CN119714451B - Method for improving flow measurement and control precision of thermal micro mass flowmeter - Google Patents
Method for improving flow measurement and control precision of thermal micro mass flowmeter Download PDFInfo
- Publication number
- CN119714451B CN119714451B CN202510226031.7A CN202510226031A CN119714451B CN 119714451 B CN119714451 B CN 119714451B CN 202510226031 A CN202510226031 A CN 202510226031A CN 119714451 B CN119714451 B CN 119714451B
- Authority
- CN
- China
- Prior art keywords
- fitting
- mass flowmeter
- flow
- signal
- function
- 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.)
- Active
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Flow Control (AREA)
Abstract
本发明提供一种提高热式微小质量流量计流量测控精度的方法，涉及流量测量技术领域，该方法包括获取热式微小质量流量计采集的气体流量信号数据，并使用多阶反馈式整型无限脉冲响应滤波器对信号数据进行滤波处理；对滤波后的信号数据进行数据校准拟合，并根据实际工况从拟合结果中筛选出所需工况下的曲面集；基于曲面集，根据实时采集的信号数据计算出对应的流量值，并基于流量值向所述热式微小质量流量计发送阀控指令，实现对气体流量的控制。本发明通过创新的滤波算法、拟合技术、模型构建及数据处理策略，有效提升热式微小质量流量计测量精度、响应速度和抗干扰能力，优化数据处理与存储，增强工况适应能力并减少阀控指令滞后影响。
The present invention provides a method for improving the flow measurement and control accuracy of a thermal micro-mass flowmeter, which relates to the field of flow measurement technology. The method includes obtaining gas flow signal data collected by the thermal micro-mass flowmeter, and using a multi-order feedback integer infinite impulse response filter to filter the signal data; performing data calibration fitting on the filtered signal data, and screening out a surface set under the required working condition from the fitting result according to the actual working condition; based on the surface set, calculating the corresponding flow value according to the real-time collected signal data, and sending a valve control instruction to the thermal micro-mass flowmeter based on the flow value to achieve control of the gas flow. The present invention effectively improves the measurement accuracy, response speed and anti-interference ability of the thermal micro-mass flowmeter through innovative filtering algorithms, fitting techniques, model construction and data processing strategies, optimizes data processing and storage, enhances the adaptability to working conditions and reduces the lag effect of valve control instructions.
Description
Technical Field
The invention relates to the technical field of flow measurement, in particular to a method for improving flow measurement and control precision of a thermal micro mass flowmeter.
Background
In the fields of industrial automation, energy management, environmental monitoring, numerous precision manufacturing, photovoltaic and semiconductor manufacturing, accurate measurement and control of gas flow is of paramount importance. Traditional flow meter technologies, such as turbine flow meters, orifice plate flow meters, and ultrasonic flow meters, play an important role in a wide range of applications, but are exposed to significant limitations in the face of extreme conditions such as high temperature, high pressure, low flow rates, or very small flow rates.
The turbine flowmeter is difficult to accurately measure micro flow due to insufficient gas kinetic energy in a low flow rate environment, is easily influenced by gas viscosity and density change, has poor measurement accuracy stability, has large pressure loss and high energy consumption when in small flow rate measurement, has extremely strict installation requirements on measurement accuracy, causes larger measurement errors due to small installation deviation, and has the advantages of increased signal attenuation and interference caused by increased ultrasonic propagation characteristics under the working conditions of complicated gas medium and large temperature and pressure fluctuation under the condition of the ultrasonic flowmeter.
Thermal micro mass flow meters have been developed that work on the principle of gas heat exchange with heating elements, with unique advantages in terms of micro flow measurement. However, the existing thermal micro mass flowmeter still has a plurality of problems to be solved. In terms of improvement of measurement accuracy, the sensor is easily influenced by factors such as ambient temperature, electromagnetic interference and the like, the measured data has large noise, the calibration method is coarse, a single calibration model cannot adapt to complex working condition changes, so that the measurement accuracy is limited, the signal processing and data calculation efficiency is poor in response speed, rapid flow change is difficult to track in time, the data processing and communication functions lack of an efficient data processing algorithm and a stable communication mechanism, the requirements of intelligent and remote control in the 4.0-era industry cannot be met, and the storage, transmission and collaborative work of mass measured data with an upper computer are difficult and heavy.
With the development of industrial technology to high precision, intellectualization, energy conservation and environmental protection, the performance of the flowmeter in each industry is expected to continuously climb, an innovative measuring and controlling method is urgently needed to break through the bottleneck of the traditional thermal micro mass flowmeter, and gas flow measurement and control with high precision, quick response, strong anti-interference and intelligent control are realized, so that the complex and changeable production requirements of the modern industry are met, the running efficiency and safety of the system are improved, the energy consumption and the cost are reduced, and the industrial automation technology is promoted to move to new steps.
Disclosure of Invention
Therefore, the embodiment of the invention provides a method for improving the flow measurement and control precision of a thermal micro mass flowmeter, which is used for solving the problems of insufficient precision, slow response speed, easiness in environmental interference, easiness in influence of coarse errors in a calibration method, large data processing calculation amount, limited storage resources, difficulty in adapting to complex fitting results, poor working condition change coping capability, influence on measurement and control precision due to valve control command hysteresis and the like in the prior art when the micro mass flow is measured.
In order to solve the above problems, an embodiment of the present invention provides a method for improving flow measurement and control accuracy of a thermal micro mass flowmeter, which is applied to a cloud server in a measurement and control system, and the method includes:
Acquiring gas flow signal data acquired by a thermal micro mass flowmeter, and filtering the signal data by using a multi-order feedback integer infinite impulse response filter;
Performing data calibration fitting on the filtered signal data, and screening a curved surface set under a required working condition from fitting results according to actual working conditions;
based on the curved surface set, calculating a corresponding flow value according to signal data acquired in real time, and sending a valve control instruction to the thermal micro mass flowmeter based on the flow value to realize control of gas flow;
The method comprises the steps of establishing a self-adaptive multi-working-condition process system model, pre-training the self-adaptive multi-working-condition process system model by utilizing transfer learning, and updating the curved surface set by utilizing a trained model, wherein the thermal micro mass flowmeter is in data interaction with a cloud server, stores a certain amount of signal data, and periodically uploads the stored signal data to the cloud server, and the cloud server performs model fitting according to a preset algorithm and sends the fitted model to the thermal micro mass flowmeter to complete model updating.
Preferably, the filtering of the signal data using a multi-order feedback integer infinite impulse response filter includes:
The original signal In (n) passes through the multi-order feedback integer infinite impulse response filter to obtain a filtered signal Out1 (n);
calculating a difference signal Sub (n) of the original signal and the filtered signal Out1 (n) =in (n) -Out1 (n);
Shaping and filtering the difference signal Sub (n) to obtain a filtered signal Out2 (n), and adding the filtered signal Out2 (n) and the filtered signal Out1 (n) to obtain a filtered signal Out (n): out (n) =out2 (n) +out1 (n);
Subtracting the filtered signal Out (n) from the original signal to obtain a difference signal Sub1 (n): sub1 (n) =in (n) -Out (n);
Shaping and filtering the difference signal Sub1 (n) to obtain a filtered signal Out3 (n), and adding the filtered signal Out3 (n) and the filtered signal Out (n) to obtain a filtered signal Out4 (n): out4 (n) =out3 (n) +outn;
Repeating the steps of calculating the difference and adding the filtered signals until a predetermined filtering order is reached.
Preferably, the process of performing data calibration fitting on the filtered signal data and screening the curved surface set under the required working condition from the fitting result according to the actual working condition includes:
Selecting a plurality of sample points collected at different temperatures, different pressures and different flows based on the filtered signal data, wherein each sample point contains analog-digital conversion values of the corresponding temperatures, pressures and flows to form a multidimensional matrix;
fitting the multi-dimensional matrix by using a hypersurface mobile variable truncated fitting algorithm to obtain a surface fitting function;
and combining the surface fitting functions to obtain a hypersurface fitting function, and screening a surface set under a required working condition from the hypersurface fitting function according to the actual working condition.
The process of fitting the multidimensional matrix by using the hypersurface mobile variable truncated fitting algorithm to obtain a surface fitting function comprises the following steps:
in each support domain, a node combination of k+1 points is extracted from a sample of sampled data of size N, where k+1<N is common Combining different nodes;
Performing least square fitting on each node combination to obtain corresponding fitting coefficients Wherein;
The fitting coefficient of each node combination is brought into a total sample, and the residual error square of each point is calculatedAnd squares all residualsArranged in ascending order, i.e;
And (3) carrying out difference value calculation on the squares of adjacent residual errors after ascending order arrangement: Wherein Representing the difference;
Setting variable cut-out value WhereinThe representation is made in a round-up form,Representing the parameter value that enables the function to take the maximum value among all the parameters of one function,The minimum function is taken;
The first h items in ascending order are selected to be summed to obtain a target value of local fitting WhereinRepresenting all parameters of a function, enabling the function to obtain a parameter value with a minimum value, taking the target value of the local fitting as a fitting regression coefficient of the supporting domain, and reconstructing a curved surface fitting function according to the fitting regression coefficient;
In each support domain to As a basis function,Constructing a surface fitting function as a coefficient matrixWhere T represents a transpose operation and applying weighted least squares solution within the support domainWherein the weighted least squares objective function is:
;
Wherein the method comprises the steps of As a function of the object to be processed,Representation pointsIs used for the weight of the (c),To support domain midpoint and support domain follow-upThe movement of the device is carried out,Representing taking an upper bound function, i.e. greater than or equal toIs used as a starting material for the engine,In order to support the radius of the domain,Representation pointsThe j-th basis function of the evaluation is located,Is indicated at the pointAnd (3) observing values, wherein m represents the number of elements in the linear basis function, n represents the number of sampling points, and a coefficient matrix is obtained by deriving an objective function.
Preferably, the process of screening the curved surface set under the required working condition from the fitting result according to the actual working condition includes:
Screening a curved surface set under a required working condition from the fitting result according to the actual working condition;
Projecting curved surfaces at different temperatures, different pressures and the same flow to obtain a correlation coefficient between the temperature and the pressure curve;
And reserving a curve segment corresponding to the absolute value of the correlation coefficient smaller than the set threshold value, and carrying out linear characterization substitution on the curve segment corresponding to the absolute value of the correlation coefficient larger than or equal to the set threshold value.
Preferably, the method further comprises the step of establishing a valve control command delay compensation model, wherein the valve control command delay compensation model is based on a working condition matrix of three parameters of pressure, opening degree and flow, and linear expressions of delay time and each parameter are obtained through least square fitting.
Preferably, the valve control command delay compensation model is expressed as:
;
Wherein, In order to predict the delay time of the delay time,In the case of a pressure force, the pressure,In order to achieve the degree of opening,In order to be a flow rate,AndIs a valve control unit related parameter.
The embodiment of the invention also provides a gas flow measurement and control system based on the thermal type micro mass flowmeter, which comprises a cloud server, wherein the cloud server is used for realizing the method for improving the flow measurement and control precision of the thermal type micro mass flowmeter.
Preferably, the system further comprises a plurality of thermal micro mass flowmeters, wherein the thermal micro mass flowmeters are used for collecting gas flow signal data, sending the signal data to the cloud server and receiving a valve control command delay compensation model trained by the cloud server.
Preferably, the thermal micro mass flowmeter includes:
the data sampling unit is used for collecting gas flow signal data;
the valve control unit is used for controlling the gas flow;
the communication unit is in communication connection with the cloud server and is used for transmitting the gas flow signal data acquired by the data sampling unit and the working state of the valve control unit, and receiving a curved surface set and a valve control command delay compensation model which are required by the flowmeter and sent by the cloud server;
The control unit is connected with the data sampling unit, the valve control unit and the communication unit, and is used for sending the gas flow signal data acquired by the data sampling unit and the working state of the valve control unit to the cloud server through the communication unit, and adaptively issuing the valve control command to the proportional valve according to a valve control command delay compensation model trained by the cloud server.
Preferably, the data sampling unit adopts a capillary tube with a set specification as a sensing element, the outer wall of the capillary tube is wound and used as windings of the heater and the detection assembly, the windings are divided into two or more groups, and the sensing element collects gas flow signal data by detecting the change of the tube wall temperature when gas flows through the capillary tube.
From the above technical scheme, the invention has the following beneficial effects:
(1) The invention adopts a multi-order feedback integer infinite impulse response filter for filtering, combines a hypersurface moving variable truncated fitting algorithm, effectively removes noise, reduces the influence of coarse errors and obviously improves the measurement precision.
(2) The invention establishes a self-adaptive multi-working condition process system model, pre-trains by utilizing transfer learning, and combines clustering and LSTM prediction to update a curved surface set, thereby being capable of rapidly adapting to complex and changeable working conditions.
Drawings
For a clearer description of embodiments of the invention or of solutions in the prior art, reference will be made to the accompanying drawings, which are intended to be used in the examples, for a clearer understanding of the characteristics and advantages of the invention, by way of illustration and not to be interpreted as limiting the invention in any way, and from which, without any inventive effort, a person skilled in the art can obtain other figures. Wherein:
FIG. 1 is a flow chart of a method for improving flow measurement and control accuracy of a thermal micro mass flowmeter according to an embodiment;
FIG. 2 is a flowchart of a method for fitting a multi-dimensional matrix to obtain a surface fitting function using a hypersurface mobile variable truncation fitting algorithm in an embodiment;
FIG. 3 is a block diagram of a measurement and control system provided in an embodiment;
fig. 4 is a schematic structural diagram of a thermal micro mass flowmeter according to an embodiment.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more clear, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
The method aims to solve the problems that in the prior art, the precision is insufficient, the response speed is low, the environment interference is easy to occur, the calibration method is easy to be influenced by coarse errors, the calculated amount of data processing is large, the storage resources are limited, the complex fitting result is difficult to adapt to, the working condition change response capability is poor, the measurement control precision is influenced by the hysteresis of a valve control instruction, and the like when the micro mass flow is measured. As shown in fig. 1, an embodiment of the present invention provides a method for improving flow measurement and control accuracy of a thermal micro mass flowmeter, which is applied to a cloud server in a measurement and control system, and the method includes:
Acquiring gas flow signal data acquired by a thermal micro mass flowmeter, and filtering the signal data by using a multi-order feedback integer infinite impulse response filter;
Performing data calibration fitting on the filtered signal data, and screening a curved surface set under a required working condition from fitting results according to actual working conditions;
Based on the curved surface set, a corresponding flow value is calculated according to the signal data acquired in real time, and a valve control instruction is sent to the thermal type micro mass flowmeter based on the flow value, so that the control of the gas flow is realized. According to the technical scheme, the method for improving the flow measurement and control precision of the thermal type micro mass flowmeter is provided, firstly, after signal data acquired by the thermal type micro mass flowmeter are acquired, the signal data are filtered by using a multi-order feedback integer infinite impulse response filter, noise can be effectively removed by the filter, environmental interference is reduced, the calculated amount is small, and the problems of insufficient precision, easiness in environmental interference and large calculated amount of data processing are solved. And then, performing calibration fitting on the filtered data, adopting a hypersurface moving variable truncated fitting algorithm and screening a curved surface set of a required working condition, avoiding the influence of coarse errors, adapting to complex fitting results, and solving the problems that the calibration method is easily influenced by the coarse errors and the storage resources are limited. And finally, calculating a flow value based on the curved surface set and the data acquired in real time, sending a valve control instruction, and simultaneously establishing a self-adaptive multi-working-condition process system model and a valve control command delay compensation model, so that the working condition change coping capacity is enhanced, the hysteresis influence of the valve control instruction is reduced, and high-precision gas flow measurement and control are realized.
The operating environment is not ideal due to the thermal type micro mass flow meter. In practice, there are various interference factors, which cause the signals output by the thermal micro mass flowmeter to be mixed with random noise. These random noises can obscure the signal acquired by the thermal micro mass flowmeter. For example, an originally clear flow signal may not accurately reflect a true flow condition due to noise interference. This has a serious impact on the reliability and accuracy of the flowmeter, making the measurement result unreliable and inaccurate, and possibly leading to incorrect guidance for subsequent control and analysis based on flow data.
In order to solve the problem, the invention adopts a multi-order feedback integer Infinite Impulse Response (IIR) filter to carry out filtering treatment on gas flow signal data acquired by the thermal type micro mass flowmeter. The method processes the original signal with noise, removes noise interference and makes the signal clear and accurate through specific design steps, thereby improving the reliability and accuracy of measurement of the thermal micro mass flowmeter. The specific design steps comprise:
The original signal In (n) passes through the multi-order feedback integer infinite impulse response filter to obtain a filtered signal Out1 (n);
Calculating a difference signal Sub (n) of the original signal and the filtered signal Out1 (n) =in (n) -Out1 (n);
Shaping and filtering the difference signal Sub (n) to obtain a filtered signal Out2 (n), and adding the filtered signal Out2 (n) and the filtered signal Out1 (n) to obtain a filtered signal Out (n): out (n) =out2 (n) +out1 (n);
Subtracting the filtered signal Out (n) from the original signal to obtain a difference signal Sub1 (n): sub1 (n) =in (n) -Out (n);
Shaping and filtering the difference signal Sub1 (n) to obtain a filtered signal Out3 (n), and adding the filtered signal Out3 (n) and the filtered signal Out (n) to obtain a filtered signal Out4 (n): out4 (n) =out3 (n) +outn;
Repeating the steps of calculating the difference and adding the filtered signals until a predetermined filtering order is reached.
Compared with the traditional IIR digital filter, the integral IIR filter has large calculated amount, and the transfer function coefficients of the integral IIR filter are integers, so that floating point operation is not needed, the calculated amount is greatly reduced, the condition that the thermal type micro mass flowmeter is used as an edge end to calculate the limited resource is met, and meanwhile, the requirement for monitoring the gas flow in real time is met.
Because the industrial field working condition has complex scene, various gases and scarce working condition data, a special method is needed to calibrate and fit the filtered signal data, and a curved surface set under the required working condition is screened from the fitting result according to the actual working condition.
Specifically, firstly, based on the filtered signal data, a plurality of sample points collected at different temperatures, different pressures and different flows are selected, wherein each sample point contains analog-digital conversion values of the corresponding temperatures, pressures and flows, and a multidimensional matrix is formed. And then fitting the multi-dimensional matrix by using a hypersurface mobile variable truncated fitting algorithm to obtain a surface fitting function. And finally, combining the surface fitting functions to obtain a hypersurface fitting function, and screening a surface set under a required working condition from the hypersurface fitting function according to the actual working condition.
Further, as shown in fig. 2, the above-mentioned process of fitting the multi-dimensional matrix using the hypersurface moving variable truncated fitting (Metasurface Moving Variable Trimmed Square, MMVTS) algorithm to obtain the surface fitting function includes:
in each support domain, a node combination of k+1 points is extracted from a sample of sampled data of size N, where k+1<N is common Different node combinations.
Performing Least Squares (LS) fitting on each node combination to obtain corresponding fitting coefficientsWherein。
The fitting coefficient of each node combination is brought into a total sample, and the residual error square of each point is calculatedAnd squares all residualsArranged in ascending order, i.e。
And (3) carrying out difference value calculation on the squares of adjacent residual errors after ascending order arrangement: Wherein Representing the difference.
Setting variable cut-out valueWhereinThe representation is made in a round-up form,Representing the parameter value that enables the function to take the maximum value among all the parameters of one function,Is a function taking a minimum value. Compared with the traditional interception constant extractionThe variable cut-off value can be used for utilizing the sample point to the greatest extent, so that the utilization rate is improved while coarse errors are eliminated.
Selecting the front in ascending orderSumming the terms to obtain a locally fitted target valueWhereinThe method is characterized in that the method comprises the steps of representing all parameters of a function, enabling the function to obtain a parameter value of a minimum value, taking a target value of local fitting as a fitting regression coefficient of the supporting domain, and reconstructing a curved surface fitting function according to the fitting regression coefficient.
Further, analog-to-digital (AD) conversion values acquired at different temperatures and different flows are selected in each support domain toAs a basis function,Constructing a surface fitting function as a coefficient matrixWhere T represents a transpose operation and applying weighted least squares solution within the support domainWherein the weighted least squares objective function is:
;
Wherein the method comprises the steps of As a function of the object to be processed,Representation pointsIs used for the weight of the (c),To support domain midpoint and support domain follow-upThe movement of the device is carried out,Representing taking an upper bound function, i.e. greater than or equal toIs used as a starting material for the engine,In order to support the radius of the domain,Representation pointsThe j-th basis function of the evaluation is located,Is indicated at the pointObservations at m represents the number of basis functions and n represents the number of sample points in the support domain that participate in the local fit. Solving coefficient matrixThe problem of the minimum value of the objective function J is converted, namely, the problem can be solved by deriving the J:
;
And (5) solving a coefficient matrix:
;
Wherein the method comprises the steps of
;
;
;
;
;
Wherein, The weighted covariance matrix is represented by a matrix of weighted covariance,The matrix of weighted feature vectors is represented,A column matrix of calibration points representing the actual gas flow,Representing a feature matrix consisting of linear basis,Representing the weight diagonal matrix of the device,The weight value representing the nth sampling point,An nth element representing an mth term in the linear base,Representing the number of elements in the linear basis function,The number of the sampling points is represented,The AD value that is acquired is represented,Indicating the actual gas flow.
Further, during the fitting process, the weight functionIs also critical. The invention adopts an exponential weight function, and has the function of enabling the nearest calibration point to have higher weight. Therefore, during fitting, calibration data closer to the target point has a larger influence on the fitting result, and the fitting accuracy is improved. For the exponential weighting functionThe value is selected according to experience adjustment in the traditional method, but the invention calculates according to the measuring range of the thermal micro mass flowmeter required by industrial production, and the calculation formula is as follows:
;
Wherein the method comprises the steps of 、Respectively the maximum and minimum of the range.The magnitude of the value affects the rate of decay of the weight,The larger the weight is, the faster the weight decays, and the more recent signal data is of interest; The smaller the opposite. When the calibration points are selected, the range is large, and the calibration points are selected more sparsely, so that the exponential weight function needs to pay more attention to the closer calibration points when the range is large. The calculation mode increases nonlinearity while meeting the inversely proportional characteristic of the range and the value, so that the fitting curved surface approximation effect is better and the method can adapt to various range demands.
Further, selecting AD values acquired under different pressures and different flows, and obtaining another curve fitting function according to the method. The hypersurface fitting function can be obtained by combining the two surface fitting functions, the hypersurface fitting function considers various variable factors and is more suitable for flow measurement under different working conditions, but hypersurface fitting simultaneously brings the problem of calculation amount storage quantity index rising, and is not suitable for edge end equipment with limited calculation capacity such as a flowmeter, therefore, at the end of MMLTS algorithm, a curved surface set under the required working conditions is screened out from fitting results according to possible working conditions actually faced, curved surface projections under different temperatures, different pressures and the same flow are carried out, correlation coefficients between the temperature and the pressure curved lines are obtained, curve segments corresponding to the absolute value of the correlation coefficients being smaller than a set threshold (set to be 0.5) are reserved, linear representation substitution is carried out on the curve segments corresponding to the absolute value of the correlation coefficients being larger than or equal to the set threshold (set to be 0.5), the dimension reduction is realized to a certain extent, and the volume of the curved surface set is compressed.
In order to improve the precision of the thermal type micro mass flowmeter and better cope with different working conditions, the invention establishes a self-adaptive multi-working-condition process system model with continuous learning capability. In the initial stage of model operation, the problem of scarce and scarce working condition data is faced, and the data volume required by model training is difficult to satisfy. To solve this problem, the present invention uses a method of transfer learning to perform model pre-training by using similar valve-controlled datasets. The method is to migrate knowledge of other related fields (similar to a valve control data set) into the current model training, and take the knowledge as an initial model of the model, so that the model has certain basic capability in an initial stage, and the model training can be started by virtue of experience of the existing related data despite insufficient data quantity.
The thermal type micro mass flowmeter and the cloud server perform data interaction, the thermal type micro mass flowmeter stores a certain amount of signal data, the stored signal data are uploaded to the cloud server at regular intervals, the cloud server performs model fitting according to a preset algorithm according to an uploaded data set, and the fitted model is issued to the thermal type micro mass flowmeter to finish model updating.
Specifically, as the thermal micro mass flowmeter continuously uploads data, a data set for the thermal micro mass flowmeter control unit valve control command hysteresis compensation task and the operating condition change prediction task is gradually formed, and the data set is continuously expanded. In order to ensure that the model always learns and predicts based on the latest and effective data, the model adopts forgetting factors to discard the overaged data. The reason for this is that some early data may not reflect the current working condition and system state over time, and by discarding old data, the model can continuously update knowledge and adapt to new conditions, so that timeliness of the model is ensured, and operations such as analysis and prediction of the working condition of the flowmeter can be continuously and accurately performed.
The data uploaded by the thermal micro mass flowmeter comprises pressure, temperature, flow and the working state of the valve control unit. Considering that the industrial process is relatively fixed, the change rule of the working state of the valve control unit can reflect the working mode of the thermal micro mass flowmeter. Based on the method, the thermal type micro mass flowmeters under the same working procedure are clustered according to the change rule of the working state of the valve control unit. For example, in chemical production, the valve control units of the thermal micro mass flowmeters in specific reaction stages are similar in operation, and these thermal micro mass flowmeters are classified into one type because the working conditions of these thermal micro mass flowmeters are similar. And the clustering result is used as a label division data set, so that different kinds of data represent different working conditions. A long short term memory network (LSTM) training model is used to predict operating condition changes. The LSTM is suitable for processing time series data, can capture long-term dependency of the data, and learn data change trend and characteristics under different working conditions. The curve set required by the thermal micro mass flowmeter can be updated in time through LSTM prediction of the working condition change. The curved surface set is related to flow measurement calculation, and flow measurement accuracy under different working conditions can be guaranteed by timely updating. The utilization rate of the limited storage space of the thermal type micro mass flowmeter is improved, a large amount of useless data is prevented from being stored, and only the curved surface set data related to the current working condition is reserved and updated. Meanwhile, the adaptability of the flowmeter to various working conditions is enhanced, so that the thermal micro mass flowmeter can rapidly adjust the measurement calculation basis according to the working condition change, the measurement and control accuracy is ensured, and the complex and various flow measurement and control requirements in industrial production are met.
In this embodiment, based on the curved surface set obtained above, a corresponding flow value is calculated according to signal data acquired in real time, and a valve control instruction is sent to the thermal micro mass flowmeter based on the flow value, so as to control the gas flow.
In addition, since there is a delay between the valve control command of the thermal type micro mass flowmeter control unit and the actual operation of the proportional valve, the delay is serious for the micro flow, but if the delay is reduced, the proportional valve adjustment speed is too high, and the service life of the proportional valve is affected. The present invention thus utilizes datasets formed for the same process as clustered. Taking different working conditions into consideration, the turn-off delay time is different, a working condition matrix is built by three parameters of pressure, opening degree and flow, the working condition matrix is used as model input, the time interval between a proportional valve control instruction and the flow reaching a set value is sent out as a label, and the delay time and the pressure can be obtained by fitting based on a least square method) Linear expression of (2)0.8 Th power of delay time and opening degree) Linear expression of (2)Delay time and flow [ ]) Linear expression of (2):
;
Through the input of the working condition matrix, the linear relation between the first variable and the delay time is well known by fitting with a least square method, but the weight relation between different variables cannot be known, so that a valve control command delay compensation model is built, and the delay time and the pressure are @) 0.8 Th power of opening degree) Flow rate [ ]) All exhibit a better linear relationship, thus defining a complex variableThe method comprises the following steps:
;
The delay time is expressed as:
;
Wherein, In order to predict the delay time of the delay time,In the case of a pressure force, the pressure,In order to achieve the degree of opening,In order to be a flow rate,AndAs a parameter related to the valve control unit,As a composite variable, the number of the variables,. The valve control unit parameters of each thermal type micro mass flowmeter are considered to be different, so that the model can be independently trained according to the uploaded historical data, the valve control command delay compensation model is adapted to specific working conditions and proportional valve parameters, and then an adjusting command is sent to the proportional valve in advance according to model prediction, so that the actual control precision of the thermal type micro mass flowmeter is improved.
Example two
As shown in fig. 3, the invention provides a measurement and control system, which comprises a cloud server for implementing the method for improving the flow measurement and control precision of the thermal micro mass flowmeter in the first embodiment. In addition, the system also comprises a plurality of thermal micro mass flowmeters, wherein the thermal micro mass flowmeters are used for collecting gas flow signal data, sending the signal data to the cloud server and receiving a valve control command delay compensation model trained by the cloud server.
Further, as shown in fig. 4, the thermal type micro mass flowmeter includes a data sampling unit, a valve control unit, a communication unit, and a control unit. The system comprises a data sampling unit, a valve control unit, a communication unit, a control unit, a cloud server, a proportional valve, a valve control command delay compensation model, a control unit and a communication unit, wherein the data sampling unit is used for collecting gas flow signal data, the valve control unit is used for controlling gas flow, the communication unit is in communication connection with the cloud server and is used for transmitting the gas flow signal data collected by the data sampling unit and the working state of the valve control unit, the curved surface set required by a flowmeter and the valve control command delay compensation model sent by the cloud server are received by the communication unit, the control unit is connected with the data sampling unit, the valve control unit and the communication unit and is used for sending the gas flow signal data collected by the data sampling unit and the working state of the valve control unit to the cloud server through the communication unit, and the valve control command delay compensation model trained by the cloud server is used for adaptively issuing the valve control command to the proportional valve.
Furthermore, the data sampling unit adopts a capillary tube with set specification (the outer diameter is 0.5mm, the inner diameter is 0.4 mm) as a sensing element, so that the sensitivity and the accuracy of measurement are improved. The outer wall of the capillary tube is wound to be used as windings of the heater and the detection assembly, the windings are divided into two or more groups, and the sensing element acquires gas flow signal data by detecting the temperature change of the tube wall when gas flows through the capillary tube. The capillary tube of the thermal micro mass flowmeter is suitable for low-flow-rate gas flow, and has wider measuring range ratio, better adaptability and higher accuracy compared with the traditional flowmeter.
The embodiment of the measurement and control system is used for implementing the method for improving the flow measurement and control precision of the thermal micro mass flowmeter, so that the embodiment of the measurement and control system can be seen as an example part of the method for improving the flow measurement and control precision of the thermal micro mass flowmeter, and therefore, the embodiment of the measurement and control system can refer to the description of the corresponding examples of each part, and in order to avoid redundancy, the description is omitted.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It is apparent that the above examples are given by way of illustration only and are not limiting of the embodiments. Other variations and modifications will be apparent to persons skilled in the art from the foregoing description. It is not necessary here nor is it exhaustive of all embodiments. And obvious variations or modifications thereof are contemplated as falling within the scope of the present invention.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510226031.7A CN119714451B (en) | 2025-02-27 | 2025-02-27 | Method for improving flow measurement and control precision of thermal micro mass flowmeter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510226031.7A CN119714451B (en) | 2025-02-27 | 2025-02-27 | Method for improving flow measurement and control precision of thermal micro mass flowmeter |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119714451A CN119714451A (en) | 2025-03-28 |
| CN119714451B true CN119714451B (en) | 2025-05-13 |
Family
ID=95090981
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202510226031.7A Active CN119714451B (en) | 2025-02-27 | 2025-02-27 | Method for improving flow measurement and control precision of thermal micro mass flowmeter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119714451B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102538887A (en) * | 2012-01-19 | 2012-07-04 | 上海华强浮罗仪表有限公司 | Thermal mass flowmeter |
| CN114252120A (en) * | 2021-12-31 | 2022-03-29 | 江苏迅创科技股份有限公司 | Micro-range thermal type gas mass flow meter |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7054767B2 (en) * | 2004-02-12 | 2006-05-30 | Eldridge Products, Inc. | Thermal mass flowmeter apparatus and method with temperature correction |
| US11073845B2 (en) * | 2019-08-26 | 2021-07-27 | Hitachi Metals, Ltd. | Parasitic flow correction method and apparatus |
| CN111595402B (en) * | 2020-05-29 | 2021-11-23 | 合肥工业大学 | Constant-temperature difference type thermal gas mass flow meter |
| CN115996040A (en) * | 2021-10-20 | 2023-04-21 | 同响科技股份有限公司 | Signal Processing Method of Impulse Response Filter |
| CN219223837U (en) * | 2022-12-28 | 2023-06-20 | 承德万达高新仪表有限公司 | Novel thermal type gas mass flowmeter |
| CN117589256A (en) * | 2023-12-01 | 2024-02-23 | 天津大学 | Combined measuring device of gas-liquid two-phase flow differential pressure flowmeter and precession flowmeter |
-
2025
- 2025-02-27 CN CN202510226031.7A patent/CN119714451B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102538887A (en) * | 2012-01-19 | 2012-07-04 | 上海华强浮罗仪表有限公司 | Thermal mass flowmeter |
| CN114252120A (en) * | 2021-12-31 | 2022-03-29 | 江苏迅创科技股份有限公司 | Micro-range thermal type gas mass flow meter |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119714451A (en) | 2025-03-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113325721B (en) | Model-free adaptive control method and system for industrial system | |
| CN111444953A (en) | A Sensor Fault Monitoring Method Based on Improved Particle Swarm Optimization Algorithm | |
| CN114742278A (en) | Building energy consumption prediction method and system based on improved LSTM | |
| CN104864984B (en) | Microresponse device thermometry based on neutral net | |
| CN114117919B (en) | Instant learning soft measurement modeling method based on sample collaborative representation | |
| CN112597431B (en) | A flow interpolation and integration method and device based on integral diffusion factor | |
| CN112560334A (en) | Method for predicting bending resilience angle of pipe based on machine learning | |
| CN115034129B (en) | NOx emission concentration soft measurement method for thermal power plant denitration device | |
| CN110957011A (en) | On-line production parameter estimation method for continuous stirred reactor under unknown time-varying measurement noise | |
| CN108595892A (en) | Soft-measuring modeling method based on time difference model | |
| CN116562171A (en) | Error assessment method for online measurement of temperature and humidity | |
| CN109523077B (en) | Wind power prediction method | |
| CN119714451B (en) | Method for improving flow measurement and control precision of thermal micro mass flowmeter | |
| CN101916394A (en) | An Online Soft-Sensing Method Based on Knowledge Fusion | |
| CN119990481B (en) | Sewage water quality prediction method based on pyramid type three-dimensional model | |
| CN112016241B (en) | Soft measurement modeling method based on mLASSO-MLP model | |
| CN118839620B (en) | On-line soft measurement method and system for water quality parameters based on optimization algorithm | |
| CN120449639A (en) | A process optimization analysis system for vacuum concentration | |
| CN113657557A (en) | Hydrological forecasting method and hydrological forecasting system based on fusion of inversion data and measured data | |
| CN110366232B (en) | Sensor transmission energy control method for remote state estimation | |
| CN116821695A (en) | Semi-supervised neural network soft measurement modeling method | |
| CN116956433A (en) | PPLN-based time sequence combination prediction model | |
| CN114756604A (en) | A prediction method of monitoring time series data based on Prophet combination model | |
| CN110750884B (en) | Instant on-line instrument checksum diagnosis method through self-adaptive support vector machine algorithm | |
| Guo | Research on enhanced calibration of biological models using hybrid particle swarm optimization and gradient descent method |
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 |