Pressurized water reactor steam generator liquid level control method and system

Abstract

The invention discloses a pressurized water reactor steam generator liquid level control method and system, and belongs to the technical field of nuclear power station control. The method comprises the steps that a liquid level set value and a measured value of the steam generator are obtained, and deviation is calculated; inputting the set value, the measured value, the deviation and the bias constant into a preset BP neural network; the network dynamically outputs three coefficients of proportion, integral and differential through forward propagation to serve as real-time parameters of a PID (Proportion Integration Differentiation) controller; the PID controller calculates and outputs a water supply flow demand value, and a water supply valve is adjusted after conversion; and calculating a root-mean-square error of liquid level control as a loss function, and updating the weight of the neural network on line by using the new liquid level deviation through a back propagation algorithm to realize adaptive optimization of PID parameters. According to the invention, the self-learning ability of the BP neural network is combined with the reliability of PID control, the control quality and stability of the liquid level of the steam generator under various power levels are significantly improved, the system can be deployed in a power plant DCS through modular packaging, and the engineering practicability is high.

Description

Technical Field

[0001] This invention belongs to the field of nuclear science and technology, specifically relating to a method and system for controlling the liquid level of a pressurized water reactor steam generator. Background Technology

[0002] The steam generator is a crucial component of a pressurized water reactor (PWR), playing a vital role in energy exchange between the primary and secondary loops. Currently, large PWR nuclear power plants commonly employ U-type steam generators (UTSGs). The main function of the steam generator is to transfer heat generated by the reactor core in the primary loop to the secondary loop, producing steam to power the turbine. Controlling the liquid level in the steam generator significantly impacts the safety of the entire nuclear power plant. Excessively high liquid levels can lead to water entering the turbine and causing blade corrosion, while excessively low levels can expose heat transfer tubes, degrading heat transfer efficiency. To ensure the safe and stable operation of the steam generator, the water level must be strictly controlled within a specified range during nuclear power plant operation. However, due to the nonlinear characteristics of the steam generator, its non-minimum phase system, and the existence of significant "false liquid levels," its water level control faces numerous technical challenges. Currently, the main method used for steam generator level control both domestically and internationally is cascade PID control. This control method can effectively solve problems such as "false levels," but its control effect mainly depends on the PID control parameters. Currently, most of the parameter tuning methods used in industry rely on pre-set rules or operator experience, which leads to problems such as long debugging cycles and multiple conflicting adjustment targets making it difficult to obtain optimal parameters, thus affecting the control effect.

[0003] With the development of current control technology, more and more advanced control solutions are being applied to the industrial field. Advanced control technology possesses dynamic adaptive capabilities and multi-objective collaborative optimization capabilities, effectively solving the problems inherent in PID control and traditional parameter tuning. Therefore, applying advanced control technology to UTSG level control is of great significance. Since actual controllers in engineering projects need to consider requirements such as computing power and stability, the designed advanced control algorithm must consider its feasibility in power plants, ensuring that the control scheme can adapt to the power plant's control equipment. Currently, nuclear power plants mainly use DCS systems for control, which have advantages such as comprehensive control functions, strong real-time data transmission, and high stability.

[0004] In summary, it is necessary to conduct advanced control research on UTSG liquid level, including advanced PID parameter tuning schemes and adaptive control technology, to improve the control effect of UTSG liquid level. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a liquid level control method and system for pressurized water reactor steam generators, which addresses the shortcomings of the prior art and solves the technical problems of traditional parameter tuning methods, such as relatively complex tuning process, long time consumption and difficulty in adapting to changes in system operating conditions.

[0006] The present invention adopts the following technical solution: A method for controlling the liquid level in a pressurized water reactor steam generator includes the following steps: S1. Obtain the liquid level setpoint and liquid level measurement value of the steam generator; S2. Calculate the liquid level deviation based on the liquid level set value and the liquid level measured value; S3. Input the liquid level setpoint, the liquid level measured value, the liquid level deviation, and an offset constant into a preset BP neural network; S4. The BP neural network is based on the forward propagation algorithm and outputs three values, which are used as the proportional coefficient, integral coefficient and derivative coefficient of the PID controller, respectively. S5. The PID controller calculates the required water flow rate based on the proportional coefficient, integral coefficient, derivative coefficient, and liquid level deviation. S6. Convert the water supply flow demand value into a valve opening command to adjust the water supply flow entering the steam generator; S7. Calculate the root mean square error of the current liquid level control as the loss function. In the next control cycle, calculate the new liquid level deviation based on the new liquid level measurement value. According to the new liquid level deviation, update the weight parameters of the BP neural network using the backpropagation algorithm. Adjust the weight parameters of the BP neural network through backpropagation to achieve adaptive updating of PID parameters until the deviation between the liquid level output value and the set value meets the control requirements.

[0007] Preferably, the BP neural network has a three-layer structure, including an input layer, a hidden layer, and an output layer, wherein the input layer has 4 neurons, the hidden layer has 10-20 neurons, and the output layer has 3 neurons.

[0008] Preferably, the number of neurons in the hidden layer of the BP neural network is 15.

[0009] Preferably, the calculation process of the forward propagation algorithm is as follows: The input signal is mapped to the weight parameters of the BP neural network to achieve the conversion of the input signal into PID parameters.

[0010] Preferably, the backpropagation algorithm uses the chain rule to calculate the weight update, and the gradient of the PID controller's output with respect to the input is replaced by the sign function sgn().

[0011] Preferably, the weight update amount is calculated as follows:

[0012] in, The inertia coefficient, For learning rate, For the current control cycle Below, the weights from the hidden layer to the output layer of the BP neural network are updated. For the current control cycle Below, the weight update amount from the input layer to the hidden layer of the BP neural network. For the current control cycle The loss function value under the following conditions This is the initial weight matrix from the hidden layer to the output layer of the BP neural network. This is the initial weight matrix from the input layer to the hidden layer of the BP neural network. This is for controlling the cycle number.

[0013] Preferably, the learning rate =0.1, the inertia coefficient =0.1.

[0014] Preferably, the water supply flow demand value output by the PID controller is the final water supply flow demand value calculated by combining the steam-water deviation received by the flow PI controller.

[0015] Preferably, the BP neural network and the PID controller form a BP-PID controller. The BP-PID controller is built and packaged into an algorithm module through the Simulink platform. The algorithm module is deployed in the DCS hardware-in-the-loop simulation platform and interacts with the nuclear power plant DCS system through a serial port. The input of the BP neural network also includes the historical deviation of the liquid level and the weight matrix of the previous moment. The output also includes the updated weight parameters, which are used for iterative calculation of PID parameters in the next control cycle.

[0016] Secondly, embodiments of the present invention provide a pressurized water reactor steam generator level control system, comprising: The data acquisition module is used to acquire the liquid level setpoint and liquid level measurement value of the steam generator; A deviation calculation module, connected to the data acquisition module, is used to calculate the liquid level deviation based on the liquid level set value and the liquid level measurement value; A BP neural network module connects the data acquisition module and the deviation calculation module, and the BP neural network module contains a preset BP neural network; The input layer of the BP neural network receives the liquid level setpoint, the liquid level measurement value, the liquid level deviation, and an offset constant. The BP neural network is configured to output three values ​​from its output layer based on the forward propagation algorithm, which serve as the proportional coefficient, integral coefficient, and derivative coefficient, respectively. The PID control module, connected to the deviation calculation module and the BP neural network module, is used to calculate and output the water supply flow demand value based on the proportional coefficient, integral coefficient, derivative coefficient, and liquid level deviation. The control execution module, connected to the PID control module, is used to convert the water supply flow demand value into a valve opening command to adjust the water supply flow. The weight adjustment module, which connects the BP neural network module and the deviation calculation module, is used to calculate the root mean square error of the current liquid level control as a loss function, and in the next control cycle, calculate a new liquid level deviation based on the new liquid level measurement value. According to the new liquid level deviation and the loss function, the weight parameters of the BP neural network are updated using the backpropagation algorithm.

[0017] Thirdly, a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the pressurized water reactor steam generator level control method described above.

[0018] Fourthly, embodiments of the present invention provide a computer-readable storage medium including a computer program, which, when executed by a processor, implements the steps of the above-described pressurized water reactor steam generator level control method.

[0019] Fifthly, a chip includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the pressurized water reactor steam generator level control method described above.

[0020] In a sixth aspect, embodiments of the present invention provide an electronic device including a computer program, which, when executed by the electronic device, implements the steps of the above-described pressurized water reactor steam generator level control method.

[0021] Compared with the prior art, the present invention has at least the following beneficial effects: A liquid level control method for pressurized water reactor steam generators deeply integrates BP neural networks and PID control. By acquiring the liquid level setpoint and measured values ​​and calculating the deviation, the BP neural network generates PID parameters. These parameters are then used to output a feedwater flow command via PID control, and the weights are updated through backpropagation to achieve parameter self-adaptation. This method overcomes the limitations of traditional PID parameters that are fixed or rely on experience for tuning. It can respond in real-time to the nonlinear characteristics of the steam generator and interference caused by false liquid levels, solving the pain points of traditional methods such as long debugging cycles, multi-objective conflicts, and difficulty in adapting to changes in operating conditions. The closed-loop process ensures real-time linkage between parameter adjustment and liquid level control, continuously reducing liquid level deviation and improving control flexibility and accuracy. It provides complete logical support for stable control under different power conditions.

[0022] Furthermore, the BP neural network is defined as a three-layer structure: 4 neurons in the input layer, 10-20 neurons in the hidden layer, and 3 neurons in the output layer. The 4 neurons in the input layer comprehensively capture the key states of the control system, ensuring the integrity of the input information. The selection of the range of 10-20 neurons in the hidden layer balances the ability to fit the nonlinear relationship of the steam generator with the computing power limitations of the nuclear power plant controller, avoiding underfitting due to an overly simple structure or increasing the computational burden and affecting real-time performance due to an overly complex structure. The 3 neurons in the output layer directly correspond to the core parameters of the PID controller, achieving a precise mapping between the input and control parameters. This structure lays a reliable hardware logic foundation for adaptive parameter adjustment, ensuring the feasibility of the BP neural network in industrial scenarios.

[0023] Furthermore, the number of hidden layer neurons in the BP neural network was determined to be 15. Through tests involving a power step decrease of 100%-50% and a liquid level setpoint step increase, it was verified that with 15 hidden layer neurons, the settling time and overshoot achieved the optimal balance: compared to 5 neurons, the settling time was significantly shortened, and the response speed to changes in operating conditions was improved; compared to more than 20 neurons, the problems of increased overshoot or prolonged settling time were avoided, while also controlling computational complexity. This setting ensures both the neural network's adaptability to false liquid levels and nonlinear characteristics of the steam generator and the real-time performance of the control algorithm, enabling the BP-PID controller to operate stably and efficiently under various operating conditions in actual nuclear power plants.

[0024] Furthermore, the forward propagation algorithm establishes a dynamic correlation between the system's real-time state and PID parameters through mapping operations between the input signal and weight parameters. It can accurately capture changes in signals such as the liquid level setpoint, measured value, and deviation, and output PID parameters adapted to the current operating conditions, achieving real-time matching of operating conditions and parameters. This operational logic breaks through the static limitations of traditional fixed parameters or experience-based tuning, enabling PID parameters to dynamically adjust with changes in steam generator power, liquid level fluctuations, and other operating conditions. This provides core computational support for parameter adaptation, improving the specificity of control parameters and the stability of control effects.

[0025] Furthermore, the backpropagation algorithm employs the chain rule to calculate the weight update and replaces the gradient of the controlled system's output with respect to the input with a sign function. The chain rule can accurately trace the correlation between the loss function and the weight parameters, enabling targeted weight adjustments and ensuring the scientific nature of parameter updates. Replacing the gradient with a sign function simplifies the calculation process, reduces the computational requirements of the nuclear power plant's DCS system, and avoids parameter update delays caused by complex gradient calculations. Simultaneously, the errors generated by the substitution can be compensated for by the learning rate, improving backpropagation efficiency without affecting control accuracy, ensuring real-time linkage between weight updates and liquid level control, and guaranteeing timely adaptive adjustments.

[0026] Furthermore, an inertia coefficient and a learning rate are introduced to construct a parameter update mechanism that balances stability and efficiency. The inertia coefficient incorporates the momentum term from the historical gradient direction, reducing oscillations during weight updates and preventing repeated weight adjustments due to fluctuations in operating conditions, thus accelerating convergence. The learning rate controls the weight update step size, preventing excessively large step sizes from hindering convergence or excessively small step sizes from causing slow convergence. This formula effectively balances the stability and efficiency of weight adjustment, enabling the BP neural network to quickly approach the optimal weight configuration and ensuring that the PID parameters are always adjusted in a direction that adapts to the current operating conditions, providing algorithmic assurance for the accuracy and stability of liquid level control.

[0027] Furthermore, the parameter combination of learning rate η=0.1 and inertia coefficient α=0.1 ensures both the efficiency of weight updates, avoiding slow convergence due to excessively small η, and the prevention of convergence difficulties caused by excessively large η; the inertia coefficient effectively suppresses oscillations in weight updates, making parameter adjustments more stable. Under different power conditions, it can achieve control effects with short settling time and small overshoot, significantly improving the robustness of the BP-PID algorithm and ensuring its stable adaptive control function under complex operating conditions of the steam generator.

[0028] Furthermore, the steam generator suffers from steam-water mismatch. A single PID control system struggles to balance liquid level and steam-water equilibrium. However, the flow PI controller, receiving the BP-PID output value and the steam-water deviation, can perform a secondary correction of the feedwater flow demand, effectively suppressing interference from false liquid levels. This overcomes the limitations of a single control logic, ensuring that feedwater flow regulation both meets liquid level control requirements and adapts to the steam-water balance, preventing significant liquid level fluctuations due to steam-water mismatch. This further improves the accuracy and stability of liquid level control, preventing excessively high liquid levels from corroding the turbine or excessively low liquid levels from damaging exposed heat transfer tubes.

[0029] Furthermore, the Simulink-based encapsulation modularizes the algorithm, facilitating serial communication with existing DCS systems in nuclear power plants without requiring significant hardware modifications, thus lowering the barrier to industrial application. The input includes historical level deviations and the previous time-series weight matrix, while the output contains updated weight parameters, enabling iterative parameter optimization and allowing the algorithm to continuously adapt to changing operating conditions. Hardware-in-the-loop deployment ensures the algorithm's feasibility in real-world industrial environments, while iterative optimization guarantees the continuity of control effectiveness, enabling the method to be stably applied in actual nuclear power plant production.

[0030] It is understood that the beneficial effects of the second to sixth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0031] In summary, the method of this invention integrates the adaptive capability of BP neural networks with the reliability of PID control, and its core advantage lies in achieving dynamic adaptation of PID parameters. It overcomes the limitations of traditional empirical tuning, can adapt to the nonlinear characteristics of steam generators and false liquid levels, exhibits small overshoot and stable regulation under different operating conditions, is compatible with nuclear power plant DCS systems, and has strong industrial applicability; it requires no complex hardware modifications, has a short commissioning cycle, and effectively ensures the safe and efficient operation of nuclear power plants.

[0032] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0033] Figure 1 It is a BP-PID neural network structure; Figure 2 It is a UTSG level BP-PID control system; Figure 3 The control process is based on a BP neural network. Figure 4 It is a BP neural network control module; Figure 5 The effects of learning rate and inertia coefficient on control performance are given, where (a) represents the effect of learning rate on control and (b) represents the effect of inertia coefficient on control. Figure 6 It is a BP neural network control module; Figure 7 The test is a step increase test of the 100%FP liquid level setpoint, where (a) is a step increase of 10% of the 100%FP liquid level setpoint, and (b) is a change in valve opening. Figure 8 The test is a step increase test of the 100%FP liquid level setpoint, where (a) is a step decrease of 10% of the 100%FP liquid level setpoint, and (b) is a change in valve opening. Figure 9The test is a step increase test of the 50%FP liquid level setpoint, where (a) is a step increase of 10% of the 50%FP liquid level setpoint, and (b) is a change in valve opening. Figure 10 The test was conducted to increase the liquid level setpoint by 20%FP, where (a) the liquid level setpoint increased by 10% by 20% and (b) the valve opening changed. Figure 11 A schematic diagram of a computer device provided in an embodiment of the present invention; Figure 12 This is a block diagram of a chip provided according to an embodiment of the present invention.

[0034] Among them, 60. Computer equipment; 61. Processor; 62. Memory; 63. Computer program; 600. Electronic device; 610. Processing unit; 620. Storage unit; 6201. Random access memory unit; 6202. Cache memory unit; 6203. Read-only memory unit; 6204. Program / utility; 6205. Program module; 630. Bus; 640. Display unit; 650. Input / output interface; 660. Network adapter; 700. External device. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] In the description of this invention, it should be understood that the terms "comprising" and "including" indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0037] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0038] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Additionally, the character " / " in this invention generally indicates that the preceding and following objects have an "or" relationship.

[0039] It should be understood that although terms such as first, second, third, etc., may be used in the embodiments of the present invention to describe the preset range, these preset ranges should not be limited to these terms. These terms are only used to distinguish the preset ranges from one another. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0040] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0041] The accompanying drawings illustrate various structural schematic diagrams according to embodiments disclosed in this invention. These drawings are not to scale, and some details have been enlarged for clarity, and some details may have been omitted. The shapes of the various regions and layers shown in the drawings, as well as their relative sizes and positional relationships, are merely exemplary and may deviate from reality due to manufacturing tolerances or technical limitations. Furthermore, those skilled in the art can design regions / layers with different shapes, sizes, and relative positions as needed.

[0042] This invention provides a liquid level control method for pressurized water reactor steam generators. It combines a BP neural network with PID control to develop a BP-PID controller, which can adaptively adjust parameters according to different operating conditions and system characteristics to meet the needs of the control system. The BP-PID controller is deployed in a hardware-in-the-loop simulation platform to test the simulation model of a UTSG. Comparison with traditional PID control algorithms verifies its good control performance, and this control algorithm can be used for control implementation in nuclear power plants.

[0043] Please see Figure 1 In a backpropagation (BP) neural network structure, I represents the input layer, M represents the hidden layer, and O represents the output layer. The number of neurons in the input layer is 4 (…). ~ These represent the setpoints in the control system. r ( k ), measured values y ( k ),deviation e ( k And bias constants, number of hidden layer neurons m The output layer has 3 neurons, representing... , and Three PID parameters. BP-PID control can dynamically adjust the three parameters of the PID controller—proportional constant, derivative constant, and integral constant—online in real time based on operating data under different working conditions, ultimately making the deviation between the output value and the set value zero, thus achieving rapid reduction of deviation.

[0044] If the output layer data is taken as PID parameters, it is represented as follows: (1) The above formula completes the calculation of the forward propagation of the BP neural network, realizing the mapping from the input signal to the PID parameters.

[0045] The weight parameters are adjusted by setting a loss function to measure the effect of the current output. The loss function is calculated using the root mean square error, as follows: (2) Backpropagation in a BP neural network adjusts the weight parameters through bias. During this adjustment process, the step size for each weight update, i.e., the learning rate, needs to be adjusted. The learning rate needs to be configured. Setting it too high may make it difficult to find the optimal solution, while setting it too low will result in slow convergence. An inertia coefficient also needs to be set. The inertia coefficient can introduce a momentum term from the historical gradient direction during weight updates, resulting in less oscillation and faster convergence. Its backpropagation calculation is as follows: The weights are updated as follows: (3) The above equation is calculated using the chain rule, as follows: (4) in, The calculation result of the PID control expression can be obtained from the incremental expression of the PID (5). The calculation is performed, and the result is shown in (6).

[0046] (5) (6) In formula (6) It is the gradient of the output of the controlled system with respect to the input, usually expressed as a sign function. sgn ( The substitution is shown in equation (7), and the error caused by the substitution is compensated by the learning rate.

[0047] (7) This update of the weights Represented as shown in equation (8), for The same applies to updates.

[0048] (8) Weights The update is represented as: (9) The above formula completes the calculation of backpropagation weight update for the BP neural network, realizing the ability of the BP neural network to adaptively learn and adjust control parameters according to different operating conditions and system characteristics.

[0049] Please see Figure 2 A UTSG level control system based on BP-PID was built, combining a BP neural network with a level PID controller, while the flow PI controller retained its original control structure.

[0050] Please see Figure 3 This invention discloses a liquid level control method for a pressurized water reactor steam generator. Upon initiation of control, the weights and other parameters of the BP neural network are first initialized. Then, the current deviation and other signals are sampled and calculated, and input into the BP neural network to calculate the output value as the current PID parameters. The new parameters are then input into the UTSG liquid level control system to begin control at the next moment. Simultaneously, the new liquid level deviation is calculated and used for backpropagation to adjust the weight parameters, thereby rapidly reducing the deviation. The specific steps are as follows: S1. Establish a BP neural network control module Please see Figure 4 A BP-PID control module was designed in Simulink, comprising a neural network calculation module and a level PID controller. External inputs include the level setpoint, measured level, level deviation, and control output (used to calculate the partial derivative of the level with respect to the controller output). The output is the water flow demand calculated by the level PID controller. The BP-PID algorithm module's main inputs also include the historical level deviation and the weight matrix from the previous time step, with outputs being the PID control parameters and the updated weight parameters.

[0051] S2. Setting the hyperparameters of the BP neural network Although BP neural network control can adapt to different systems and operating conditions for rapid control, its internal parameter settings have a significant impact on the control effect. These mainly include the learning rate, inertia coefficient, and initial weight parameters. First, the impact of different learning rates and inertia coefficients on the control was tested. Figure 4 The constructed BP-PID control algorithm was applied to an offline control system, and experiments were conducted to test the step response of the liquid level setpoint increasing by 10% at 100% FP. Specific results are as follows: Figure 5 As shown.

[0052] Please see Figure 5 To compare the impact of changes in learning rate and inertia coefficient on control performance under the same operating conditions, since the overshoot does not change significantly during the change process, only the settling time is selected as the evaluation index. Figure 5 (a) shows the comparison results of changing the learning rate while keeping the inertia coefficient constant. It can be seen that the control effect improves as the learning rate decreases, but the effect does not change much when the learning rate is less than 0.1. At the same time, setting the learning rate too small may reduce the learning efficiency of the neural network and affect the control effect. Figure 5 (b) shows the comparison results of changing the inertia coefficient while keeping the learning rate constant. It can be seen that the control effect improves as the inertia coefficient decreases, but the effect is not significant when the inertia coefficient is less than 0.1. Based on the above results, the learning rate η = 0.1 and the inertia coefficient α = 0.1 were finally selected.

[0053] In addition to the parameters mentioned above, the number of hidden layers in a BP neural network also significantly affects the control performance. The control effect under different numbers of hidden layers was verified using an offline control system. The operating conditions were selected as a 100%-50% power step decrease and a 20% step increase from the 100% power level setpoint. The test results are shown in Tables 1 and 2.

[0054] Table 1. The Influence of the Number of Hidden Layers on Power Step Tests

[0055] Table 2. The Influence of the Number of Hidden Layers on the Step Test of Liquid Level Setpoint

[0056] Table 1 shows the power step drop test results from 100% to 50%. Since the liquid level fluctuation range did not change significantly, only the settling time was used as the evaluation criterion. Table 1-1 shows that the control effect gradually improves as the number of hidden layers increases from 5 to 15, but the control effect remains relatively unchanged when the number of hidden layers is between 5 and 20. The control effect increases significantly when the number of hidden layers increases to 30, but gradually deteriorates as the number of hidden layers continues to increase. Table 2 shows that the settling time is relatively long when the number of hidden layers is 5. Between 10 and 20 layers, the settling time and overshoot are not significantly different. As the number of hidden layers continues to increase, the overshoot or settling time will increase significantly. Based on the comprehensive analysis of the results of the two test conditions, a hidden layer number between 10 and 20 is more suitable. Therefore, 15 was ultimately selected as the number of hidden layers for BP.

[0057] S3. Design the BP-PID control module interface The above steps realize the development of the BP-PID control algorithm module and the setting of the BP neural network hyperparameters. The designed BP-PID algorithm module is built in Simulink, combining the BP-PID control algorithm with a flow PI controller to form a... Figure 6 The UTSG level control algorithm module shown.

[0058] Please see Figure 6 This paper describes a UTSG level control algorithm based on BP-PID designed in the Simulink platform. It mainly consists of three parts: a BP-PID calculation module, a flow PI controller, and an interface module. The BP-PID calculation module is the control algorithm constructed in step S1. The flow controller calculates the final feedwater flow requirement value by receiving the output value of the BP-PID module and the steam-water deviation. The feedwater flow requirement value is then used by a valve opening calculation program to calculate the specific valve opening for control. The interface module handles external interaction, receiving three input signals: level setpoint, level measurement value, and steam-water deviation. The level deviation required in BP-PID control is calculated from the level setpoint and level measurement value, requiring no external input. The output signal of this module is the feedwater flow requirement value.

[0059] In another embodiment of the present invention, a pressurized water reactor steam generator level control system is provided. This system can be used to implement the above-mentioned pressurized water reactor steam generator level control method. Specifically, the pressurized water reactor steam generator level control system includes a data acquisition module, a deviation calculation module, a BP neural network module, a PID control module, a control execution module, and a weight adjustment module.

[0060] The data acquisition module is used to acquire the liquid level setpoint and liquid level measurement value of the steam generator. A deviation calculation module, connected to the data acquisition module, is used to calculate the liquid level deviation based on the liquid level set value and the liquid level measurement value; A BP neural network module connects the data acquisition module and the deviation calculation module, and the BP neural network module contains a preset BP neural network; The input layer of the BP neural network receives the liquid level setpoint, the liquid level measurement value, the liquid level deviation, and an offset constant. The BP neural network is configured to output three values ​​from its output layer based on the forward propagation algorithm, which serve as the proportional coefficient, integral coefficient, and derivative coefficient, respectively. The PID control module, connected to the deviation calculation module and the BP neural network module, is used to calculate and output the water supply flow demand value based on the proportional coefficient, integral coefficient, derivative coefficient, and liquid level deviation. The control execution module, connected to the PID control module, is used to convert the water supply flow demand value into a valve opening command to adjust the water supply flow. The weight adjustment module, which connects the BP neural network module and the deviation calculation module, is used to calculate the root mean square error of the current liquid level control as a loss function, and in the next control cycle, calculate a new liquid level deviation based on the new liquid level measurement value. According to the new liquid level deviation and the loss function, the weight parameters of the BP neural network are updated using the backpropagation algorithm.

[0061] This invention provides a terminal device comprising a processor and a memory. The memory stores a computer program, which includes program instructions. The processor executes the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or other general-purpose processors, graphics processing units (GPUs), tensor processing units (TPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions to achieve a corresponding method flow or corresponding function. The processor described in this embodiment can be used for the operation of a pressurized water reactor steam generator level control method, including: The process involves: acquiring the setpoint and measured level of the steam generator; calculating the level deviation based on the setpoint and measured level; inputting the setpoint, measured level, level deviation, and a bias constant into a preset BP neural network; the BP neural network, based on a forward propagation algorithm, outputting three values ​​as the proportional, integral, and derivative coefficients of the PID controller; the PID controller calculating and outputting the required feedwater flow rate based on the proportional, integral, and derivative coefficients and the level deviation; converting the required feedwater flow rate into a valve opening command to adjust the feedwater flow rate into the steam generator; calculating the root mean square error of the current level control as a loss function; calculating a new level deviation based on the new measured level in the next control cycle; and updating the weight parameters of the BP neural network using a backpropagation algorithm based on the new level deviation. This backpropagation adjustment of the BP neural network's weight parameters achieves adaptive updating of the PID parameters until the deviation between the level output value and the setpoint meets the control requirements.

[0062] Please see Figure 5The terminal device is a computer device. In this embodiment, the computer device 60 includes a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When executed by the processor 61, the computer program 63 implements the pressurized water reactor steam generator level control method of the embodiment. To avoid repetition, these details are not elaborated here. Alternatively, when executed by the processor 61, the computer program 63 implements the functions of each model / unit in the pressurized water reactor steam generator level control system of the embodiment. To avoid repetition, these details are not elaborated here.

[0063] Computer device 60 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. Computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will understand that... Figure 5 This is merely an example of computer device 60 and does not constitute a limitation on computer device 60. It may include more or fewer components than shown, or combine certain components, or different components. For example, computer device may also include input / output devices, network access devices, buses, etc.

[0064] The processor 61 may be a Central Processing Unit (CPU), or other general-purpose processors, graphics processing units (GPUs), tensor processing units (TPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0065] The memory 62 can be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 can also be an external storage device of the computer device 60, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. equipped on the computer device 60.

[0066] Furthermore, the memory 62 may include both internal storage units of the computer device 60 and external storage devices. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 can also be used to temporarily store data that has been output or will be output.

[0067] Please see Figure 6 The terminal device is an electronic device 600, which is manifested in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including storage unit 620 and processing unit 610), a display unit 640, etc.

[0068] The storage unit stores program code, which can be executed by the processing unit 610 to perform the steps described in the method section of this specification according to various exemplary embodiments of the present invention. For example, the processing unit 610 can perform actions such as... Figure 1 The steps are shown in the figure.

[0069] Storage unit 620 may include a readable medium in the form of a volatile storage unit, such as random access memory (RAM) 6201 and / or cache memory 6202, and may further include a read-only memory (ROM) 6203.

[0070] Storage unit 620 may also include a program / utility 6204 having a set (at least one) program module 6205, such program module 6205 including but not limited to: operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.

[0071] Bus 630 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the multiple bus structures.

[0072] Electronic device 600 can also communicate with one or more external devices 700 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable a user to interact with electronic device 600, and / or with any device that enables electronic device 600 to communicate with one or more other computing devices (e.g., router, modem). This communication can be performed via input / output interface 650. Furthermore, electronic device 600 can also communicate with one or more networks (e.g., local area network, wide area network, and / or public network, such as the Internet) via network adapter 660. Network adapter 660 can communicate with other modules of electronic device 600 via bus 630. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.

[0073] Example 4 This invention also provides a storage medium, specifically a computer-readable storage medium, which is a memory device in a terminal device for storing programs and data. It is understood that the computer-readable storage medium here can include both built-in storage media in the terminal device and extended storage media supported by the terminal device; it can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for loading and execution by a processor, which can be one or more computer programs (including program code). More specific examples of the computer-readable storage medium include: an electrical connection with one or more wires, a portable disk, a hard disk, random access memory, read-only memory, erasable programmable read-only memory, optical fiber, portable compact disk read-only memory, optical storage device, magnetic storage device, or any suitable combination thereof.

[0074] Computer-readable storage media also include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable storage medium can also be any readable medium other than a readable storage medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium can be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, radio frequency, etc., or any suitable combination thereof.

[0075] Program code for performing the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the user's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0076] One or more instructions stored in a computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the pressurized water reactor steam generator level control method in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor to perform the following steps: The process involves: acquiring the setpoint and measured level of the steam generator; calculating the level deviation based on the setpoint and measured level; inputting the setpoint, measured level, level deviation, and a bias constant into a preset BP neural network; the BP neural network, based on a forward propagation algorithm, outputting three values ​​as the proportional, integral, and derivative coefficients of the PID controller; the PID controller calculating and outputting the required feedwater flow rate based on the proportional, integral, and derivative coefficients and the level deviation; converting the required feedwater flow rate into a valve opening command to adjust the feedwater flow rate into the steam generator; calculating the root mean square error of the current level control as a loss function; calculating a new level deviation based on the new measured level in the next control cycle; and updating the weight parameters of the BP neural network using a backpropagation algorithm based on the new level deviation. This backpropagation adjustment of the BP neural network's weight parameters achieves adaptive updating of the PID parameters until the deviation between the level output value and the setpoint meets the control requirements.

[0077] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0078] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0079] Based on the BP-PID control algorithm, it is deployed in the DCS hardware-in-the-loop simulation platform. The BP-PID-based UTSG level control algorithm designed in the Simulink platform is encapsulated as an algorithm module, and then data is exchanged with the DCS system through a serial port.

[0080] The control effect was tested under the following conditions: a 10% increase or decrease in the liquid level setpoint at 100% FP; a 10% increase in the liquid level setpoint at 50% FP, 30% FP, and 20% FP; and a step decrease in power from 100% FP to 50% FP and from 100% FP to 30% FP. Figure 7 and Figure 8 The effect of liquid level step rise and fall test at 100% FP.

[0081] The liquid level setpoint step increase test was performed at 50%FP and 20%FP as follows: Figure 9 and Figure 10 As shown.

[0082] Table 3 Comparison of Control Indicators for Step Changes in Liquid Level

[0083] Figures 7 to 10 To compare the control effect of step disturbances in the liquid level setpoint after 100 seconds of steady-state operation at 100%, 50%, 30%, and 20% power levels, Table 3 shows the specific control indicators. The results show that the BP-PID control effect has zero overshoot in the 100% FP liquid level step rise / fall, with a slightly longer settling time than the NSGA-II tuned effect. In the 50% FP liquid level step rise, the settling time is shorter than the tuned parameter, but the overshoot is slightly larger. In the 30% FP liquid level rise step, the control effect is better than the tuned parameter in both settling time and overshoot. In the 20% FP liquid level rise step, the overshoot is even better, but the settling time is slightly longer. In the above control results, the valve fluctuation is relatively small due to the relatively stable control. In the 100% FP liquid level rise / fall experiment, due to the large demand for feedwater flow, there is a significant increase in the FLCV opening; the FLCV changes are relatively small at other power levels.

[0084] The results above show that the BP-PID control effect is significantly better than the original PID control. Compared with the parameters tuned by NSGA-II, it also improves in some operating conditions, proving that the BP-PID algorithm has a good effect in UTSG level control.

[0085] In summary, this invention provides a liquid level control method and system for pressurized water reactor steam generators, effectively solving the problems of traditional PID parameters relying on experience for tuning, long debugging cycles, and difficulty in adapting to changes in operating conditions. By combining a BP neural network with PID control, adaptive updates of PID parameters are achieved. Under different power conditions such as 100%, 50%, 30%, and 20% FP, the liquid level control overshoot is significantly lower than that of traditional PID control, the settling time is adapted to industrial needs, and valve fluctuations are small and operation is stable. Compared with NSGA-II tuning parameters, the control effect is superior under certain operating conditions, and it can adapt to the nonlinear characteristics of steam generators and false liquid level interference. It can be deployed in nuclear power plant DCS systems to ensure the safe and stable operation of steam generators and has good industrial applicability.

[0086] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0087] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0088] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0089] In the embodiments provided by this invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0090] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0091] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0092] If the integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random-access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0093] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus, and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can 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, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0094] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0095] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0096] The above content is only for illustrating the technical concept of the present invention and should not be construed as limiting the scope of protection of the present invention. Any modifications made to the technical solution based on the technical concept proposed in this invention shall fall within the scope of protection of the claims of this invention.

Claims

1. A method for controlling the liquid level in a pressurized water reactor steam generator, characterized in that, Includes the following steps: S1. Obtain the liquid level setpoint and liquid level measurement value of the steam generator; S2. Calculate the liquid level deviation based on the liquid level set value and the liquid level measured value; S3. Input the liquid level setpoint, the liquid level measured value, the liquid level deviation, and an offset constant into a preset BP neural network; S4. The BP neural network is based on the forward propagation algorithm and outputs three values, which are used as the proportional coefficient, integral coefficient and derivative coefficient of the PID controller, respectively. S5. The PID controller calculates the required water flow rate based on the proportional coefficient, integral coefficient, derivative coefficient, and liquid level deviation. S6. Convert the water supply flow demand value into a valve opening command to adjust the water supply flow entering the steam generator; S7. Calculate the root mean square error of the current liquid level control as the loss function. In the next control cycle, calculate the new liquid level deviation based on the new liquid level measurement value. According to the new liquid level deviation, update the weight parameters of the BP neural network using the backpropagation algorithm. Adjust the weight parameters of the BP neural network through backpropagation to achieve adaptive updating of PID parameters until the deviation between the liquid level output value and the set value meets the control requirements.

2. The pressurized water reactor steam generator level control method according to claim 1, characterized in that, The BP neural network has a three-layer structure, including an input layer, a hidden layer, and an output layer. The input layer has 4 neurons, the hidden layer has 10-20 neurons, and the output layer has 3 neurons.

3. The pressurized water reactor steam generator level control method according to claim 2, characterized in that, The hidden layer of the BP neural network has 15 neurons.

4. The pressurized water reactor steam generator level control method according to claim 1, characterized in that, The calculation process of the forward propagation algorithm is as follows: The input signal is mapped to the weight parameters of the BP neural network to achieve the conversion of the input signal into PID parameters.

5. The pressurized water reactor steam generator level control method according to claim 1, characterized in that, The backpropagation algorithm uses the chain rule to calculate the weight update, and the gradient of the PID controller's output with respect to the input is replaced by the sign function sgn().

6. The pressurized water reactor steam generator level control method according to claim 5, characterized in that, The weight update amount is calculated as follows: in, The inertia coefficient, For learning rate, For the current control cycle Below, the weights from the hidden layer to the output layer of the BP neural network are updated. For the current control cycle Below, the weight update amount from the input layer to the hidden layer of the BP neural network. For the current control cycle The loss function value under the following conditions This is the initial weight matrix from the hidden layer to the output layer of the BP neural network. This is the initial weight matrix from the input layer to the hidden layer of the BP neural network. This is for controlling the cycle number.

7. The pressurized water reactor steam generator level control method according to claim 6, characterized in that, The learning rate =0.1, the inertia coefficient =0.

1.

8. The method for controlling the liquid level of a pressurized water reactor steam generator according to claim 1, characterized in that, The water supply flow demand value output by the PID controller is the final water supply flow demand value calculated by combining the steam-water deviation received by the flow PI controller.

9. The pressurized water reactor steam generator level control method according to claim 1, characterized in that, The BP neural network and the PID controller form a BP-PID controller. The BP-PID controller is built and packaged into an algorithm module through the Simulink platform. The algorithm module is deployed in the DCS hardware-in-the-loop simulation platform and interacts with the nuclear power plant DCS system through a serial port. The input of the BP neural network also includes the historical deviation of the liquid level and the weight matrix of the previous moment. The output also includes the updated weight parameters, which are used for iterative calculation of PID parameters in the next control cycle.

10. A liquid level control system for a pressurized water reactor steam generator, characterized in that, include: The data acquisition module is used to acquire the liquid level setpoint and liquid level measurement value of the steam generator; A deviation calculation module, connected to the data acquisition module, is used to calculate the liquid level deviation based on the liquid level set value and the liquid level measurement value. A BP neural network module connects the data acquisition module and the deviation calculation module, and the BP neural network module contains a preset BP neural network; The input layer of the BP neural network receives the liquid level setpoint, the liquid level measurement value, the liquid level deviation, and an offset constant. The BP neural network is configured to output three values ​​from its output layer based on the forward propagation algorithm, which serve as the proportional coefficient, integral coefficient, and derivative coefficient, respectively. The PID control module, connected to the deviation calculation module and the BP neural network module, is used to calculate and output the water supply flow demand value based on the proportional coefficient, integral coefficient, derivative coefficient, and liquid level deviation. The control execution module, connected to the PID control module, is used to convert the water supply flow demand value into a valve opening command to adjust the water supply flow. The weight adjustment module, which connects the BP neural network module and the deviation calculation module, is used to calculate the root mean square error of the current liquid level control as a loss function, and in the next control cycle, calculate a new liquid level deviation based on the new liquid level measurement value. According to the new liquid level deviation and the loss function, the weight parameters of the BP neural network are updated using the backpropagation algorithm.

Images