Pumped storage speed regulator control method and device based on data driving, equipment and medium

Abstract

The invention discloses a pumped storage speed regulator control method and device based on data driving, equipment and a medium. Processing historical operation data by adopting a random forest model, and predicting a generator rotor angle, a relative deviation of a generator rotor rotating speed, a relative deviation of a water turbine torque and a guide vane opening degree deviation at a future moment; an error output state in a prediction time domain and a feedback control increment in a control time domain are used as independent variables to construct a target function of a hydraulic turbine set output error; constructing a cost function of the target function by taking the generator rotor angle, the relative deviation of the generator rotor rotating speed, the relative deviation of the water turbine torque, the guide vane opening deviation and the control voltage of the speed regulator at the future moment as independent variables, and optimizing the target function; the control voltage enabling the cost function to be minimum is calculated to serve as the control voltage of the speed regulator at the future moment, mechanism analysis does not need to be conducted on a speed regulator system, dynamic changes of the system can be reflected in real time, and the future state and performance of the system are accurately predicted.

Description

Technical Field

[0001] The present invention relates to energy storage speed control technology, and in particular to a data-driven pumped storage speed regulator control method, device, equipment and medium. Background Art

[0002] Speed ​​regulators are widely used in various industrial equipment and systems, such as pumped storage power stations, fans, pumps, etc., to adjust the operating speed of the equipment to meet different process requirements.

[0003] Existing speed governor control systems rely on accurate modeling and then optimal control design. However, pumped storage units need to operate under two different operating conditions: water pumps and turbines. This places higher demands on the regulation performance and control methods of the speed governor. Under turbine operating conditions, the speed governor needs to control the frequency, opening or mechanical power of the unit; while under pump operating conditions, the speed governor is mainly responsible for adjusting the guide vane opening. The complex operating conditions of the speed governor system cover a variety of states under normal, abnormal, emergency, maintenance and special application scenarios. In actual applications, corresponding measures need to be taken according to the characteristics of different operating conditions to ensure the stable operation and performance optimization of the speed governor system. It is difficult to establish an accurate mechanism model and to adapt to complex and changing operating conditions and load changes. Summary of the invention

[0004] The present invention provides a data-driven pumped storage speed governor control method, device, equipment and medium, which do not require mechanism analysis of the speed governor system. By analyzing a large amount of historical data and real-time monitoring data, it can reflect the dynamic changes of the system in real time and more accurately predict the future state and performance of the system.

[0005] In a first aspect, the present invention provides a data-driven pumped storage speed regulator control method, comprising:

[0006] Obtain historical operating data of the turbine unit;

[0007] The historical operation data is processed by using a random forest model to predict the relative deviation of the generator rotor angle, the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation at a future moment;

[0008] The objective function of the output error of the turbine unit is constructed with the error output state in the prediction time domain and the feedback control increment in the control time domain as independent variables;

[0009] The cost function of the objective function is constructed by taking the generator rotor angle, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation and the control voltage of the speed regulator at the future time as independent variables;

[0010] The objective function is optimized, and the control voltage that minimizes the cost function is calculated as the control voltage of the speed regulator at a future moment.

[0011] Optionally, before the random forest model is used to process the historical operation data, the method further includes:

[0012] Preprocessing the historical operation data to obtain preprocessed data;

[0013] The preprocessed data is subjected to principal component analysis to determine the principal components of the historical operation data.

[0014] Optionally, the random forest model includes multiple decision trees, and the historical operation data is processed using the random forest model, including:

[0015] Performing multiple sampling with replacement on the principal components of the historical operation data to obtain multiple sampling samples;

[0016] Input a sample into a corresponding decision tree to obtain the regression prediction result of the decision tree;

[0017] The average values ​​of the regression prediction results of multiple decision trees are calculated to obtain the relative deviation of the generator rotor angle, generator rotor speed, turbine torque and guide vane opening deviation at the future moment.

[0018] Optionally, the objective function is:

[0019]

[0020] Among them, X e (k+j|k) is the error output state in the prediction time domain, u e (k+j-1|k) is the feedback control increment in the control domain, and Q and R are weight matrices.

[0021] Optionally, the cost function is:

[0022] F(δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k))

[0023] Among them, δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n(k+j|k) represents the generator rotor angle, relative deviation of generator rotor speed, relative deviation of turbine torque, guide vane opening deviation and control voltage of the governor at control time step k+j.

[0024] Optionally, optimizing the objective function includes:

[0025] Converting the optimization problem of the objective function into a constrained quadratic programming problem;

[0026] Solving the quadratic programming problem to obtain a feedback control input increment sequence in a control time domain;

[0027] The first item in the feedback control input increment sequence is taken as the actual control input increment to act on the system, and the rest are discarded.

[0028] Optionally, optimizing the objective function includes:

[0029] The objective function is optimized by using the original dual neural network, and the control voltage that minimizes the cost function is calculated as the control voltage of the speed regulator at the future moment.

[0030] In a second aspect, the present invention further provides a data-driven pumped storage speed regulator control device, comprising:

[0031] A data acquisition module, used to acquire historical operation data of the water turbine unit;

[0032] A data prediction module, used to process the historical operation data using a random forest model to predict the generator rotor angle, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation at a future moment;

[0033] An objective function building module is used to build an objective function of the output error of the hydro-turbine unit by taking the error output state in the prediction time domain and the feedback control increment in the control time domain as independent variables;

[0034] A cost function construction module is used to construct a cost function of the objective function with the generator rotor angle, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation and the control voltage of the speed regulator at a future time as independent variables;

[0035] The optimization calculation module is used to optimize the objective function and calculate the control voltage that minimizes the cost function as the control voltage of the speed regulator at a future time.

[0036] In a third aspect, the present invention further provides an electronic device, comprising:

[0037] one or more processors;

[0038] A storage device for storing one or more programs;

[0039] When the one or more programs are executed by the one or more processors, the one or more processors implement the data-driven pumped storage speed regulator control method provided in the first aspect of the present invention.

[0040] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the data-driven pumped-storage speed regulator control method provided in the first aspect of the present invention.

[0041] The data-driven pumped storage speed governor control method provided by the present invention uses a random forest model to process historical operation data, predicts the relative deviation of the generator rotor angle, the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation at future times, and constructs an objective function of the turbine unit output error with the error output state in the prediction time domain and the feedback control increment in the control time domain as independent variables, and constructs a cost function of the objective function with the generator rotor angle, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation and the control voltage of the speed governor at future times as independent variables, optimizes the objective function, and calculates the control voltage that minimizes the cost function as the control voltage of the speed governor at future times. It only needs to collect the operation data of the turbine unit and perform data driving, and there is no need to perform mechanism analysis on the speed governor system. By analyzing a large amount of historical data and real-time monitoring data, it can reflect the dynamic changes of the system in real time and more accurately predict the future state and performance of the system.

[0042] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0044] Figure 1 A flow chart of a data-driven pumped storage speed regulator control method provided by the present invention;

[0045] Figure 2 A schematic diagram of the structure of a data-driven pumped storage speed governor control device provided by the present invention;

[0046] Figure 3 The present invention provides a schematic structural diagram of an electronic device.

[0047] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0048] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0049] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0050] Figure 1 This is a flow chart of a data-driven pumped storage speed regulator control method provided by the present invention. This embodiment is applicable to the complex and changeable working conditions and load changes of the pumped storage unit. The method can be executed by the data-driven pumped storage speed regulator control device provided by the embodiment of the present invention. The device can be implemented by software and / or hardware, and is usually configured in an electronic device, such as Figure 1 As shown, the data-driven pumped storage speed governor control method includes the following steps:

[0051] S101. Obtain historical operation data of a hydro turbine unit.

[0052] In the embodiment of the present invention, historical operation data of the water turbine unit is obtained, for example, historical data of unit speed, power, flow, guide vane opening, pressure, etc., which is not limited in the embodiment of the present invention.

[0053] S102. Use a random forest model to process historical operation data to predict the generator rotor angle, relative deviation of the generator rotor speed, relative deviation of the turbine torque and guide vane opening deviation at future times.

[0054] In the embodiment of the present invention, after obtaining the historical operation data of the turbine group, the historical operation data can be preprocessed, including processing data packet loss, abnormal data and other issues. For example, common data anomalies include duplicate values, abnormal values ​​and missing values. For excluding abnormal values, a mathematical method of limiting standard deviation can be used to eliminate them. Suppose there are n variables X in the data set. If the data (represents the matrix X m That is, the data in the i-th row and j-th column of the m-th variable in the data set) satisfies That is to determine the data The abnormal data are eliminated, and 0.08 is a reference constant, which can be adjusted according to actual needs.

[0055] After preprocessing the historical operation data, principal component analysis is performed on the preprocessed data to determine the principal components of the historical operation data.

[0056] For example, for the historical operation data at a certain sampling time, if the historical operation data at the sampling time includes L variables a1, a2, ..., a L , then the data matrix of historical operation data at n sampling moments is:

[0057]

[0058] The data matrix is ​​normalized as follows:

[0059]

[0060] in

[0061] Calculate the correlation coefficient matrix of historical operation data at n sampling moments:

[0062]

[0063] The correlation coefficient is:

[0064]

[0065] Among them, i,j=1,…,L.

[0066] The eigenvalues ​​(l1, l2, ..., l L ) and the corresponding eigenvectors v1,v2,…,v L, so that where U = [v1, v2, …, v L ].

[0067] Calculate the contribution rate of each component. The contribution rate calculation formula is as follows:

[0068]

[0069] The components with contribution rates higher than the threshold are taken as the main components, and the remaining costs are removed. For example, the threshold may be 99.8%.

[0070] In an embodiment of the present invention, a random forest model is used to process historical operation data to predict the relative deviation of the generator rotor angle, the generator rotor speed, the relative deviation of the turbine torque, and the guide vane opening deviation at future times. Random forest is an integrated learning algorithm, a tree classifier composed of multiple decision trees. In an embodiment of the present invention, the principal components of the historical operation data are sampled with replacement multiple times to obtain multiple sampling samples, and a sampling sample is input into a corresponding decision tree to obtain the regression prediction result of the decision tree, and the average value of the regression prediction results of multiple decision trees is calculated to obtain the relative deviation of the generator rotor angle, the generator rotor speed, the relative deviation of the turbine torque, and the guide vane opening deviation at future times.

[0071] For a given input variable X and an independent and identically distributed random variable θ, the random forest predictor can be expressed as h(X,θ). Multiple decision trees are determined according to different random variables θ, and the prediction result is obtained by weighted average output. Assuming that the input (X) and output variable (Y) to be tested are independently distributed, the average generalization error of each predictor is defined as E X,Y [Yh(X)] 2 .

[0072] By predicting the number of trees k [h(θ,X k )] Take the average value and get E when the number of trees k tends to infinity. X,Y [Y-avh k (x,θ k )] 2 →E X,Y [YE θ (x,θ k )] 2 , the average generalization error of each tree is:

[0073] P E * (tree)=E θ E X,Y [Yh(X,θ)] 2

[0074] The generalization error of random forest is expressed as:

[0075] P E * (forest)=E X,Y {E θ [Yh(X,θ)]} 2

[0076] =E θ E' θ E X,Y [Yh(X,θ)][Yh(X,θ')]

[0077] The above formula can be written as E θ E' θ [ρ(θ,θ')d(θ)d(θ')], where For independent θ and θ', the weighted correlation coefficient between the residuals Yh(X,θ) and Yh(X,θ') is defined as:

[0078]

[0079] Then get If for all θ, EY = E X h(X,θ), we have:

[0080]

[0081] S103, constructing an objective function of the output error of the hydro-turbine unit with the error output state in the prediction time domain and the feedback control increment in the control time domain as independent variables.

[0082] In the embodiment of the present invention, data-driven predictive control design is performed, and the control time domain is set to N c and the prediction time domain is N p The control time interval of the whole system is T = nT0, where T0 is the minimum time unit corresponding to the simulation. At time kT, the signal control time set within the future NT time can be expressed as {kT, (k+1)T, ..., (k+N c -1)T}.

[0083] The key control object of the speed regulator system is speed, and the control input is the control voltage of the speed regulator. The control input can be expressed as:

[0084]

[0085] U n (k+i|k) represents the control input for control time step k+i set when the control time step is k.

[0086] Since the prediction time domain may be larger than the control time domain, for i ≥ N c Moment, U n (k+i|k) can be expressed as:

[0087] U n (k+i|k)=U n (k+N c -1|k),N c ≤i≤N p

[0088] The prediction model of the speed regulator system can be expressed as a mathematical formula:

[0089] X n (k+1|k)=f(δ n (k|k),ω n (k|k),m n (k|k),y(k|k),U n (k|k))

[0090] δ n (k|k),ω n (k|k),m n (k|k),y(k|k),U n (k|k) represent the relative deviation of the generator rotor angle, the generator rotor speed, the relative deviation of the turbine torque, the guide vane opening deviation and the control input, respectively.

[0091] The prediction model of the governor system generator rotor angle, relative deviation of generator rotor speed, relative deviation of turbine torque and guide vane opening deviation in the prediction time domain is:

[0092] X n (k+j+1|k)

[0093] =f(δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k))

[0094] Among them, δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k) represents the generator rotor angle, relative deviation of generator rotor speed, relative deviation of turbine torque, guide vane opening deviation and control input of the governor system at control time step k+j respectively. The output of the model is the governor state X at control time step k+j+1. n(k+j+1|k).

[0095] The mathematical formula of the speed regulator optimization strategy can be expressed as:

[0096]

[0097] xT n (k+j+1|k)

[0098] =f(δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k))

[0099] U min ≤U n (k)≤U max

[0100] Among them, α is the weight coefficient, U max and U min represent the upper and lower limits of the control input U(k), respectively, and F(δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k)) is the cost function.

[0101] The cost function can be expressed as an algorithm for optimizing the deviation of the system state quantity and the control quantity. Therefore, when designing speed control, the following objective function is introduced:

[0102]

[0103] Among them, X e (k+j|k) is the error output state in the prediction time domain, u e (k+j-1|k) is the feedback control increment in the control time domain, and Q and R are weight matrices. The first term on the right side of the equation reflects the system's ability to track the target, and the second term reflects the constraint on the control increment.

[0104] S104, constructing a cost function of the objective function with the generator rotor angle at a future time, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation and the control voltage of the speed regulator as independent variables.

[0105] As mentioned above, the cost function is:

[0106] F(δ n (k+j|k),ω n(k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k))

[0107] Among them, δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k) represents the generator rotor angle, relative deviation of generator rotor speed, relative deviation of turbine torque, guide vane opening deviation and control voltage of the governor at control time step k+j.

[0108] In order to prevent the occurrence of infeasible solutions during the optimization process and to avoid excessive increments in the system control quantity, which may cause sudden changes in the control quantity of the controlled system, soft constraints need to be introduced in the above formula. Therefore, the cost function is transformed into the following form:

[0109]

[0110] Among them, N P is the prediction time domain, N c To control the time domain, N needs to be satisfied P ≥1,1≤N c ≤N P , ρ is the weight coefficient, and ε is the relaxation factor.

[0111] S105, optimizing the objective function, and calculating the control voltage that minimizes the cost function as the control voltage of the speed regulator at a future time.

[0112] In the embodiment of the present invention, the above objective function is optimized, and the control voltage that minimizes the cost function is calculated as the control voltage of the speed regulator at the future moment.

[0113] Exemplarily, in some embodiments of the present invention, the optimization problem of the objective function is converted into a constrained quadratic programming problem, the quadratic programming problem is solved to obtain a feedback control input increment sequence in the control time domain, the first item in the feedback control input increment sequence is taken as the actual control input increment to act on the system, and the rest are all discarded.

[0114] Specifically, for model predictive control, the choice of parameters in the cost function has a great impact on the entire control process. Matrices Q and R are often set as diagonal matrices. For matrix Q, the value of its diagonal elements reflects the sensitivity of a state variable error in the controller. The larger the value in the Q matrix, the higher the sensitivity of the corresponding state variable increment. The value of the diagonal elements in the matrix R reflects the energy consumption of the control increment. The larger the value, the greater the energy consumption of the corresponding control increment. The size of the prediction time domain and the control time domain affects the response speed of the system. The larger the value, the longer the response speed.

[0115] In order to solve the optimization problem of the above objective function, a new state vector is constructed for the discrete state equation of the speed regulation error model:

[0116]

[0117] This gives a new state space expression:

[0118]

[0119] in:

[0120]

[0121] At the current time k, the future state ξ(k+i|t), i=1,2,…,N P It can be obtained by iterating the above formula:

[0122]

[0123] The system prediction output expression can be obtained:

[0124] Y=Ψξ(k)+ΘΔU

[0125] in:

[0126]

[0127]

[0128] ΔU=[Δu(k)Δu(k+1)…Δu(k+N c )] T

[0129] Substituting the predicted output expression into the constraint, we can obtain:

[0130] J=(YY ref ) T Q Q (YY ref )+ΔU T R R ΔU

[0131] in, represents the Kronecker inner product, and the objective function can be rewritten as:

[0132] J=[ΔU(t) T ,ε] T H t [ΔU(t) T ,ε]+G t [ΔU(t) T ,ε]

[0133] in:

[0134]

[0135] In this way, the objective function is transformed into a standard quadratic programming problem, and the positive definite matrix H t is the Hessian matrix, G t Used to describe the linear part. At this point, the transformation of the objective function is completed. The design of constraints is discussed below.

[0136] For the control of the speed regulator system, the constraints of the control quantity and the control increment are very important. At the same time, the model predictive control algorithm is very advantageous for solving the control quantity with constraints. Therefore, for the above objective function, the constraints of the control quantity and the control increment also need to be considered. The expression is:

[0137] u min (t+k)≤u(t+k)≤u max (t+k)

[0138] k=0,1,…,N c -1

[0139] Δu min (t+k)≤Δu(t+k)≤Δu max (t+k)

[0140] k=0,1,…,N c -1

[0141] In solving quadratic programming problems, constraints usually appear in the form of multiplication of the variables to be solved and matrices. Therefore, it is necessary to convert the expression of the constraints to facilitate solving the converted quadratic programming problem. The following is the conversion process:

[0142] For the control quantity and control increment, there is the following relationship:

[0143] u(t+k)=u(t+k-1)+Δu(t+k)

[0144] set up:

[0145]

[0146] Such constraints can be written as follows:

[0147] U min ≤A*ΔU t +U t ≤U max

[0148] At this point, the optimization problem of model predictive control is transformed into the solution of the following constrained quadratic programming problem:

[0149] J=[ΔU(t) T ,ε] T H t [ΔU(t) T ,ε]+G t [ΔU(t) T ,ε]

[0150] Subject to

[0151] ΔU min ≤ΔU t ≤ΔU max

[0152] U min ≤AΔU t +U t ≤U max

[0153] In each control cycle, after solving the above quadratic programming problem, a series of feedback control input increments in the control time domain can be obtained as follows:

[0154]

[0155] In each cycle of solving, a series of control input increments can be obtained, but only the first item in the sequence is taken as the actual control input increment to act on the system, and the rest are discarded. After entering the next cycle, the above solving process is repeated, and the cycle is repeated until the speed regulator system reaches the target speed.

[0156] In another embodiment of the present invention, the primal-dual neural network is used to solve the above optimization problem. The process can be summarized as follows: first, the key parameters W, A, c and b are determined; then, based on these parameters, the parameter matrix M and vector p required by the primal-dual neural network can be constructed; then, the vector ξ containing the optimization system control signal x is substituted into the PDNN dynamic model to solve the optimal output signal ξ * ; Since the optimal output signal ξ *The dual decision vector λ is included in * , so from ξ * By extracting the first n elements corresponding to the control input signal x dimension, the optimal control input sequence is obtained.

[0157] The advantage of the method adopted is that it transforms the quadratic programming problem into the solution of a set of first-order differential equations, avoiding the process of solving the inverse of the matrix in the traditional quadratic programming solution method, reducing the algorithm complexity and improving the operation speed.

[0158] The data-driven pumped storage speed governor control method provided by the present invention uses a random forest model to process historical operation data, predicts the relative deviation of the generator rotor angle, the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation at future times, and constructs an objective function of the turbine unit output error with the error output state in the prediction time domain and the feedback control increment in the control time domain as independent variables, and constructs a cost function of the objective function with the generator rotor angle, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation and the control voltage of the speed governor at future times as independent variables, optimizes the objective function, and calculates the control voltage that minimizes the cost function as the control voltage of the speed governor at future times. It only needs to collect the operation data of the turbine unit and perform data driving, and there is no need to perform mechanism analysis on the speed governor system. By analyzing a large amount of historical data and real-time monitoring data, it can reflect the dynamic changes of the system in real time and more accurately predict the future state and performance of the system.

[0159] Figure 2 A schematic diagram of a pumped storage speed governor control device based on data drive provided by the present invention is shown in FIG. Figure 2 As shown, the data-driven pumped storage speed governor control device includes:

[0160] The data acquisition module 201 is used to acquire the historical operation data of the water turbine unit;

[0161] A data prediction module 202 is used to process the historical operation data using a random forest model to predict the relative deviation of the generator rotor angle, the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation at a future time;

[0162] An objective function construction module 203 is used to construct an objective function of the output error of the hydro-turbine unit by taking the error output state in the prediction time domain and the feedback control increment in the control time domain as independent variables;

[0163] A cost function construction module 204 is used to construct a cost function of the objective function with the generator rotor angle, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation and the control voltage of the speed regulator at a future time as independent variables;

[0164] The optimization calculation module 205 is used to optimize the objective function and calculate the control voltage that minimizes the cost function as the control voltage of the speed regulator at a future time.

[0165] In some embodiments of the present invention, the data-driven pumped storage speed regulator control device further includes:

[0166] A data preprocessing module, used for preprocessing the historical operation data before processing the historical operation data using a random forest model to obtain preprocessed data;

[0167] The principal component analysis module is used to perform principal component analysis on the preprocessed data to determine the principal components of the historical operation data.

[0168] In some embodiments of the present invention, the random forest model includes multiple decision trees, and the data prediction module 202 includes:

[0169] A sampling submodule, used for performing multiple replacement sampling on the principal components of the historical operation data to obtain multiple sampling samples;

[0170] The decision tree processing submodule is used to input a sample into a corresponding decision tree to obtain the regression prediction result of the decision tree;

[0171] The average value calculation submodule is used to calculate the average value of the regression prediction results of multiple decision trees to obtain the relative deviation of the generator rotor angle, generator rotor speed, turbine torque and guide vane opening deviation at the future moment.

[0172] In some embodiments of the present invention, the objective function is:

[0173]

[0174] Among them, X e (k+j|k) is the error output state in the prediction time domain, u e (k+j-1|k) is the feedback control increment in the control domain, and Q and R are weight matrices.

[0175] In some embodiments of the present invention, the cost function is:

[0176] F(δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k))

[0177] Among them, δ n(k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k) represents the generator rotor angle, relative deviation of generator rotor speed, relative deviation of turbine torque, guide vane opening deviation and control voltage of the governor at control time step k+j.

[0178] In some embodiments of the present invention, the optimization calculation module 205 includes:

[0179] A problem conversion submodule, used for converting the optimization problem of the objective function into a constrained quadratic programming problem;

[0180] A solving submodule, used for solving the quadratic programming problem to obtain a feedback control input increment sequence in a control time domain;

[0181] The control input increment determination submodule is used to take the first item in the feedback control input increment sequence as the actual control input increment to act on the system, and discard all the rest.

[0182] In some embodiments of the present invention, the optimization calculation module 205 includes:

[0183] The optimization submodule is used to optimize the objective function by using the original dual neural network, and calculate the control voltage that minimizes the cost function as the control voltage of the speed regulator at a future time.

[0184] The above-mentioned data-driven pumped-storage speed regulator control device can execute the data-driven pumped-storage speed regulator control method provided in the aforementioned embodiment of the present invention, and has the corresponding functional modules and beneficial effects for executing the data-driven pumped-storage speed regulator control method.

[0185] Figure 3 A schematic diagram of the structure of an electronic device provided by the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0186] like Figure 3As shown, the electronic device includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0187] A number of components in the electronic device are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0188] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a data-driven pumped storage speed governor control method.

[0189] In some embodiments, the data-driven pumped-storage speed regulator control method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on an electronic device via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the data-driven pumped-storage speed regulator control method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform a data-driven pumped-storage speed regulator control method in any other appropriate manner (e.g., by means of firmware).

[0190] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0191] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0192] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0193] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0194] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0195] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0196] An embodiment of the present invention further provides a computer program product, including a computer program, which, when executed by a processor, implements a data-driven pumped storage speed regulator control method as provided in any embodiment of the present application.

[0197] In the process of implementation, the computer program product can be written in one or more programming languages ​​or a combination thereof to perform the computer program code of the present invention, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, using an Internet service provider to connect through the Internet).

[0198] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0199] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A data-driven pumped storage speed governor control method, characterized in that: include: Obtain historical operating data of the turbine unit; The historical operation data is processed by using a random forest model to predict the relative deviation of the generator rotor angle, the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation at a future moment; The objective function of the output error of the turbine unit is constructed with the error output state in the prediction time domain and the feedback control increment in the control time domain as independent variables; The cost function of the objective function is constructed by taking the generator rotor angle, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation and the control voltage of the speed regulator at the future time as independent variables; The objective function is optimized, and the control voltage that minimizes the cost function is calculated as the control voltage of the speed regulator at a future moment.

2. The data-driven pumped storage speed governor control method according to claim 1 is characterized in that: Before the random forest model is used to process the historical operation data, the method further includes: Preprocessing the historical operation data to obtain preprocessed data; The preprocessed data is subjected to principal component analysis to determine the principal components of the historical operation data.

3. The data-driven pumped storage speed governor control method according to claim 2 is characterized in that: The random forest model includes multiple decision trees. The random forest model is used to process the historical operation data, including: Performing multiple sampling with replacement on the principal components of the historical operation data to obtain multiple sampling samples; Input a sample into a corresponding decision tree to obtain the regression prediction result of the decision tree; The average values ​​of the regression prediction results of multiple decision trees are calculated to obtain the relative deviation of the generator rotor angle, generator rotor speed, turbine torque and guide vane opening deviation at the future moment.

4. The data-driven pumped storage speed governor control method according to claim 1 is characterized in that: The objective function is: Among them, X e (k+j|k) is the error output state in the prediction time domain, u e (k+j-1|k) is the feedback control increment in the control domain, and Q and R are weight matrices.

5. The data-driven pumped storage speed governor control method according to claim 1, characterized in that: The cost function is: F(δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k)) Among them, δ n (k+j|k),ω n (k+j|k),m n (k+j|k),y(k+j|k),U n (k+j|k) represents the generator rotor angle, relative deviation of generator rotor speed, relative deviation of turbine torque, guide vane opening deviation and control voltage of the governor at control time step k+j.

6. The data-driven pumped storage speed governor control method according to claim 1 is characterized in that: Optimizing the objective function includes: Converting the optimization problem of the objective function into a constrained quadratic programming problem; Solving the quadratic programming problem to obtain a feedback control input increment sequence in a control time domain; The first item in the feedback control input increment sequence is taken as the actual control input increment to act on the system, and the rest are discarded.

7. The data-driven pumped storage speed governor control method according to claim 1 is characterized in that: Optimizing the objective function includes: The objective function is optimized by using the original dual neural network, and the control voltage that minimizes the cost function is calculated as the control voltage of the speed regulator at the future moment.

8. A data-driven pumped storage speed governor control device, characterized in that: include: A data acquisition module, used to acquire historical operation data of the water turbine unit; A data prediction module, used to process the historical operation data using a random forest model to predict the generator rotor angle, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation at a future moment; An objective function building module is used to build an objective function of the output error of the hydro-turbine unit by taking the error output state in the prediction time domain and the feedback control increment in the control time domain as independent variables; A cost function construction module is used to construct a cost function of the objective function with the generator rotor angle, the relative deviation of the generator rotor speed, the relative deviation of the turbine torque and the guide vane opening deviation and the control voltage of the speed regulator at a future moment as independent variables; The optimization calculation module is used to optimize the objective function and calculate the control voltage that minimizes the cost function as the control voltage of the speed regulator at a future moment.

9. An electronic device, characterized in that: include: one or more processors; A storage device for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the data-driven pumped storage speed regulator control method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, a data-driven pumped storage speed regulator control method as described in any one of claims 1 to 7 is implemented.

Images