Power generation prediction method, device and equipment for thermal power plant

Abstract

The invention discloses a thermal power plant power generation prediction method, device and equipment, and belongs to the technical field of power generation prediction.The method comprises the steps that historical power utilization characteristic data of a target power supply area corresponding to a thermal power plant to be predicted is collected and cleaned, and then the cleaned historical power utilization characteristic data is obtained; on the basis of the historical power consumption characteristic data of the target power supply area, constructing a sample power consumption characteristic vector corresponding to the target power supply area and daily power consumption corresponding to the sample power consumption characteristic vector; and a neural network is adopted to learn the sample electricity consumption feature vector and the daily electricity consumption corresponding to the sample electricity consumption feature vector to obtain a daily electricity consumption prediction model, so that the daily electricity consumption prediction model can be used for electricity consumption demand prediction, prediction is performed from the demand side, the finally obtained result is more accurate, and the user experience is improved. The problems that in the prior art, power generation prediction precision of a thermal power plant is low, and the requirement for professional knowledge is high are effectively solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power generation prediction, and in particular relates to a method, device and equipment for predicting power generation in a thermal power plant. Background Art

[0002] A thermal power plant, also known as a coal-fired power plant, is a plant that uses the heat energy generated by fuel combustion to heat water into steam, drive the steam turbine to rotate, and then drive the generator to generate electricity. As an important source of electricity supply, the prediction of the power generation of a thermal power plant is of great significance for the dispatch of the power system. Traditional thermal power plant power generation prediction methods mainly rely on empirical formulas and the professional knowledge of staff, and have problems such as low prediction accuracy and poor real-time performance. Therefore, it is of great significance to study a high-precision and real-time thermal power plant power generation prediction method. Summary of the invention

[0003] The present invention provides a method for predicting power generation in a thermal power plant, which is used to solve the problem that the traditional technology relies on empirical formulas and professional knowledge of staff to predict power generation, resulting in low prediction accuracy and poor real-time performance.

[0004] In a first aspect, the present invention provides a method for predicting power generation in a thermal power plant, comprising:

[0005] For the thermal power plant to be predicted, historical power consumption characteristic data of the target power supply area corresponding to the thermal power plant to be predicted is collected, and the historical power consumption characteristic data is cleaned to obtain the cleaned historical power consumption characteristic data;

[0006] Based on the historical power consumption characteristic data of the target power supply area, construct a sample power consumption characteristic vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption characteristic vector;

[0007] A neural network is used to learn the sample electricity consumption feature vector and the daily electricity consumption corresponding to the sample electricity consumption feature vector to obtain a daily electricity consumption prediction model;

[0008] Collecting the real-time power consumption feature vector corresponding to the target power supply area, and using the daily power consumption prediction model to analyze the real-time power consumption feature vector to determine the predicted daily power consumption;

[0009] The line loss electricity from the thermal power plant to be predicted to the target power supply area is determined, and the predicted power generation corresponding to the thermal power plant to be predicted is determined based on the line loss electricity and the predicted daily electricity consumption.

[0010] Furthermore, historical power consumption characteristic data of the target power supply area corresponding to the thermal power plant to be predicted is collected, including:

[0011] Collect the daily average temperature, daily average humidity, rainfall, season type, date type and daily power consumption of the target power supply area corresponding to the thermal power plant to be predicted, and obtain the historical power consumption characteristic data of the target power supply area;

[0012] The date types include working days, holidays, and non-working days that are not holidays.

[0013] Furthermore, after cleaning the historical power consumption characteristic data, the cleaned historical power consumption characteristic data is obtained, including: filtering out the historical power consumption characteristic data with missing values ​​and / or abnormal values ​​to obtain the cleaned historical power consumption characteristic data.

[0014] Furthermore, based on the historical power consumption characteristic data of the target power supply area, a sample power consumption characteristic vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption characteristic vector are constructed, including:

[0015] For any historical electricity consumption characteristic data of the target power supply area, the daily average temperature, daily average humidity, rainfall, season type, date type and daily electricity consumption in the previous N days in the historical electricity consumption characteristic data are used as the sample electricity consumption feature vector corresponding to the target power supply area, and the daily electricity consumption in the historical electricity consumption characteristic data is used as the daily electricity consumption corresponding to the sample electricity consumption feature vector.

[0016] Furthermore, a neural network is used to learn the sample power consumption feature vector and the daily power consumption corresponding to the sample power consumption feature vector to obtain a daily power consumption prediction model, including:

[0017] Constructing a BP neural network and an error function value corresponding to the BP neural network, and randomly initializing network parameters of the BP neural network;

[0018] Using the sample electricity consumption feature vector as input data of the BP neural network, using the daily electricity consumption corresponding to the sample electricity consumption feature vector as the expected label value of the BP neural network, and obtaining the error function value of the BP neural network;

[0019] According to the error function value of the BP neural network, the network parameters of the BP neural network are updated to obtain the BP neural network after the network parameters are updated;

[0020] For the BP neural network after the network parameters are updated, the error function value of the BP neural network is re-obtained;

[0021] It is determined whether the re-acquired error function value is less than a preset error threshold. If so, the training is completed and the current BP neural network is used as a daily power consumption prediction model. Otherwise, the step of updating the network parameters of the BP neural network is returned.

[0022] Furthermore, the error function value corresponding to the construction of the BP neural network is:

[0023]

[0024] Among them, RMSE represents the error function value, y i It represents the actual output obtained by taking the electric feature vector of the i-th sample as input; d i represents the expected label value corresponding to the i-th sample electricity consumption feature vector, and n represents the total number of input sample electricity consumption feature vectors.

[0025] Furthermore, the network parameters of the BP neural network are updated, including:

[0026]

[0027] Among them, ω t represents the network parameters in the tth training process, ω t-1 represents the network parameters in the t-1th training process, η t represents the learning rate during the t-th training process, ξ t represents the first update control factor in the t-th training process, λ t represents the second update control factor in the t-th training process, ε represents the smoothing factor, α1 represents the first exponential decay parameter, α2 represents the second exponential decay parameter, ξ t-1 represents the first update control factor in the t-1th training process, represents the gradient of the error function value relative to the weight parameter during the t-1th training process, λ t-1 Represents the second update control factor in the t-1th training process.

[0028] Furthermore, the learning rate in the t-th training process is:

[0029]

[0030] Among them, η t-1 represents the learning rate during the t-1th training process, ΔRMSE = RMSE t -RMSE t-1 , RMSE t represents the error function value during the t-th training process, E t-1 represents the error function value in the t-1th training process, γ1 represents the first constant term between (0,1), γ2 represents the second constant term between (0,1), and γ2 is less than γ1.

[0031] In a second aspect, an embodiment of the present invention provides a thermal power plant power generation prediction device, comprising: a data acquisition module, a sample construction module, a data learning module, a power consumption prediction module, and a power generation prediction module;

[0032] The data collection module is used to collect historical power consumption characteristic data of the target power supply area corresponding to the thermal power plant to be predicted, and clean the historical power consumption characteristic data to obtain the cleaned historical power consumption characteristic data;

[0033] The sample construction module is used to construct a sample power consumption feature vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption feature vector based on the historical power consumption feature data of the target power supply area;

[0034] The data learning module is used to use a neural network to learn the sample power consumption feature vector and the daily power consumption corresponding to the sample power consumption feature vector to obtain a daily power consumption prediction model;

[0035] The power consumption prediction module is used to collect the real-time power consumption feature vector corresponding to the target power supply area, and use the daily power consumption prediction model to analyze the real-time power consumption feature vector to determine the predicted daily power consumption;

[0036] The power generation prediction module is used to determine the line loss power from the thermal power plant to be predicted to the target power supply area, and determine the predicted power generation corresponding to the thermal power plant to be predicted based on the line loss power and the predicted daily power consumption.

[0037] In a third aspect, the present invention provides a thermal power plant power generation prediction device, including a processor and a memory;

[0038] The memory stores computer-executable instructions;

[0039] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the method for predicting power generation in a thermal power plant as described in the first aspect.

[0040] The present invention provides a method, device and equipment for predicting power generation in a thermal power plant. The method collects historical power consumption characteristic data of a target power supply area corresponding to a thermal power plant to be predicted, cleans the historical power consumption characteristic data, and obtains the cleaned historical power consumption characteristic data. Based on the historical power consumption characteristic data of the target power supply area, a sample power consumption characteristic vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption characteristic vector are constructed. A neural network is used to learn the sample power consumption characteristic vector and the daily power consumption corresponding to the sample power consumption characteristic vector to obtain a daily power consumption prediction model, so that the daily power consumption prediction model can be used to predict power consumption demand, and prediction is performed from the demand side, so that the final result is more accurate, which effectively solves the problems of low power generation prediction accuracy and high professional knowledge requirements of thermal power plants in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0042] Figure 1 A schematic flow chart of a method for predicting power generation in a thermal power plant provided in an embodiment of the present invention.

[0043] Figure 2 A schematic structural diagram of a thermal power plant power generation prediction device provided in an embodiment of the present invention.

[0044] Figure 3 A schematic structural diagram of a thermal power plant power generation prediction device provided in an embodiment of the present invention.

[0045] Among them, 201-data acquisition module, 202-sample construction module, 203-data learning module, 204-power consumption prediction module, 205-power generation prediction module, 301-memory, 302-processor, 303-bus.

[0046] The above drawings have shown clear embodiments of the present invention, which will be described in more detail below. These drawings and text descriptions are not intended to limit the scope of the present invention in any way, but to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0047] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of devices and methods consistent with some aspects of the present invention as detailed in the appended claims.

[0048] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0049] like Figure 1 As shown, an embodiment of the present invention provides a method for predicting power generation in a thermal power plant, comprising:

[0050] S101, for the thermal power plant to be predicted, collecting historical power consumption characteristic data of the target power supply area corresponding to the thermal power plant to be predicted, and cleaning the historical power consumption characteristic data to obtain the cleaned historical power consumption characteristic data;

[0051] In the prediction of power generation of thermal power plants, the historical power consumption characteristic data of the target power supply area is an important basis for accurate prediction. The following are the steps to collect and process the historical power consumption characteristic data of the target power supply area for the thermal power plant to be predicted: Determine the source of historical power consumption data of the target power supply area, which usually includes the database of the power company, the smart meter system, the distribution network monitoring system, etc. The collected data should include but are not limited to the following types: Power load data: including power consumption at different time scales (such as hourly, daily, and monthly). Power consumption period characteristics: such as power consumption during peak hours and valley hours. Weather data: temperature, humidity, wind speed, etc. in the target area, because they affect power consumption. Holiday information: because holidays may affect power consumption patterns. However, it is worth noting that these data types are only used as examples, and other types of data can also be included, and the collected data should cover a sufficiently long time span to ensure the representativeness of the data and the accuracy of trend analysis.

[0052] Data cleaning can include: Missing value processing: Check the missing values ​​in the data set and choose an appropriate method to handle them, such as interpolation, using the average of the data before and after, etc. Outlier detection: Identify outliers in the data, which may be caused by data recording errors, equipment failures, or other factors. Outliers can be detected by statistical methods (such as box plots) or machine learning methods. Data verification: Ensure the consistency and accuracy of the data, such as checking the continuity and rationality of time series data. Data normalization: Normalize or standardize the data to eliminate the impact of dimensions and facilitate subsequent analysis. Duplicate data processing: Delete duplicate data records. Data conversion: Convert non-numeric data (such as holiday information) into numerical data that can be used for model analysis. Thus, the data can be converted into learnable data.

[0053] S102: constructing a sample power consumption feature vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption feature vector based on the historical power consumption feature data of the target power supply area;

[0054] Optionally, historical electricity consumption feature data can be analyzed to determine features that are highly correlated with daily electricity consumption, such as temperature, humidity, season, weekdays / weekends, and historical electricity consumption over the same period. Statistical methods (such as Pearson correlation coefficient, Spearman rank correlation coefficient) or machine learning models (such as random forest, gradient boosting machine) can also be used to evaluate the importance of features. In this way, more important data can be determined from multiple types of data to reduce the complexity of data processing.

[0055] The factors affecting electricity consumption on a certain day and the electricity consumption of the previous N days can be used as sample electricity consumption feature vectors, and the daily electricity consumption corresponding to the certain day can be used as the label value, so that big data learning can be realized, and thus data prediction can be achieved.

[0056] S103, using a neural network to learn the sample electricity consumption feature vector and the daily electricity consumption corresponding to the sample electricity consumption feature vector to obtain a daily electricity consumption prediction model;

[0057] A convolutional neural network or a BP (Back Propagation) neural network can be selected to learn the sample electricity consumption feature vector and the daily electricity consumption corresponding to the sample electricity consumption feature vector, so as to obtain a daily electricity consumption prediction model with data prediction capability. The BP neural network is a multi-layer feedforward neural network trained by the error back propagation algorithm. It is one of the most widely used artificial neural networks, especially in the fields of function approximation, pattern recognition, classification, and prediction. The BP neural network has the advantages of simple structure and high prediction accuracy. Therefore, the embodiment of the present invention preferably uses the BP neural network to perform data relationship.

[0058] Optionally, intelligent optimization algorithms such as gradient descent method, particle swarm algorithm, etc. may be used to update the network parameters of the BP neural network, thereby obtaining a daily electricity consumption prediction model.

[0059] It is worth noting that after obtaining the daily electricity consumption prediction model, the daily electricity consumption prediction model can be updated according to subsequent data to make the prediction more and more accurate.

[0060] S104, collecting the real-time power consumption feature vector corresponding to the target power supply area, and using the daily power consumption prediction model to analyze the real-time power consumption feature vector to determine the predicted daily power consumption;

[0061] It is worth noting that the data structure of the real-time power consumption feature vector should be the same as that of the sample power consumption feature vector to ensure that the daily power consumption prediction model can effectively identify the data.

[0062] After obtaining the predicted daily electricity consumption, the demand of the target power supply area is determined, and the power generation can be predicted based on this.

[0063] S105, determining the line loss from the thermal power plant to be predicted to the target power supply area, and determining the predicted power generation corresponding to the thermal power plant to be predicted based on the line loss and the predicted daily power consumption.

[0064] For example: based on the unit line loss, the line loss in the entire transportation process can be determined, and finally the line loss electricity can be determined. The sum of the line loss electricity and the predicted daily electricity consumption can be considered as the minimum power generation of the thermal power plant. At the same time, electricity consumption generally fluctuates, so a margin can also be set, and the sum of the margin, line loss electricity and predicted daily electricity consumption can be used as the predicted power generation corresponding to the thermal power plant to be predicted.

[0065] A method for predicting power generation in a thermal power plant provided in an embodiment of the present invention collects historical power consumption characteristic data of a target power supply area corresponding to a thermal power plant to be predicted, and cleans the historical power consumption characteristic data to obtain the cleaned historical power consumption characteristic data; based on the historical power consumption characteristic data of the target power supply area, a sample power consumption characteristic vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption characteristic vector are constructed; a neural network is used to learn the sample power consumption characteristic vector and the daily power consumption corresponding to the sample power consumption characteristic vector to obtain a daily power consumption prediction model, so that the daily power consumption prediction model can be used to predict power consumption demand, and prediction is performed from the demand side, so that the final result is more accurate, which effectively solves the problems of low power generation prediction accuracy and high professional knowledge requirements of thermal power plants in the prior art, and improves the real-time performance of power generation prediction.

[0066] In an embodiment of the present invention, the historical power consumption characteristic data of the target power supply area corresponding to the thermal power plant to be predicted is collected, including:

[0067] Collect the daily average temperature, daily average humidity, rainfall, season type, date type and daily power consumption of the target power supply area corresponding to the thermal power plant to be predicted, and obtain the historical power consumption characteristic data of the target power supply area;

[0068] The date types include working days, holidays, and non-working days other than holidays, and the season types may be spring, summer, autumn, and winter.

[0069] In the subsequent forecasting process, the daily average temperature, daily average humidity and rainfall can all be obtained through weather forecasts to achieve prediction.

[0070] In an embodiment of the present invention, after cleaning the historical power consumption characteristic data, the cleaned historical power consumption characteristic data is obtained, including: filtering out the historical power consumption characteristic data with missing values ​​and / or abnormal values ​​to obtain the cleaned historical power consumption characteristic data.

[0071] It is worth noting that the above-mentioned cleaning operation of the historical electricity consumption characteristic data is only a preferred method of the embodiment of the present invention, and other operations can also be used to clean the data.

[0072] In an embodiment of the present invention, based on the historical power consumption characteristic data of the target power supply area, a sample power consumption characteristic vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption characteristic vector are constructed, including:

[0073] For any historical electricity consumption characteristic data of the target power supply area, the daily average temperature, daily average humidity, rainfall, season type, date type and daily electricity consumption in the previous N days in the historical electricity consumption characteristic data are used as the sample electricity consumption feature vector corresponding to the target power supply area, and the daily electricity consumption in the historical electricity consumption characteristic data is used as the daily electricity consumption corresponding to the sample electricity consumption feature vector.

[0074] In the embodiment of the present invention, a neural network is used to learn the sample power consumption feature vector and the daily power consumption corresponding to the sample power consumption feature vector to obtain a daily power consumption prediction model, including:

[0075] Constructing a BP neural network and an error function value corresponding to the BP neural network, and randomly initializing network parameters of the BP neural network;

[0076] The network parameters of the BP neural network can be its weight parameters and / or threshold parameters. By training these parameters, the BP neural network can be applied to a certain data recognition task.

[0077] Using the sample electricity consumption feature vector as input data of the BP neural network, using the daily electricity consumption corresponding to the sample electricity consumption feature vector as the expected label value of the BP neural network, and obtaining the error function value of the BP neural network;

[0078] According to the error function value of the BP neural network, the network parameters of the BP neural network are updated to obtain the BP neural network after the network parameters are updated;

[0079] For the BP neural network after the network parameters are updated, the error function value of the BP neural network is re-obtained;

[0080] It is determined whether the re-acquired error function value is less than a preset error threshold. If so, the training is completed and the current BP neural network is used as a daily power consumption prediction model. Otherwise, the step of updating the network parameters of the BP neural network is returned.

[0081] In the embodiment of the present invention, the error function value corresponding to the construction of the BP neural network is:

[0082]

[0083] Among them, RMSE represents the error function value, y i It represents the actual output obtained by taking the electric feature vector of the i-th sample as input; d i represents the expected label value corresponding to the i-th sample electricity consumption feature vector, and n represents the total number of input sample electricity consumption feature vectors.

[0084] The error function value provided by the embodiment of the present invention can effectively evaluate the training degree of the BP neural network, so as to determine whether the training can be terminated based on it.

[0085] In the embodiment of the present invention, updating the network parameters of the BP neural network includes:

[0086]

[0087]

[0088] Among them, ω t represents the network parameters in the tth training process, ω t-1 represents the network parameters in the t-1th training process, η t represents the learning rate during the t-th training process, ξ t represents the first update control factor in the t-th training process, λ t represents the second update control factor in the t-th training process, ε represents the smoothing factor, α1 represents the first exponential decay parameter, α2 represents the second exponential decay parameter, ξ t-1 represents the first update control factor in the t-1th training process, represents the gradient of the error function value relative to the weight parameter during the t-1th training process, λ t-1 Represents the second update control factor in the t-1th training process.

[0089] The network parameter updating method provided in the embodiment of the present invention can use the concept of momentum to adjust parameters, thereby achieving the technical effect of shortening the training time and avoiding falling into the local minimum, which helps to accurately find the optimal parameter combination.

[0090] In this embodiment of the present invention, the learning rate in the t-th training process is:

[0091]

[0092] Among them, η t-1 represents the learning rate during the t-1th training process, ΔRMSE = RMSE t -RMSE t-1 , RMSE t represents the error function value during the t-th training process, Et-1 represents the error function value in the t-1th training process, γ1 represents the first constant term between (0, 1), γ2 represents the second constant term between (0, 1), and γ2 is less than γ1.

[0093] The prior art mainly relies on fixed proportional coefficients for adjustment, which will make the adjustment range of the learning rate too large, causing the network to oscillate, and the accuracy of the final training model cannot be guaranteed. The adaptive adjustment of the learning rate provided in the embodiment of the present invention appropriately reduces the adjustment range of the learning rate according to the size of the relative error, and performs training at a more reasonable learning rate, which can ensure that the training process of the daily power consumption prediction model is carried out at the optimal learning rate and avoid oscillation.

[0094] like Figure 2 As shown, the embodiment of the present invention provides a thermal power plant power generation prediction device, including: a data acquisition module 201, a sample construction module 202, a data learning module 203, a power consumption prediction module 204 and a power generation prediction module 205;

[0095] The data collection module 201 is used to collect historical power consumption characteristic data of the target power supply area corresponding to the thermal power plant to be predicted, and clean the historical power consumption characteristic data to obtain the cleaned historical power consumption characteristic data;

[0096] The sample construction module 202 is used to construct a sample power consumption feature vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption feature vector based on the historical power consumption feature data of the target power supply area;

[0097] The data learning module 203 is used to use a neural network to learn the sample power consumption feature vector and the daily power consumption corresponding to the sample power consumption feature vector to obtain a daily power consumption prediction model;

[0098] The power consumption prediction module 204 is used to collect the real-time power consumption feature vector corresponding to the target power supply area, and use the daily power consumption prediction model to analyze the real-time power consumption feature vector to determine the predicted daily power consumption;

[0099] The power generation prediction module 205 is used to determine the line loss from the thermal power plant to be predicted to the target power supply area, and determine the predicted power generation corresponding to the thermal power plant to be predicted based on the line loss and the predicted daily power consumption.

[0100] Figure 2 The thermal power plant power generation prediction device provided in the illustrated embodiment can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0101] Figure 3 A schematic diagram of the structure of a thermal power plant power generation prediction device provided by an embodiment of the present invention. Figure 3 The processing device 300 may include a memory 301 and a processor 302. Exemplarily, the memory 301 and the processor 302 are interconnected via a bus 303.

[0102] The memory 301 stores computer-executable instructions;

[0103] The processor 302 executes the computer-executable instructions stored in the memory, so that the processor executes any of the above-mentioned thermal power plant power generation prediction methods.

[0104] Figure 3 The thermal power plant power generation prediction device shown can execute the technical solution shown in the above method embodiment, and its implementation principle and beneficial effects are similar, which will not be repeated here.

[0105] An embodiment of the present invention provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement any of the above-mentioned thermal power plant power generation prediction methods.

[0106] An embodiment of the present invention may also provide a computer program product, including a computer program, which, when executed by a processor, implements any of the above-mentioned methods for predicting power generation in a thermal power plant.

[0107] All or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a readable memory. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned memory (storage medium) includes: read-only memory (English: read-only memory, abbreviated: ROM), RAM, flash memory, hard disk, solid state drive, magnetic tape (English: magnetic tape), floppy disk (English: floppy disk), optical disc (English: optical disc) and any combination thereof.

[0108] The embodiments of the present invention are described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processing unit of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processing unit of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0109] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0110] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0111] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art that are not disclosed by the present invention. The specification and examples are to be considered exemplary only, and the true scope and spirit of the present invention is indicated by the claims.

[0112] It should be understood that the present invention is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for predicting power generation in a thermal power plant, characterized in that: include: For the thermal power plant to be predicted, historical power consumption characteristic data of the target power supply area corresponding to the thermal power plant to be predicted is collected, and the historical power consumption characteristic data is cleaned to obtain the cleaned historical power consumption characteristic data; Based on the historical power consumption characteristic data of the target power supply area, construct a sample power consumption characteristic vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption characteristic vector; A neural network is used to learn the sample electricity consumption feature vector and the daily electricity consumption corresponding to the sample electricity consumption feature vector to obtain a daily electricity consumption prediction model; Collecting the real-time power consumption feature vector corresponding to the target power supply area, and using the daily power consumption prediction model to analyze the real-time power consumption feature vector to determine the predicted daily power consumption; The line loss electricity from the thermal power plant to be predicted to the target power supply area is determined, and the predicted power generation corresponding to the thermal power plant to be predicted is determined based on the line loss electricity and the predicted daily electricity consumption.

2. The method for predicting power generation in a thermal power plant according to claim 1, characterized in that: Collect historical power consumption characteristic data of the target power supply area corresponding to the thermal power plant to be predicted, including: Collect the daily average temperature, daily average humidity, rainfall, season type, date type and daily power consumption of the target power supply area corresponding to the thermal power plant to be predicted, and obtain the historical power consumption characteristic data of the target power supply area; The date types include working days, holidays, and non-working days that are not holidays.

3. The method for predicting power generation in a thermal power plant according to claim 2, characterized in that: After cleaning the historical power consumption characteristic data, the cleaned historical power consumption characteristic data is obtained, including: filtering out the historical power consumption characteristic data with missing values ​​and / or abnormal values ​​to obtain the cleaned historical power consumption characteristic data.

4. The method for predicting power generation in a thermal power plant according to claim 2, characterized in that: Based on the historical power consumption characteristic data of the target power supply area, a sample power consumption characteristic vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption characteristic vector are constructed, including: For any historical electricity consumption characteristic data of the target power supply area, the daily average temperature, daily average humidity, rainfall, season type, date type and daily electricity consumption in the previous N days in the historical electricity consumption characteristic data are used as the sample electricity consumption feature vector corresponding to the target power supply area, and the daily electricity consumption in the historical electricity consumption characteristic data is used as the daily electricity consumption corresponding to the sample electricity consumption feature vector.

5. The method for predicting power generation in a thermal power plant according to claim 4, characterized in that: A neural network is used to learn the sample electricity consumption feature vector and the daily electricity consumption corresponding to the sample electricity consumption feature vector to obtain a daily electricity consumption prediction model, including: Constructing a BP neural network and an error function value corresponding to the BP neural network, and randomly initializing network parameters of the BP neural network; Using the sample electricity consumption feature vector as input data of the BP neural network, using the daily electricity consumption corresponding to the sample electricity consumption feature vector as the expected label value of the BP neural network, and obtaining the error function value of the BP neural network; According to the error function value of the BP neural network, the network parameters of the BP neural network are updated to obtain the BP neural network after the network parameters are updated; For the BP neural network after the network parameters are updated, the error function value of the BP neural network is re-obtained; It is determined whether the re-acquired error function value is less than a preset error threshold. If so, the training is completed and the current BP neural network is used as a daily power consumption prediction model. Otherwise, the step of updating the network parameters of the BP neural network is returned.

6. The method for predicting power generation in a thermal power plant according to claim 5, characterized in that: The error function value corresponding to constructing the BP neural network is: Among them, RMSE represents the error function value, y i It represents the actual output obtained by taking the electric feature vector of the i-th sample as input; d i represents the expected label value corresponding to the i-th sample electricity consumption feature vector, and n represents the total number of input sample electricity consumption feature vectors.

7. The method for predicting power generation in a thermal power plant according to claim 5, characterized in that: Update the network parameters of the BP neural network, including: Among them, ω t represents the network parameters in the tth training process, ω t-1 represents the network parameters in the t-1th training process, η t represents the learning rate during the t-th training process, ξ t represents the first update control factor in the tth training process, λ t represents the second update control factor in the t-th training process, ε represents the smoothing factor, α1 represents the first exponential decay parameter, α2 represents the second exponential decay parameter, ξ t-1 represents the first update control factor in the t-1th training process, represents the gradient of the error function value relative to the weight parameter during the t-1th training process, λ t-1 Represents the second update control factor in the t-1th training process.

8. The method for predicting power generation in a thermal power plant according to claim 7, characterized in that: The learning rate in the t-th training process is: Among them, η t-1 represents the learning rate during the t-1th training process, ΔRMSE = RMSE t -RMSE t-1 , RMSE t represents the error function value in the tth training process, E t-1 represents the error function value in the t-1th training process, γ1 represents the first constant term between (0,1), γ2 represents the second constant term between (0,1), and γ2 is less than γ1.

9. A thermal power plant power generation prediction device, characterized in that: include: Data collection module, sample construction module, data learning module, power consumption prediction module and power generation prediction module; The data collection module is used to collect historical power consumption characteristic data of the target power supply area corresponding to the thermal power plant to be predicted, and clean the historical power consumption characteristic data to obtain the cleaned historical power consumption characteristic data; The sample construction module is used to construct a sample power consumption feature vector corresponding to the target power supply area and a daily power consumption corresponding to the sample power consumption feature vector based on the historical power consumption feature data of the target power supply area; The data learning module is used to use a neural network to learn the sample power consumption feature vector and the daily power consumption corresponding to the sample power consumption feature vector to obtain a daily power consumption prediction model; The power consumption prediction module is used to collect the real-time power consumption feature vector corresponding to the target power supply area, and use the daily power consumption prediction model to analyze the real-time power consumption feature vector to determine the predicted daily power consumption; The power generation prediction module is used to determine the line loss power from the thermal power plant to be predicted to the target power supply area, and determine the predicted power generation corresponding to the thermal power plant to be predicted based on the line loss power and the predicted daily power consumption.

10. A thermal power plant power generation prediction device, characterized in that: including a processor and a memory; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the processor executes the method for predicting power generation in a thermal power plant according to any one of claims 1 to 9.

Images