A data-driven steam generator water level independent modeling system and method

Abstract

The present invention discloses a data-driven independent modeling system and method for steam generator water level, which specifically includes: after reading externally input steam generator related parameter data, data processing is performed on the data and data output is performed through two paths, one path is output to a transfer function dynamic parameter model, and the other path is output to a transfer function model, the transfer function dynamic parameter model uses the input data to calculate through a neural network and then outputs the real-time parameter value required for the transfer function calculation, the transfer function model performs real-time calculation based on the input data and the real-time parameter value, and finally outputs the water level value of the steam generator. The present invention realizes independent modeling of the steam generator water level in a data-driven manner, while improving the simulation accuracy of the steam generator water level, still maintaining the technical effect of facilitating controller design, and is helpful to carry out research on digital twin technology for complex systems.

Description

Technical Field

[0001] The present invention relates to the technical field of nuclear power system equipment simulation, and in particular to a data-driven independent modeling system and method for steam generator water level. Background Art

[0002] Steam generators are key equipment in pressurized water reactor nuclear power plants. They are used to transfer the heat energy generated by the fission reaction in the nuclear reactor core to the secondary circuit through the primary circuit coolant, generate high-temperature and high-pressure steam to drive the steam turbine to do work and use the generator to convert mechanical energy into electrical energy. If the water level of the steam generator in the secondary circuit of a nuclear power plant is not properly controlled, frequent reactor shutdowns will occur. This type of shutdown due to violation of the water level safety limit is very common when the reactor is operating at a low power level. At this time, the system exhibits strong non-minimum phase system characteristics and the flow signal is no longer credible. Therefore, systematic research on steam generators and improving modeling accuracy to timely track and feedback their operating status and properly control them can effectively reduce the number of costly shutdowns.

[0003] There are two main methods for modeling the water level of steam generators: one is a nonlinear differential time-domain model based on the laws of conservation of energy, momentum and mass. It can relatively accurately simulate the dynamic changes in the water level during the actual operation of the steam generator. The other is transfer function modeling based on thermal hydraulic experimental identification. The steam generator water level transfer function model proposed by Irving is widely used in the control algorithm design process. The model consists of several linear models, each of which is established at a specific power operating point. The change process between different powers is treated as a linear process, reducing the complexity of the mathematical model.

[0004] However, both of the above two modeling methods have shortcomings. Mechanism modeling is not suitable for the design of automatic control systems due to the high complexity of its mathematical model. The transfer function model is relatively simple, but it is difficult to accurately fit the actual operating data from low load to high load. Therefore, the data-driven independent modeling method of steam generator water level proposed in the present invention will make up for the shortcomings of the above two methods, while ensuring that the mathematical model is convenient for controller design and improving the simulation accuracy of the model. And it does not rely on various modeling and simulation software, and can independently develop programs to realize dynamic simulation of steam generator water level.

[0005] The existing patent CN118409514A discloses a simulation method for an evaporator water level control system, the method comprising: obtaining a pre-constructed evaporator water level control model; based on a real operating condition signal, controlling the start of the evaporator water level control model; during the operation of the evaporator water level control model, updating the response parameters of the water supply valve regulating unit and the water pump regulating unit, and obtaining the value of the water supply valve opening output by the evaporator water level control model; if it is determined based on the value of the water supply valve opening that a preset stability condition is met, then terminating the simulation of the evaporator water level control model.

[0006] The existing patent CN110879620A discloses a nuclear power vertical steam generator liquid level control method and system, which introduces a fractional-order PID controller and provides a parameter setting method, which can reduce tracking errors and meet the water level control requirements of the steam generator.

[0007] None of the above prior arts involve modeling the water level value of the steam generator using a transfer function model and a transfer function dynamic parameter model.

[0008] In view of the above technical problems, the present invention is proposed. Summary of the invention

[0009] Based on the above technical problems, the purpose of the present invention is to propose a data-driven independent modeling system and method for steam generator water level, so as to solve the problem that the steam generator water level modeling method used for controller design in the prior art cannot meet the simulation accuracy requirements when the nuclear reactor power changes significantly.

[0010] To achieve the above objectives, the present invention proposes a data-driven independent modeling system and method for steam generator water level. The specific technical solution is as follows:

[0011] A data-driven independent modeling system for steam generator water level, comprising a data reading module, a data processing module connected to the output end of the data reading module, a dynamic parameter model module and a transfer function model module connected to the output end of the data processing module, and a result output module connected to the output end of the transfer function model module, wherein the output end of the dynamic parameter model module is connected to the input end of the transfer function model module;

[0012] Wherein, the data reading module is used to read the steam generator related parameter data input externally;

[0013] The data processing module is used to process data and output data through two paths, one path is output to the dynamic parameter model module, and the other path is output to the transfer function model module;

[0014] The dynamic parameter model module is used to obtain the real-time parameter values ​​required for transfer function calculation using input data, and output the real-time parameter values ​​to the transfer function model module;

[0015] The transfer function model module is used to perform real-time calculations based on input data and real-time parameter values;

[0016] The result output module is used to output the water level value of the steam generator.

[0017] Furthermore, the steam generator related parameters include feed water flow, steam flow, power, primary circuit temperature change and feed water temperature change.

[0018] Furthermore, the feed water flow and the steam flow are output to the transfer function model module, and the steam flow, the power, the primary circuit temperature change and the feed water temperature change are output to the dynamic parameter model module.

[0019] Furthermore, the transfer function structure in the transfer function model module includes an integral link G1 / s, an inertia link G2 / (1+τ2s) and an oscillation link G3s(1 / τ 2 +4π 2 / T 2 +2s / τ+s 2 );

[0020] Among them, G1 is a coefficient independent of the load, G2 is the gain coefficient of the inertia link, G3 is the gain coefficient of the oscillation link, s is the complex parameter in the Laplace transform, τ is the damping time constant, τ2 is the inertia time constant, and T is the oscillation period.

[0021] Furthermore, the transfer function equation of the steam generator water level in the transfer function model is:

[0022] y(s)=(G1 / s-G2 / (1+τ2s))(q e (s)-q v (s))+(G3s(1 / τ 2 +4π 2 / T 2 +2s / τ+s 2 )q e (s)

[0023] Among them, q e (s) is the water flow rate, q v (s) is the steam flow rate.

[0024] Furthermore, the state equation after the transfer function equation of the steam generator water level is discretized is:

[0025] x1(t)=x1(t-Δt)+Δt G1[q e (t-Δt)–q v (t-Δt)]

[0026] x2(t)=x2(t-Δt)–α2x2(t-Δt)+β2[q e (t-Δt)-q v (t-Δt)]

[0027] x3(t)=x3(t-Δt)–α3x3(t-Δt)+x4(t-Δt)+β3q e (t-Δt)

[0028] x4(t)=x4(t-Δt)–α4x3(t-Δt)

[0029] y(t)=x1(t)+x2(t)+x3(t)

[0030] Where α2 = 1 – e –Δt / τ2 , β2=G2(1–e –Δt / τ2 ), α3=2–2e –Δt / τ cos2πΔt / T,α4=1+e –2Δt / τ –2e –Δt / τ cos(2πΔt / T), β3=(G3T / 2π)e –Δt / τ sin(2πΔt / T);

[0031] The matrix calculation method can be used to quickly solve the state equation.

[0032] Furthermore, the process of constructing the transfer function dynamic parameter model in the dynamic parameter model module includes:

[0033] The neural network algorithm in machine learning is used to construct the dynamic parameter model of the transfer function;

[0034] Obtain training data sets through experimental calculations of physical models or high-precision mechanism simulation models;

[0035] The transfer function dynamic parameter model is trained using the training data set, and the model is modified according to the training results.

[0036] Furthermore, the input layer parameters of the transfer function dynamic parameter model are set to power, steam flow, primary circuit temperature change and feed water temperature change, and the output layer parameters are set to the parameters to be solved G2, G3, τ2, T, τ in the transfer function.

[0037] The present invention also proposes a data-driven independent modeling method for steam generator water level, which is implemented by the data-driven independent modeling system for steam generator water level, and comprises:

[0038] Read the externally input steam generator related parameter data;

[0039] After the data is processed, two transmission paths are performed, and one path of data is directly input into the transfer function model;

[0040] Input another channel of data into the transfer function dynamic parameter model, and output the real-time parameter value required by the transfer function;

[0041] The transfer function model performs real-time calculations based on the input data and real-time parameter values, and outputs the water level value of the steam generator.

[0042] The present invention proposes a computer-readable storage medium, which includes a stored computer program. When the computer program is executed by a processor, the device where the storage medium is located is controlled to execute the steps of the above-mentioned data-driven steam generator water level independent modeling method.

[0043] Based on the above technical solution, the data-driven independent modeling system and method for steam generator water level proposed in the present invention has at least the following beneficial effects compared with the prior art:

[0044] 1. The present invention proposes a data-driven independent modeling system and method for steam generator water level, which can improve the accuracy of steam generator water level simulation by realizing independent modeling of steam generator water level in a data-driven manner.

[0045] 2. The present invention proposes a data-driven independent modeling system and method for steam generator water level, which has a simple system structure and modeling model and is convenient for controller design.

[0046] 3. The present invention proposes a data-driven independent modeling system and method for steam generator water level. After reading the relevant parameters of the steam generator, data transmission and calculation are performed in two ways, which can improve the efficiency of steam generator water level modeling.

[0047] 4. The present invention proposes a data-driven independent modeling system and method for steam generator water level, which solves the problem of independent modeling of steam generator water level by combining parameter modeling and real-time function calculation. The research method of this technical solution is helpful to carry out research on digital twin technology for complex systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0049] Figure 1 A structural schematic diagram of a data-driven steam generator water level independent modeling system proposed by the present invention;

[0050] Figure 2 A schematic diagram of the composition of a transfer function in a data-driven steam generator water level independent modeling system proposed by the present invention;

[0051] Figure 3 A schematic diagram of the structure of a dynamic parameter model of a transfer function in a data-driven steam generator water level independent modeling system proposed by the present invention;

[0052] Figure 4 A schematic flow chart of a data-driven independent modeling method for steam generator water level proposed in the present invention. DETAILED DESCRIPTION

[0053] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0054] The present invention is further described in detail below in conjunction with specific embodiments, and these embodiments cannot be understood as limiting the scope of protection claimed by the present invention. The term "including" when used indicates the existence of a feature, but does not exclude the existence or addition of one or more other features; the orientation or position relationship indicated by the terms "lateral", "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inside", "outside", etc. is based on the orientation or position relationship shown in the drawings, which is only for the convenience of description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention; in addition, the terms "first" and "second" are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance.

[0055] In the description, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be directly connected, or indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. In addition, in the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0056] Example 1

[0057] This embodiment proposes a data-driven independent modeling system for steam generator water level, which realizes independent modeling of steam generator water level in a data-driven manner. Figure 1 As shown, the system includes a data reading module, a data processing module, a dynamic parameter model module, a transfer function model module and a result output module. The output end of the data reading module is connected to the input end of the data processing module, the output end of the data processing module is connected to the input end of the dynamic parameter model module and the input end of the transfer function model module, the output end of the dynamic parameter model module is connected to the input end of the transfer function model module, and the output end of the transfer function model module is connected to the input end of the result output module.

[0058] Among them, the data reading module is used to read the steam generator related parameter data input externally; the data processing module is used to process the read external input data, and output it to the dynamic parameter model module and the transfer function model module through two channels respectively; the dynamic parameter model module is used to use the input data to obtain the real-time parameters required for the transfer function calculation, and output the obtained real-time parameters to the transfer function model module; the transfer function model module is used to perform real-time calculation based on the input data and real-time parameters; the result output module is used to output the water level value of the steam generator according to the calculation result of the transfer function model module.

[0059] The present invention proposes a data-driven independent modeling system for steam generator water level, which realizes independent modeling of steam generation water level in a data-driven manner, thereby improving the simulation accuracy of steam generator water level while still maintaining the technical effect of facilitating controller design.

[0060] The following is a detailed description of a data-driven independent modeling system for steam generator water level proposed by the present invention.

[0061] In this embodiment, the external input data read by the data reading module includes the water flow rate q e (t), steam flow rate q v (t), power P, primary circuit temperature change Δq p (t), feed water temperature change Δq f (t) and other parameter values.

[0062] In this embodiment, the data processing module processes the data and outputs it through two channels. One channel is directly transmitted to the transfer function model module, and the other channel is transmitted to the dynamic parameter model module as an input signal. The data is distributed according to demand. For example, the input data transmitted to the transfer function model module includes the water flow rate q e (t) and steam flow rate q v(t); the input data passed to the dynamic parameter model module include power P, steam flow q v (t), primary circuit temperature change Δq p (t) and feed water temperature change Δq f (t), the variables passed to the dynamic parameter model module must be additionally normalized.

[0063] In this embodiment, the transfer function model module includes a constructed steam generator water level transfer function model.

[0064] See also Figure 2 As shown in Figure 2, the composition of the steam generator water level transfer function model is as follows:

[0065] Changes in the power generation demand of nuclear power plants will lead to changes in the steam flow demand of steam turbines, and also require changes in the feedwater flow of steam generators. Many factors affect the water level of steam generators. The transfer function structure of the steam generator water level model is determined by breaking down the influencing factors, including:

[0066] Component 1: Points

[0067] The changes in the flow rates of feed water flowing into the steam generator and steam flowing out of the steam generator will change the actual water loading of the steam generator, thereby causing changes in the water level. The integral link is used to reflect the dynamic changes of the process. The integral link is designed to be G1 / s to simulate the mass volume effect of the steam generator.

[0068] Composition 2: Inertia link

[0069] When the steam flow rate decreases, the steam pressure at the top of the steam generator will increase, causing the steam bubbles inside the steam generator to collapse and rupture, resulting in a lower initial water level. When the feed water flow rate increases, the newly flowing cold water will encounter the steam-water mixture inside the steam generator and convert its steam part into liquid water, which will also cause the water level of the steam generator to decrease. This is exactly the opposite of the expected effect. A decrease in steam flow or an increase in feed water flow means an increase in the mass of water in the steam generator. The expected water level should rise, but in fact the water level drops. This phenomenon of steam generator water level contraction and expansion, which is contrary to expectations, uses the inertia link to reflect its dynamic change process. The inertia link is designed as G2 / (1+τ2s) to simulate the thermal effect of the expansion and contraction of the steam generator water level.

[0070] Composition 3: Oscillation

[0071] The steam of the steam generator is condensed after the steam turbine does work, and enters the steam generator through the water supply pipe to complete the cycle. The water supply inlet is located in the upper middle part of the steam generator, and flows into the bottom along the outer periphery of the tube bundle through the descending channel, and then flows back through the U-shaped tube bundle to take away the heat of the first circuit. The water in this circulation process has a large momentum under relatively small flow rate changes. When the water supply flow rate suddenly decreases, the water in the descending channel still maintains a certain circulation flow rate at the beginning of the change, and then slowly decreases. Therefore, the water level of the steam generator will experience mechanical oscillation, which will disappear within a certain period of time. Therefore, an oscillation link is added to describe the process of this phenomenon. The oscillation link is designed as G3s(1 / τ 2 +4π 2 / T 2 +2s / τ+s 2 ), simulating the water level oscillation effect of the steam generator, where τ is the damping time constant and T is the oscillation period.

[0072] In summary, the structure of the steam generator water level transfer function is determined to be composed of three parts: "integral link" + "inertia link" + "oscillation link".

[0073] According to the above steam generator water level transfer function structure, the transfer function equation of the steam generator water level y(s) after Laplace transformation at time t is analyzed as follows:

[0074] y(s)=(G1 / s-G2 / (1+τ2s))(q e (s)-q v (s))+(G3s(1 / τ 2 +4π 2 / T 2 +2s / τ+s 2 )q e (s)

[0075] Among them, q e (s) is the water flow rate, q v (s) is the steam flow rate. Based on data and theoretical analysis, it can be seen that the coefficient G1 in the transfer function is a constant independent of the load, and the coefficients G2, G3, τ2, T, τ are respectively related to the "reactor power P", "steam flow rate q v (t)", "Primary circuit temperature change Δq p (t)”, “Feedwater temperature change Δq f (t)”, G2 is the gain coefficient of the inertia link, G3 is the gain coefficient of the oscillation link, s is the complex parameter in the Laplace transform, τ is the damping time constant, τ2 is the inertia time constant, and T is the oscillation period.

[0076] The above transfer function equation is converted into the form of state space equation for discretization, and the underlying calculation is carried out without relying on the independent operation of the existing simulation modeling software. The discretized state equation is as follows:

[0077] x1(t)=x1(t-Δt)+Δt G1[q e (t-Δt)–q v (t-Δt)]

[0078] x2(t)=x2(t-Δt)–α2x2(t-Δt)+β2[q e (t-Δt)-q v (t-Δt)]

[0079] x3(t)=x3(t-Δt)–α3x3(t-Δt)+x4(t-Δt)+β3q e (t-Δt)

[0080] x4(t)=x4(t-Δt)–α4x3(t-Δt)

[0081] y(t)=x1(t)+x2(t)+x3(t)

[0082] Where α2 = 1 – e –Δt / τ2 , β2=G2(1–e –Δt / τ2 ), α3=2–2e –Δt / τ cos2πΔt / T,α4=1+e –2Δt / τ –2e –Δt / τ cos(2πΔt / T), β3=(G3T / 2π)e –Δt / τ sin(2πΔt / T). α2, α3, α4, β2, and β3 are parameters formed after the transfer function equation is converted into the state equation and have no specific physical meaning.

[0083] The matrix calculation method is used to quickly solve the above equations.

[0084] In this embodiment, the dynamic parameter model module includes a transfer function dynamic parameter model, and the input data is used as the input signal of the transfer function dynamic parameter model. After calculation by the neural network, the real-time parameter value required for the transfer function calculation is output, and the real-time parameters include coefficients G2, G3, τ2, T, τ.

[0085] See also Figure 3 As shown, it is a schematic diagram of the structure of the dynamic parameter model of the transfer function. The construction process of the dynamic parameter model of the transfer function is as follows:

[0086] First, a dynamic parameter model of the transfer function is designed. According to the nonlinear time-varying characteristics of the real physical world, a neural network algorithm in machine learning is selected to construct a multivariate nonlinear multi-output regression model to realize dynamic fitting of various parameter curves contained in the transfer function, thereby upgrading the staged linearization approximation under the existing technical scheme to a full-state nonlinear approximation.

[0087] Specifically, the neural network algorithm in machine learning is used to construct the transfer function dynamic parameter model, and its input layer parameters are set as power P, steam flow q v (t), primary circuit temperature change Δq p (t) and feed water temperature change Δq f (t), the output layer parameters are set to the parameters to be solved in the transfer function G2, G3, τ2, T, τ". According to existing experience, the number of hidden layers and the number of neurons in each layer (N×M) are initially set to (3×6), and are subsequently adjusted according to experiments until the best effect is obtained.

[0088] Then, a data set for training the dynamic parameter model of the transfer function is obtained; the water level change curve of the steam generator under the condition of a step change at a point in the evaporator feed water flow or the evaporator steam flow under different power states is collected through physical experiments or high-precision mechanism model simulation. The Laplace Transform method is used to solve the parameter values ​​in the transfer function model. A data table is constructed to record the corresponding relationship between various types of data to form a data set for training the dynamic parameter model of the transfer function. The data set is shown in Table 1 below.

[0089] Table 1. Transfer function dynamic parameter model training data set

[0090]

[0091]

[0092] Finally, train the transfer function dynamic parameter model; define the model structure, determine the number of neural network layers, activation functions and the number of neurons in each layer; compile the model, select the loss function, optimization algorithm and evaluation indicators; then train the model, use the above data set to train the model, set the batch size, number of iterations, callback function, learning rate, regularization parameters, etc. and adjust them in time according to the training progress; finally, optimize the model, obtain the model output under a given state and calculate the transfer function model step response curve based on the state and output value, and compare it with the physical model experimental results or high-precision mechanism simulation model results under the same input conditions, evaluate the performance of the multivariate nonlinear multi-output regression model and decide whether to correct the model.

[0093] Specifically, all training data sets obtained through experimental calculations of physical models or high-precision mechanism simulation models are used as training samples to train the above neural network model. After supervised learning is completed, the model is saved as a callable dynamic parameter model. Select a specific input state, substitute the output of the neural network model "G2, G3, τ2, T, τ" into the transfer function equation, and obtain the step response curve of the steam flow or feed water flow. Compare with the results of the physical model experiment or the results of the high-precision mechanism simulation model under the same input conditions to decide whether to continue to optimize the model.

[0094] In this embodiment, the transfer function model module performs real-time calculation using input data and the real-time parameter value output by the dynamic parameter model module. After the calculation is completed, the result output module outputs the water level value of the steam generator.

[0095] Example 2

[0096] This embodiment proposes a data-driven independent modeling method for steam generator water level, which is implemented by the data-driven independent modeling system for steam generator water level proposed in Example 1. Figure 4 As shown, the method specifically includes:

[0097] Read the externally input steam generator related parameter data;

[0098] Specifically, the program is initialized first, and then the data of the steam generator related parameters input from the outside are read. The steam generator related parameters include the feed water flow rate q e (t), steam flow rate q v (t), power P, primary circuit temperature change Δq p (t), feed water temperature change Δq f (t) etc.

[0099] After the data is processed, two transmission paths are performed, and one path of data is directly input into the transfer function model;

[0100] Specifically, after the input data is processed, the data is distributed according to demand and output through two channels, where the water flow rate q e (t) and steam flow rate q v (t) output to the transfer function model for subsequent real-time calculation;

[0101] Input another channel of data into the transfer function dynamic parameter model, and output the real-time parameter value required by the transfer function;

[0102] Specifically, for power P and steam flow q v (t), primary circuit temperature change Δq p (t) and feed water temperature change Δq f(t) After normalization, the input signal is input into the trained transfer function dynamic parameter model, and after calculation by the neural network, the real-time parameter values ​​required for transfer function calculation are output. The real-time parameters include G2, G3, τ2, T, τ.

[0103] The transfer function model performs real-time calculations based on the input data and real-time parameter values, and outputs the water level value of the steam generator.

[0104] Specifically, the transfer function model uses the input data and the real-time parameter value output by the transfer function dynamic parameter model to perform real-time calculations. After the calculation is completed, the water level value of the steam generator is output. The steam generator water level transfer function structure consists of an integral link G1 / s, an inertia link G2 / (1+τ2s) and an oscillation link G3s(1 / τ 2 +4π 2 / T 2 +2s / τ+s 2 ) is composed of three parts. Through the analysis of the above three parts, the transfer function equation of the steam generator water level y(s) after Laplace transformation at time t is obtained as follows:

[0105] y(s)=(G1 / s-G2 / (1+τ2s))(q e (s)-q v (s))+(G3s(1 / τ 2 +4π 2 / T 2 +2s / τ+s 2 )q e (s)

[0106] Among them, q e (s) is the water flow rate, q v (s) is the steam flow rate, G1 is a constant independent of the load, G2 is the gain coefficient of the inertia link, G3 is the gain coefficient of the oscillation link, s is the complex parameter in the Laplace transform, τ is the damping time constant, τ2 is the inertia time constant, and T is the oscillation period.

[0107] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A data-driven independent modeling system for steam generator water level, characterized by: It comprises a data reading module, a data processing module connected to the output end of the data reading module, a dynamic parameter model module and a transfer function model module connected to the output end of the data processing module, and a result output module connected to the output end of the transfer function model module, wherein the output end of the dynamic parameter model module is connected to the input end of the transfer function model module; Wherein, the data reading module is used to read the steam generator related parameter data input externally; The data processing module is used to process data and output data through two paths, one path is output to the dynamic parameter model module, and the other path is output to the transfer function model module; The dynamic parameter model module is used to obtain the real-time parameter values ​​required for transfer function calculation using input data, and output the real-time parameter values ​​to the transfer function model module; The transfer function model module is used to perform real-time calculations based on input data and real-time parameter values; The result output module is used to output the water level value of the steam generator.

2. The data-driven independent modeling system for steam generator water level according to claim 1, characterized in that: The steam generator related parameters include feed water flow, steam flow, power, primary circuit temperature change and feed water temperature change.

3. The data-driven independent modeling system for steam generator water level according to claim 2, characterized in that: The feed water flow and the steam flow are output to the transfer function model module, and the steam flow, the power, the primary circuit temperature change and the feed water temperature change are output to the dynamic parameter model module.

4. The data-driven independent modeling system for steam generator water level according to claim 1, characterized in that: The transfer function structure in the transfer function model module includes an integral link G1 / s, an inertia link G2 / (1+τ2s) and an oscillation link G3s(1 / τ 2 +4π 2 / T 2 +2s / τ+s 2 ); Among them, G1 is a coefficient independent of the load, G2 is the gain coefficient of the inertia link, G3 is the gain coefficient of the oscillation link, s is the complex parameter in the Laplace transform, τ is the damping time constant, τ2 is the inertia time constant, and T is the oscillation period.

5. The data-driven independent modeling system for steam generator water level according to claim 4, characterized in that: The transfer function equation of the steam generator water level in the transfer function model is: y(s)=(G1 / s-G2 / (1+τ2s))(q e (s)-q v (s))+(G3s(1 / τ 2 +4π 2 / T 2 +2s / τ+s 2 )q e (s) Among them, q e (s) is the water flow rate, q v (s) is the steam flow rate.

6. The data-driven independent modeling system for steam generator water level according to claim 5, characterized in that: The state equation after discretization of the transfer function equation of the steam generator water level is: x1(t)=x1(t-Δt)+Δt G1[q e (t-Δt)–q v (t-Δt)] x2(t)=x2(t-Δt)–α2x2(t-Δt)+β2[q e (t-Δt)-q v (t-Δt)] x3(t)=x3(t-Δt)–α3x3(t-Δt)+x4(t-Δt)+β3q e (t-Δt) x4(t)=x4(t-Δt)–α4x3(t-Δt) y(t)=x1(t)+x2(t)+x3(t) where α2 = 1–e –Δt / τ2 , β2 = G2(1–e –Δt / τ2 ), α3 = 2–2e –Δt / τ cos2πΔt / T, α4 = 1+e –2Δt / τ –2e –Δt / τ cos(2πΔt / T), β3 = (G3T / 2π)e –Δt / τ sin(2πΔt / T); The matrix calculation method can be used to quickly solve the state equation.

7. The data-driven independent modeling system for steam generator water level according to claim 3, characterized in that: The construction process of the transfer function dynamic parameter model in the dynamic parameter model module includes: The neural network algorithm in machine learning is used to construct the dynamic parameter model of the transfer function; Obtain training data sets through experimental calculations of physical models or high-precision mechanism simulation models; The transfer function dynamic parameter model is trained using the training data set, and the model is modified according to the training results.

8. The data-driven independent modeling system for steam generator water level according to claim 7, characterized in that: The input layer parameters of the transfer function dynamic parameter model are set to power, steam flow, primary circuit temperature change and feed water temperature change, and the output layer parameters are set to the parameters to be solved G2, G3, τ2, T, τ in the transfer function.

9. A data-driven independent modeling method for steam generator water level, the method being implemented by a data-driven independent modeling system for steam generator water level according to any one of claims 1 to 8, characterized in that: The method comprises: Read the steam generator related parameter data input externally; After the data is processed, two transmission paths are performed, and one path of data is directly input into the transfer function model; Input another channel of data into the transfer function dynamic parameter model, and output the real-time parameter value required by the transfer function; The transfer function model performs real-time calculations based on the input data and real-time parameter values, and outputs the water level value of the steam generator.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, which, when executed by a processor, controls the device where the storage medium is located to execute the steps of the data-driven steam generator water level independent modeling method according to claim 9.

Images