Water turbine generator set water guide bearing temperature control method, device, equipment and medium

Abstract

The application discloses a kind of water turbine generator unit water guide bush temperature control method, device, equipment and medium, local feature is extracted from original time series data using convolutional neural network, capture short-term dependence in temperature change, local feature is handled using long short-term memory network, capture long-term dependence in temperature data, while combining channel attention network, adaptively recalibrate the channel weight of local feature, to further improve the feature expression capability of model, for water guide bush temperature prediction model, using data-driven model prediction control and multi-model switching control strategy, design optimization problem, based on gradient descent method Rolling Optimal Control, improve temperature control precision, and then improve the safety of hydropower station.

Description

TECHNICAL FIELD

[0001] The present application relates to temperature control technology, in particular to a water turbine generator set water guide bush temperature control method, device, equipment and medium. BACKGROUND

[0002] In the process of hydropower generation, the water turbine is an important key part. Among them, the water guide bush is one of the key components of the water turbine, and its stability directly affects the performance and service life of the water turbine.

[0003] The water guide bush temperature of the water turbine is one of the important parameters for measuring the running state and health condition of the water turbine. The water guide bush temperature that is too high may cause the lubricating oil to deteriorate, the material to deform, and even the equipment to fail, thereby affecting the normal operation and power generation efficiency of the hydropower station. Therefore, accurately controlling the water guide bush temperature of the water turbine is of great significance to ensure the safe operation of the hydropower station and improve economic benefits.

[0004] The traditional water guide bush temperature control method mainly relies on statistical models and empirical formulas, and these methods often perform poorly when dealing with nonlinear and non-stationary time series data, with limited control accuracy. SUMMARY

[0005] The present application provides a water turbine generator set water guide bush temperature control method, device, equipment and medium to improve temperature control accuracy and improve the safety of the hydropower station.

[0006] In a first aspect, the present application provides a water turbine generator set water guide bush temperature control method, comprising:

[0007] Collecting historical temperature data of the water guide bush of the water turbine generator set;

[0008] Inputting the historical temperature data into a water guide bush temperature prediction model for processing to predict the predicted temperature of the water guide bush at the next control time, wherein the water guide bush temperature prediction model integrates convolutional neural networks, channel attention networks and long short-term memory networks;

[0009] Based on the predicted temperature, the control parameters of the water turbine generator set are optimized using model predictive control.

[0010] Optionally, before inputting the historical temperature data into the water guide bush temperature prediction model for processing, it further comprises:

[0011] Cleaning the historical temperature data to remove abnormal data;

[0012] Resampling the cleaned historical temperature data to a fixed time interval;

[0013] Linearly interpolating the resampled historical temperature data to align the data.

[0014] Optionally, the historical temperature data is input into a water guide bush temperature prediction model for processing to predict a predicted temperature of the water guide bush at a next control time, including:

[0015] The historical temperature data is input into the convolutional neural network for processing to obtain local features of the historical temperature data.

[0016] The local features are input into the channel attention network for processing to obtain a channel attention weight vector, the channel attention weight vector including channel attention weights of each channel of the local features.

[0017] The local features are multiplied by the channel attention weight vector to obtain channel attention features.

[0018] The channel attention features are input into the long short-term memory network for processing to obtain time sequence features of the historical temperature data.

[0019] A predicted temperature of the water guide bush at a next control time is predicted based on the time sequence features.

[0020] Optionally, based on the predicted temperature, a model predictive control is used to optimize control parameters of the hydroelectric generator set, including:

[0021] The unit load and the cooling liquid level are taken as inputs of the model predictive control to calculate a cost function value of the model predictive control.

[0022] It is determined whether the cost function value is greater than a threshold value.

[0023] If yes, the unit load and the cooling liquid level are updated based on the predicted temperature, parameters of the model predictive control are adaptively learned, and the step of taking the unit load and the cooling liquid level as inputs of the model predictive control to calculate the cost function value of the model predictive control is executed until the cost function value is less than or equal to the threshold value or the number of iterations reaches a preset number of times.

[0024] If no, the control parameters of the hydroelectric generator set are output.

[0025] Optionally, the unit load and the cooling liquid level are updated based on the predicted temperature, including:

[0026] A Jacobian matrix of the water guide bush temperature prediction model is calculated based on the predicted temperature.

[0027] The unit load and the cooling liquid level are updated by using a gradient descent method based on the Jacobian matrix.

[0028] Optionally, based on the predicted temperature, the model predictive control is used to optimize the control parameter of the hydro-generator unit into an optimization problem of a target function:

[0029]

[0030] |Δu(t)|≤Δu max ;

[0031] u min ≤u(t)≤u max ;

[0032]

[0033] wherein, is the target function, a and b are weight parameters, the historical input is U p (t)=[u(t),u(t-1),...,u(t-l u )],the historical output is Y p =[y(t),y(t-1),...,y(t-l y )],and R(t)=[r(t+1),r(t+2),...,r(t+T p )] is the reference output, is the predicted output, U(t)=[u(t),u(t+1),...,u(t+T c -1)] is the optimal control input, and ΔU(t)=[Δu(t),Δu(t+1),...,Δu(t+T c -1)] is the incremental control movement.

[0034] Optionally, a Jacobian matrix of the water guide bush temperature prediction model is calculated based on the predicted temperature, comprising:

[0035] The derivative of the output of the water guide bush temperature prediction model is calculated as the Jacobian matrix of the water guide bush temperature prediction model.

[0036] In a second aspect, the present application further provides a water guide bush temperature control device for a hydro-generator unit, comprising:

[0037] a data acquisition module for acquiring historical temperature data of the water guide bush of the hydro-generator unit;

[0038] a temperature prediction module for inputting the historical temperature data into a water guide bush temperature prediction model for processing to predict a predicted temperature of the water guide bush at a next control time, wherein the water guide bush temperature prediction model integrates a convolutional neural network, a channel attention network and a long short-term memory network;

[0039] a parameter optimization module, configured to optimize a control parameter of the hydro-generator unit based on the predicted temperature by using model predictive control.

[0040] In a third aspect, the present application provides an electronic device, comprising:

[0041] one or more processors;

[0042] a storage device configured to store one or more programs;

[0043] When the one or more programs are executed by the one or more processors, the one or more processors implement the hydro-generator unit water guide bush temperature control method provided in the first aspect of the present application.

[0044] In a fourth aspect, the present application provides a computer readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the hydro-generator unit water guide bush temperature control method provided in the first aspect of the present application.

[0045] The hydro-generator unit water guide bush temperature control method provided by the present application extracts local features from original time series data by using a convolutional neural network, captures short-term dependencies in temperature changes, processes local features by using a long short-term memory network, captures long-term dependencies in temperature data, and combines a channel attention network to adaptively recalibrate channel weights of local features, thereby further improving the feature expression capability of the model. For the water guide bush temperature prediction model, a data-driven model predictive control and multi-model switching control strategy are used to design an optimization problem, and a gradient descent method is used for rolling optimal control, thereby improving the temperature control precision and further improving the safety of the hydropower station.

[0046] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0048] Figure 1 A flowchart of a hydro-generator unit water guide bush temperature control method provided by the present application;

[0049] Figure 2A structural schematic diagram of a water guide bush temperature prediction model provided by the application is provided.

[0050] Figure 3 A structural schematic diagram of a convolutional neural network provided by the application is provided.

[0051] Figure 4 A processing process schematic diagram of a channel attention network provided by the application is provided.

[0052] Figure 5 A structural schematic diagram of a long short-term memory network provided by the application is provided.

[0053] Figure 6 A framework structural schematic diagram of a multi-model MPC provided by the application is provided.

[0054] Figure 7 A structural schematic diagram of a water guide bush temperature control device of a hydro-generator unit provided by the application is provided.

[0055] Figure 8 A structural schematic diagram of an electronic device provided by the application is provided.

[0056] Through the above-mentioned drawings, the specific embodiments of the present application have been shown, and will be described in more detail hereinafter. These drawings and textual descriptions are not intended to limit the scope of the concept of the present application by any means, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0057] In order to make the technical personnel in the technical field better understand the present application scheme, the technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the scope of protection of the present application.

[0058] It should be noted that the terms "first", "second" and the like in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not necessarily limit to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0059] Figure 1 A flowchart of a water guide bearing temperature control method for a hydroelectric generating set is provided in the present application. The embodiment can be applicable to processing nonlinear and non-stationary time series data, and improving the accuracy of water guide bearing temperature control. The method can be executed by a water guide bearing temperature control device for a hydroelectric generating set provided by the present application. The device can be realized by software and / or hardware, and is usually configured in an electronic device, such as a computer. Figure 1 The water guide bearing temperature control method includes the following steps:

[0060] S101, collect historical temperature data of the water guide bearing of the hydroelectric generating set.

[0061] In the embodiment of the present application, a certain pumped storage power station is taken as the research object, and the historical temperature data includes the water guide bearing temperature of the hydroelectric generating set from May 2023 to May 2024, a total of 24537. Due to data loss, sensor failure, input error and other reasons, there are certain abnormal data in the historical temperature data, which will lead to unstable model training, inaccurate prediction results, incorrect prediction trend and many other problems, so it is very important to clean the abnormal data. The mean, standard deviation and other statistical quantities are used to identify and process abnormal data, and the data more than twice the mean are considered as abnormal data and cleaned to increase the stability and reliability of the training data.

[0062] Due to different data collection frequencies of the water guide bearing temperature of the hydroelectric generating set, data loss and other reasons, the data at different time points are not synchronized, so data alignment is needed to ensure the consistency of the data in the time dimension, so that the model can effectively use these data for training and prediction. First, resample the data to ensure that the data is uniform to a fixed time interval to ensure the consistency of the data in the time dimension, so that the subsequent interpolation and model training are more accurate and effective. Through observation and analysis of the training data, the present application adopts the downward sampling method to fix the data with shorter and non-uniform time intervals to a sampling time T=1h.

[0063] Considering that the bearing temperature changes relatively smoothly, linear interpolation is adopted for data alignment and resampling. Assuming that there are known data points y1 and y2 at time points t1 and t2, and there is a missing data point y at time t, where t1<t<t2, the linear interpolation formula is:

[0064]

[0065] After the data alignment is completed, in order to improve the accuracy and robustness of the univariate time series prediction, the input data is set to the observation values of the previous five time points of the current time, and the output is the observation value of the next time. The data processing mode is shown in the following two formulas:

[0066] X t =[y t-5 ,y t-4 ,y t-3 ,y t-2 ,y t-1 ]

[0067] Y t =y t

[0068] Wherein, X t is the input data of the network model, Y t is the output of the network model, y t is the bearing temperature data of the next time, [y t-5 , y t-4 , y t-3 , y t-2 , y t-1 ] represents the bearing temperature history data of the previous five time points of the current time.

[0069] S102, input the historical temperature data into the water guide bearing temperature prediction model for processing, to predict the predicted temperature of the water guide bearing at the next control time.

[0070] Figure 2 The structure diagram of a water guide bearing temperature prediction model provided by the present application is shown in FIG. 1. Figure 2 As shown in the embodiment of the present application, the water guide bearing temperature prediction model fuses the convolutional neural network (CNN), the channel attention network (SEA) and the long short-term memory network (LSTM).

[0071] Specifically, as shown in FIG. 2, the water guide bearing temperature prediction model provided by the present application is composed of an input layer, a convolutional neural network (CNN) layer, a channel attention network (SEA) layer and a long short-term memory network (LSTM) layer. Figure 2As shown, the historical temperature data is input into the convolutional neural network for processing to obtain local features of the historical temperature data, the local features are input into the channel attention network (SE-Attention in the figure) for processing to obtain a channel attention weight vector, the channel attention weight vector includes channel attention weights of each channel of the local features, the local features are multiplied by the channel attention weight vector to obtain channel attention features, the channel attention features are input into the long short-term memory network for processing to obtain time series features of the historical temperature data, and the time series features are used to predict the predicted temperature of the water guide bush at the next control time.

[0072] Figure 3 A structural diagram of a convolutional neural network provided by the present application is shown in Figure 1. Figure 3 As shown, the convolutional neural network is a branch widely used in the field of deep learning, which can efficiently extract features from data, has invariance, noise resistance and other advantages. The basic structure of the convolutional neural network includes an input layer, a convolutional layer, a pooling layer, a fully connected layer and an output layer, etc., wherein the convolutional layer is responsible for convolution operation on data and extracts potential features, and the pooling layer is used for downsampling and compression of network parameters. By alternately using the convolutional layer and the pooling layer, the convolutional neural network can maximize the extraction of potential features of the input data, while reducing the error caused by manual feature extraction.

[0073] When processing the water turbine generator bush temperature time series problem, the convolutional neural network has the following advantages. First, the convolutional neural network can enhance the local feature capture capability and effectively extract the information of sudden events in the local time series. Second, the convolutional neural network has high parameter efficiency. Since the convolution kernel remains unchanged in the calculation, it has the characteristics of shared weights, combined with the characteristics of local connection, the parameter amount of the model is reduced, and the training efficiency is higher. In addition, the convolutional neural network has parallel computing capability, which can simultaneously extract features in multiple local regions. The calculation at different time steps is independent and does not need to be performed in sequence, thereby speeding up the training. Finally, the convolution operation of the convolutional neural network can smooth the interference when processing the interference noise or abnormal value in one-dimensional data, while retaining the effective information, thereby enhancing the robustness of the model.

[0074] Figure 4 A processing process diagram of a channel attention network provided by the present application is shown in Figure 2. Figure 4 As shown, the channel attention network is a channel type attention network that introduces an attention mechanism in the channel dimension. Figure 4Different colors represent different weights, so that the importance of each feature channel is different, so that the convolutional neural network can selectively focus on the channel with larger weight. This mechanism effectively solves the problem of feature loss caused by different channel proportions in the convolution and pooling process. The channel attention network mainly consists of three parts: squeezing, excitation and attention. First, through the squeezing operation, the feature is compressed in the spatial dimension, and the local feature with the size of HxWxC is converted into a feature vector with the size of 1x1xC, so as to obtain a global receptive field. The number of channels of this vector matches the number of channels of the input local feature, which represents the global response distribution on the feature channel. This step is realized by global average pooling. Next is the excitation operation, which generates weights for each feature channel by introducing parameters W. W is a learnable multilayer perceptron, which is reduced in dimension to reduce the number of parameters, and the normalized weights are obtained through the Sigmoid function, which explicitly models the correlation between feature channels. Finally, through the attention operation, the weights output by the excitation are regarded as the importance of each feature channel, which are weighted to the original feature (i.e. local feature) through the way of multiplying the channel width, and the re-labeling in the channel dimension is completed.

[0075] Figure 5 A structure diagram of a long short-term memory network provided by the application is shown in Figure 5 The long short-term memory network is an improved model based on a recurrent neural network (RNN), which adopts a three-layer gate design to better solve the problems of RNN gradient disappearance and gradient explosion. The long short-term memory network makes appropriate adjustments to the calculation of the hidden state and adds a cell state C as a memory unit to associate information with a longer time span. It also forgets part of the memory of the previous information, which has superior prediction ability. The gate structure and memory unit of the long short-term memory network can help it effectively process long time series to achieve long-term preservation of historical information of time series. The calculation formula of the long short-term memory network is:

[0076] f t =σ(W fx x t +W fh h t-1 +b f )

[0077] i t =σ(W ix x t +W ih h t-1 +b i )

[0078] g t =tanh(W gx xt +W gh h t-1 +b g )

[0079] O t =σ(W ox x t +W oh h t-1 +b o )

[0080] C t =g t i t +C t-1 f t

[0081] h t =tanh(C t )O t

[0082] wherein W is a weight value; b is a bias; sigma is a Sigmoid function; C t-1 is a cell state of a previous moment; h t-1 is an intermediate state of the previous moment; x t is an input of a current moment; f t is an output of a forget gate; i t and O t are outputs of an input gate and an output gate respectively; g t is an output of an input node; C t is a cell state; h t is an intermediate state.

[0083] The three gate units are a forget gate, an input gate and an output gate respectively, wherein the forget gate decides which past information needs to be retained to avoid useless information continuing to pass back. The forget gate performs weighted calculation on h t-1 and x t , and then performs nonlinear mapping to a value of 0-1 through Sigmoid to obtain f t , which is used for a certain degree of memory and forgetting of the cell state C t-1 of the previous time step, wherein the weight value of W fh is not shared. The function of the input gate is to read data and decide whether to update the data to the memory unit. The Tanh layer generates candidate new information to be input, and the sigma layer generates a value of 0-1 to determine the degree of update of the new information to the cell state. The function of the output gate is to pass the processed data to the next moment, and a value of 0-1 is generated through the Sigmoid function to determine how much the cell state is filtered before output. The cell state is first activated through the Tanh function, and then activated h tMultiplying, the final input to the next time step of the hidden state h t , and lead to h t And after a certain weight to get this step output.

[0084] It can be seen that the activation function of the input-output gate adopts the Tanh function, because the second derivative of Tanh is not equal to zero in a large interval, so it can prevent gradient disappearance. Its convergence speed is usually faster, the derivative is simple, the calculation resources are less, and the monotonicity makes the problem easy to convert into a convex optimization problem. In addition, the core of the long short-term memory network structure is the cell state number. The cell state is like a conveyor belt throughout the entire process, which changes little, while the hidden state changes greatly, so the cell state is mainly used to remember previous information. The update process of the cell is: the cell state from the previous time step is first filtered by the f t Gate to retain the information, and then through the i t Gate to control the new information to be injected, and then C t Output to the next time step, and after activation, it is used for output. In general, the cell state mainly undergoes the filtering and adding processes, so the overall change is not large.

[0085] The long short-term memory network has the following advantages: first, it performs better in modeling and feature extraction of long-term dependency data. Second, the long short-term memory network can adapt to variable-length sequence input and flexibly process data of different lengths. In addition, the long short-term memory network solves the problems of gradient explosion and gradient disappearance to some extent. In the application of hydro-generator bearing temperature, the long short-term memory network can effectively capture and remember the data dependency relationship in a long time span, and is suitable for processing long-term trends and periodic changes in time series data. Convolutional neural networks are good at extracting local features from data and can capture local patterns and trends in input data through convolution operations. Before introducing the convolutional neural network into the long short-term memory network model, the convolutional neural network is used to extract features, and then the long short-term memory network is used for time series modeling, which can improve the prediction accuracy. This method can better capture the complex patterns and trends in the hydro-generator bearing temperature data.

[0086] To obtain the best prediction model parameters, the present application designs a control variable method to gradually optimize the model, and tests the influence of increasing the depth of the model by gradually increasing the number of layers of the long short-term memory network on the prediction effect. First, fix the number of layers of the convolutional neural network, test the influence of the long short-term memory network with different number of layers and neurons on the prediction result, so as to obtain the best long short-term memory network parameters; then fix the long short-term memory network structure, adjust the design of the convolutional neural network to obtain the best parameters.

[0087] S103、based on the predicted temperature, the control parameters of the hydro-generating unit are optimized by model predictive control.

[0088] Model predictive control (MPC) is a model-based control strategy widely used in industrial process control and optimization. Although MPC has been proven to be effective in industrial processes, it still has some limitations. As industrial processes are becoming large-scale, complex and highly coupled, different system parameters or structures can lead to different operating modes and bring multi-mode control problems. There are different operating parameters for each mode, and if the control strategy does not match the corresponding mode, the stability of the system will be reduced, and even serious production accidents and losses will occur. Some improved MPC methods designed to handle multi-mode processes are called multi-model MPC or switching MPC. These methods establish multiple models for each operating mode and design a model switching strategy. Figure 6 A schematic diagram of a multi-model MPC framework provided by the present application is shown in Figure 6 As shown, the switching MPC of multi-mode control is usually composed of three parts: a prediction model group, a controller group and a decision unit. The working process is as follows: first, a series of prediction models need to be established to track the behavior of the controlled system under different operating modes. Then, a switching strategy is designed in the decision unit to distinguish different operating modes. On this basis, the optimal controller can be obtained by solving the problem to ensure the stability of the entire process under each operating mode.

[0089] 1. The unit load and the cooling liquid level are taken as the inputs of the model predictive control, and the cost function value of the model predictive control is calculated. In addition, a reference trajectory, an optimization termination condition threshold and an iteration number are set.

[0090] 2. Determine whether the cost function value is greater than the threshold value.

[0091] 3. If yes, update the unit load and the cooling liquid level based on the predicted temperature, adaptively learn the parameters of the model predictive control, and return to execute the step of taking the unit load and the cooling liquid level as the inputs of the model predictive control to calculate the cost function value of the model predictive control until the cost function value is less than or equal to the threshold value, or the iteration number reaches a preset number.

[0092] 4. If no, output the control parameters of the hydro-generating unit.

[0093] Specifically, the goal of MPC is to find a suitable control signal u(t) through online optimization, so that the output of the system can track the reference trajectory r(t) as closely as possible. In the present application, given the prediction and control horizons T p and T c , the optimal problem can be expressed as an optimization problem of the objective function:

[0094]

[0095] |Δu(t)|≤Δu max

[0096] u min ≤u(t)≤u max

[0097]

[0098] where, is the objective function, a and b are weight parameters, the historical input is U p (t)=[u(t),u(t-1),...,u(t-l u )],the historical output is Y p =[y(t),y(t-1),...,y(t-l y )],and R(t)=[r(t+1),r(t+2),...,r(t+T p )] is the reference output, is the predicted output, U(t)=[u(t),u(t+1),...,u(t+T c -1)] is the optimal control input, and ΔU(t)=[Δu(t),Δu(t+1),...,Δu(t+T c -1)] is the incremental control movement.

[0099] g(·) represents a prediction model based on the SE-CNN-LSTM network, which can be expressed as:

[0100]

[0101] The generator set load and the cooling liquid level are selected as the inputs of the MPC controller, and the reasons are as follows: first, during the operation of the generator, if the load of the generator set is too heavy, it will cause the generator to generate too much heat, thereby causing the bearing bush temperature to be too high, at this time, the load of the generator set needs to be adjusted to reduce the overload phenomenon, thereby reducing the bearing bush temperature. Secondly, the generator needs to be cooled, and if the cooling liquid level is insufficient, it may lead to insufficient cooling of the bearing bush, thereby causing the bearing bush temperature to be too high, at this time, the cooling liquid level needs to be checked, and the cooling liquid needs to be added in time and the cooling system needs to be maintained.

[0102] In the embodiment of the present application, since the proposed prediction method is based on deep learning, its mathematical expression cannot be directly expressed, which brings challenges to solving optimization problems. In order to reduce the number of iterations and quickly find the optimal result, an adaptive gradient descent method is selected to solve the optimization problem:

[0103] Uk+1 (t) = U k (t) + ΔU k (t)

[0104]

[0105] where η1>0 is the learning rate and k is the iteration number. The derivative of the objective function J(t) can be rewritten as:

[0106]

[0107] Thus,

[0108]

[0109] It can be found that the key to handle the optimization problem is to calculate the derivative of the prediction model output, i.e. the Jacobian matrix Because the activation functions used in the water guide bush temperature prediction model are continuous and differentiable Sigmoid function and Tanh function, it is feasible to calculate the Jacobian matrix of the water guide bush temperature prediction model. In addition, numerical differentiation is also used as an alternative method to ensure that each element in the Jacobian matrix is available.

[0110] For the gradient descent method, the constraints can be managed by the projected gradient method, i.e. projecting the optimization variables onto the allowed hyper-space. Since the constraints in the proposed optimization problem are box constraints (constant or linear constraints), we have:

[0111] U k+1 (t) = P1(U k (t) + ΔU k (t))

[0112]

[0113] where is the projection function of an element of the vector u k (t), Δu k (t), which can be defined as

[0114]

[0115] where u max is the upper bound constraint of u k (t), u min is the lower bound constraint of u k (t), and Δu max is the incremental control input Δu kThe bound of (t). When it comes to system state constraints, how to handle control input constraints, add a prediction method to the prediction model to ensure that the constraints are met. In order to speed up the optimization process, we propose an improved adaptive gradient descent method, which uses the natural logarithm to attenuate the learning rate, thereby improving the convergence speed of the method:

[0116]

[0117] Where alpha is the attenuation rate of each iteration, and k is the iteration number.

[0118] By solving the optimization problem, the optimal control input sequence can be obtained, and then the first element of U(t) is applied as the control signal in the system to ensure that the output of the system can accurately track the reference trajectory under different emission modes.

[0119] The water turbine generator set water guide bush temperature control method provided by the application extracts local features from original time series data using a convolutional neural network, captures short-term dependencies in temperature changes, processes local features using a long short-term memory network, captures long-term dependencies in temperature data, and combines channel attention networks to adaptively recalibrate channel weights of local features, thereby further improving the feature expression capability of the model. For the water guide bush temperature prediction model, a data-driven model predictive control and multi-model switching control strategy is used to design an optimization problem, and a gradient descent method is used for rolling optimal control to improve temperature control accuracy and improve the safety of the hydropower station.

[0120] Figure 7 The structure diagram of the water turbine generator set water guide bush temperature control device provided by the application is shown in Figure 7 As shown in the figure, the water turbine generator set water guide bush temperature control device comprises:

[0121] The data acquisition module 201 is configured to acquire historical temperature data of the water guide bush of the water turbine generator set.

[0122] The temperature prediction module 202 is configured to input the historical temperature data into a water guide bush temperature prediction model for processing to predict the predicted temperature of the water guide bush at the next control time, wherein the water guide bush temperature prediction model integrates a convolutional neural network, a channel attention network and a long short-term memory network.

[0123] The parameter optimization module 203 is configured to optimize the control parameters of the water turbine generator set based on the predicted temperature using model predictive control.

[0124] In some embodiments of the application, the water turbine generator set water guide bush temperature control device further comprises:

[0125] a data cleaning module, configured to clean the historical temperature data before inputting the historical temperature data into the water guide bearing temperature prediction model to remove abnormal data;

[0126] a resampling module, configured to resample the cleaned historical temperature data to a fixed time interval;

[0127] a data alignment module, configured to perform linear interpolation on the resampled historical temperature data to perform data alignment.

[0128] In some embodiments of the present application, the temperature prediction module 202 comprises:

[0129] a local feature extraction submodule, configured to input the historical temperature data into the convolutional neural network to obtain local features of the historical temperature data;

[0130] an attention processing submodule, configured to input the local features into the channel attention network to obtain a channel attention weight vector, the channel attention weight vector comprising channel attention weights of each channel of the local features;

[0131] a multiplication submodule, configured to multiply the local features by the channel attention weight vector to obtain channel attention features;

[0132] a time series feature extraction submodule, configured to input the channel attention features into the long short-term memory network to obtain time series features of the historical temperature data;

[0133] a temperature prediction submodule, configured to predict a predicted temperature of the water guide bearing at a next control time based on the time series features.

[0134] In some embodiments of the present application, the parameter optimization module 203 comprises:

[0135] a cost function calculation submodule, configured to calculate a cost function value of the model predictive control by taking the unit load and the cooling liquid level as inputs of the model predictive control;

[0136] a judgment submodule, configured to judge whether the cost function value is greater than a threshold value;

[0137] a return execution submodule, configured to, when the cost function value is greater than the threshold value, update the unit load and the cooling liquid level based on the predicted temperature, adaptively learn parameters of the model predictive control, and return to execute the step of calculating the cost function value of the model predictive control by taking the unit load and the cooling liquid level as inputs of the model predictive control until the cost function value is less than or equal to the threshold value or the number of iterations reaches a preset number of times;

[0138] The parameter output sub-module is configured to output the control parameter of the hydroelectric generating set when the cost function value is less than or equal to a threshold.

[0139] In some embodiments of the present application, the return execution sub-module comprises:

[0140] The Jacobian matrix calculation unit is configured to calculate a Jacobian matrix of the water guide bush temperature prediction model based on the predicted temperature.

[0141] The parameter update unit is configured to update the unit load and the cooling liquid level by using a gradient descent method based on the Jacobian matrix.

[0142] In some embodiments of the present application, the optimization of the control parameter of the hydroelectric generating set based on the predicted temperature by using the model prediction control is converted into an optimization problem of an objective function:

[0143]

[0144] | Δu(t) | ≤ Δu max ;

[0145] u min ≤ u(t) ≤ u max ;

[0146]

[0147] wherein, is the objective function, a and b are weight parameters, the historical input is U p (t) = [u(t), u(t-1),..., u(t-l u )], the historical output is Y p = [y(t), y(t-1),..., y(t-l y )], and R(t) = [r(t+1), r(t+2),..., r(t+T p -1)] is the reference output, is the predicted output, U(t) = [u(t), u(t+1),..., u(t+T c -1)] is the optimal control input, and ΔU(t) = [Δu(t), Δu(t+1),..., Δu(t+T c -1)] is the incremental control movement.

[0148] In some embodiments of the present application, the Jacobian matrix calculation unit comprises:

[0149] The calculation sub-unit is configured to calculate the derivative of the output of the water guide bush temperature prediction model as the Jacobian matrix of the water guide bush temperature prediction model.

[0150] The water turbine generator set water guide bearing temperature control device can execute the water turbine generator set water guide bearing temperature control method provided by the foregoing embodiments of the application, and has corresponding functional modules and beneficial effects of executing the water turbine generator set water guide bearing temperature control method.

[0151] Figure 8 A structure diagram of an electronic device is provided. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular telephones, smart phones, wearable devices (e.g., headsets, glasses, watches, etc.), and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not meant to limit implementations of the applications described and / or claimed in this document.

[0152] As shown in Figure 8 The electronic device includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., connected to the at least one processor 11 in communication, wherein the memory stores a computer program executable by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or loaded into the random access memory (RAM) 13 from the storage unit 18. In the RAM 13, various programs and data required for 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.

[0153] A plurality 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 magnetic 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 telecommunications networks.

[0154] The processor 11 can be various general and / or special purpose processing components having 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 specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 performs various methods and processes described above, such as the hydro-generator set water guide bush temperature control method.

[0155] In some embodiments, the hydro-generator set water guide bush temperature control method can be implemented as a computer program tangibly embodied in a computer readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device via the ROM 12 and / or the communication unit 19. When the computer program is loaded onto the RAM 13 and executed by the processor 11, one or more steps of the hydro-generator set water guide bush temperature control method described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the hydro-generator set water guide bush temperature control method by any other suitable means, such as by means of firmware.

[0156] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (PLD), a computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0157] Computer programs used to implement the methods of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed, implements the functions / acts specified in the flowcharts and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, and partially on a machine or a remote machine or a server.

[0158] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. A computer-readable storage medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of a program of instructions in a transitory signal, a portable computer diskette, 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 disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0159] To provide for interaction with a user, the systems and techniques described here can 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 a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0160] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can 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), blockchain network, and the Internet.

[0161] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. Servers can be cloud servers, also known as cloud computing servers or cloud hosts, which are a host product in the cloud computing service system to solve the defects of large management difficulty and weak business scalability in traditional physical hosts and VPS services.

[0162] The embodiment of the present application also provides a computer program product, comprising a computer program which, when executed by a processor, implements the water turbine generator set water guide bearing temperature control method as provided in any embodiment of the present application.

[0163] The computer program product can be written in one or more programming languages or combinations of languages including object-oriented languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, 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 the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0164] It should be understood that the various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the scope of the present application. For example, the steps recited in the present application can be performed in parallel, in series, or in a different order, and the present application is not limited in this regard.

[0165] The above detailed description does not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements, and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for controlling the temperature of a water turbine generator set water guide bushing, characterized by, The method comprises the following steps: collecting historical temperature data of a water guide bearing bush of a hydroelectric generating set; inputting the historical temperature data into a water guide bearing bush temperature prediction model for processing to predict a predicted temperature of the water guide bearing bush at a next control time, wherein the water guide bearing bush temperature prediction model integrates a convolutional neural network, a channel attention network and a long short-term memory network; optimizing control parameters of the hydroelectric generating set based on the predicted temperature using model predictive control; optimizing control parameters of the hydroelectric generating set based on the predicted temperature using model predictive control, comprising: taking unit load and cooling liquid level as input of model predictive control to calculate a cost function value of the model predictive control; determining whether the cost function value is greater than a threshold value; the threshold value is an optimization termination condition threshold value; if yes, updating the unit load and the cooling liquid level based on the predicted temperature, adaptively learning parameters of the model predictive control, and returning to perform the step of taking unit load and cooling liquid level as input of model predictive control to calculate a cost function value of the model predictive control until the cost function value is less than or equal to the threshold value, or the number of iterations reaches a preset number of times; if no, outputting the control parameters of the hydroelectric generating set.

2. The method of claim 1, wherein the temperature of the water turbine generator unit water guide bushing is controlled by, Before inputting the historical temperature data into the water guide bearing bush temperature prediction model for processing, the method further comprises the following steps: performing data cleaning on the historical temperature data to remove abnormal data; resampling the cleaned historical temperature data to a fixed time interval; performing linear interpolation on the resampled historical temperature data to align the data.

3. The method of claim 1, wherein the temperature of the water guide bushing is controlled by controlling the temperature of the water guide bushing by controlling the temperature of the water guide bushing. Inputting the historical temperature data into the water guide bearing bush temperature prediction model for processing to predict a predicted temperature of the water guide bearing bush at a next control time, comprising: inputting the historical temperature data into the convolutional neural network for processing to obtain local features of the historical temperature data; inputting the local features into the channel attention network for processing to obtain a channel attention weight vector, the channel attention weight vector comprising channel attention weights of each channel of the local features; multiplying the local features by the channel attention weight vector to obtain channel attention features; inputting the channel attention features into the long short-term memory network for processing to obtain time sequence features of the historical temperature data; predicting a predicted temperature of the water guide bearing bush at a next control time based on the time sequence features.

4. The method of claim 1, wherein the temperature of the water turbine generator guide bearing is controlled by, Updating the unit load and the cooling liquid level based on the predicted temperature, comprising: calculating a Jacobian matrix of the water guide bearing bush temperature prediction model based on the predicted temperature; updating the unit load and the cooling liquid level using a gradient descent method based on the Jacobian matrix.

5. The method of claim 4, wherein the temperature of the water turbine generator guide bearing is controlled by, Based on the predicted temperature, optimizing the control parameters of the hydroelectric generating set using model predictive control converts into an optimization problem of an objective function: ； ； ； ； ； wherein, is the objective function, and is a weight parameter, the historical input is , the historical output is , while is the reference output, is the predicted output, is the optimal control input, is the incremental control move; is the predicted horizon, is the control horizon, t is a time point.

6. The method of claim 5, wherein the temperature of the water guide bushing is controlled by controlling the temperature of the water guide bushing by controlling the temperature of the water guide bushing. calculating a Jacobian matrix of the water guide bearing bush temperature prediction model based on the predicted temperature, comprising: calculating a derivative of an output of the water guide bearing bush temperature prediction model as the Jacobian matrix of the water guide bearing bush temperature prediction model.

7. A water turbine generator unit water guide bush temperature control device for executing the water turbine generator unit water guide bush temperature control method according to any one of claims 1 to 6, characterized by The method comprises the following steps: A data collection module is configured to collect historical temperature data of a water guide bearing bush of a hydroelectric generating set. A temperature prediction module is configured to input the historical temperature data into a water guide bearing bush temperature prediction model for processing to predict a predicted temperature of the water guide bearing bush at a next control time, wherein the water guide bearing bush temperature prediction model integrates a convolutional neural network, a channel attention network, and a long short-term memory network. A parameter optimization module is configured to optimize a control parameter of the hydroelectric generating set based on the predicted temperature using model predictive control.

8. An electronic device, comprising: The method comprises: one or more processors; a memory 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 water guide bearing bush temperature control method of any one of claims 1-6.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the water guide bearing bush temperature control method of any one of claims 1-6. The program is executed by the processor to implement the water guide bearing bush temperature control method of any one of claims 1-6.

Images