the condition that only the average efficiency value is concerned and the fluctuation is ignored, or only the fluctuation is concerned and the efficiency level is ignored, provide more comprehensive and reasonable evaluation basis for subsequent evaporation control parameter optimization, and help to find the evaporation control parameter which has good performance in the aspects of efficiency and stability.
Further, as shown in fig. 3, step S3 includes:
And S31, setting the initial evaporation control parameters as initial solutions, arranging the initial evaporation control parameters from large to small according to the evaporation fitness, and determining an initial solution sequence.
And S32, setting the first P solutions of the initial solution sequence as optimal solutions, setting the last Q solutions as inferior solutions, and obtaining P optimal solutions and Q inferior solutions, wherein Q is N times of P, P, Q are integers, and N is an integer greater than 5.
And S33, carrying out random clustering on the Q inferior solutions according to the P superior solutions to obtain P solution sets, wherein the number of the inferior solutions in each solution set is the same.
And S34, regarding the P solution sets, setting the inferior solution with the minimum adaptability in the solution sets as a difference solution, obtaining P difference Jie Shi adaptability, and setting P dynamic optimizing directions according to the P optimal solution adaptability and the P difference solution adaptability, wherein the dynamic optimizing directions are optimal or inferior-avoiding adjustment.
And S35, according to the P dynamic optimizing directions, according to the adaptive optimizing step length, the inferior solutions in the P solutions are adjusted to obtain P updated solution sets, the P updated solution sets are identified, if the adaptability of the inferior solutions in the updated solution sets is greater than the adaptability of Jie Shi, the superior solutions in the solution sets are replaced by the inferior solutions, and if the adjusted inferior solutions do not meet the evaporation control parameter space, the adjustment is not carried out.
And S36, continuing iterative optimization based on the dynamic optimization direction and the adaptive optimization step length until the preset iteration times are reached, outputting P current solution sets, selecting the current solution set with the largest sum of the fitness among the P current solution sets as an optimal solution set, and outputting the optimal solution of the optimal solution set as the optimal evaporation control parameter.
Specifically, a plurality of initial evaporation control parameters and corresponding evaporation fitness are acquired. Setting the initial evaporation control parameters as initial solutions, and then using a sorting algorithm (such as bubbling sorting, quick sorting and the like) to sort the initial evaporation control parameters according to the evaporation fitness from large to small, so as to obtain an initial solution sequence. The initial solution sequence is obtained through the sequencing, so that initial evaporation control parameters with good evaporation adaptability can be arranged in front, a basis is provided for distinguishing the optimal solution and the inferior solution subsequently, the possible optimal solution area can be positioned quickly, and the optimizing efficiency is improved.
The first P solutions of the initial solution sequence are set as optimal solutions, the evaporation fitness corresponding to the solutions is relatively high, the solutions are better in all the current initial solutions, the last Q solutions are set as inferior solutions, the evaporation fitness of the solutions is relatively low, P optimal solutions and Q inferior solutions are determined, wherein Q is N times of P, P, Q are integers, and N is an integer larger than 5. The solution space can be further analyzed and processed in a mode of distinguishing the optimal solution and the inferior solution, and the information of the optimal solution is utilized to improve the inferior solution, so that better evaporation control parameters are gradually found.
And carrying out random clustering on the Q inferior solutions by using a random number generator, randomly dividing the Q inferior solutions into P groups, wherein each group corresponds to one optimal solution, and respectively integrating the optimal solution and the corresponding multiple inferior solutions to obtain P solution sets. Each solution set contains one optimal solution and N inferior solutions (n=q/P). The random clustering can enable the inferior solutions to be uniformly distributed in different solution sets, avoid too centralization or dispersion of the inferior solutions in some solution sets, and provide a reasonable grouping basis for subsequent adjustment of the inferior solutions in different solution sets based on the optimal solutions.
For each solution set, traversing the inferior solutions in the solution sets, and finding the inferior solution with the minimum fitness as the difference solution. And then comparing the difference Jie Shi fitness and the optimal Jie Shi fitness of each solution set, and setting P dynamic optimizing directions according to the P optimal solution fitness and the P difference solution fitness. The dynamic optimization direction may be to adjust the inferior solution toward the optimal solution (i.e., toward optimal adjustment) or to avoid adjusting the inferior solution toward the inferior solution (i.e., away from inferior adjustment). By determining the difference solutions and setting the dynamic optimizing directions, a reasonable optimizing strategy can be formulated according to the specific conditions of each solution set, so that the optimizing process is more targeted, and optimizing efficiency and accuracy are improved.
And according to the determined dynamic optimizing direction, adjusting the inferior solutions in each solution set according to the adaptive optimizing step length to form an updated solution set. If the adaptation degree of the inferior solutions in the update solution set is greater than the Jie Shi fitness degree of the superior solutions, the superior solutions are replaced by the inferior solutions. In the adjustment process, it is necessary to check the parameter ranges before and after the adjustment, and if the adjusted inferior solution does not satisfy the evaporation control parameter space (for example, the steam supply amount exceeds the maximum value allowed by the device), the adjustment is not performed. According to the dynamic optimizing direction and the adaptive optimizing step length, the inferior solution can be gradually improved in the solution space, so that the inferior solution is developed to a better direction. Meanwhile, the validity and feasibility of parameter adjustment are ensured by limiting the adjusted inferior solution in the evaporation control parameter space.
The predetermined iteration number is the number of iterations in a preset optimizing process. Iterative optimization is carried out based on the dynamic optimizing direction and the adaptive optimizing step length, and each iteration is carried out according to the determined dynamic optimizing direction, and the inferior solutions in each solution set are adjusted according to the adaptive optimizing step length to form an updated solution set. After a predetermined number of iterations is reached, the sum of fitness of all solutions in each of the current P sets of solutions is calculated. And selecting a solution set with the maximum sum of fitness as an optimal solution set, and outputting an optimal solution in the optimal solution set as an optimal evaporation control parameter. Through repeated iterative optimization, the optimal solution can be searched more comprehensively in the solution space, and the optimal solution in the optimal solution set with the maximum sum of fitness is finally selected as the optimal evaporation control parameter, so that the probability of finding the global optimal evaporation control parameter is improved, and the evaporation operation of the single-effect evaporator is optimized.
Further, in step S34, setting P dynamic optimization directions according to the P optimal solution fitness and the P differential solution fitness includes:
Step S341, randomly selecting a first solution set from the P solution sets, and obtaining a first optimal solution fitness, a first poor solution fitness and a plurality of first poor solution fitness of the first solution set.
Step S342, calculating a first bad solution fitness average value of the plurality of first bad solution fitness values, respectively calculating absolute values of the first optimal solution fitness value, the deviation of the first bad solution fitness value and the first bad solution fitness average value, and determining a first optimal solution deviation value and a first bad solution deviation value.
Step S343, setting a first dynamic optimizing direction according to the first optimal solution deviation value and the first inferior solution deviation value, and adding the first dynamic optimizing direction to the P dynamic optimizing directions, wherein if the first optimal solution deviation value is greater than or equal to the first inferior solution deviation value, the first dynamic optimizing direction is optimizing adjustment, and if the first optimal solution deviation value is smaller than the first inferior solution deviation value, the first dynamic optimizing direction is inferior adjustment avoidance.
Specifically, the specific process of determining the dynamic optimizing direction comprises the steps of randomly selecting a first solution set from P solution sets by using a random number generator, and acquiring a first optimal solution fitness, a first poor solution fitness and a plurality of first poor solution fitness from the first solution set. The first optimal solution fitness is the fitness corresponding to the optimal solution in the first solution set, namely the maximum fitness, the first poor solution fitness is the fitness corresponding to the poor solution in the first solution set, namely the minimum fitness, and the first poor solution fitness is the fitness corresponding to other solutions except the optimal solution and the poor solution.
And adding all the inferior solutions in the first solution set, and dividing the sum by the number of the inferior solutions to obtain a first inferior solution fitness average value, wherein the first inferior solution fitness average value reflects the average level of the inferior Jie Shi fitness in the first solution set. And then, calculating the absolute value of the deviation between the first optimal solution fitness and the first inferior solution fitness mean value, marking the absolute value as a first optimal solution deviation value, and calculating the absolute value of the deviation between the first inferior solution fitness and the first inferior solution fitness mean value, marking the absolute value as a first inferior solution deviation value. The deviation degree of the optimal solution and the difference solution relative to the average level of the inferior solution can be quantized by calculating the average value of the adaptability of the first inferior solution and the absolute value of the deviation of the adaptability of the first inferior solution and the optimal solution and the difference Jie Shi, and a quantization basis is provided for accurately setting the dynamic optimizing direction.
Comparing the calculated first optimal solution deviation value with the first inferior solution deviation value, setting the first dynamic optimizing direction as optimal adjustment if the first optimal solution deviation value is larger than or equal to the first inferior solution deviation value, and setting the first dynamic optimizing direction as inferior adjustment if the first optimal solution deviation value is smaller than the first inferior solution deviation value. By setting the first dynamic optimizing direction in the mode, a reasonable optimizing strategy can be formulated according to the deviation condition of the optimal solution and the difference solution in the first solution set relative to the average level of the inferior solution, so that optimizing efficiency is improved, and the optimal solution tends to be more rapidly or is far away from the inferior solution.
The operations of steps S341 to S343 are performed for each of the P solution sets, and a corresponding dynamic optimization direction is set for each solution set.
Further, obtaining the adaptation optimizing step length includes:
And S3-1, determining a first optimizing fitness in a first direction according to the first dynamic optimizing direction, wherein the first optimizing fitness is a first optimizing fitness or a first difference fitness.
And S3-2, respectively calculating the ratio of a plurality of first inferior solution fitness to the first optimizing fitness to determine a plurality of compensation coefficients, wherein if the first optimizing fitness is the first optimizing fitness, the compensation coefficient is the ratio of the first inferior solution fitness to the first optimizing fitness, and if the first optimizing fitness is the first inferior solution fitness, the compensation coefficient is the ratio of the first inferior solution fitness to the first inferior solution fitness.
And step S3-3, optimizing the initial optimizing step according to the plurality of compensation coefficients, determining a first step set, and adding the first step set to the adaptive optimizing step.
Specifically, for the selected first solution set, a first optimizing fitness of the first direction is determined according to the determined first dynamic optimizing direction, and if the first dynamic optimizing direction is the optimizing adjustment, the first optimizing fitness is the first optimizing solution fitness, and if the first dynamic optimizing direction is the avoiding adjustment, the first optimizing fitness is the first difference solution fitness.
Obtaining a plurality of first inferior solution fitness, respectively calculating the ratio of the plurality of first inferior solution fitness to the first optimizing fitness, and determining a plurality of compensation coefficients, wherein the compensation coefficients reflect the proportional relation between the inferior solution and an optimizing reference solution (an optimal solution or a difference solution) in terms of fitness and are used for optimizing an initial optimizing step length. When the first optimizing fitness is the first optimizing solution fitness, namely the optimizing reference solution is the optimizing solution, and the compensation coefficient is the ratio of the first inferior solution fitness to the first optimizing solution fitness. When the first optimizing fitness is the first differential solution fitness, that is, the optimizing reference solution is the differential solution, the compensation coefficient is the ratio of the first differential solution fitness to the first inferior solution fitness. The relation between the inferior solution and the optimizing reference solution on the fitness can be quantized by calculating the compensation coefficient, and a basis is provided for adjusting the optimizing step length according to the conditions of different inferior solutions, so that the suitability and the accuracy of optimizing step length setting are improved.
The first step length set is an optimized optimizing step length set obtained by adjusting the initial optimizing step length according to a plurality of compensation coefficients, and comprises optimizing step lengths optimized for each inferior solution. The initial optimizing step length is the basic parameter adjustment amplitude set at the beginning of the optimizing process, and comprises the initial adjustment amplitude of evaporation control parameters such as steam supply quantity, liquid inlet flow, liquid outlet flow, steam temperature, steam pressure and the like. For each solution set, optimizing the initial optimizing step by using a plurality of compensation coefficients, namely multiplying the compensation coefficients by the initial optimizing step, and determining a plurality of first step sizes, wherein each first step size corresponds to one inferior solution in the solution set. Summarizing the plurality of first step sizes to obtain a first step size set, and adding the first step size set to the adaptive optimizing step size so as to facilitate the subsequent adjustment and optimization of the plurality of inferior solutions. By optimizing the initial optimizing step length according to the compensation coefficient, the optimizing step length can be more suitable for the conditions of different inferior solutions, and optimizing precision is improved. Different inferior solutions obtain different optimizing step sizes according to the adaptability relation between the inferior solutions and the optimizing reference solution, so that the optimizing process is more reasonable and efficient.
In summary, the evaporation energy-saving self-adaptive regulation and control method under multi-parameter analysis provided by the embodiment of the application has the following technical effects:
The evaporation control parameter space of the single-effect evaporator is obtained, the subsequent optimization range is determined, and possible parameter combinations can be comprehensively covered by a random selection mode meeting uniform distribution constraint, so that the one-sided performance of parameter selection is avoided, and more possibilities are provided for finding the global optimal solution. By means of the evaporation efficiency prediction by combining the solution properties, the environment parameters and the expected concentration indexes of the brine to be evaporated, the complex situation in the actual evaporation process can be reflected more truly. And the predicted evaporation efficiency is evaluated to obtain a plurality of evaporation fitness, so that the quality degree of different parameter combinations in an actual evaporation scene is quantized, and a guide is provided for subsequent parameter optimization. Based on the evaporation fitness obtained in the previous step, parameter optimization is performed under the constraint of a parameter space so as to find the optimal evaporation control parameter. The introduction of the dynamic optimizing direction and the adaptive optimizing step length can more flexibly cope with different environment and condition changes, and the optimal evaporation control parameters can be found more efficiently and accurately. And in a preset time zone, controlling the single-effect evaporator to perform evaporation operation according to the optimal evaporation control parameters, and applying the obtained optimal evaporation control parameters to actual evaporation operation to ensure that the evaporator operates according to the optimal parameter combination, thereby realizing an energy-saving and efficient evaporation process.
Overall, the embodiment of the application can comprehensively consider the evaporation control parameters, synthesize multiple factors to carry out accurate efficiency prediction and adaptability evaluation, and find the optimal parameters through dynamic optimization, thereby improving the evaporation efficiency of the single-effect evaporator, realizing the aim of energy saving and consumption reduction, being better suitable for various changes in the chemical production process, improving the stability of the production process and guaranteeing the stability of the product quality.
In a second embodiment, as shown in fig. 4, the embodiment of the present application provides an evaporation energy-saving adaptive regulation and control system under multi-parameter analysis, where the system includes:
The initial evaporation control parameter selection module 10 is configured to obtain an evaporation control parameter space of the single-effect evaporator, and randomly select a plurality of initial evaporation control parameters that satisfy a uniform distribution constraint.
The evaporation efficiency prediction evaluation module 20 is configured to execute evaporation efficiency prediction of the plurality of initial evaporation control parameters under a plurality of nodes in a predetermined time zone based on the solution attribute, the environmental parameter and the desired concentration index of the brine to be evaporated, obtain a plurality of predicted evaporation efficiency sets, and evaluate to obtain a plurality of evaporation fitness.
And the evaporation control parameter optimizing module 30 is configured to perform evaporation control parameter optimization according to the plurality of evaporation fitness degrees with the evaporation control parameter space as a constraint, and output an optimal evaporation control parameter, where the parameter optimization is performed according to a dynamic optimizing direction and an adaptive optimizing step length, and the dynamic optimizing direction is a trend optimization adjustment or a degradation avoidance adjustment.
And the evaporation control module 40 is used for controlling the single-effect evaporator to perform evaporation operation according to the optimal evaporation control parameter in the preset time zone.
Further, the initial evaporation control parameter selection module 10 according to the embodiment of the present application is configured to execute the following steps:
The method comprises the steps of obtaining an evaporation control parameter space of a single-effect evaporator, wherein the evaporation control parameter comprises a steam supply quantity, a liquid inlet flow, a liquid outlet flow, a steam temperature and a steam pressure, randomly selecting a first evaporation control parameter in the evaporation control parameter space to be set as a first initial evaporation control parameter, randomly selecting a second evaporation control parameter again, setting the second evaporation control parameter as a second initial evaporation control parameter if the Euclidean distance between the second evaporation control parameter and the first initial evaporation control parameter is larger than a preset distance threshold value, continuously randomly selecting a third evaporation control parameter, and iteratively selecting the third evaporation control parameter as a third initial evaporation control parameter until the preset quantity is met and outputting the multiple initial evaporation control parameters if the Euclidean distance between the third evaporation control parameter and the first initial evaporation control parameter is larger than the preset distance threshold value and the Euclidean distance between the third evaporation control parameter and the second initial evaporation control parameter is larger than the preset distance threshold value.
Further, the evaporation efficiency prediction evaluation module 20 according to the embodiment of the present application is further configured to perform the following steps:
The method comprises the steps of collecting solution properties and environment parameters of brine to be evaporated, wherein the solution properties comprise solution types, initial concentration and initial temperature, the environment parameters comprise environment temperature and environment humidity, taking the solution properties, the environment parameters and an expected concentration index as condition constraints, taking the property characteristics of a single-effect evaporator as equipment constraints, taking evaporation control as guidance, acquiring a sample evaporation control parameter set and a sample evaporation time set through big data retrieval, marking evaporation efficiencies under different sample evaporation control parameters and different sample evaporation times to obtain a sample evaporation efficiency set, training an integrated learning operator based on the sample evaporation control parameter set, the sample evaporation time set and the sample evaporation efficiency set until convergence, obtaining an evaporation efficiency prediction plug-in, wherein the integrated learning operator at least comprises a random forest, a BP neural network and a support vector machine, determining a plurality of evaporation times according to a plurality of nodes in the preset time zone, utilizing the evaporation efficiency prediction plug-in to execute evaporation efficiency prediction of the plurality of initial evaporation control parameters under the plurality of evaporation times, and outputting a plurality of prediction evaporation efficiency sets, wherein the evaporation time is the time interval between the nodes and the initial time.
Further, the evaporation efficiency prediction evaluation module 20 according to the embodiment of the present application is further configured to perform the following steps:
The method comprises the steps of collecting solution properties and environment parameters of brine to be evaporated, wherein the solution properties comprise solution types, initial concentration and initial temperature, the environment parameters comprise environment temperature and environment humidity, taking the solution properties, the environment parameters and an expected concentration index as condition constraints, taking the property characteristics of a single-effect evaporator as equipment constraints, taking evaporation control as guidance, acquiring a sample evaporation control parameter set and a sample evaporation time set through big data retrieval, marking evaporation efficiencies under different sample evaporation control parameters and different sample evaporation times to obtain a sample evaporation efficiency set, training an integrated learning operator based on the sample evaporation control parameter set, the sample evaporation time set and the sample evaporation efficiency set until convergence, obtaining an evaporation efficiency prediction plug-in, wherein the integrated learning operator at least comprises a random forest, a BP neural network and a support vector machine, determining a plurality of evaporation times according to a plurality of nodes in the preset time zone, utilizing the evaporation efficiency prediction plug-in to execute evaporation efficiency prediction of the plurality of initial evaporation control parameters under the plurality of evaporation times, and outputting a plurality of prediction evaporation efficiency sets, wherein the evaporation time is the time interval between the nodes and the initial time.
Further, the evaporation control parameter optimizing module 30 according to the embodiment of the present application is further configured to perform the following steps:
Setting an initial evaporation control parameter as an initial solution, arranging a plurality of initial evaporation control parameters from large to small according to the evaporation fitness, and determining an initial solution sequence; setting the first P solutions of the initial solution sequence as optimal solutions, setting the last Q solutions as inferior solutions, obtaining P optimal solutions and Q inferior solutions, wherein Q is N times of P, P, Q is an integer, N is an integer larger than 5, carrying out random clustering on the Q inferior solutions according to the P optimal solutions, obtaining P update solution sets, wherein the quantity of the inferior solutions in each solution set is the same, setting the inferior solution with the minimum adaptability in the solution set as a differential solution for the P solution sets, obtaining P difference Jie Shi degrees, setting P dynamic optimizing directions according to the P optimal solution adaptability and the P difference solution adaptability, wherein the dynamic optimizing directions are in trend adjustment or avoidance adjustment, carrying out adjustment on the inferior solutions in the P solution sets according to the adaptive step length, identifying the P update solution sets, carrying out iteration adjustment until the current optimal solution is not satisfied by the P update solution sets, and carrying out the iteration adjustment until the current optimal solution set is not satisfied by the optimal solution set, and the current optimal solution set is controlled until the current optimal solution is not satisfied by the optimal solution set, and the current optimal solution set is controlled and the optimal solution set is not satisfied.
Further, setting P dynamic optimizing directions according to the P optimal solution fitness and the P differential solution fitness, and executing the steps includes:
The method comprises the steps of randomly selecting a first solution set from P solution sets, obtaining first optimal solution fitness, first difference solution fitness and a plurality of first bad solution fitness, calculating first bad solution fitness means of the plurality of first bad solution fitness, calculating absolute values of deviations of the first optimal solution fitness, the first difference solution fitness and the first bad solution fitness means respectively, determining a first optimal solution deviation value and a first bad solution deviation value, setting a first dynamic optimizing direction according to the first optimal solution deviation value and the first bad solution deviation value, and adding the first dynamic optimizing direction to the P dynamic optimizing directions, wherein if the first optimal solution deviation value is larger than or equal to the first bad solution deviation value, the first dynamic optimizing direction is in optimal adjustment, and if the first optimal solution deviation value is smaller than the first bad solution deviation value, the first dynamic optimizing direction is in bad adjustment.
Further, the evaporation control parameter optimizing module 30 according to the embodiment of the present application is further configured to perform the following steps:
The method comprises the steps of determining a first optimizing fitness of a first direction according to the first dynamic optimizing direction, calculating the ratio of a plurality of first inferior solution fitness to the first optimizing fitness respectively, determining a plurality of compensation coefficients, wherein the compensation coefficient is the ratio of the first inferior solution fitness to the first optimizing fitness if the first optimizing fitness is the first optimizing fitness, the compensation coefficient is the ratio of the first inferior solution fitness to the first optimizing fitness if the first optimizing fitness is the first poor solution fitness, the compensation coefficient is the ratio of the first inferior solution fitness to the first inferior solution fitness if the first optimizing fitness is the first poor solution fitness, optimizing the initial optimizing step according to the plurality of compensation coefficients, determining a first step set, and adding the first step set to the adapting optimizing step.
The foregoing detailed description of the evaporation energy-saving adaptive control method under multi-parameter analysis will clearly enable those skilled in the art to know the evaporation energy-saving adaptive control system under multi-parameter analysis in this embodiment, and for the system disclosed in the second embodiment, the system has corresponding functional modules and beneficial effects as corresponding to the method disclosed in the first embodiment, and relevant places refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411923679.1A CN119689871B (en) | 2024-12-25 | 2024-12-25 | Evaporation energy-saving adaptive control method and system based on multi-parameter analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411923679.1A CN119689871B (en) | 2024-12-25 | 2024-12-25 | Evaporation energy-saving adaptive control method and system based on multi-parameter analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119689871A true CN119689871A (en) | 2025-03-25 |
| CN119689871B CN119689871B (en) | 2025-09-05 |
Family
ID=95042505
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411923679.1A Active CN119689871B (en) | 2024-12-25 | 2024-12-25 | Evaporation energy-saving adaptive control method and system based on multi-parameter analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119689871B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120805506A (en) * | 2025-09-08 | 2025-10-17 | 中科信创技术有限公司 | Auxiliary design optimization method and system for cable laying |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6505475B1 (en) * | 1999-08-20 | 2003-01-14 | Hudson Technologies Inc. | Method and apparatus for measuring and improving efficiency in refrigeration systems |
| KR20120085140A (en) * | 2011-10-04 | 2012-07-31 | 주식회사 켐에쎈 | Svrc model predicting evaporation heat of pure organic compound |
| US8996141B1 (en) * | 2010-08-26 | 2015-03-31 | Dunan Microstaq, Inc. | Adaptive predictive functional controller |
| WO2022244389A1 (en) * | 2021-05-20 | 2022-11-24 | Mitsubishi Electric Corporation | Robust adaptive dynamic mode decomposition for modeling, prediction, and control of high dimensional physical systems |
| CN117577208A (en) * | 2023-11-28 | 2024-02-20 | 深圳职业技术大学 | A method for optimizing process parameters of evaporation process based on effective performance evaluation index |
| WO2024126785A1 (en) * | 2022-12-16 | 2024-06-20 | Katholieke Universiteit Leuven | Hybrid modelling and control of industrial batch distillation processes |
| CN118643906A (en) * | 2024-07-13 | 2024-09-13 | 徐州丰成盐化工有限公司 | Evaporation energy-saving optimization method and system based on intelligent control |
-
2024
- 2024-12-25 CN CN202411923679.1A patent/CN119689871B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6505475B1 (en) * | 1999-08-20 | 2003-01-14 | Hudson Technologies Inc. | Method and apparatus for measuring and improving efficiency in refrigeration systems |
| US8996141B1 (en) * | 2010-08-26 | 2015-03-31 | Dunan Microstaq, Inc. | Adaptive predictive functional controller |
| KR20120085140A (en) * | 2011-10-04 | 2012-07-31 | 주식회사 켐에쎈 | Svrc model predicting evaporation heat of pure organic compound |
| WO2022244389A1 (en) * | 2021-05-20 | 2022-11-24 | Mitsubishi Electric Corporation | Robust adaptive dynamic mode decomposition for modeling, prediction, and control of high dimensional physical systems |
| WO2024126785A1 (en) * | 2022-12-16 | 2024-06-20 | Katholieke Universiteit Leuven | Hybrid modelling and control of industrial batch distillation processes |
| CN117577208A (en) * | 2023-11-28 | 2024-02-20 | 深圳职业技术大学 | A method for optimizing process parameters of evaporation process based on effective performance evaluation index |
| CN118643906A (en) * | 2024-07-13 | 2024-09-13 | 徐州丰成盐化工有限公司 | Evaporation energy-saving optimization method and system based on intelligent control |
Non-Patent Citations (2)
| Title |
|---|
| YU ZHENG，等: "Multi-Algorithm Hybrid Optimization of Back Propagation (BP) Neural Networks for Reference Crop Evapotranspiration Prediction Models", WATER, 31 December 2023 (2023-12-31), pages 1 - 13 * |
| 丁绪东，等: "数据驱动的蒸发器在线建模方法", 计算机与现代化, no. 327, 31 December 2022 (2022-12-31), pages 22 - 31 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120805506A (en) * | 2025-09-08 | 2025-10-17 | 中科信创技术有限公司 | Auxiliary design optimization method and system for cable laying |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119689871B (en) | 2025-09-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN119689871B (en) | Evaporation energy-saving adaptive control method and system based on multi-parameter analysis | |
| CN105807609A (en) | Multi-parameter mine spraying heat exchange efficiency calculation method-based spraying system | |
| CN118909763B (en) | A fermentation tank pressure self-adaptive regulating device | |
| CN118262810A (en) | A polyethylene performance prediction method based on extreme gradient boosting tree model | |
| CN118963150A (en) | A control parameter optimization method and device for multi-stage reactor | |
| CN118551308A (en) | Edible fungus breeding growth process monitoring method and system based on data analysis | |
| CN119089803B (en) | Sizing process parameter optimization method based on particle swarm optimization | |
| CN111967689A (en) | Model and method for wind power generation prediction by combining multivariate and stepwise linear regression and artificial neural network | |
| Marenbach et al. | Data-driven structured modelling of a biotechnological fed-batch fermentation by means of genetic programming | |
| CN118795855B (en) | A kind of can production control method | |
| Li et al. | A nonlinear PID-enhanced adaptive latent factor analysis model | |
| CN112836431A (en) | Fault prediction method for penicillin fermentation process based on PSO-LSTM | |
| CN120180809A (en) | Multi-objective optimization method, equipment and medium for flat permanent magnet synchronous linear motor | |
| CN119721192A (en) | An optimization method for artificial intelligence to perform tasks | |
| CN115496516A (en) | Power plant coal burning big data analysis method based on self-adaptive support vector machine | |
| CN110376892A (en) | A kind of aircraft automatic calibrating method | |
| Crnkić et al. | Simulation-based optimization of capital greenhouse gas emissions in water distribution systems using Particle Swarm Optimization | |
| CN120340717B (en) | A platinum group nanomaterial catalyst formulation optimization method and system | |
| CN119720723B (en) | Population selection method and electronic device for optimizing natural gas liquefaction process parameters | |
| Li et al. | Moisture Prediction in Tobacco Drying Process Based on the NRBO-GRU Algorithm | |
| CN121128952A (en) | A predictive control method for a wire drying machine based on multi-model fusion | |
| Yu et al. | Development of Corn Drying Process Optimization System Based on LSTM and NSGA‐II Algorithms | |
| Han et al. | Research on Performance Optimization of Cooling Fan Based on GA-BP Neural Network | |
| Liu et al. | Optimizing parameters of fuzzy controller based on genetic algorithm | |
| CN121210132A (en) | Intelligent rendering systems and methods based on cloud computing and AI |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |