Thermoelectricity cooperative control method and system, and storage medium

Abstract

The present application relates to the technical field of thermoelectric system control, and particularly relates to a thermoelectric cooperative control method and system and a storage medium, the method comprising: calculating photovoltaic power generation output electric energy and heat storage value of a thermal power plant through thermoelectric cooperative information, and obtaining a thermoelectric cooperative control target function according to the photovoltaic power generation output electric energy and the heat storage value of the thermal power plant; calculating photovoltaic power generation power supply coefficients and thermal power plant power supply coefficients based on the thermoelectric cooperative control target function; constructing a power supply prediction model to predict power supply information; and controlling the photovoltaic power generation and the thermal power plant based on the photovoltaic and thermal power plant power supply coefficients and the power supply prediction results, which effectively solves the problem of resource waste caused by insufficient system regulation capacity due to the randomness and intermittence of photovoltaic power generation and the high proportion of thermal power, thereby effectively optimizing resource allocation, effectively managing the power system through power supply prediction, and effectively promoting sustainable resource development.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of thermoelectric system control, and particularly relates to a thermoelectric cooperative control method and system and a storage medium. BACKGROUND

[0002] In recent years, with the rapid development of renewable energy, photovoltaic power generation, as an important part, has been widely used in the world. Photovoltaic power generation, with its clean and low-carbon characteristics, effectively alleviates the environmental pollution problem caused by traditional fossil energy. However, due to the significant influence of weather conditions on photovoltaic power generation, its output power presents randomness and intermittency, which brings great challenges to the stable operation of the power system.

[0003] At the same time, as a traditional energy supply mode, the thermal power plant has stable base load power supply capability in the power system, but due to the operation mode of "heat determines electricity", the thermal power plant has insufficient regulation flexibility and cannot quickly respond to the load changes caused by photovoltaic power generation fluctuations.

[0004] The information disclosed in this BACKGROUND section is only intended to enhance the understanding of the general background of the present disclosure and is not intended to be a recognition or any form of suggestion that this information constitutes prior art. SUMMARY

[0005] The present application provides a thermoelectric cooperative control method, system and storage medium, which can effectively solve the problems in the background art.

[0006] In order to achieve the above purpose, the technical scheme adopted by the present application is:

[0007] A thermoelectric cooperative control method, the method comprising:

[0008] Collecting thermoelectric cooperative information, the thermoelectric cooperative information including photovoltaic power plant working information and thermal power plant working information;

[0009] Based on the photovoltaic power plant working information and the thermal power plant working information respectively, calculating photovoltaic power generation output electric energy and thermal power plant heat storage value, obtaining photovoltaic output electric energy and thermal heat storage value;

[0010] Based on the photovoltaic power generation output electric energy and the thermal power plant heat storage value, constructing a thermoelectric cooperative target function, and outputting a photovoltaic power generation power supply coefficient and a thermal power plant power supply coefficient according to the thermoelectric cooperative target function, to obtain a thermoelectric pre-control result;

[0011] Constructing a power supply prediction model, training the power supply prediction model by using a machine learning algorithm, and obtaining a power supply prediction result;

[0012] According to the power supply prediction result, the actual power supply output is evaluated, and a power supply evaluation result is obtained. Based on the power supply evaluation result, the power supply prediction model is optimized.

[0013] According to the thermoelectricity pre-regulation result, the photovoltaic power plant and the thermal power plant are balanced and regulated in power supply in combination with the power supply prediction result.

[0014] Further, the photovoltaic power generation output electric energy is calculated based on the photovoltaic power plant working information, including:

[0015] According to the photovoltaic power plant working information, the real-time light intensity and the real-time temperature of the photovoltaic cell panel are extracted.

[0016] The historical thermoelectricity cooperative working information is collected, and the test standard light intensity, the photovoltaic standard maximum output electric energy, and the cell panel standard temperature are called to calculate the photovoltaic output electric energy according to the historical thermoelectricity cooperative working information. The historical thermoelectricity cooperative information includes historical photovoltaic power plant working information and historical thermal power plant working information.

[0017] The photovoltaic power generation output electric energy P at the t period PV，t The calculation formula is as follows:

[0018]

[0019] Where P PV，t is the photovoltaic power generation output electric energy at the t period, P STC is the photovoltaic standard maximum output electric energy, G AC，t is the real-time light intensity at the t period, G STC is the test standard light intensity, T C，t is the real-time temperature of the photovoltaic cell panel at the t period, T CC is the cell panel standard temperature.

[0020] Further, the thermal storage value of the thermal power plant is calculated based on the thermal power plant working information, including:

[0021] According to the thermal power plant working information, the real-time heat storage power of the heat storage device and the real-time heat release power of the heat storage device are called. The thermal power plant working information is traversed, the thermal equipment data is called, and the heat storage equipment heat storage efficiency and heat release efficiency are obtained according to the thermal equipment data.

[0022] According to the historical thermoelectricity cooperative working information, the heat loss rate of the heat storage device is obtained.

[0023] The thermal storage value E of the thermal power plant at the t period HS，t The calculation formula is as follows:

[0024]

[0025] wherein E HS，t is the heat storage value of the heat storage device at time t, k1 is the heat loss rate of the heat storage device, k2 is the heat storage efficiency of the heat storage device at time interval Δt, k3 is the heat release efficiency of the heat storage device at time interval Δt, E HS-C，t is the real-time heat storage power of the heat storage device, E HS-F，t is the real-time heat release power of the heat storage device.

[0026] Further, the photovoltaic power supply coefficient and the heat and power plant power supply coefficient are output according to the heat and power coordination target function, and a heat and power pre-control result is obtained, including:

[0027] The power plant operation cycle and load requirement are obtained by traversing the historical heat and power coordination working information, and the photovoltaic power plant operation cost and the heat and power plant operation cost, the photovoltaic power generation loss cost and the heat and power plant loss cost, and the heat storage rated value of the heat and power plant are called.

[0028] Market electricity price data are collected, the grid power supply price is obtained according to the market electricity price data, and the photovoltaic power supply price and the heat and power plant power supply price are calculated according to the grid power supply price and the heat storage rated value of the heat and power plant respectively, and the calculation formula is as follows:

[0029]

[0030] wherein PR PV，t is the photovoltaic power supply price, PR HS，t is the heat and power plant power supply price, P PV，t represents the photovoltaic power output electrical energy at time t, P STC is the photovoltaic standard maximum output electrical energy, PR t is the grid power supply price, E HS，t is the heat storage value of the heat and power plant at time t, E HS，E is the heat storage rated value of the heat and power plant.

[0031] The photovoltaic power supply coefficient and the heat and power plant power supply coefficient are calculated according to the heat and power coordination target function, and the heat and power coordination target function is as follows:

[0032]

[0033] wherein S represents economic benefit, P PV，t represents the photovoltaic power output electrical energy at time t, E HS，t represents the heat storage value of the heat and power plant at time t, PR PV，t is the photovoltaic power supply price, PR HS，t is the heat and power plant power supply price, k4 is the photovoltaic power supply coefficient, k5 is the heat and power plant power supply coefficient, CPV，t C is the operating cost of the photovoltaic power plant HS，t L is the operating cost of the thermal power plant PV L is the photovoltaic power generation loss cost HS T is the loss cost of the thermal power plant, and P is the operating period of the power plant L is the load requirement represents a constraint condition, that is, the power generation capacity of the power plant cannot exceed the load requirement;

[0034] According to the photovoltaic power generation power supply coefficient and the thermal power plant power supply coefficient, a thermal power pre-control result is obtained.

[0035] Further, the power supply prediction model is constructed, and a machine learning algorithm is used to train the power supply prediction model to obtain a power supply prediction result; including:

[0036] Collect historical meteorological information and historical power supply information, and divide the historical meteorological information and historical power supply information into a power supply prediction training set and a power supply prediction test set according to the historical meteorological information and historical power supply information;

[0037] A power supply prediction model is constructed, and the power supply prediction model is trained and optimized according to the power supply prediction training set and the power supply prediction test set, respectively, to obtain a model training and optimization result, and the power supply prediction model includes a convolutional neural network model and a time series model;

[0038] According to the model training and optimization result, a convolutional neural network is used to extract meteorological features from the historical meteorological information, and the meteorological features are flattened to obtain a one-dimensional meteorological feature vector;

[0039] According to the model training and optimization result, a time series model is constructed to integrate and fuse the one-dimensional meteorological feature vector and historical power supply information, and to predict future power supply.

[0040] Further, the convolutional neural network is used to extract meteorological features from the historical meteorological information, and the meteorological features are flattened to obtain a one-dimensional meteorological feature vector, including:

[0041] The historical meteorological information is normalized, the normalized historical meteorological information is format-converted, time series format data is converted into image format, meteorological data at each time point is taken as a pixel value of an image, and a standard historical meteorological information image is obtained;

[0042] A convolutional neural network model is constructed, and the standard historical meteorological information image is identified based on the model training and optimization result, and key meteorological features are extracted to obtain a multi-dimensional meteorological feature map;

[0043] The multi-dimensional meteorological feature map is subjected to a flattening operation to obtain a one-dimensional meteorological feature vector, and the flattening operation is used to interface with the time series model.

[0044] Further, the time series model is constructed to integrate and fuse the one-dimensional meteorological feature vector and historical data, and to predict future power supply, including:

[0045] The time series model is a long short-term memory network model:

[0046] An input layer receives the one-dimensional meteorological feature vector and the historical power supply information;

[0047] An LSTM layer is responsible for processing power supply data and meteorological features that change over time series, and predicting future power supply;

[0048] A decision layer maps the features output by the LSTM layer to the output space of power supply prediction, i.e., converts the features of time series data into the results of power supply prediction;

[0049] An output layer outputs the power supply prediction results.

[0050] Further, according to the power supply prediction results, the actual power supply output is evaluated to obtain a power supply evaluation result, and based on the power supply evaluation result, the power supply prediction model is optimized, including:

[0051] Real-time monitoring of the power supply data, and error analysis of the power supply prediction results and actual power supply results, to obtain a power supply error analysis result;

[0052] Based on the power supply error analysis result, the power supply prediction model defects are identified, and the power supply prediction model is iteratively corrected according to the power supply prediction model defects.

[0053] A thermoelectricity collaborative control system, the system comprising:

[0054] A thermoelectricity data acquisition module acquires thermoelectricity collaborative information, the thermoelectricity collaborative information including photovoltaic power plant working information and thermoelectricity power plant working information;

[0055] A thermoelectricity data calculation module calculates photovoltaic power output electrical energy and thermoelectricity power plant heat storage values based on the photovoltaic power plant working information and the thermoelectricity power plant working information, respectively, to obtain photovoltaic output electrical energy and thermoelectricity heat storage values;

[0056] A collaborative coefficient output module constructs a thermoelectricity collaborative target function based on the photovoltaic power output electrical energy and the thermoelectricity heat storage values, and outputs photovoltaic power supply coefficients and thermoelectricity power plant supply coefficients according to the thermoelectricity collaborative target function, to obtain thermoelectricity pre-control results;

[0057] The power supply prediction module constructs a power supply prediction model, trains the power supply prediction model by using a machine learning technique, and obtains a power supply prediction result;

[0058] The prediction model optimization module evaluates an actual power supply output according to the power supply prediction result, obtains a power supply evaluation result, and optimizes the power supply prediction model based on the power supply evaluation result;

[0059] The thermoelectricity collaborative control module performs power supply balance regulation and control on the photoelectric power plant and the thermoelectricity power plant according to the thermoelectricity pre-regulation and control result and the power supply prediction result.

[0060] A thermoelectricity collaborative control storage medium, which is a computer readable medium, wherein the thermoelectricity collaborative control storage medium, when executed by a processor, implements the steps of the thermoelectricity collaborative control storage method according to any one of the claims.

[0061] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the specific embodiments of the present application can be implemented according to the content of the description, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0062] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or the prior art description. Obviously, the drawings in the following description are only some embodiments described in the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0063] Figure 1 The flowchart of the thermoelectricity collaborative control method is shown in the figure.

[0064] Figure 2 The flowchart of the power supply prediction model construction is shown in the figure.

[0065] Figure 3 The structure diagram of the thermoelectricity collaborative control system is shown in the figure. DETAILED DESCRIPTION

[0066] The technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments.

[0067] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used in the description herein is for describing particular embodiments only and is not intended to be limiting of the application. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0068] Embodiment one;

[0069] As Figure 1 shown, the application provides a thermoelectricity cooperative control method, the method comprising:

[0070] S10: collecting thermoelectricity cooperative information, the thermoelectricity cooperative information comprising photovoltaic power plant working information and thermal power plant working information;

[0071] S20: calculating photovoltaic power generation output electric energy and thermal power plant heat storage value based on the photovoltaic power plant working information and the thermal power plant working information respectively, and obtaining the photovoltaic output electric energy and the thermal power heat storage value;

[0072] S30: constructing a thermoelectricity cooperative target function based on the photovoltaic power generation output electric energy and the thermal power plant heat storage value, and outputting photovoltaic power generation power supply coefficient and thermal power plant power supply coefficient according to the thermoelectricity cooperative target function, and obtaining thermoelectricity pre-regulation and control result;

[0073] S40: constructing a power supply prediction model, training the power supply prediction model by using a machine learning algorithm, and obtaining a power supply prediction result;

[0074] S50: evaluating actual power supply output according to the power supply prediction result, obtaining a power supply evaluation result, and optimizing the power supply prediction model based on the power supply evaluation result;

[0075] S60: performing power supply balance regulation and control on the photovoltaic power plant and the thermal power plant according to the thermoelectricity pre-regulation and control result and the power supply prediction result.

[0076] Specifically, first, the operating status of the photovoltaic power plant and the thermal power plant is obtained through the sensor network, and the thermal power coordination information is collected using various functional sensors such as solar radiation sensors, temperature sensors, etc., including solar radiation intensity, photovoltaic panel temperature, boiler operating status, steam flow, heat carrier temperature, etc. According to the working information of the photovoltaic power plant and the thermal power plant, such as radiation intensity and panel temperature, boiler state, etc., the photovoltaic power output and the thermal power storage value are calculated. According to the photovoltaic power output and the thermal power storage value, a thermal power coordination objective function is constructed using, for example, linear programming, which is mainly used to optimize the energy output and scheduling of photovoltaic and thermal power plants to maximize cost-effectiveness and minimize environmental impact. Then, a machine learning model is designed to predict future power supply conditions, and the model prediction results are compared with the actual power supply conditions. Through error analysis, the accuracy of the model is evaluated, and the model parameters are adjusted, such as improving the learning rate, adding hidden layers, or adjusting input features. According to the prediction results and actual demand, the output of photovoltaic power generation and thermal power plants is adjusted. It should be noted that the dynamic adjustment of the power supply coefficient needs to be based on real-time data and prediction results to be flexible, and a reasonable feedback and adjustment mechanism should be set up to respond to emergencies and changes in demand.

[0077] Through the technical scheme of the present application, the problem of insufficient system regulation ability caused by the high proportion of thermal power due to the randomness and intermittency of photovoltaic power generation is effectively solved, thereby effectively optimizing resource allocation and managing the power system through power supply prediction, greatly promoting sustainable development of resources.

[0078] Further, the photovoltaic power output is calculated based on the working information of the photovoltaic power plant, including:

[0079] According to the working information of the photovoltaic power plant, the real-time illumination intensity and the real-time temperature of the photovoltaic cell panel are extracted;

[0080] Historical thermal power coordination working information is collected, and the test standard illumination intensity, the photovoltaic standard maximum output power, and the cell panel standard temperature are called to calculate the photovoltaic output power based on the historical thermal power coordination working information. The historical thermal power coordination information includes historical photovoltaic power plant working information and historical thermal power plant working information;

[0081] The photovoltaic power output P PV，t The calculation formula is as follows:

[0082]

[0083] Where P PV，t is the photovoltaic power output at time t, P STC is the photovoltaic standard maximum output power, G AC，t is the real-time illumination intensity at time t, and G STCTo test the standard light intensity, T C，t is the real-time temperature of the photovoltaic panel during period t, T CC is the standard temperature of the solar panel.

[0084] As a preferred embodiment of the above, first, based on the obtained working information of the photovoltaic power plant, the real-time light intensity and the real-time temperature of the photovoltaic panel are extracted therefrom, and the historical working information of the photovoltaic power plant is collected, including the historical light intensity, the temperature data of the photovoltaic panel, the photovoltaic power generation, etc. The test standard light intensity can be calculated based on the average value of the historical light intensity data or other statistical indicators. The photovoltaic standard maximum output power is calculated by the maximum conversion efficiency of the photovoltaic panel and the power generation capacity under standard environmental conditions such as standard light intensity and standard temperature. Generally, 25°C is taken as the standard temperature of the panel, and the photovoltaic output power is calculated according to the photovoltaic power generation output power calculation formula.

[0085] Furthermore, the thermal power plant heat storage value is calculated based on the thermal power plant operating information, including:

[0086] According to the working information of the thermal power plant, the real-time heat storage power and the real-time heat release power of the heat storage device are retrieved, the working information of the thermal power plant is traversed, the thermoelectric equipment data is retrieved, and the heat storage efficiency and heat release efficiency of the heat storage device are obtained according to the thermoelectric equipment data;

[0087] Obtaining the heat loss rate of the heat storage device based on historical thermoelectric collaborative operation information;

[0088] Thermal power plant heat storage value E during period t HS，t The calculation formula is as follows:

[0089]

[0090] Among them, E HS，t is the heat storage value of the thermal power plant during period t, k1 is the heat loss rate of the heat storage device, k2 is the heat storage efficiency of the heat storage device during period Δt, k3 is the heat release efficiency of the heat storage device during period Δt, E HS-C，t The real-time heat storage power of the heat storage device, E HS-F，t Provides real-time heat release to the heat storage device.

[0091] In this embodiment, first, according to the collected thermal power plant working information, the energy transfer rate from the heat source to the heat storage device is measured and recorded, the real-time heat storage power of the heat storage device is obtained, then the rate of energy release of the heat storage device to the external system or environment is recorded, the real-time heat release power of the heat storage device is obtained, the operation data of the thermal power plant is collected and recorded, including the boiler temperature, steam pressure, etc., which will be used to evaluate the performance of the heat storage device, the heat storage efficiency and heat release efficiency of the heat storage device are obtained based on the historical thermal power plant working information, and the heat dissipation loss rate of the heat storage device is calculated, which reflects the energy loss due to incomplete insulation of the equipment during the heat storage process, and finally the thermal storage value of the thermal power plant is calculated by the thermal storage value calculation formula of the thermal power plant.

[0092] Further, according to the photovoltaic power supply coefficient and the thermal power plant power supply coefficient output by the thermal power coordination target function, the thermal power pre-control result is obtained, including:

[0093] The power plant operation cycle and load requirement are obtained by traversing the historical thermal power coordination working information, the photovoltaic power plant operation cost and the thermal power plant operation cost, the photovoltaic power generation loss cost and the thermal power plant loss cost, and the thermal power plant storage rated value are retrieved;

[0094] Market electricity price data is collected, the grid power supply price is obtained according to the market electricity price data, and the photovoltaic power supply price and the thermal power plant power supply price are calculated according to the grid power supply price and the thermal power plant heat storage rated value respectively, the calculation formula is as follows:

[0095]

[0096] Among them, PR PV，t is the photovoltaic power supply price, PR HS，t is the thermal power plant power supply price, P PV，t represents the photovoltaic power output electrical energy at t period, P STC is the standard maximum output electrical energy of photovoltaic, PR t is the grid power supply price, E HS，t is the thermal storage value of the thermal power plant at t period, E HS，E is the thermal storage rated value of the thermal power plant;

[0097] The photovoltaic power supply coefficient and the thermal power plant power supply coefficient are calculated according to the thermal power coordination target function, and the thermal power coordination target function is as follows:

[0098]

[0099] Among them, S represents economic benefit, P PV，t represents the photovoltaic power output electrical energy at t period, E HS，t represents the thermal storage value of the thermal power plant at t period, PR PV，t is the photovoltaic power supply price, PRHS，t k4 is the power supply coefficient of photovoltaic power generation, k5 is the power supply coefficient of thermal power plant, C PV，t HS，t is the operation cost of thermal power plant, L PV HS is the loss cost of thermal power plant, T is the operation cycle of power plant, P L is the load requirement, represents the constraint condition, that is, the power generation capacity of power plant cannot exceed the load requirement;

[0100] According to the power supply coefficient of photovoltaic power generation and the power supply coefficient of thermal power plant, the thermal power pre-control result is obtained.

[0101] Specifically, first, the historical thermal power cooperative working information is traversed, the operation cost, loss cost and thermal power storage rated value of photovoltaic power plant and thermal power plant are called, the current and expected market electricity price data are collected, which will be used to calculate the power supply price and evaluate the economic benefit, the grid power supply price is obtained according to the collected market electricity price data, and the power supply price of photovoltaic power generation is determined by combining the photovoltaic power output, the standard maximum output power of photovoltaic power, the thermal power storage value of thermal power plant and the rated value of thermal power storage of thermal power plant, the power supply price of thermal power plant is obtained, the power supply coefficient of photovoltaic power generation and the power supply coefficient of thermal power plant are calculated by adopting optimization algorithms such as linear programming, genetic algorithm and the like, and finally the actual output power of photovoltaic power generation is calculated by the power supply coefficient of photovoltaic power generation and the output power of photovoltaic power generation, and the calculation formula is P PV =k4·P PV，t , P PV，t is the output power of photovoltaic power generation, k4 is the power supply coefficient of photovoltaic power generation, and the actual output power of thermal power plant is obtained by the power supply coefficient of thermal power plant and the thermal power storage value of thermal power plant, and the calculation formula is P HS =k5·E HS，t , E HS，t is the thermal power storage value of thermal power plant, k5 is the power supply coefficient of thermal power plant, and the pre-control of photovoltaic power generation and thermal power plant is carried out according to the actual output power of the two, it should be noted that the thermal power cooperative control objective function is to optimize the operation strategy of the whole thermal power cooperative system, and is not only used to maximize the economic benefit, which is only a key target, and in specific application, it may also include stability and environmental protection and other trade-off factors, and in the present expression, it is mainly used to realize the optimization of economic benefit by using power generation income, operation cost and system loss.

[0102] Further, as shown in Figure 2 , a power supply prediction model is constructed, a machine learning algorithm is adopted to train the power supply prediction model, and a power supply prediction result is obtained; including:

[0103] ​​S41: Collect historical meteorological information and historical power supply information, and divide the historical meteorological information and historical power supply information into a power supply prediction training set and a power supply prediction test set;

[0104] S42: Build a power supply prediction model, train and optimize the power supply prediction model according to the power supply prediction training set and the power supply prediction test set, and obtain model training optimization results. The power supply prediction model includes a convolutional neural network model and a time series model.

[0105] S43: Based on the model training optimization results, a convolutional neural network is used to extract meteorological features from historical meteorological information, and the meteorological features are flattened to obtain a one-dimensional meteorological feature vector;

[0106] S44: Based on the model training optimization results, a time series model is constructed to integrate the one-dimensional meteorological feature vector with the historical power supply information, and predict future power supply.

[0107] As a preferred embodiment of the above, first collect historical meteorological information such as temperature, humidity, wind speed, sunshine duration, etc. from the relevant meteorological departments and power system databases, and historical power supply information such as hourly power supply and load conditions, clean and normalize the data, process missing values ​​and outliers, ensure data quality, use an appropriate ratio such as 80% training set and 20% test set, divide the data into power supply prediction training set and power supply prediction test set, so as to train and evaluate the model, use convolutional neural network model and time series model to build power supply prediction model, use power supply prediction training set to test convolutional neural network model and time series model respectively. The trained convolutional neural network model is then used to extract features from historical meteorological information. By capturing local features in the data, the multidimensional feature map output by the convolutional neural network model is flattened and converted into a one-dimensional meteorological feature vector. The one-dimensional meteorological feature vector and historical power supply information are combined to construct a time series model. The time series model is used to analyze the fused data and predict future power supply conditions. It should be noted that overfitting is prone to occur, especially when processing large amounts of meteorological data. Regularization, cross-validation and other means can be used to avoid the model from over-learning the training data and ensure its generalization ability on the test set.

[0108] Furthermore, a convolutional neural network is used to extract meteorological features from historical meteorological information, and the meteorological features are flattened to obtain a one-dimensional meteorological feature vector, including:

[0109] Normalize the historical meteorological information, convert the format of the normalized historical meteorological information, convert the time series format data into image format, use the meteorological data at each time point as a pixel value of the image, and obtain a standard historical meteorological information map;

[0110] A convolutional neural network model is constructed, and based on the model training optimization result, a standard historical meteorological information map is recognized, and key meteorological features are extracted to obtain a multi-dimensional meteorological feature map;

[0111] The multi-dimensional meteorological feature map is flattened to obtain a one-dimensional meteorological feature vector, and the flattening operation is used to interface with the time series model.

[0112] In this embodiment, first, all historical meteorological data is normalized to eliminate the influence of different measurement scales, and a normalization method such as the min-max scaling method is usually used to scale the data to between 0 and 1, and then the time series format data is converted into an image format, specifically, each time point in the time series data can be regarded as a pixel point in the image, and the gray value of each pixel represents the meteorological data value such as temperature, humidity, etc. at the time point, so that a complete time period of data will be converted into an image, and then a convolutional neural network model is designed, which includes multiple convolutional layers and pooling layers. The convolutional layer extracts local features in the image by applying multiple filters, and the pooling layer reduces the spatial dimension of the features, enhances the anti-interference ability of the model, and reduces the calculation amount. The trained convolutional neural network model is used to perform convolution and pooling operation on the standard historical meteorological information map to extract key meteorological features. The feature extraction result is a multi-dimensional meteorological feature map, which contains high-level abstract features learned from the original input image. The multi-dimensional meteorological feature map is flattened to convert it into a one-dimensional meteorological feature vector. Flattening is the process of converting a multi-dimensional data structure into a one-dimensional array. This step is to interface the output of the CNN with the input format of the time series model.

[0113] Further, a time series model is constructed to integrate and fuse the one-dimensional meteorological feature vector and historical data, and to predict future power supply, including:

[0114] The time series model is a long short-term memory network model:

[0115] The input layer receives the one-dimensional meteorological feature vector and historical power supply information.

[0116] The LSTM layer is responsible for processing power supply data and meteorological features that change over time, and predicting future power supply.

[0117] The decision layer maps the features output by the LSTM layer to the output space of the power supply prediction, i.e. converts the features of the time series data into the results of the power supply prediction.

[0118] The output layer outputs the power supply prediction result.

[0119] Specifically, the meteorological feature vector extracted by the convolutional neural network contains key information of meteorological data at multiple time points. This vector serves as the input feature, while historical power supply data such as hourly power generation, load conditions, and power equipment operating status serve as another input for the time series model. The input data is then normalized to eliminate the influence of differences between different features. The input layer receives one-dimensional meteorological feature vectors and historical power supply information as inputs for the model, with the input dimension being the number of time steps and the number of features. Each time step includes meteorological information and historical power supply information. The LSTM layer captures long-term and short-term dependencies through memory cells, adapting to changes in power supply and meteorological data over time. One or more LSTM layers are used, each containing multiple neurons. The specific number of neurons can be adjusted based on experimental results and computational capacity. The output features of the LSTM layer are then mapped to the output space of the power supply prediction. The decision layer is used to convert the time series features processed by the LSTM into the results of the power supply prediction. A linear activation function is typically used to produce numerical output for power supply prediction. Finally, the output layer calculates the final power supply prediction result based on the output of the decision layer. The output is usually a single numerical value representing the power generation or load prediction for a future period.

[0120] Further, according to the power supply prediction result, the actual power supply output is evaluated to obtain a power supply evaluation result. Based on the power supply evaluation result, the power supply prediction model is optimized, including:

[0121] Real-time monitoring of power supply data and error analysis of power supply prediction results and actual power supply results are performed to obtain power supply error analysis results.

[0122] Based on the power supply error analysis results, the power supply prediction model defects are identified, and the power supply prediction model is iteratively corrected according to the power supply prediction model defects.

[0123] As a preferred embodiment of the above embodiment, the sensor is used to monitor the power supply data, including real-time power generation, power grid load, energy storage device state and other key parameters. Using the trained power supply prediction model, the future power supply prediction results are generated according to the real-time weather data and historical power supply data. The prediction results are compared with the actual power supply data to calculate the prediction error. Common error indicators include mean square error, root mean square error, and absolute percentage error. The error is classified according to types such as large bias and large fluctuation to identify situations where the prediction error is large under certain time periods or certain weather conditions. For example, some seasonal weather changes or sudden weather events may cause the model error to increase. According to the error analysis results, identify the defects of the power supply prediction model, such as overfitting, insufficient input features, time delay, etc. Based on the power supply error analysis results, the model is iteratively corrected. Whenever the prediction error of the model exceeds the preset threshold, the retraining or fine-tuning of the model is triggered in a timely manner. It should be noted that iterative correction of the model is an effective method to improve prediction accuracy, but the optimization process needs to be ensured to be timely. If the feedback is delayed or the optimization period is too long, it may cause power supply prediction errors, which may affect the power grid dispatching decision.

[0124] Embodiment two;

[0125] As Figure 3 shown, based on the same inventive concept as the thermal power cooperative control method in the preceding embodiments, the present application also provides a thermal power cooperative control system, the system comprises:

[0126] a thermal power data acquisition module for acquiring thermal power cooperative information, including photovoltaic power plant working information and thermal power plant working information;

[0127] a thermal power data calculation module for calculating photovoltaic power output and thermal power plant heat storage value based on photovoltaic power plant working information and thermal power plant working information respectively, and obtaining photovoltaic output power and thermal power heat storage value;

[0128] a cooperative coefficient output module for constructing a thermal power cooperative target function based on photovoltaic power output and thermal power heat storage value, and outputting photovoltaic power supply coefficient and thermal power plant power supply coefficient according to the thermal power cooperative target function, and obtaining thermal power pre-control result;

[0129] a power supply prediction module for constructing a power supply prediction model, training the power supply prediction model using machine learning technology, and obtaining a power supply prediction result;

[0130] a prediction model optimization module for evaluating the actual power supply output based on the power supply prediction result, obtaining a power supply evaluation result, and optimizing the power supply prediction model based on the power supply evaluation result;

[0131] The thermoelectricity collaborative control module balances and controls the power supply of the photoelectric power plant and the thermoelectricity power plant according to the thermoelectricity pre-control result and the power supply prediction result.

[0132] The adjustment system in the present application can effectively realize a thermoelectricity collaborative control method, and the technical effects thereof are as described in the above embodiment, which will not be repeated here.

[0133] Embodiment three;

[0134] Based on the same inventive concept as the thermoelectricity collaborative control method in the above embodiment, the present application further provides a thermoelectricity collaborative control storage medium, which is a computer readable medium, wherein the thermoelectricity collaborative control storage medium is executed by a processor to realize the steps of the thermoelectricity collaborative control storage method.

[0135] Similarly, the above optimization scheme of the system can also correspondingly realize the optimization effect of the method in embodiment one, which will not be repeated here.

[0136] Although the present application is described in combination with specific features and embodiments thereof, it is obvious that various modifications and combinations can be made without departing from the spirit and scope of the present application. Accordingly, the present specification and drawings are merely exemplary illustrations of the present application defined in the appended claims, and any and all modifications, variations, combinations or equivalents that fall within the scope of the present application are intended to be covered. Obviously, those skilled in the art can make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalents, the present application is intended to include these modifications and variations.

Claims

1. A thermoelectric coordinated control method, characterized in that: The method comprises: Collecting thermal power synergy information, wherein the thermal power synergy information includes photovoltaic power plant operating information and thermal power plant operating information; Calculating the photovoltaic power generation output power and the thermal power plant heat storage value based on the photovoltaic power plant operating information and the thermal power plant operating information, respectively, to obtain the photovoltaic output power and the thermal power plant heat storage value; Constructing a thermoelectric synergy objective function based on the photovoltaic power generation output electric energy and the thermal power plant heat storage value, and outputting a photovoltaic power generation power supply coefficient and a thermal power plant power supply coefficient according to the thermoelectric synergy objective function to obtain a thermoelectric pre-control result; The photovoltaic power supply coefficient and the thermal power plant power supply coefficient are calculated according to the thermoelectric synergy objective function, and the thermoelectric synergy objective function is as follows: ； in Indicates economic benefits, express The photovoltaic power generation output electric energy during the period, express The thermal power plant heat storage value for the period, The photovoltaic power supply price is The electricity price for the thermal power plant, is the photovoltaic power supply coefficient, is the power supply coefficient of the thermal power plant, is the operating cost of the photovoltaic power plant, is the operating cost of the thermal power plant, is the photovoltaic power generation loss cost, is the loss cost of the thermal power plant, is the power plant operation cycle, is the load requirement, Represents the constraint condition, that is, the power plant's generating capacity cannot exceed the load requirement; Building a power supply prediction model, training the power supply prediction model using a machine learning algorithm, and obtaining power supply prediction results; Evaluate the actual power output according to the power supply prediction result, obtain a power supply evaluation result, and optimize the power supply prediction model based on the power supply evaluation result; According to the thermal power pre-regulation result, combined with the power supply prediction result, power supply balance regulation is performed on the photovoltaic plant and the thermal power plant.

2. The heat and power coordinated control method according to claim 1, characterized in that: Calculating photovoltaic power output electric energy based on the photovoltaic power plant operating information further includes: Extracting real-time light intensity and real-time temperature of photovoltaic panels according to the photovoltaic power plant working information; Collect historical thermal and electric collaborative work information, and calculate photovoltaic output power based on the test standard light intensity, photovoltaic standard maximum output power, and solar panel standard temperature, wherein the historical thermal and electric collaborative work information includes historical photovoltaic power plant operation information and historical thermal power plant operation information; Photovoltaic power generation output power during the period The calculation formula is as follows: ； in for The photovoltaic power generation output electric energy during the period, is the maximum output power of the photovoltaic standard, for The real-time light intensity during the period, is the test standard light intensity, for The real-time temperature of the photovoltaic panels during the period, is the standard temperature of the solar panel.

3. The heat and power coordinated control method according to claim 2, characterized in that: Calculating the thermal power plant heat storage value based on the thermal power plant operating information further includes: Retrieving the real-time heat storage power and the real-time heat release power of the heat storage device according to the working information of the thermal power plant, traversing the working information of the thermal power plant, retrieving the thermoelectric equipment data, and obtaining the heat storage efficiency and heat release efficiency of the heat storage device according to the thermoelectric equipment data; obtaining a heat dissipation loss rate of a heat storage device according to the historical thermoelectric cooperative operation information; Thermal power plant heat storage value during the period The calculation formula is as follows: ； in, for The thermal power plant heat storage value for the period, is the heat loss rate of the heat storage device, for The heat storage efficiency of the heat storage device during the period, for The heat release efficiency of the heat storage device during the period, The real-time heat storage power of the heat storage device, The heat storage device releases heat power in real time.

4. The heat and power coordinated control method according to claim 2 or 3, characterized in that: Outputting the photovoltaic power supply coefficient and the thermal power plant power supply coefficient according to the thermoelectric synergy objective function to obtain a thermoelectric pre-regulation result further includes: Traversing the historical thermal power collaborative operation information, obtaining the power plant operation cycle and load requirements, and retrieving the photovoltaic power plant operation cost and the thermal power plant operation cost, the photovoltaic power generation loss cost and the thermal power plant loss cost, and the thermal power plant heat storage rating; Collect market electricity price data, obtain the grid power supply price based on the market electricity price data, and calculate the photovoltaic power supply price and the thermal power plant power supply price based on the grid power supply price and the thermal power plant heat storage rating, respectively. The calculation formula is as follows: ； ； in, The photovoltaic power supply price is The electricity price for the thermal power plant, express The photovoltaic power generation output electric energy during the period, is the maximum output power of the photovoltaic standard, is the electricity price of the power grid, for The thermal power plant heat storage value for the period, providing a thermal storage rating for the thermal power plant; A thermal power pre-regulation result is obtained according to the photovoltaic power generation power supply coefficient and the thermal power plant power supply coefficient.

5. The heat and power coordinated control method according to claim 1, characterized in that: The constructing of the power supply prediction model, training the power supply prediction model using a machine learning algorithm, and obtaining a power supply prediction result further includes: Collecting historical meteorological information and historical power supply information, and dividing the historical meteorological information and historical power supply information into a power supply prediction training set and a power supply prediction test set; Constructing a power supply prediction model, training and optimizing the power supply prediction model according to the power supply prediction training set and the power supply prediction test set, respectively, to obtain a model training optimization result, wherein the power supply prediction model includes a convolutional neural network model and a time series model; According to the model training optimization results, a convolutional neural network is used to extract meteorological features from the historical meteorological information, and the meteorological features are flattened to obtain a one-dimensional meteorological feature vector; According to the model training optimization results, a time series model is constructed to integrate the one-dimensional meteorological feature vector and the historical power supply information, and to predict future power supply.

6. The heat and power coordinated control method according to claim 5, characterized in that: The extracting meteorological features from the historical meteorological information using a convolutional neural network and flattening the meteorological features to obtain a one-dimensional meteorological feature vector further includes: Normalizing the historical meteorological information, converting the format of the normalized historical meteorological information, converting the time series format data into an image format, using the meteorological data at each time point as a pixel value of the image, and obtaining a standard historical meteorological information graph; Constructing a convolutional neural network model, identifying the standard historical meteorological information map based on the model training optimization results, and extracting key meteorological features to obtain a multidimensional meteorological feature map; A flattening operation is performed on the multidimensional meteorological feature map to obtain a one-dimensional meteorological feature vector, wherein the flattening operation is used to interface with the time series model.

7. The heat and power coordinated control method according to claim 6, characterized in that: Constructing a time series model to integrate and fuse the one-dimensional meteorological feature vector with historical data and predicting future power supply further includes: The time series model is a long short-term memory network model: An input layer, receiving the one-dimensional meteorological feature vector and the historical power supply information; The LSTM layer is responsible for processing power supply data and meteorological characteristics that change over time and predicting future power supply; The decision layer maps the features output by the LSTM layer to the output space of the power supply forecast, that is, converts the features of the time series data into the results of the power supply forecast; Output layer, outputs power supply prediction results.

8. The heat and power coordinated control method according to claim 7, characterized in that: The method further comprises: evaluating an actual power supply output according to the power supply prediction result, obtaining a power supply evaluation result, and optimizing the power supply prediction model based on the power supply evaluation result. monitoring the power supply data in real time, and performing error analysis between the power supply prediction result and the actual power supply result to obtain a power supply error analysis result; Based on the power supply error analysis result, a power supply prediction model defect is identified, and the power supply prediction model is iteratively corrected according to the power supply prediction model defect.

9. A heat and power coordinated control system, characterized in that: The system comprises: A thermoelectric data acquisition module collects thermoelectric synergy information, including photovoltaic power plant operating information and thermal power plant operating information; a thermoelectric data calculation module, which calculates the photovoltaic power generation output power and the thermal power plant heat storage value based on the photovoltaic power plant working information and the thermal power plant working information, and obtains the photovoltaic output power and the thermoelectric heat storage value; a synergy coefficient output module, which constructs a thermoelectric synergy objective function based on the photovoltaic power generation output power and the thermoelectric heat storage value, and outputs the photovoltaic power generation power supply coefficient and the thermoelectric power plant power supply coefficient according to the thermoelectric synergy objective function to obtain a thermoelectric pre-control result; The photovoltaic power supply coefficient and the thermal power plant power supply coefficient are calculated according to the thermoelectric synergy objective function, and the thermoelectric synergy objective function is as follows: ； in Indicates economic benefits, express The photovoltaic power generation output electric energy during the period, express The thermal power plant heat storage value for the period, The photovoltaic power supply price is The electricity price for the thermal power plant, is the photovoltaic power supply coefficient, is the power supply coefficient of the thermal power plant, is the operating cost of the photovoltaic power plant, is the operating cost of the thermal power plant, is the photovoltaic power generation loss cost, is the loss cost of the thermal power plant, is the power plant operation cycle, is the load requirement, Represents the constraint condition, that is, the power plant's generating capacity cannot exceed the load requirement; The power supply prediction module constructs a power supply prediction model, uses machine learning technology to train the power supply prediction model, and obtains power supply prediction results; The prediction model optimization module evaluates the actual power supply output according to the power supply prediction result, obtains a power supply evaluation result, and optimizes the power supply prediction model based on the power supply evaluation result; The heat and power coordinated control module performs power supply balance control on the photovoltaic plant and the thermal power plant according to the heat and power pre-control result and the power supply prediction result.

10. A thermoelectric coordinated control storage medium, characterized in that: The heat-electricity coordinated control storage medium is a computer-readable medium, wherein when the heat-electricity coordinated control storage medium is executed by a processor, the steps of the heat-electricity coordinated control method according to any one of claims 1 to 8 are implemented.

Images