CN118170184B - A temperature control method and system for a reactor based on artificial intelligence - Google Patents

Abstract

The invention provides a temperature control method and a temperature control system for a reaction kettle based on artificial intelligence. Belongs to the technical field of intelligent regulation and control. Monitoring the temperature field of the three-dimensional, real-time and micro-section inside the reaction kettle through a multi-mode temperature sensing array and a heat flow analysis model; and inputting the collected multidimensional temperature data to a controller; the controller generates an countermeasure network based on the built-in depth, the generator model generates a temperature control strategy by utilizing multidimensional temperature data and operation behaviors, and the judging device model evaluates the temperature control strategy according to actual temperature feedback; and carrying out real-time self-adaptive adjustment on the temperature control strategy generated by the depth generation countermeasure network and the real-time temperature data through the fuzzy logic reasoning system. By combining the multi-mode temperature sensing array and the heat flow analysis model, the three-dimensional, real-time and accurate monitoring of the temperature field of the micro-section inside the reaction kettle is realized, and the refinement degree of temperature control is improved.

Description

A temperature control method and system for a reactor based on artificial intelligence

Technical Field

The invention proposes a temperature control method and system for a reactor based on artificial intelligence, belonging to the technical field of intelligent regulation.

Background Art

As the core equipment of chemical reactions, the temperature control of reactors is crucial to ensure product quality, improve production efficiency and ensure safe production. Traditional reactor temperature control methods usually rely on simple temperature sensors and fixed control algorithms, which makes it difficult to achieve accurate, real-time and micro-segment temperature control. In addition, traditional control methods often have difficulty in making quick and accurate responses when faced with complex and changing reaction environments and conditions, thus affecting the operation and safety of the reactor.

In recent years, with the rapid development of artificial intelligence technology, more and more fields have begun to try to introduce artificial intelligence into temperature control. Technologies such as deep learning and generative adversarial networks have shown strong capabilities in processing complex data and optimizing control strategies, providing new possibilities for reactor temperature control. However, how to effectively apply artificial intelligence technology to reactor temperature control is still a problem that needs to be solved.

At present, although some artificial intelligence-based reactor temperature control methods have been proposed, most of them only focus on single temperature monitoring or control strategy generation, lacking comprehensive monitoring of the three-dimensional, real-time, micro-segment temperature field inside the reactor and real-time adaptive adjustment of the temperature control strategy. Therefore, developing a method that can comprehensively utilize advanced means such as multi-modal temperature sensor arrays, deep learning technology, and fuzzy logic reasoning systems to achieve accurate, real-time, and adaptive control of reactor temperature has important practical application value and broad market prospects.

Summary of the invention

The present invention provides a temperature control method and system for a reactor based on artificial intelligence to solve the problems mentioned in the above background technology:

The present invention proposes a temperature control method for a reactor based on artificial intelligence, characterized in that the method comprises:

S1. Use a multi-modal temperature sensor array and a heat flow analysis model to perform three-dimensional, real-time, micro-segment temperature field monitoring inside the reactor; and input the collected multi-dimensional temperature data into the controller;

S2, the controller is based on a built-in deep generative adversarial network. The generator model uses multi-dimensional temperature data and operation behavior to generate a temperature control strategy, and the judgement model evaluates the temperature control strategy based on actual temperature feedback;

S3, using the fuzzy logic reasoning system to adaptively adjust the temperature control strategy generated by the deep generative adversarial network and the real-time temperature data in real time;

S4. Based on the temperature control strategy after deep generative adversarial network and fuzzy logic fine-tuning, the controller controls the variable frequency heating device and intelligent cooling unit to adjust the temperature of the reactor.

Furthermore, the S1 includes:

A multi-modal temperature sensor array is deployed inside the reactor to continuously monitor the temperature inside the reactor in real time.

Each sensor node in the multi-modal temperature sensing array synchronously records the temperature value of its respective location and forms a data matrix that comprehensively reflects the temperature distribution of each point in the kettle;

The temperature data collected by each sensor is mapped to three-dimensional space coordinates through the preset sensor coordinate system, and the three-dimensional temperature field model inside the reactor is constructed by using the interpolation algorithm in combination with the spatial relationship between the sensors.

The heat flow analysis model is used to simulate and calculate the heat transfer process inside the reactor. The real-time temperature data directly measured by the multi-modal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set.

The integrated and processed multi-dimensional temperature data is transmitted to the controller in real time through a multi-channel transmission protocol.

Furthermore, the S2 includes:

The controller preprocesses the received multi-dimensional temperature data, combines the preprocessed multi-dimensional temperature data with the operation behavior data of the equipment, and converts them into feature vectors that can be recognized by deep learning;

The processed multi-dimensional temperature data and operational behavior characteristics are used as inputs and input into the generator model built into the deep generative adversarial network;

The generator model generates new temperature control instructions in real time according to the current temperature and operating conditions. The controller converts the temperature control strategy generated by the generator into specific equipment operation instructions and sends them to the temperature control system of the reactor for execution to adjust the actual temperature inside the reactor.

After implementing the new temperature control strategy, the temperature changes of each micro-section inside the reactor are continuously monitored in real time through the multi-modal temperature sensor array to obtain actual temperature feedback data;

The judgement model in the deep generative adversarial network receives the actual temperature feedback data, compares the actual temperature data with the preset threshold range of the temperature field, and outputs an evaluation score;

If the evaluation given by the judge is lower or higher than the expected threshold, the generator will adjust and optimize itself according to the feedback signal of the judge and generate a new control strategy again.

Furthermore, the S3 includes:

Acquire multi-dimensional real-time temperature data in real time. After receiving real-time or historical temperature data and operation behavior parameters, the deep generative adversarial network generates a preliminary temperature control strategy;

Establish a rule-based fuzzy logic reasoning system, and input the preliminary temperature control strategy generated by the deep generative adversarial network and the real-time temperature data into the fuzzy logic reasoning system;

The fuzzy logic reasoning system maps the real-time temperature data to the corresponding fuzzy sets and applies the predefined fuzzy logic rule base for reasoning;

The fuzzy logic reasoning system obtains one or a group of adaptively adjusted temperature control strategies based on the relationship between real-time temperature data and the expected temperature range, as well as the deep generative adversarial network strategy.

Furthermore, the S4 includes:

The controller receives the temperature control strategy instruction after adaptive adjustment outputted from the fuzzy logic inference system and parses the instruction;

Based on the parameters obtained by the analysis, the controller determines whether the reactor needs to be heated or cooled;

If heating is required, a control signal is sent to the variable frequency heating device to instruct it to heat the reactor according to a predetermined heating rate and duration;

If cooling is required, the controller sends a control signal to the intelligent cooling unit, instructing it to cool the reactor according to a predetermined cooling rate and duration;

During the control process, the controller continuously monitors the temperature changes inside the reactor and compares it with the target temperature; if there is a deviation between the actual temperature and the target temperature, the controller will adjust the control parameters of the heating and cooling devices in real time according to the adjustment strategy output by the fuzzy logic reasoning system.

The present invention proposes a temperature control system for a reactor based on artificial intelligence, the system comprising:

Temperature acquisition module: Through the multi-modal temperature sensor array and heat flow analysis model, the temperature field inside the reactor is monitored in three dimensions, real time, and in micro-segments; and the collected multi-dimensional temperature data is input into the controller;

Strategy generation module: The controller is based on a built-in deep generative adversarial network. The generator model uses multi-dimensional temperature data and operation behavior to optimize the temperature control strategy, and the judge model evaluates the temperature control strategy based on actual temperature feedback.

Adjustment module: Use the fuzzy logic reasoning system to perform real-time adaptive adjustment on the temperature control strategy generated by the deep generative adversarial network and the real-time temperature data;

Control execution module: Based on the temperature control strategy after deep generative adversarial network and fuzzy logic fine-tuning, the controller controls the variable frequency heating device and intelligent cooling unit to control the temperature of the reactor.

Furthermore, the temperature acquisition module includes:

Deployment module: deploy a multi-modal temperature sensor array inside the reactor to continuously monitor the temperature inside the reactor in real time through the multi-modal temperature sensor array;

Matrix generation module: Each sensor node in the multi-modal temperature sensing array synchronously records the temperature value of its respective location and forms a data matrix that fully reflects the temperature distribution of each point in the kettle;

Model building module: Through the preset sensor coordinate system, the temperature data collected by each sensor is mapped to the three-dimensional space coordinates, and combined with the spatial relationship between the sensors, the interpolation algorithm is used to build a three-dimensional temperature field model inside the reactor;

Simulation module: Through the heat flow analysis model, the heat transfer process inside the reactor is simulated and calculated. The real-time temperature data directly measured by the multi-modal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set;

Transmission module: transmits the integrated and processed multi-dimensional temperature data to the controller in real time through a multi-channel transmission protocol.

Furthermore, the strategy generation module includes:

Preprocessing module: The controller preprocesses the received multi-dimensional temperature data, combines the preprocessed multi-dimensional temperature data with the operation behavior data of the equipment, and converts it into a feature vector that can be recognized by deep learning;

Input module: takes the processed multi-dimensional temperature data and operation behavior characteristics as input and inputs them into the generator model built into the deep generative adversarial network;

Temperature adjustment module: The generator model generates new temperature control instructions in real time according to the current temperature and operating conditions. The controller converts the temperature control strategy generated by the generator into specific equipment operation instructions, sends them to the temperature control system of the reactor for execution, and adjusts the actual temperature inside the reactor;

Feedback data acquisition module: After executing the new temperature control strategy, continue to monitor the temperature changes of each micro-section inside the reactor in real time through the multi-modal temperature sensor array to obtain actual temperature feedback data;

Evaluation score output module: The judgement model in the deep generative adversarial network receives the actual temperature feedback data, compares the actual temperature data with the preset threshold range of the temperature field, and outputs the evaluation score;

Secondary generation module: If the evaluation given by the judge is lower or higher than the expected threshold, the generator will self-adjust and optimize according to the feedback signal of the judge and generate a new control strategy again.

Furthermore, the adjustment module includes:

Preliminary strategy generation module: real-time acquisition of multi-dimensional real-time temperature data. After receiving real-time or historical temperature data and operation behavior parameters, the deep generative adversarial network generates a preliminary temperature control strategy;

System establishment module: Establish a rule-based fuzzy logic reasoning system, and input the preliminary temperature control strategy generated by the deep generative adversarial network and the real-time temperature data into the fuzzy logic reasoning system;

Reasoning module: The fuzzy logic reasoning system maps the real-time temperature data to the corresponding fuzzy set and applies the pre-defined fuzzy logic rule base for reasoning;

Adaptive adjustment module: The fuzzy logic reasoning system obtains one or a group of adaptive temperature control strategies based on the relationship between real-time temperature data and the expected temperature range, as well as the deep generative adversarial network strategy.

Furthermore, the control execution module includes:

Parsing module: The controller receives the temperature control strategy instruction after adaptive adjustment output from the fuzzy logic reasoning system and parses the instruction;

Judgment module: Based on the analyzed parameters, the controller determines whether the reactor needs to be heated or cooled;

Heating module: If heating is required, a control signal is sent to the variable frequency heating device to instruct it to heat the reactor according to a predetermined heating rate and duration;

Cooling module: If cooling is required, the controller sends a control signal to the intelligent cooling unit, instructing it to cool the reactor according to a predetermined cooling rate and duration;

Real-time adjustment module: During the control process, the controller continuously monitors the temperature changes inside the reactor and compares it with the target temperature; if there is a deviation between the actual temperature and the target temperature, the controller will adjust the control parameters of the heating and cooling devices in real time according to the adjustment strategy output by the fuzzy logic reasoning system.

The beneficial effects of the present invention are as follows: by combining a multimodal temperature sensor array with a heat flow analysis model, accurate three-dimensional, real-time, and micro-segment temperature field monitoring inside the reactor is achieved, thereby improving the refinement of temperature control; by adopting a deep generative adversarial network, the temperature control strategy can be dynamically generated and optimized according to real-time multi-dimensional temperature data and equipment operation behavior, thereby enhancing the system's adaptive ability and prediction accuracy; by introducing a fuzzy logic reasoning system, the control strategy generated by deep learning is further adaptively adjusted in real time, making temperature control more flexible and intelligent, and able to cope with the nonlinear and time-varying characteristics of complex chemical reactions; by using a variable frequency heating device The linkage control of the device and the intelligent cooling unit can optimize energy use and reduce energy consumption while ensuring that the reactor temperature is stable within the set range; by establishing a closed-loop feedback control system from data acquisition, strategy generation, strategy evaluation to strategy implementation and re-optimization, the reactor temperature is always in a precisely controllable state, which is beneficial to improving product quality, reducing production anomalies and ensuring process safety; this technical solution integrates multiple cutting-edge technologies such as multimodal sensing technology, heat flow analysis, deep learning, fuzzy logic, etc., to form a highly integrated intelligent temperature control system with autonomous optimization function, which provides advanced solutions for reactor temperature control in chemical, pharmaceutical and other industries.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a schematic diagram of the method of the present invention;

FIG. 2 is a schematic diagram of the system of the present invention.

DETAILED DESCRIPTION

In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

In the following description, many specific details are set forth to facilitate a full understanding of the present invention. The embodiments described are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

One embodiment of the present invention, as shown in FIG1 , is a temperature control method for a reactor based on artificial intelligence, the method comprising:

S1. Use a multi-modal temperature sensor array and a heat flow analysis model to perform three-dimensional, real-time, micro-segment temperature field monitoring inside the reactor; and input the collected multi-dimensional temperature data into the controller;

S2, the controller is based on a built-in deep generative adversarial network. The generator model uses multi-dimensional temperature data and operation behavior to generate the optimal temperature control strategy, and the judge model evaluates the temperature control strategy based on actual temperature feedback;

S3, using the fuzzy logic reasoning system to adaptively adjust the temperature control strategy generated by the deep generative adversarial network and the real-time temperature data in real time;

S4. Based on the temperature control strategy after deep generative adversarial network and fuzzy logic fine-tuning, the controller controls the variable frequency heating device and intelligent cooling unit to adjust the temperature of the reactor.

The working principle of the above technical solution is as follows: first, a multimodal temperature sensor array is used to monitor the temperature field inside the reactor in three dimensions, real time, and micro-segment. At the same time, combined with the heat flow analysis model, the temperature distribution inside the reactor can be obtained more accurately; the controller uses the built-in deep generative adversarial network, in which the generator model generates the optimal temperature control strategy based on multi-dimensional temperature data and operation behavior. At the same time, the judge model evaluates the generated temperature control strategy based on the actual temperature feedback to ensure that the generated strategy can effectively control the temperature of the reactor; the fuzzy logic reasoning system is used to perform real-time adaptive adjustment of the temperature control strategy generated by the deep generative adversarial network and the real-time temperature data. Through fuzzy logic reasoning, the temperature control strategy can be flexibly adjusted according to the current actual situation to cope with different working conditions and environmental changes; according to the temperature control strategy generated by the deep generative adversarial network and fine-tuned by fuzzy logic, the controller adjusts the variable frequency heating device and the intelligent cooling unit to control the temperature of the reactor. By dynamically adjusting the working state and parameters of the heating and cooling devices, the temperature of the reactor can be accurately controlled to ensure the stability and efficiency of the reaction process.

The effects of the above technical solution are as follows: through the multimodal temperature sensor array and the heat flow analysis model, the three-dimensional, real-time, micro-segment temperature field monitoring inside the reactor can be performed, realizing real-time perception and monitoring of the temperature changes in the reaction process, and improving the control accuracy of the reaction process; using the deep generative adversarial network, the optimal temperature control strategy is generated based on multi-dimensional temperature data and operation behavior, and a more adaptable and optimized temperature control strategy can be generated according to real-time data and operation conditions, thereby improving the intelligence level of temperature control; with the help of the fuzzy logic reasoning system, the temperature control strategy is adaptively adjusted in real time, and the temperature control strategy can be flexibly adjusted according to the temperature changes inside the reactor and the actual operation conditions, ensuring the adaptability to different working conditions and environmental changes; based on the temperature control strategy generated by the deep generative adversarial network and fine-tuned by fuzzy logic, the controller can accurately adjust the variable frequency heating device and the intelligent cooling unit to achieve precise control of the reactor temperature, ensuring the stability and efficiency of the reaction process.

In one embodiment of the present invention, the S1 includes:

A multi-modal temperature sensor array is deployed inside the reactor to continuously monitor the temperature inside the reactor in real time.

Each sensor node in the multi-modal temperature sensing array synchronously records the temperature value of its respective location and forms a data matrix that comprehensively reflects the temperature distribution of each point in the kettle;

Through the preset sensor coordinate system, the temperature data collected by each sensor is mapped to the three-dimensional space coordinates, and combined with the spatial relationship between the sensors, an interpolation algorithm is used to construct a three-dimensional temperature field model of the reactor at the micro-segment level.

The heat flow analysis model is used to simulate and calculate the heat transfer process inside the reactor, including conduction, convection and radiation effects. The real-time temperature data directly measured by the multi-modal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set.

The integrated and processed multi-dimensional temperature data is transmitted to the controller in real time through a multi-channel transmission protocol.

The working principle of the above technical solution is as follows: deploy a multimodal temperature sensor array inside the reactor, and continuously monitor the temperature in real time through each sensor node. Each sensor node synchronously records the temperature value of its location and forms a data matrix reflecting the temperature distribution of each point in the reactor; the temperature data collected by each sensor is mapped to the three-dimensional space coordinates through the preset sensor coordinate system. Using the spatial relationship and interpolation algorithm between sensors, a three-dimensional temperature field model of the reactor is constructed at the micro-segment level; the heat transfer process inside the reactor is simulated and calculated through the heat flow analysis model, including conduction, convection and radiation effects. The real-time temperature data directly measured by the multimodal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set; the integrated and processed multi-dimensional temperature data is transmitted to the controller in real time through a multi-channel transmission protocol. The controller receives and processes these data, performs intelligent temperature control according to the preset algorithm and control strategy, and realizes precise adjustment of the reactor temperature.

The effects of the above technical solution are as follows: by deploying a multi-modal temperature sensor array, continuous real-time monitoring of the temperature inside the reactor can be achieved, and a data matrix that fully reflects the temperature distribution of each point in the reactor is formed through synchronous recording of sensor nodes, providing a comprehensive understanding of the temperature distribution of the reactor; by using the preset sensor coordinate system and interpolation algorithm, a three-dimensional temperature field model that is detailed to the micro-segment level is constructed, which can more accurately describe the temperature distribution inside the reactor and provide a more accurate basis for subsequent temperature control; the real-time temperature data measured by the multi-modal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set, which can more comprehensively reflect the temperature state inside the reactor and provide more reference basis for temperature control; the integrated and processed multi-dimensional temperature data is transmitted to the controller in real time through a multi-channel transmission protocol, so as to realize real-time monitoring and precise adjustment of the reactor temperature, improve the real-time and intelligent level of temperature control, and ensure the stability and efficiency of the reaction process.

In one embodiment of the present invention, the S2 includes:

The controller preprocesses the received multi-dimensional temperature data, which may include data cleaning, missing value filling, normalization and other operations, combines the preprocessed multi-dimensional temperature data with the equipment's operating behavior data (such as stirring speed, heating power, etc.), and converts it into a feature vector that can be recognized by deep learning;

The processed multi-dimensional temperature data and operational behavior characteristics are used as inputs and input into the generator model built into the deep generative adversarial network;

The generator model learns how to generate the optimal temperature control strategy that is most likely to lead to the target temperature field state according to the input data during the training process through the back propagation algorithm and the optimization algorithm;

The generator model generates new temperature control instructions in real time according to the current temperature and operating conditions, such as heating rate adjustment, cooling rate adjustment, etc. The controller converts the temperature control strategy generated by the generator into specific equipment operation instructions, sends them to the temperature control system of the reactor for execution, and adjusts the actual temperature inside the reactor;

After implementing the new temperature control strategy, the temperature changes of each micro-section inside the reactor are continuously monitored in real time through the multi-modal temperature sensor array to obtain actual temperature feedback data;

The judgement model in the deep generative adversarial network receives the actual temperature feedback data, compares the actual temperature data with the preset threshold range of the temperature field, and outputs an evaluation score;

If the evaluation given by the judge is lower or higher than the expected threshold, the generator will adjust and optimize itself according to the feedback signal of the judge and generate a new control strategy again.

The working principle of the above technical solution is as follows: pre-processed multi-dimensional temperature data and operation behavior characteristics are used as input and passed into the generator model. The generator model learns how to generate the optimal temperature control strategy that is most likely to lead to the target temperature field state based on these input data through the back propagation algorithm and the optimization algorithm during the training process; the generator model generates new temperature control instructions in real time according to the current temperature and operating conditions, such as heating rate adjustment, cooling rate adjustment, etc. These control instructions will be converted into specific equipment operation instructions and sent to the temperature control system of the reactor for execution to adjust the actual temperature inside the reactor; after executing the new temperature control strategy, the controller monitors the temperature changes of each micro-segment inside the reactor in real time through the multi-modal temperature sensor array to obtain the actual temperature feedback data. The judge model receives the actual temperature feedback data, compares the actual temperature data with the preset threshold interval of the temperature field, and outputs an evaluation score to reflect the quality of the current strategy; if the evaluation given by the judge is lower or higher than the expected threshold, the generator will self-adjust and optimize according to the feedback signal of the judge, and generate a new control strategy again to continuously optimize the effect and performance of the temperature control strategy.

The effects of the above technical solution are: by learning and optimizing multi-dimensional temperature data and operating behaviors through a deep learning model, a more accurate temperature control strategy can be generated, so that the temperature inside the reactor can reach the target temperature more stably; the solution realizes automatic temperature control strategy generation and optimization, without human intervention, and can automatically adjust the control strategy according to real-time temperature feedback data, thereby improving the efficiency and accuracy of temperature control; by real-time monitoring and adjustment of the temperature control strategy, temperature fluctuations and deviations can be minimized to the greatest extent, ensuring the stability of the reaction process, thereby improving production efficiency and product quality; the optimized temperature control strategy can more effectively utilize energy, reduce energy waste, reduce production costs, and is conducive to sustainable development; through the training and optimization of deep generative adversarial networks, the system can adapt to different reaction conditions and environmental changes, thereby improving the robustness and adaptability of the system.

In one embodiment of the present invention, S3 includes:

Acquire multi-dimensional real-time temperature data in real time. After receiving real-time or historical temperature data and operation behavior parameters, the deep generative adversarial network generates a preliminary temperature control strategy;

Establish a rule-based fuzzy logic reasoning system, which contains multiple fuzzy sets and membership functions. These functions are used to describe different temperature ranges and the degree of control actions (such as "slight heating", "moderate cooling", etc.), and input the preliminary temperature control strategy generated by the deep generative adversarial network and the real-time temperature data into the fuzzy logic reasoning system;

The fuzzy logic reasoning system maps the real-time temperature data to the corresponding fuzzy sets and applies the predefined fuzzy logic rule base for reasoning;

The fuzzy logic reasoning system obtains one or a group of adaptively adjusted temperature control strategies based on the relationship between real-time temperature data and the expected temperature range, as well as the deep generative adversarial network strategy.

The working principle of the above technical solution is as follows: the system obtains multi-dimensional temperature data in real time, including temperature information at different locations inside the reactor; after receiving real-time or historical temperature data and operating behavior parameters, the deep generative adversarial network generates a preliminary temperature control strategy. This step uses the ability of deep learning to generate temperature control strategies based on historical data and current conditions; the system establishes a rule-based fuzzy logic reasoning system, which contains multiple fuzzy sets and membership functions to describe different temperature ranges and the degree of control actions. For example, "slight heating", "moderate cooling", etc.; the system inputs the preliminary temperature control strategy generated by the deep generative adversarial network and the real-time temperature data into the fuzzy logic reasoning system; the fuzzy logic reasoning system uses a pre-defined fuzzy logic rule base for reasoning based on the relationship between the real-time temperature data and the expected temperature range, as well as the deep generative adversarial network strategy. The reasoning process takes into account the current temperature situation and the expected temperature range to generate one or a group of adaptively adjusted temperature control strategies.

The effects of the above technical solutions are as follows: by acquiring multi-dimensional temperature data in real time, the system can timely understand the changes in the temperature inside the reactor and provide accurate data support for subsequent temperature control; the deep generative adversarial network can generate a preliminary temperature control strategy based on real-time or historical temperature data and operating behavior parameters. Such strategy generation is intelligent and has strong learning ability, and can better adapt to different process conditions and changes; the rule-based fuzzy logic reasoning system can combine the preliminary temperature control strategy generated by the deep generative adversarial network with the real-time temperature data, and perform reasoning based on the pre-defined fuzzy logic rule base. This fuzzy logic reasoning system can handle the ambiguity and uncertainty of real-time data, and improve the accuracy and stability of temperature control; through the fuzzy logic reasoning system, the system can obtain one or a group of adaptively adjusted temperature control strategies based on the relationship between real-time temperature data and the expected temperature range, as well as the deep generative adversarial network strategy. This can effectively maintain the internal temperature of the reactor within the set target range, improving the stability and controllability of the production process.

In one embodiment of the present invention, the S4 includes:

The controller receives the adaptively adjusted temperature control strategy instructions output by the fuzzy logic inference system, which includes precise control instructions for the current temperature state inside the reactor, and parses the instructions to determine the required heating or cooling rate, duration and other related parameters.

Based on the parameters obtained by the analysis, the controller determines whether the reactor needs to be heated or cooled;

If heating is required, a control signal is sent to the variable frequency heating device, instructing it to heat the reactor at a predetermined heating rate and duration; the variable frequency heating device can adjust its output power according to the control signal, thereby achieving precise control of the internal temperature of the reactor.

If cooling is required, the controller sends a control signal to the intelligent cooling unit, instructing it to cool the reactor according to a predetermined cooling rate and duration; the intelligent cooling unit can adjust its working mode according to the control signal, such as adjusting the flow rate or temperature of the cooling medium, to ensure the stability and uniformity of the temperature inside the reactor.

During the control process, the controller continuously monitors the temperature changes inside the reactor and compares it with the target temperature; if there is a deviation between the actual temperature and the target temperature, the controller will adjust the control parameters of the heating and cooling devices in real time according to the adjustment strategy output by the fuzzy logic reasoning system.

The working principle of the above technical solution is as follows: the controller first receives the adaptively adjusted temperature control strategy instruction output from the fuzzy logic reasoning system. The instruction contains precise control instructions for the current internal temperature state of the reactor, and parses it to determine the required heating or cooling rate, duration and other related parameters; based on the parameters obtained by the analysis, the controller determines whether the reactor needs to be heated or cooled; if heating is required, the controller sends a control signal to the variable frequency heating device, instructing it to heat the reactor according to the predetermined heating rate and duration. The variable frequency heating device can adjust its output power according to the control signal to achieve precise control of the internal temperature of the reactor. If cooling is required, the controller sends a control signal to the intelligent cooling unit, instructing it to cool the reactor according to the predetermined cooling rate and duration. The intelligent cooling unit can adjust its working mode according to the control signal, such as adjusting the flow rate or temperature of the cooling medium, to ensure the stability and uniformity of the internal temperature of the reactor; during the control process, the controller continuously monitors the temperature changes inside the reactor and compares it with the target temperature. If there is a deviation between the actual temperature and the target temperature, the controller will adjust the control parameters of the heating and cooling devices in real time according to the adjustment strategy output by the fuzzy logic reasoning system to achieve precise control and stable maintenance of the temperature.

The effects of the above technical solution are: the temperature control strategy instructions after adaptive adjustment output by the fuzzy logic reasoning system, as well as the analysis and real-time adjustment of the parameters by the controller, can achieve precise control of the temperature inside the reactor. This helps to ensure that the required temperature conditions in the production process are met, and improves the stability and controllability of the production process; the use of variable frequency heating devices and intelligent cooling units can adjust the output power and working mode according to actual needs, and achieve effective use of energy. This not only reduces production costs, but also helps to reduce energy consumption and reduce the impact on the environment, in line with the requirements of sustainable development; by real-time monitoring of the temperature changes inside the reactor and timely adjustment of the control parameters, the target temperature can be reached more quickly and maintained stable. This helps to shorten the production cycle, improve production efficiency, and thus increase production capacity and market competitiveness; the application of adaptive temperature control strategies and intelligent control systems reduces the need for manual operation. This reduces the risk of human error and improves the stability and reliability of the production process.

One embodiment of the present invention, as shown in FIG2 , is a temperature control system for a reactor based on artificial intelligence, the system comprising:

Temperature acquisition module: Through the multi-modal temperature sensor array and heat flow analysis model, the temperature field inside the reactor is monitored in three dimensions, real time, and in micro-segments; and the collected multi-dimensional temperature data is input into the controller;

Strategy generation module: The controller is based on a built-in deep generative adversarial network. The generator model uses multi-dimensional temperature data and operation behavior to generate the optimal temperature control strategy, and the judgement model evaluates the temperature control strategy based on actual temperature feedback;

Adjustment module: Use the fuzzy logic reasoning system to perform real-time adaptive adjustment on the temperature control strategy generated by the deep generative adversarial network and the real-time temperature data;

Control execution module: Based on the temperature control strategy after deep generative adversarial network and fuzzy logic fine-tuning, the controller controls the variable frequency heating device and intelligent cooling unit to control the temperature of the reactor.

The working principle of the above technical solution is as follows: first, a multimodal temperature sensor array is used to monitor the temperature field inside the reactor in three dimensions, real time, and micro-segment. At the same time, combined with the heat flow analysis model, the temperature distribution inside the reactor can be obtained more accurately; the controller uses the built-in deep generative adversarial network, in which the generator model generates the optimal temperature control strategy based on multi-dimensional temperature data and operation behavior. At the same time, the judge model evaluates the generated temperature control strategy based on the actual temperature feedback to ensure that the generated strategy can effectively control the temperature of the reactor; the fuzzy logic reasoning system is used to perform real-time adaptive adjustment of the temperature control strategy generated by the deep generative adversarial network and the real-time temperature data. Through fuzzy logic reasoning, the temperature control strategy can be flexibly adjusted according to the current actual situation to cope with different working conditions and environmental changes; according to the temperature control strategy generated by the deep generative adversarial network and fine-tuned by fuzzy logic, the controller adjusts the variable frequency heating device and the intelligent cooling unit to control the temperature of the reactor. By dynamically adjusting the working state and parameters of the heating and cooling devices, the temperature of the reactor can be accurately controlled to ensure the stability and efficiency of the reaction process.

The effects of the above technical solution are as follows: through the multimodal temperature sensor array and the heat flow analysis model, the three-dimensional, real-time, micro-segment temperature field monitoring inside the reactor can be performed, realizing real-time perception and monitoring of the temperature changes in the reaction process, and improving the control accuracy of the reaction process; using the deep generative adversarial network, the optimal temperature control strategy is generated based on multi-dimensional temperature data and operation behavior, and a more adaptable and optimized temperature control strategy can be generated according to real-time data and operation conditions, thereby improving the intelligence level of temperature control; with the help of the fuzzy logic reasoning system, the temperature control strategy is adaptively adjusted in real time, and the temperature control strategy can be flexibly adjusted according to the temperature changes inside the reactor and the actual operation conditions, ensuring the adaptability to different working conditions and environmental changes; based on the temperature control strategy generated by the deep generative adversarial network and fine-tuned by fuzzy logic, the controller can accurately adjust the variable frequency heating device and the intelligent cooling unit to achieve precise control of the reactor temperature, ensuring the stability and efficiency of the reaction process.

In one embodiment of the present invention, the temperature acquisition module includes:

Deployment module: deploy a multi-modal temperature sensor array inside the reactor to continuously monitor the temperature inside the reactor in real time through the multi-modal temperature sensor array;

Matrix generation module: Each sensor node in the multi-modal temperature sensing array synchronously records the temperature value of its respective location and forms a data matrix that fully reflects the temperature distribution of each point in the kettle;

Model building module: Through the preset sensor coordinate system, the temperature data collected by each sensor is mapped to the three-dimensional space coordinates, and combined with the spatial relationship between the sensors, an interpolation algorithm is used to build a three-dimensional temperature field model of the reactor at the micro-segment level;

Simulation module: Through the heat flow analysis model, the heat transfer process inside the reactor is simulated and calculated, including conduction, convection and radiation effects. The real-time temperature data directly measured by the multi-modal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set.

Transmission module: transmits the integrated and processed multi-dimensional temperature data to the controller in real time through a multi-channel transmission protocol.

The working principle of the above technical solution is as follows: deploy a multimodal temperature sensor array inside the reactor, and continuously monitor the temperature in real time through each sensor node. Each sensor node synchronously records the temperature value of its location and forms a data matrix reflecting the temperature distribution of each point in the reactor; the temperature data collected by each sensor is mapped to the three-dimensional space coordinates through the preset sensor coordinate system. Using the spatial relationship and interpolation algorithm between sensors, a three-dimensional temperature field model of the reactor is constructed at the micro-segment level; the heat transfer process inside the reactor is simulated and calculated through the heat flow analysis model, including conduction, convection and radiation effects. The real-time temperature data directly measured by the multimodal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set; the integrated and processed multi-dimensional temperature data is transmitted to the controller in real time through a multi-channel transmission protocol. The controller receives and processes these data, performs intelligent temperature control according to the preset algorithm and control strategy, and realizes precise adjustment of the reactor temperature.

The effects of the above technical solution are as follows: by deploying a multi-modal temperature sensor array, continuous real-time monitoring of the temperature inside the reactor can be achieved, and a data matrix that fully reflects the temperature distribution of each point in the reactor is formed through synchronous recording of sensor nodes, providing a comprehensive understanding of the temperature distribution of the reactor; by using the preset sensor coordinate system and interpolation algorithm, a three-dimensional temperature field model that is detailed to the micro-segment level is constructed, which can more accurately describe the temperature distribution inside the reactor and provide a more accurate basis for subsequent temperature control; the real-time temperature data measured by the multi-modal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set, which can more comprehensively reflect the temperature state inside the reactor and provide more reference basis for temperature control; the integrated and processed multi-dimensional temperature data is transmitted to the controller in real time through a multi-channel transmission protocol, so as to realize real-time monitoring and precise adjustment of the reactor temperature, improve the real-time and intelligent level of temperature control, and ensure the stability and efficiency of the reaction process.

In one embodiment of the present invention, the strategy generation module includes:

Preprocessing module: The controller preprocesses the received multi-dimensional temperature data, which may include data cleaning, missing value filling, normalization and other operations. The preprocessed multi-dimensional temperature data is combined with the operation behavior data of the equipment (such as stirring speed, heating power, etc.) and converted into feature vectors that can be recognized by deep learning.

Input module: takes the processed multi-dimensional temperature data and operation behavior characteristics as input and inputs them into the generator model built into the deep generative adversarial network;

The generator model learns how to generate the optimal temperature control strategy that is most likely to lead to the target temperature field state according to the input data during the training process through the back propagation algorithm and the optimization algorithm;

Temperature adjustment module: The generator model generates new temperature control instructions in real time according to the current temperature and operating conditions, such as heating rate adjustment, cooling rate adjustment, etc. The controller converts the temperature control strategy generated by the generator into specific equipment operation instructions, sends them to the temperature control system of the reactor for execution, and adjusts the actual temperature inside the reactor;

Feedback data acquisition module: After executing the new temperature control strategy, continue to monitor the temperature changes of each micro-section inside the reactor in real time through the multi-modal temperature sensor array to obtain actual temperature feedback data;

Evaluation score output module: The judgement model in the deep generative adversarial network receives the actual temperature feedback data, compares the actual temperature data with the preset threshold range of the temperature field, and outputs the evaluation score; this score reflects the quality of the current strategy;

Secondary generation module: If the evaluation given by the judge is lower or higher than the expected threshold, the generator will self-adjust and optimize according to the feedback signal of the judge and generate a new control strategy again.

The working principle of the above technical solution is as follows: pre-processed multi-dimensional temperature data and operation behavior characteristics are used as input and passed into the generator model. The generator model learns how to generate the optimal temperature control strategy that is most likely to lead to the target temperature field state based on these input data through the back propagation algorithm and the optimization algorithm during the training process; the generator model generates new temperature control instructions in real time according to the current temperature and operating conditions, such as heating rate adjustment, cooling rate adjustment, etc. These control instructions will be converted into specific equipment operation instructions and sent to the temperature control system of the reactor for execution to adjust the actual temperature inside the reactor; after executing the new temperature control strategy, the controller monitors the temperature changes of each micro-segment inside the reactor in real time through the multi-modal temperature sensor array to obtain the actual temperature feedback data. The judge model receives the actual temperature feedback data, compares the actual temperature data with the preset threshold interval of the temperature field, and outputs an evaluation score to reflect the quality of the current strategy; if the evaluation given by the judge is lower or higher than the expected threshold, the generator will self-adjust and optimize according to the feedback signal of the judge, and generate a new control strategy again to continuously optimize the effect and performance of the temperature control strategy.

The effects of the above technical solution are: by learning and optimizing multi-dimensional temperature data and operating behaviors through a deep learning model, a more accurate temperature control strategy can be generated, so that the temperature inside the reactor can reach the target temperature more stably; the solution realizes automatic temperature control strategy generation and optimization, without human intervention, and can automatically adjust the control strategy according to real-time temperature feedback data, thereby improving the efficiency and accuracy of temperature control; by real-time monitoring and adjustment of the temperature control strategy, temperature fluctuations and deviations can be minimized to the greatest extent, ensuring the stability of the reaction process, thereby improving production efficiency and product quality; the optimized temperature control strategy can more effectively utilize energy, reduce energy waste, reduce production costs, and is conducive to sustainable development; through the training and optimization of deep generative adversarial networks, the system can adapt to different reaction conditions and environmental changes, thereby improving the robustness and adaptability of the system.

In one embodiment of the present invention, the adjustment module includes:

Preliminary strategy generation module: real-time acquisition of multi-dimensional real-time temperature data. After receiving real-time or historical temperature data and operation behavior parameters, the deep generative adversarial network generates a preliminary temperature control strategy;

System establishment module: Establish a rule-based fuzzy logic reasoning system, which includes multiple fuzzy sets and membership functions. These functions are used to describe different temperature ranges and the degree of control actions (such as "slight heating", "moderate cooling", etc.), and input the preliminary temperature control strategy generated by the deep generative adversarial network and the real-time temperature data into the fuzzy logic reasoning system;

Reasoning module: The fuzzy logic reasoning system maps the real-time temperature data to the corresponding fuzzy set and applies the pre-defined fuzzy logic rule base for reasoning;

Adaptive adjustment module: The fuzzy logic reasoning system obtains one or a group of adaptive temperature control strategies based on the relationship between real-time temperature data and the expected temperature range, as well as the deep generative adversarial network strategy.

The working principle of the above technical solution is as follows: the system obtains multi-dimensional temperature data in real time, including temperature information at different locations inside the reactor; after receiving real-time or historical temperature data and operating behavior parameters, the deep generative adversarial network generates a preliminary temperature control strategy. This step uses the ability of deep learning to generate temperature control strategies based on historical data and current conditions; the system establishes a rule-based fuzzy logic reasoning system, which contains multiple fuzzy sets and membership functions to describe different temperature ranges and the degree of control actions. For example, "slight heating", "moderate cooling", etc.; the system inputs the preliminary temperature control strategy generated by the deep generative adversarial network and the real-time temperature data into the fuzzy logic reasoning system; the fuzzy logic reasoning system uses a pre-defined fuzzy logic rule base for reasoning based on the relationship between the real-time temperature data and the expected temperature range, as well as the deep generative adversarial network strategy. The reasoning process takes into account the current temperature situation and the expected temperature range to generate one or a group of adaptively adjusted temperature control strategies.

The effects of the above technical solutions are as follows: by acquiring multi-dimensional temperature data in real time, the system can timely understand the changes in the temperature inside the reactor and provide accurate data support for subsequent temperature control; the deep generative adversarial network can generate a preliminary temperature control strategy based on real-time or historical temperature data and operating behavior parameters. Such strategy generation is intelligent and has strong learning ability, and can better adapt to different process conditions and changes; the rule-based fuzzy logic reasoning system can combine the preliminary temperature control strategy generated by the deep generative adversarial network with the real-time temperature data, and perform reasoning based on the pre-defined fuzzy logic rule base. This fuzzy logic reasoning system can handle the ambiguity and uncertainty of real-time data, and improve the accuracy and stability of temperature control; through the fuzzy logic reasoning system, the system can obtain one or a group of adaptively adjusted temperature control strategies based on the relationship between real-time temperature data and the expected temperature range, as well as the deep generative adversarial network strategy. This can effectively maintain the internal temperature of the reactor within the set target range, improving the stability and controllability of the production process.

In one embodiment of the present invention, the control execution module includes:

Analysis module: The controller receives the adaptively adjusted temperature control strategy instructions output by the fuzzy logic inference system, which contains precise control instructions for the current temperature state inside the reactor, and analyzes the instructions; determines the required heating or cooling rate, duration and other related parameters.

Judgment module: Based on the analyzed parameters, the controller determines whether the reactor needs to be heated or cooled;

Heating module: If heating is required, a control signal is sent to the variable frequency heating device, instructing it to heat the reactor at a predetermined heating rate and duration; the variable frequency heating device can adjust its output power according to the control signal, thereby achieving precise control of the internal temperature of the reactor.

Cooling module: If cooling is required, the controller sends a control signal to the intelligent cooling unit, instructing it to cool the reactor at a predetermined cooling rate and duration; the intelligent cooling unit can adjust its working mode according to the control signal, such as adjusting the flow rate or temperature of the cooling medium, to ensure the stability and uniformity of the temperature inside the reactor.

Real-time adjustment module: During the control process, the controller continuously monitors the temperature changes inside the reactor and compares it with the target temperature; if there is a deviation between the actual temperature and the target temperature, the controller will adjust the control parameters of the heating and cooling devices in real time according to the adjustment strategy output by the fuzzy logic reasoning system.

The working principle of the above technical solution is as follows: the controller first receives the adaptively adjusted temperature control strategy instruction output from the fuzzy logic reasoning system. The instruction contains precise control instructions for the current internal temperature state of the reactor, and parses it to determine the required heating or cooling rate, duration and other related parameters; based on the parameters obtained by the analysis, the controller determines whether the reactor needs to be heated or cooled; if heating is required, the controller sends a control signal to the variable frequency heating device, instructing it to heat the reactor according to the predetermined heating rate and duration. The variable frequency heating device can adjust its output power according to the control signal to achieve precise control of the internal temperature of the reactor. If cooling is required, the controller sends a control signal to the intelligent cooling unit, instructing it to cool the reactor according to the predetermined cooling rate and duration. The intelligent cooling unit can adjust its working mode according to the control signal, such as adjusting the flow rate or temperature of the cooling medium, to ensure the stability and uniformity of the internal temperature of the reactor; during the control process, the controller continuously monitors the temperature changes inside the reactor and compares it with the target temperature. If there is a deviation between the actual temperature and the target temperature, the controller will adjust the control parameters of the heating and cooling devices in real time according to the adjustment strategy output by the fuzzy logic reasoning system to achieve precise control and stable maintenance of the temperature.

The effects of the above technical solution are: the temperature control strategy instructions after adaptive adjustment output by the fuzzy logic reasoning system, as well as the analysis and real-time adjustment of the parameters by the controller, can achieve precise control of the temperature inside the reactor. This helps to ensure that the required temperature conditions in the production process are met, and improves the stability and controllability of the production process; the use of variable frequency heating devices and intelligent cooling units can adjust the output power and working mode according to actual needs, and achieve efficient use of energy. This not only reduces production costs, but also helps to reduce energy consumption and reduce the impact on the environment, in line with the requirements of sustainable development; by real-time monitoring of the temperature changes inside the reactor and timely adjustment of the control parameters, the target temperature can be reached more quickly and maintained stable. This helps to shorten the production cycle, improve production efficiency, and thus increase production capacity and market competitiveness; the application of adaptive temperature control strategies and intelligent control systems reduces the need for manual operation. This reduces the risk of human error and improves the stability and reliability of the production process.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims (2)

1. A temperature control method for a reactor based on artificial intelligence, characterized in that the method comprises: S1. Use a multi-modal temperature sensor array and a heat flow analysis model to perform three-dimensional, real-time, micro-segment temperature field monitoring inside the reactor; and input the collected multi-dimensional temperature data into the controller; S2, the controller is based on a built-in deep generative adversarial network. The generator model uses multi-dimensional temperature data and operation behavior to generate a temperature control strategy, and the judgement model evaluates the temperature control strategy based on actual temperature feedback; S3, using the fuzzy logic reasoning system to adaptively adjust the temperature control strategy generated by the deep generative adversarial network and the real-time temperature data in real time; S4. Based on the temperature control strategy after deep generative adversarial network and fuzzy logic fine-tuning, the controller controls the variable frequency heating device and the intelligent cooling unit to adjust the temperature of the reactor; Said S1 comprises: A multi-modal temperature sensor array is deployed inside the reactor to continuously monitor the temperature inside the reactor in real time. Each sensor node in the multi-modal temperature sensing array synchronously records the temperature value of its respective location and forms a data matrix that comprehensively reflects the temperature distribution of each point in the kettle; The temperature data collected by each sensor is mapped to three-dimensional space coordinates through the preset sensor coordinate system, and the three-dimensional temperature field model inside the reactor is constructed by using the interpolation algorithm in combination with the spatial relationship between the sensors. The heat flow analysis model is used to simulate and calculate the heat transfer process inside the reactor. The real-time temperature data directly measured by the multi-modal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set. The integrated and processed multi-dimensional temperature data is transmitted to the controller in real time through a multi-channel transmission protocol; The S2 comprises: The controller preprocesses the received multi-dimensional temperature data, combines the preprocessed multi-dimensional temperature data with the operation behavior data of the equipment, and converts them into feature vectors that can be recognized by deep learning; The processed multi-dimensional temperature data and operational behavior characteristics are used as inputs and input into the generator model built into the deep generative adversarial network; The generator model generates new temperature control instructions in real time according to the current temperature and operating conditions. The controller converts the temperature control strategy generated by the generator into specific equipment operation instructions and sends them to the temperature control system of the reactor for execution to adjust the actual temperature inside the reactor. After implementing the new temperature control strategy, the temperature changes of each micro-section inside the reactor are continuously monitored in real time through the multi-modal temperature sensor array to obtain actual temperature feedback data; The judgement model in the deep generative adversarial network receives the actual temperature feedback data, compares the actual temperature data with the preset threshold range of the temperature field, and outputs an evaluation score; If the evaluation given by the judge is lower or higher than the expected threshold, the generator will adjust and optimize itself according to the feedback signal of the judge and generate a new control strategy again; The S3 includes: Acquire multi-dimensional real-time temperature data in real time. After receiving real-time or historical temperature data and operation behavior parameters, the deep generative adversarial network generates a preliminary temperature control strategy; Establish a rule-based fuzzy logic reasoning system, and input the preliminary temperature control strategy generated by the deep generative adversarial network and the real-time temperature data into the fuzzy logic reasoning system; The fuzzy logic reasoning system maps the real-time temperature data to the corresponding fuzzy sets and applies the predefined fuzzy logic rule base for reasoning; The fuzzy logic reasoning system obtains one or a group of adaptively adjusted temperature control strategies based on the relationship between real-time temperature data and the expected temperature range and the deep generative adversarial network strategy; The S4 comprises: The controller receives the temperature control strategy instruction after adaptive adjustment outputted from the fuzzy logic inference system and parses the instruction; Based on the parameters obtained by the analysis, the controller determines whether the reactor needs to be heated or cooled; If heating is required, a control signal is sent to the variable frequency heating device to instruct it to heat the reactor according to a predetermined heating rate and duration; If cooling is required, the controller sends a control signal to the intelligent cooling unit, instructing it to cool the reactor according to a predetermined cooling rate and duration; During the control process, the controller continuously monitors the temperature changes inside the reactor and compares it with the target temperature; if there is a deviation between the actual temperature and the target temperature, the controller will adjust the control parameters of the heating and cooling devices in real time according to the adjustment strategy output by the fuzzy logic reasoning system. 2. A temperature control system for a reactor based on artificial intelligence, characterized in that the system comprises: Temperature acquisition module: Through the multi-modal temperature sensor array and heat flow analysis model, the temperature field inside the reactor is monitored in three dimensions, real time, and in micro-segments; and the collected multi-dimensional temperature data is input into the controller; Strategy generation module: The controller is based on a built-in deep generative adversarial network. The generator model generates temperature control strategies using multi-dimensional temperature data and operation behaviors, and the judgement model evaluates the temperature control strategies based on actual temperature feedback. Adjustment module: Use the fuzzy logic reasoning system to perform real-time adaptive adjustment on the temperature control strategy generated by the deep generative adversarial network and the real-time temperature data; Control execution module: Based on the temperature control strategy after deep generative adversarial network and fuzzy logic fine-tuning, the controller controls the variable frequency heating device and intelligent cooling unit to control the temperature of the reactor; The temperature acquisition module comprises: Deployment module: deploy a multi-modal temperature sensor array inside the reactor to continuously monitor the temperature inside the reactor in real time through the multi-modal temperature sensor array; Matrix generation module: Each sensor node in the multi-modal temperature sensing array synchronously records the temperature value of its respective location and forms a data matrix that fully reflects the temperature distribution of each point in the kettle; Model building module: Through the preset sensor coordinate system, the temperature data collected by each sensor is mapped to the three-dimensional space coordinates, and combined with the spatial relationship between the sensors, the interpolation algorithm is used to build a three-dimensional temperature field model inside the reactor; Simulation module: Through the heat flow analysis model, the heat transfer process inside the reactor is simulated and calculated. The real-time temperature data directly measured by the multi-modal temperature sensor array is combined with the temperature data calculated by the heat flow analysis model to form a multi-dimensional comprehensive temperature data set; Transmission module: transmits the integrated and processed multi-dimensional temperature data to the controller in real time through a multi-channel transmission protocol; The strategy generation module comprises: Preprocessing module: The controller preprocesses the received multi-dimensional temperature data, combines the preprocessed multi-dimensional temperature data with the equipment's operation behavior data, and converts it into a feature vector that can be recognized by deep learning; Input module: takes the processed multi-dimensional temperature data and operation behavior characteristics as input and inputs them into the generator model built into the deep generative adversarial network; Temperature adjustment module: The generator model generates new temperature control instructions in real time according to the current temperature and operating conditions. The controller converts the temperature control strategy generated by the generator into specific equipment operation instructions, sends them to the temperature control system of the reactor for execution, and adjusts the actual temperature inside the reactor. Feedback data acquisition module: After executing the new temperature control strategy, continue to monitor the temperature changes of each micro-section inside the reactor in real time through the multi-modal temperature sensor array to obtain actual temperature feedback data; Evaluation score output module: The judgement model in the deep generative adversarial network receives the actual temperature feedback data, compares the actual temperature data with the preset threshold range of the temperature field, and outputs the evaluation score; Secondary generation module: If the evaluation given by the judge is lower or higher than the expected threshold, the generator will adjust and optimize itself according to the feedback signal of the judge and generate a new control strategy again; The adjustment module comprises: Preliminary strategy generation module: real-time acquisition of multi-dimensional real-time temperature data. After receiving real-time or historical temperature data and operation behavior parameters, the deep generative adversarial network generates a preliminary temperature control strategy; System establishment module: Establish a rule-based fuzzy logic reasoning system, and input the preliminary temperature control strategy generated by the deep generative adversarial network and the real-time temperature data into the fuzzy logic reasoning system; Reasoning module: The fuzzy logic reasoning system maps the real-time temperature data to the corresponding fuzzy set and applies the pre-defined fuzzy logic rule base for reasoning; Adaptive adjustment module: The fuzzy logic reasoning system obtains one or a group of adaptive temperature control strategies based on the relationship between real-time temperature data and the expected temperature range, as well as the deep generative adversarial network strategy; The control execution module includes: Parsing module: The controller receives the adaptively adjusted temperature control strategy instruction output from the fuzzy logic reasoning system and parses the instruction; Judgment module: Based on the analyzed parameters, the controller determines whether the reactor needs to be heated or cooled; Heating module: If heating is required, a control signal is sent to the variable frequency heating device to instruct it to heat the reactor according to a predetermined heating rate and duration; Cooling module: If cooling is required, the controller sends a control signal to the intelligent cooling unit, instructing it to cool the reactor according to a predetermined cooling rate and duration; Real-time adjustment module: During the control process, the controller continuously monitors the temperature changes inside the reactor and compares it with the target temperature; if there is a deviation between the actual temperature and the target temperature, the controller will adjust the control parameters of the heating and cooling devices in real time according to the adjustment strategy output by the fuzzy logic reasoning system.