Prediction method and device for generating capacity of coal-fired power plant

Abstract

The invention relates to the technical field of computers, in particular to a method and device for predicting the generating capacity of a coal-fired power plant, and the method comprises the steps: obtaining coal type data in the power plant in a preset time period, carrying out the preprocessing of the coal type data, obtaining the preprocessing data, inputting the preprocessing data into a trained power plant electric quantity prediction model, and carrying out the prediction of the generating capacity of the power plant. And obtaining an electric quantity prediction result. Therefore, the problem of how to accurately calculate the generating capacity of the coal-fired power plant can be solved. The method can better serve related work such as production plan arrangement and power grid access scheduling of the coal-fired power plant, so that resource allocation and coping strategies can be made in advance based on relatively accurate predicted electric quantity.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method and device for predicting power generation of a coal-fired power plant. Background Art

[0002] For the stable operation of the power system, the power forecast of coal-fired power plants is crucial. Accurate forecasting can help the grid dispatching department to make power balance plans in advance and reasonably arrange the start and stop and power adjustment of other conventional energy generation (such as nuclear power generation, hydropower generation, etc.), thereby ensuring the frequency and voltage stability of the power grid. Summary of the invention

[0003] The present invention aims to solve the problem of how to accurately predict the power generation of a coal-fired power plant and provides a method for predicting the power generation of a coal-fired power plant.

[0004] In a first aspect, an embodiment of the present invention provides a method for predicting power generation of a coal-fired power plant, comprising:

[0005] Obtain data on coal types in power plants within a preset time period;

[0006] Preprocessing the coal type data to obtain preprocessed data;

[0007] The preprocessed data is input into the trained power plant power forecasting model to obtain the power forecasting result.

[0008] In a second aspect, an embodiment of the present invention further provides a device for predicting and measuring power generation of a coal-fired power plant, comprising:

[0009] An acquisition module is used to acquire coal type data in a power plant within a preset time period;

[0010] A preprocessing module, used for preprocessing the coal type data to obtain preprocessed data;

[0011] The prediction module is used to input the preprocessed data into the trained power plant power prediction model to obtain the power prediction result.

[0012] In a third aspect, an embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in any embodiment of the present invention is implemented.

[0013] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to execute the method described in any embodiment of the present invention.

[0014] The embodiment of the present invention provides a method for predicting the power generation of a coal-fired power plant, which obtains coal type data in the power plant within a preset time period, preprocesses the coal type data to obtain preprocessed data, and inputs the preprocessed data into a trained power plant power prediction model to obtain a power prediction result. Therefore, the present invention can solve the problem of how to accurately predict the power generation of a coal-fired power plant. It can better serve the production planning arrangement, grid access scheduling and other related work of coal-fired power plants, so that they can make resource allocation and response strategies in advance based on relatively accurate predicted power. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 This is a flow chart of a method for predicting power generation of a coal-fired power plant provided by one embodiment of the present invention;

[0017] Figure 2 It is a structural diagram of a device for predicting power generation of a coal-fired power plant provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions 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 part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0019] Please refer to Figure 1 The embodiment of the present invention provides a method for predicting power generation of a coal-fired power plant, the method comprising:

[0020] Step 100, obtaining coal type data in the power plant within a preset time period;

[0021] Step 102: preprocess the coal type data to obtain preprocessed data;

[0022] Step 104: input the preprocessed data into the trained power plant power forecasting model to obtain a power forecasting result.

[0023] In an embodiment of the present invention, the coal type data in the power plant within a preset time period is obtained, the coal type data is preprocessed to obtain preprocessed data, and the preprocessed data is input into a trained power plant power forecasting model to obtain a power forecast result. Therefore, the present invention can solve the problem of how to accurately calculate the power generation of a coal-fired power plant. It can better serve the production planning arrangement, grid access scheduling and other related work of coal-fired power plants, so that they can make resource allocation and response strategies in advance based on relatively accurate predicted power.

[0024] In one embodiment of the present invention, the coal type data includes temperature data, moisture data, carbon content data and hydrogen content data.

[0025] In the embodiment of the present invention, the coal type data specifically covers multi-dimensional information such as internal temperature parameters of the coal body, humidity index, carbon element mass fraction data, and hydrogen element mass fraction data. These data can provide strong data support for the model to predict electricity.

[0026] In one embodiment of the present invention, a trained power plant power prediction model is obtained according to the following steps:

[0027] Acquire a historical data set; wherein the historical data set includes the power generation of historical coal-fired power plants and historical coal type data; the historical data set includes a training set and a prediction set;

[0028] Taking the power generation of historical coal-fired power plants and historical coal type data as input, the initial power plant power forecasting model is trained with the power generation of historical coal-fired power plants to obtain a trained power plant power forecasting model.

[0029] In one embodiment of the present invention, training the initial power plant power forecasting model includes:

[0030] For each iteration training of the initial power plant power forecasting model update, the training set is input into the initial power plant power forecasting model to obtain the forecast result of the current iteration training;

[0031] Obtain the loss of the current iterative training based on the prediction results and the historical power generation of the coal-fired power plant;

[0032] Back propagation is performed based on the loss to update the parameters of the initial power plant power forecasting model to complete the current iterative training.

[0033] In the embodiment of the present invention, in the initial stage of constructing the power plant power forecasting model, it is necessary to widely collect the power generation related data of the historical coal-fired power plant and the historical coal type data matching it from many different types of data sources, among which the monitoring system database of the historical coal-fired power plant is one of the most critical data sources. The acquired historical data set needs to be properly and reasonably segmented. The training set is mainly responsible for constructing the power plant power forecasting model and optimizing and improving it, while the forecasting set is used to accurately evaluate and effectively verify the performance of the model after all the model training work is completed. When performing the division operation on the training set and the forecasting set, a variety of methods such as random sampling or division according to time sequence can be flexibly selected. For example, when the data presents the arrangement form of a time series, a large proportion of data in the front position can be selected as a training set, such as selecting 70% to 80% of the data volume, and setting the remaining small part of the data as a forecasting set, so that it can effectively simulate the situation of using historical data to train the model and predict future power in actual scenarios.

[0034] The historical coal-fired power plant power generation data and historical coal type data are determined as the key feature elements of the input, and the historical coal-fired power plant power generation (here usually refers to the actual subsequent power generation observation value that corresponds to the power generation data) is set as the target variable, and then the initial power plant power forecasting model is trained. This training process is fundamentally a process of optimizing and adjusting the model parameters, and its purpose is to enable the model to accurately grasp the complex and subtle internal relationship and regular characteristics between power generation and coal type data. For different types of power plant power forecasting models, there are obvious differences in the specific training process. Taking the linear regression model in the traditional machine learning model as an example, its training process mainly focuses on accurately determining the regression coefficient based on the given training data information, using optimization algorithms such as the least squares method, so as to minimize the error function between the predicted value and the actual value. Error functions such as the mean square error (MSE) are commonly used measurement indicators.

[0035] For the long short-term memory network (LSTM), a typical representative of deep learning models, its training process is relatively more complicated. First, a series of preprocessing operations should be carried out on the historical coal data, such as data cleaning to remove noise and outliers, and standardization operations to make the data within a specific numerical range. After completing the preprocessing, the data is input into the LSTM model in strict accordance with the order of the time series. At each time step, the model will carry out a complex calculation process based on the input data information and the current hidden state, and update the cell state and hidden state with the help of a unique gating mechanism. During the entire training process, the backpropagation algorithm (BPTT) is used to calculate the corresponding gradient value, and then the gradient descent algorithm or its various variants (such as Adagrad, Adadelta, Adam, etc.) is used to update and adjust the weight parameters of the model, so as to achieve the goal of minimizing the loss function. If the classification task is involved, the cross entropy loss function will be used, and the mean square error function will be used in the regression task scenario.

[0036] In the process of model training, some hyperparameters must be reasonably set, such as learning rate, number of iterations, and number of hidden layers. The specific selection of these hyperparameters will have a critical impact on the final training results and overall performance of the model, so it usually takes many repeated experiments and careful tuning processes to determine the most suitable hyperparameter combination. For example, a more mature method such as grid search or random search can be used to perform a comprehensive traversal search operation within the pre-given hyperparameter value range, and comprehensively evaluate and analyze the performance of the model on the validation set under different hyperparameter combinations, and finally select the hyperparameter combination with the best performance and apply it to the formal training process of the model. By continuously carrying out iterative training on the training set, repeatedly adjusting and optimizing the model parameters, until the loss function value of the model on the training set shows a convergence trend or reaches the pre-set training stop condition, the trained power plant power forecasting model can be successfully obtained. The model can make a more accurate prediction and analysis of the future power situation based on the input coal data information.

[0037] In one embodiment of the present invention, pre-processing includes one or more of the following processing methods:

[0038] Data cleaning processing;

[0039] Abnormal data elimination processing;

[0040] Data standardization processing;

[0041] Data interpolation and filling processing.

[0042] In the embodiment of the present invention, the operation of data cleaning focuses on implementing a rigorous review and verification process for the collected data, and its purpose is to accurately screen out data that has errors, duplications, incomplete content or does not meet the requirements of the format, so as to improve the overall quality of the data and make the data more suitable for the needs of subsequent analysis, modeling and other related operations. In the data set, there are often some data points that are significantly deviated from the vast majority of data, which are abnormal data. There are many reasons for the generation of abnormal data, which may be caused by errors in the measurement link, system failures or errors in input. Eliminating these abnormal data can effectively avoid their adverse effects on data analysis, model training and other work, thereby ensuring that the final results have higher reliability. Data standardization processing is of great significance. Given that data with different characteristics usually have different dimensions and value ranges, data standardization is to convert these different data into a unified scale according to specific rules, so as to more conveniently carry out data analysis and comparison between different features. Especially in the application scenario of machine learning algorithms, many models have clear requirements for data standardization, and this processing is more critical. Data interpolation and filling processing plays a unique role. When there are missing values ​​in the data set, in order to avoid unnecessary reduction in the amount of data and maintain the integrity of the data as much as possible, the interpolation method can be used to infer the value of the missing position according to the rules presented by the existing data, and then complete the filling work.

[0043] In one embodiment of the present invention, the loss of the current iterative training is calculated by the following formula:

[0044]

[0045] Where K is the loss of the current iteration training, n is the number of samples in the training set, To predict the results, is the electricity generated by historical coal-fired power plants.

[0046] In one embodiment of the present invention, the trained power plant electricity prediction model is a long short-term memory network.

[0047] In the embodiment of the present invention, the Long Short-Term Memory (LSTM) is a recurrent neural network (RNN) architecture in deep learning. The core innovation of LSTM lies in its unique cell state and gate unit structure. The cell state is like an information highway that runs through the entire sequence processing process, which can store and transmit information for a long time, so that the network will not easily lose important information when processing long sequences. Around the cell state, LSTM designs three main gate units: Forget Gate, Input Gate and Output Gate.

[0048] The forget gate determines which information in the cell state at the previous moment should be retained or discarded. By comprehensively considering the input data and the hidden state at the previous moment, it outputs a value between 0 and 1, corresponding to the retention ratio of each dimension of the cell state. The input gate is responsible for controlling which parts of the current input information will be updated to the cell state. It also calculates based on the input data and the hidden state to determine the content and extent of the update. The output gate determines which information in the cell state will be output as the hidden state at the current moment, thereby affecting the calculation at the next moment.

[0049] According to an embodiment of another aspect, the present invention provides a device for predicting power generation of a coal-fired power plant. Figure 2 A schematic block diagram of a device for predicting power generation of a coal-fired power plant according to an embodiment is shown. It can be understood that the device can be implemented by any device, equipment, platform and device cluster with computing and processing capabilities. Figure 2 As shown, the device includes: an acquisition unit 200, a preprocessing module 202 and a prediction module 204. The main functions of each component unit are as follows:

[0050] The acquisition module 200 is used to acquire the coal type data in the power plant within a preset time period;

[0051] A preprocessing module 202 is used to preprocess the coal type data to obtain preprocessed data;

[0052] The prediction module 204 is used to input the pre-processed data into the trained power plant power prediction model to obtain a power prediction result.

[0053] In one embodiment of the present invention, the coal type data includes temperature data, humidity data, carbon content data and hydrogen content data.

[0054] In one embodiment of the present invention, the method further comprises obtaining the trained power plant power prediction model according to the following steps:

[0055] Acquire a historical data set; wherein the historical data set includes the power generation of historical coal-fired power plants and historical coal type data; the historical data set includes a training set and a prediction set;

[0056] The initial power plant electricity forecasting model is trained with the power generation of the historical coal-fired power plant and the historical coal type data as input and the power generation of the historical coal-fired power plant as output to obtain the trained power plant electricity forecasting model.

[0057] In one embodiment of the present invention, the training of the initial power plant power forecasting model includes:

[0058] For each iteration training of the initial power plant power forecasting model update, the training set is input into the initial power plant power forecasting model to obtain the forecast result of the current iteration training;

[0059] Obtaining a loss of current iterative training according to the prediction result and the power generation of the historical coal-fired power plant;

[0060] Back propagation is performed based on the loss to update the parameters of the initial power plant electricity forecasting model to complete the current iterative training.

[0061] In one embodiment of the present invention, the pretreatment includes one or more of the following treatments:

[0062] Data cleaning processing;

[0063] Abnormal data elimination processing;

[0064] Data standardization processing;

[0065] Data interpolation and filling processing.

[0066] In one embodiment of the present invention, the loss of the current iterative training is calculated by the following formula:

[0067]

[0068] Where K is the loss of the current iteration training, n is the number of samples in the training set, is the prediction result, is the power generation of the historical coal-fired power plant.

[0069] In one embodiment of the present invention, the trained power plant electricity prediction model is a long short-term memory network.

[0070] It is to be understood that the structure illustrated in the embodiment of the present invention does not constitute a specific limitation on a device for predicting and measuring the power generation of a coal-fired power plant. In other embodiments of the present invention, a device for predicting and measuring the power generation of a coal-fired power plant may include more or fewer components than shown in the figure, or combine some components, or split some components, or arrange the components differently. The components shown in the figure may be implemented in hardware, software, or a combination of software and hardware.

[0071] The information interaction, execution process and other contents between the modules in the above-mentioned device are based on the same concept as the embodiment of the method of the present invention. For the specific contents, please refer to the description in the embodiment of the method of the present invention, and no further description is given here.

[0072] An embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, a method for predicting power generation of a coal-fired power plant in any embodiment of the present invention is implemented.

[0073] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the processor executes a method for predicting power generation of a coal-fired power plant in any embodiment of the present invention.

[0074] Specifically, a system or device equipped with a storage medium can be provided, on which software program codes that implement the functions of any of the above-mentioned embodiments are stored, and a computer (or CPU or MPU) of the system or device can be enabled to read and execute the program codes stored in the storage medium.

[0075] In this case, the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, and thus the program code and the storage medium storing the program code constitute a part of the present invention.

[0076] The storage medium embodiments for providing the program code include a floppy disk, a hard disk, a magneto-optical disk, an optical disk (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), a magnetic tape, a non-volatile memory card, and a ROM. Alternatively, the program code can be downloaded from a server computer by a communication network.

[0077] In addition, it should be clear that the functions of any of the above embodiments can be implemented not only by executing the program code read by the computer, but also by enabling an operating system operating on the computer to complete part or all of the actual operations based on instructions from the program code.

[0078] In addition, it can be understood that the program code read from the storage medium is written to a memory provided in an expansion board inserted into the computer or to a memory provided in an expansion module connected to the computer, and then based on the instructions of the program code, a CPU installed on the expansion board or expansion module is enabled to perform part or all of the actual operations, thereby realizing the functions of any of the above-mentioned embodiments.

[0079] It should be noted that, in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0080] A person of ordinary skill in the art can understand that all or part of the steps of implementing the above method embodiments can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiments; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various media that can store program codes.

[0081] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for predicting power generation of a coal-fired power plant, characterized in that: include: Obtain data on coal types in power plants within a preset time period; Preprocessing the coal type data to obtain preprocessed data; The preprocessed data is input into the trained power plant power forecasting model to obtain the power forecasting result.

2. The method according to claim 1, characterized in that The coal type data includes temperature data, humidity data, carbon content data and hydrogen content data.

3. The method according to claim 2, characterized in that The method further comprises obtaining the trained power plant power prediction model according to the following steps: Acquire a historical data set; wherein the historical data set includes historical coal-fired power plant power generation and historical coal type data; the historical data set includes a training set and a prediction set; Taking the power generation of the historical coal-fired power plant and the historical coal type data as input and the power generation of the historical coal-fired power plant as output, the initial power plant power forecasting model is trained to obtain the trained power plant power forecasting model.

4. The method according to claim 3, characterized in that The training of the initial power plant power forecasting model includes: For each iteration training of the initial power plant power forecasting model update, the training set is input into the initial power plant power forecasting model to obtain the forecast result of the current iteration training; Obtaining a loss of current iterative training according to the prediction result and the power generation of the historical coal-fired power plant; Back propagation is performed based on the loss to update the parameters of the initial power plant electricity forecasting model to complete the current iterative training.

5. The method according to claim 4, characterized in that The pretreatment includes one or more of the following treatments: Data cleaning processing; Abnormal data elimination processing; Data standardization processing; Data interpolation and filling processing.

6. The method according to claim 5, characterized in that The loss of the current iteration training is calculated by the following formula: Where K is the loss of the current iteration training, n is the number of samples in the training set, is the prediction result, is the power generation of the historical coal-fired power plant.

7. The method according to any one of claims 1 to 6, characterized in that The trained power plant electricity prediction model is a long short-term memory network.

8. A device for predicting power generation of a coal-fired power plant, characterized in that: include: An acquisition module is used to acquire coal type data in a power plant within a preset time period; A preprocessing module, used for preprocessing the coal type data to obtain preprocessed data; The prediction module is used to input the preprocessed data into the trained power plant power prediction model to obtain the power prediction result.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed in a computer, the computer is caused to execute the method according to any one of claims 1 to 7.

Images