Photovoltaic power generation equipment group regulation and control method and system

Abstract

The invention is suitable for the technical field of photovoltaic power generation, and particularly relates to a photovoltaic power generation equipment group regulation and control method and system, and the method comprises the steps: constructing a photovoltaic power generation prediction model and a power consumption prediction model, constructing training data based on photovoltaic power generation data, historical power consumption data and environment condition data, and carrying out the model training; acquiring real-time environmental conditions, and outputting photovoltaic power generation prediction data and power consumption prediction data; and extracting the total power consumption in each time period, carrying out power generation group combination based on the total power consumption, and carrying out dynamic adjustment on the photovoltaic power generation equipment. According to the method, data analysis is carried out on weather data, illumination conditions and power grid loads, a prediction model is constructed, the power generation amount is calculated based on the prediction model, on the premise that the power generation amount is guaranteed, dynamic regulation and control are carried out on all photovoltaic devices, the photovoltaic devices are divided into fixed power generation devices and dynamic power generation devices, power supply stability is guaranteed, and the energy conversion rate is increased.

Description

Technical Field

[0001] The present invention belongs to the technical field of photovoltaic power generation, and in particular relates to a method and system for controlling a group of photovoltaic power generation equipment. Background Art

[0002] Photovoltaic power generation refers to a technology that uses the photovoltaic effect to directly convert sunlight into electrical energy. The core component of photovoltaic power generation is a solar panel, which is usually composed of multiple solar cells (such as silicon-based photovoltaic cells) connected in series or in parallel. When sunlight shines on a solar cell, the semiconductor material inside it (such as monocrystalline silicon, polycrystalline silicon, amorphous silicon or thin film material) will undergo a photoelectric effect.

[0003] In the current photovoltaic power generation process, all photovoltaic power generation equipment is always working, and the power generation is always at the maximum value. When the electricity demand is at a low point, if the power generation is carried out at the maximum power, the excess electricity needs to be stored in the form of energy storage. The cost of energy storage equipment is very high. If the power generation and power consumption can be basically balanced, the cost will be greatly reduced, and the number of working state switching times of photovoltaic power generation equipment will be reduced, thereby reducing the loss to photovoltaic power generation equipment. Summary of the invention

[0004] The purpose of the present invention is to provide a method for controlling a group of photovoltaic power generation equipment, aiming to solve the problem that the existing technology cannot ensure the balance between power generation and power consumption, thereby increasing the energy storage cost.

[0005] The present invention is implemented as follows: a method for controlling a group of photovoltaic power generation equipment, the method comprising:

[0006] Collecting photovoltaic power generation data, historical power consumption data and environmental condition data, wherein the environmental condition data at least includes temperature data, humidity data and light intensity data;

[0007] Build photovoltaic power generation prediction models and electricity consumption prediction models, build training data based on photovoltaic power generation data, historical electricity consumption data, and environmental condition data, and conduct model training;

[0008] Acquire real-time environmental conditions, and output photovoltaic power generation prediction data and power consumption prediction data based on the trained photovoltaic power generation prediction model and power consumption prediction model, respectively. The photovoltaic power generation prediction data includes power generation data of each photovoltaic power generation group in each time period, and the power consumption prediction data includes power consumption data of each power consumption equipment group in each time period;

[0009] The total electricity consumption in each time period is extracted, the power generation groups are combined based on the total electricity consumption, and the photovoltaic power generation equipment is dynamically adjusted based on the results of the power generation group combination.

[0010] Preferably, the steps of constructing a photovoltaic power generation prediction model and a power consumption prediction model, constructing training data based on photovoltaic power generation data, historical power consumption data and environmental condition data, and performing model training specifically include:

[0011] Constructing photovoltaic power generation prediction model and electricity consumption prediction model, both of which are neural network models;

[0012] Based on the type of training data required by the model, extract corresponding training data from photovoltaic power generation data, historical electricity consumption data, and environmental condition data;

[0013] The training data is divided into a training set and a test set, and the photovoltaic power generation prediction model and the electricity consumption prediction model are trained respectively to obtain the trained photovoltaic power generation prediction model and the electricity consumption prediction model.

[0014] Preferably, the step of acquiring real-time environmental conditions and outputting photovoltaic power generation prediction data and electricity consumption prediction data based on the trained photovoltaic power generation prediction model and electricity consumption prediction model, respectively, specifically includes:

[0015] Based on the location of each photovoltaic power generation equipment and power-consuming equipment, network information query is carried out to obtain real-time environmental conditions;

[0016] Extract power generation influencing factor data from real-time environmental conditions, import them into the photovoltaic power generation prediction model, and generate photovoltaic power generation prediction data;

[0017] The data of factors affecting electricity consumption are extracted from real-time environmental conditions, imported into the electricity consumption prediction model, and the electricity consumption prediction data is generated.

[0018] Preferably, the steps of extracting the total power consumption in each time period, combining power generation groups based on the total power consumption, and dynamically adjusting the photovoltaic power generation equipment based on the result of the power generation group combination specifically include:

[0019] The power consumption forecast data and the photovoltaic power generation forecast data are divided according to a preset time gradient to determine the total power consumption in each time gradient and the corresponding stage power generation data, wherein the stage power generation data records the stage power generation of each photovoltaic power generation device in the time period;

[0020] Combine the stage power generation based on the total power consumption within the time gradient to determine the power generation combination mode in which the total power generation can meet the preset power generation conditions;

[0021] All power generation combinations within each time gradient are globally combined, the results of the global combination are statistically analyzed, the global combination with the least equipment switching times is selected, and the dynamic control plan within each stage is determined.

[0022] Preferably, within the dynamic control scheme, the selected photovoltaic power generation equipment is used as fixed power generation equipment, and the unselected photovoltaic power generation equipment is used as dynamic power generation equipment to monitor the real-time electricity demand. When the real-time electricity demand reaches a preset proportion of the real-time power generation, the dynamic power generation equipment is started to generate electricity.

[0023] Another object of the present invention is to provide a photovoltaic power generation equipment group control system, the system comprising:

[0024] A data acquisition module, used to collect photovoltaic power generation data, historical power consumption data and environmental condition data, wherein the environmental condition data at least includes temperature data, humidity data and light intensity data;

[0025] Model training module, used to build photovoltaic power generation prediction model and electricity consumption prediction model, build training data based on photovoltaic power generation data, historical electricity consumption data and environmental condition data, and perform model training;

[0026] A data prediction module is used to obtain real-time environmental conditions and output photovoltaic power generation prediction data and power consumption prediction data based on the trained photovoltaic power generation prediction model and power consumption prediction model, respectively. The photovoltaic power generation prediction data includes power generation data of each photovoltaic power generation group in each time period, and the power consumption prediction data includes power consumption data of each power consumption equipment group in each time period;

[0027] The dynamic adjustment module is used to extract the total power consumption in each time period, combine the power generation groups based on the total power consumption, and dynamically adjust the photovoltaic power generation equipment based on the results of the power generation group combination.

[0028] Preferably, the model training module includes:

[0029] A model building unit, used to build a photovoltaic power generation prediction model and a power consumption prediction model, both of which are neural network models;

[0030] A training data extraction unit, used to extract corresponding training data from photovoltaic power generation data, historical power consumption data and environmental condition data based on the type of training data required by the model;

[0031] The model testing unit is used to divide the training data into a training set and a test set, train the photovoltaic power generation prediction model and the electricity consumption prediction model respectively, and obtain the trained photovoltaic power generation prediction model and the electricity consumption prediction model.

[0032] Preferably, the data prediction module includes:

[0033] A condition query unit is used to query network information based on the location of each photovoltaic power generation equipment and power consumption equipment to obtain real-time environmental conditions;

[0034] A power generation prediction unit is used to extract power generation influencing factor data from real-time environmental conditions, import the data into a photovoltaic power generation prediction model, and generate photovoltaic power generation prediction data;

[0035] The power consumption prediction unit is used to extract power consumption influencing factor data from real-time environmental conditions, import it into the power consumption prediction model, and generate power consumption prediction data.

[0036] Preferably, the dynamic adjustment module includes:

[0037] The prediction data segmentation unit is used to segment the power consumption prediction data and the photovoltaic power generation prediction data according to a preset time gradient, and determine the total power consumption in each time gradient and the corresponding stage power generation data, wherein the stage power generation data records the stage power generation of each photovoltaic power generation device in the time period;

[0038] An equipment combination unit, used to combine the stage power generation based on the total power consumption within the time gradient, and determine a power generation combination mode in which the total power generation can meet the preset power generation conditions;

[0039] The equipment dispatching unit is used to globally combine all power generation combinations within each time gradient, collect statistics on the results of the global combination, select the global combination with the least number of equipment switching times, and determine the dynamic control plan within each stage.

[0040] Preferably, within the dynamic control scheme, the selected photovoltaic power generation equipment is used as fixed power generation equipment, and the unselected photovoltaic power generation equipment is used as dynamic power generation equipment to monitor the real-time electricity demand. When the real-time electricity demand reaches a preset proportion of the real-time power generation, the dynamic power generation equipment is started to generate electricity.

[0041] The present invention provides a method for controlling a group of photovoltaic power generation equipment, which performs data analysis on weather data, lighting conditions and grid load, constructs a prediction model, calculates the power generation based on the prediction model, and dynamically controls all photovoltaic equipment while ensuring the power generation, dividing them into fixed power generation equipment and dynamic power generation equipment, thereby ensuring stable power supply and improving energy conversion efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 A flow chart of a photovoltaic power generation equipment group control method provided by an embodiment of the present invention;

[0043] Figure 2 A flowchart of the steps of constructing a photovoltaic power generation prediction model and a power consumption prediction model provided by an embodiment of the present invention, constructing training data based on photovoltaic power generation data, historical power consumption data and environmental condition data, and performing model training;

[0044] Figure 3 A flowchart of the steps of obtaining real-time environmental conditions and outputting photovoltaic power generation prediction data and electricity consumption prediction data respectively based on the trained photovoltaic power generation prediction model and electricity consumption prediction model provided by an embodiment of the present invention;

[0045] Figure 4 A flowchart of the steps of extracting the total power consumption in each time period, combining power generation groups based on the total power consumption, and dynamically adjusting the photovoltaic power generation equipment based on the result of the power generation group combination provided by an embodiment of the present invention;

[0046] Figure 5 An architecture diagram of a photovoltaic power generation equipment group control system provided by an embodiment of the present invention;

[0047] Figure 6 An architecture diagram of a model training module provided in an embodiment of the present invention;

[0048] Figure 7 An architecture diagram of a data prediction module provided by an embodiment of the present invention;

[0049] Figure 8 An architectural diagram of a dynamic adjustment module provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0051] It is understood that the terms "first", "second", etc. used in this application may be used herein to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, without departing from the scope of this application, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script.

[0052] like Figure 1 FIG. 1 is a flow chart of a photovoltaic power generation equipment group control method provided by an embodiment of the present invention, the method comprising:

[0053] S100, collecting photovoltaic power generation data, historical power consumption data and environmental condition data, wherein the environmental condition data at least includes temperature data, humidity data and light intensity data.

[0054] In this step, photovoltaic power generation data, historical electricity consumption data and environmental condition data are collected. In the historical power generation process, the power generation conditions of each photovoltaic power generation equipment are recorded, including the power generation power and the power generation amount. The environmental conditions at each moment are also recorded, including the temperature data, humidity data and light intensity data that affect the operation of the photovoltaic power generation equipment. Similarly, the power consumption of electrical equipment under different working conditions is also different. They are recorded to obtain historical electricity consumption data. The photovoltaic power generation data, historical electricity consumption data and environmental condition data all belong to the same time period, such as within a year.

[0055] S200, constructing a photovoltaic power generation prediction model and a power consumption prediction model, constructing training data based on photovoltaic power generation data, historical power consumption data, and environmental condition data, and performing model training.

[0056] In this step, a photovoltaic power generation prediction model and a power consumption prediction model are constructed. The prediction model is constructed and trained using the collected power generation data, power consumption data and environmental condition data. The photovoltaic power generation prediction model can predict the power generation data under the real-time environmental condition data. Similarly, based on the environmental conditions, the power consumption prediction model can also predict the power consumption. The above photovoltaic power generation prediction model and power consumption prediction model both use a neural network model, and a decision tree can also be used. Based on the environmental conditions that affect photovoltaic power generation and the power generation and power consumption under the corresponding conditions, two sets of training data are constructed, which are used to train the photovoltaic power generation prediction model and the power consumption prediction model respectively, and their accuracy is tested. After the test is completed, the photovoltaic power generation prediction model and the power consumption prediction model can be put into use.

[0057] S300, obtaining real-time environmental conditions, and outputting photovoltaic power generation prediction data and power consumption prediction data respectively based on the trained photovoltaic power generation prediction model and power consumption prediction model, wherein the photovoltaic power generation prediction data includes power generation data of each photovoltaic power generation group in each time period, and the power consumption prediction data includes power consumption data of each power consumption equipment group in each time period.

[0058] In this step, real-time environmental conditions are obtained. For photovoltaic power generation, the main factors affecting the power generation are temperature and light intensity. For humidity, its influence is not direct, but indirectly affects the power generation by affecting the dirt layer and insulation on the surface of the photovoltaic panel. When predicting the power generation data of each photovoltaic power generation device, an online query is performed according to the current time and the predicted time to be predicted, and the various environmental conditions within the corresponding time are determined, and imported into the photovoltaic power generation prediction model. By using the photovoltaic power generation prediction model for prediction, the power generation of each photovoltaic power generation device in each time period can be obtained, and the power generation can be determined by integration. Similarly, the environmental factors that affect the power consumption are collected and imported into the power consumption prediction model. Based on the power consumption prediction model, the predicted power consumption at each moment is determined to obtain the power consumption prediction data.

[0059] S400, extracting the total power consumption in each time period, combining power generation groups based on the total power consumption, and dynamically adjusting the photovoltaic power generation equipment based on the result of the power generation group combination.

[0060] In this step, the total power consumption in each time period is extracted, and the power consumption forecast data and the photovoltaic power generation forecast data are divided according to a preset gradient to divide them into data corresponding to multiple time periods. For example, if one hour is a time period, the power consumption forecast data and the photovoltaic power generation forecast data in each hour are generated accordingly. For power consumption and power generation, if the two can maintain a balanced state for a long time, there is no need to spend a lot of cost to add power storage equipment. However, if each photovoltaic power generation equipment is frequently started and stopped for a long time, on the one hand, it will cause unnecessary energy loss, and on the other hand, it will increase the aging speed of the photovoltaic power generation equipment. In order to solve this problem, in the present invention, the power generation equipment in each time period is analyzed, and the photovoltaic power generation equipment is analyzed. Combination, such as running four photovoltaic power generation equipment at the same time, their power generation can just meet the electricity consumption in that time period, but this is only the optimal solution within that time period. When the time is further extended, such as including multiple time periods in a day, then at the critical moment between time periods, there is a high probability that the power generation equipment will be switched to work. Therefore, it is necessary to conduct a unified analysis of the global power generation equipment to minimize the start and stop times of photovoltaic power generation equipment. According to the global analysis results, some photovoltaic power generation equipment will be used as fixed power generation equipment, and other photovoltaic power generation equipment will be used as dynamic power generation equipment. Fixed power generation equipment will work continuously, and dynamic power generation equipment will only work when there is a power consumption gap, which will greatly reduce the start and stop times of photovoltaic power generation equipment and ensure the overall supply and demand balance.

[0061] like Figure 2As shown, as a preferred embodiment of the present invention, the steps of constructing a photovoltaic power generation prediction model and a power consumption prediction model, constructing training data based on photovoltaic power generation data, historical power consumption data and environmental condition data, and performing model training specifically include:

[0062] S201, constructing a photovoltaic power generation prediction model and a power consumption prediction model, both of which are neural network models.

[0063] In this step, a photovoltaic power generation prediction model and a power consumption prediction model are constructed. A multi-layer perceptron (MLP) model is constructed using the Keras library. The model contains an input layer, three hidden layers (the number of neurons in each layer decreases), and an output layer. The hidden layer uses the ReLU activation function, and the output layer uses the linear activation function. The loss function of the model is set to mean square error (suitable for regression problems), the optimizer is Adam (adaptive learning rate optimizer), and the evaluation indicator is specified as mean absolute error (MAE).

[0064] S202, based on the type of training data required by the model, extract corresponding training data from photovoltaic power generation data, historical electricity consumption data and environmental condition data.

[0065] S203, dividing the training data into a training set and a test set, training the photovoltaic power generation prediction model and the electricity consumption prediction model respectively, and obtaining the trained photovoltaic power generation prediction model and the electricity consumption prediction model.

[0066] In this step, the training data is extracted, the batch size (32), the maximum number of training rounds (1000 rounds) and the proportion of the validation set (10%) are defined, the early stopping strategy is enabled (the training is stopped when the validation set loss has not decreased for 50 consecutive rounds, and the optimal model weight is retained) to prevent overfitting, and the model is trained using the training set data. During the training process, the model will update the weights according to the loss function and the optimizer, and at the end of each training round, the evaluation indicators on the validation set are calculated. The trained model is used to predict the training set and the test set respectively to obtain the predicted power generation or power consumption. The mean absolute error (MAE) between the predicted results and the actual power generation on the training set and the test set is calculated as the performance indicator of the model on the training set and the test set. Based on the same method, the training of the power consumption prediction model is completed.

[0067] like Figure 3 As shown, as a preferred embodiment of the present invention, the step of obtaining real-time environmental conditions and outputting photovoltaic power generation prediction data and electricity consumption prediction data based on the trained photovoltaic power generation prediction model and electricity consumption prediction model, respectively, specifically includes:

[0068] S301, perform online information query based on the location of each photovoltaic power generation equipment and power consumption equipment to obtain real-time environmental conditions.

[0069] In this step, the location information of each photovoltaic power generation equipment and power-consuming equipment is obtained, and an online query is performed based on the above location information to query the environmental condition information corresponding to the location, which at least includes environmental condition data within the predicted time, such as temperature data, humidity data and light intensity data.

[0070] S302, extracting power generation influencing factor data from real-time environmental conditions, importing the data into a photovoltaic power generation prediction model, and generating photovoltaic power generation prediction data.

[0071] In this step, data on factors affecting power generation are extracted from real-time environmental conditions. For photovoltaic power generation equipment, temperature, humidity and light intensity all affect power generation. Therefore, it is necessary to extract the corresponding temperature, humidity and light intensity from the real-time environmental conditions. The above data are all continuous data. For example, the temperature, humidity and light intensity at any time within a day are obtained and imported into the photovoltaic power generation prediction model to generate photovoltaic power generation prediction data.

[0072] S303, extracting power consumption influencing factor data from real-time environmental conditions, importing the data into a power consumption prediction model, and generating power consumption prediction data.

[0073] In this step, data on factors affecting electricity consumption are extracted from real-time environmental conditions. Similarly, temperature, humidity and light intensity can be used as factors affecting electricity consumption. The above factors can also be increased or decreased and imported into the electricity consumption prediction model to generate electricity consumption prediction data.

[0074] like Figure 4 As shown, as a preferred embodiment of the present invention, the steps of extracting the total power consumption in each time period, combining power generation groups based on the total power consumption, and dynamically adjusting the photovoltaic power generation equipment based on the result of the power generation group combination specifically include:

[0075] S401, dividing the power consumption forecast data and the photovoltaic power generation forecast data according to the preset time gradient, determining the total power consumption within each time gradient and the corresponding stage power generation data, wherein the stage power generation data records the stage power generation of each photovoltaic power generation equipment within the time period.

[0076] In this embodiment, the electricity consumption forecast data and the photovoltaic power generation forecast data are segmented according to a preset time gradient. The time gradient is determined according to the total forecast time. For example, if the forecast time is 24 hours, the time gradient can be set to ten minutes or thirty minutes. If the forecast time is one week, the time gradient can be set to two hours. The data are segmented according to the time gradient to obtain the total electricity consumption in each time period and the stage power generation in the corresponding time period.

[0077] S402, combining the stage power generation based on the total power consumption within the time gradient, and determining a power generation combination mode in which the total power generation can meet the preset power generation conditions.

[0078] In this embodiment, the stage power generation is combined based on the total power consumption within the time gradient. The total power consumption within a certain time period can be met through a variety of combinations. For example, there are 10 groups of photovoltaic power generation equipment with a total power consumption of 5 units. The power generation of the 10 groups of photovoltaic power generation equipment within the time period is 1 unit. Then any five photovoltaic power generation equipment combined together can meet the power demand within the time period. Then, each photovoltaic power generation equipment is combined so that the total power generation of the combined photovoltaic power generation equipment in the corresponding time period is greater than the power consumption in the time period. If the total power consumption is P, then in order to ensure the stability of power supply, the total power generation of the combined photovoltaic power generation equipment is at least 1.1P, and the power generation combination mode within each time period is determined.

[0079] S403, globally combine all power generation combination modes within each time gradient, count the results of the global combination, select the global combination with the least number of equipment switching times, and determine the dynamic control plan within each stage.

[0080] In this embodiment, all power generation combination modes within each time gradient are globally combined. If there are multiple combinations of photovoltaic power generation equipment within a time period, then there are multiple time periods within the prediction time, and there is a combination relationship between the time periods. In this case, the combination modes within the entire prediction time are globally combined. For example, there are a total of ten time periods within the prediction time, and the numbers of combinations within the ten time periods are 4, 5, 2, 7, 8, 2, 4, 5, 6 and 8 respectively. Then, there are 4*5*2*7*8*2*4*5*6*8 global combinations. In order to reduce the number of device switching times within the prediction time, each global combination is analyzed, and the number of device switching times under each global combination is counted. The definition of device switching is as follows: from the first time period to the next time period In each time period, turning off and on photovoltaic power generation equipment is counted as an independent equipment switch. For example, the photovoltaic power generation equipment turned on in the first time period is A, C and D, and the photovoltaic power generation equipment turned on in the second time period is A, D and E. Then photovoltaic power generation equipment A and D are continuously started, photovoltaic power generation equipment C is turned off, and photovoltaic power generation equipment E is started. The number of equipment switches is 2. The global combination with the least number of equipment switches is selected. Then the photovoltaic power generation equipment working in each time period has been determined. In each time period, the selected photovoltaic power generation equipment is used as a fixed power generation equipment, and the unselected photovoltaic power generation equipment is used as a dynamic power generation equipment. The real-time electricity demand is monitored. When the real-time electricity demand reaches the preset proportion of the real-time power generation, the dynamic power generation equipment is started to generate electricity.

[0081] like Figure 5 As shown, a photovoltaic power generation equipment group control system provided by an embodiment of the present invention includes:

[0082] The data collection module 100 is used to collect photovoltaic power generation data, historical power consumption data and environmental condition data, and the environmental condition data at least includes temperature data, humidity data and light intensity data.

[0083] In this system, the data acquisition module 100 collects photovoltaic power generation data, historical power consumption data and environmental condition data. In the historical power generation process, the power generation conditions of each photovoltaic power generation equipment are recorded, including the power generation power and the power generation amount. The environmental conditions at each moment are also recorded, including the temperature data, humidity data and light intensity data that affect the operation of the photovoltaic power generation equipment. Similarly, the power consumption of electrical equipment under different working conditions is also different. They are recorded to obtain historical power consumption data. The photovoltaic power generation data, historical power consumption data and environmental condition data all belong to the same time period, such as within a year.

[0084] The model training module 200 is used to construct a photovoltaic power generation prediction model and a power consumption prediction model, construct training data based on photovoltaic power generation data, historical power consumption data and environmental condition data, and perform model training.

[0085] In this system, the model training module 200 constructs a photovoltaic power generation prediction model and a power consumption prediction model.

[0086] A prediction model is constructed and trained using the collected power generation data, power consumption data and environmental condition data. The photovoltaic power generation prediction model can predict the power generation data under the real-time environmental condition data. Similarly, based on the environmental conditions, the power consumption prediction model can also predict the power consumption. The above photovoltaic power generation prediction model and power consumption prediction model both use neural network models, and a decision tree can also be used. Based on the environmental conditions that affect photovoltaic power generation and the power generation and power consumption under the corresponding conditions, two sets of training data are constructed, which are used to train the photovoltaic power generation prediction model and the power consumption prediction model respectively, and their accuracy is tested. After the test is completed, the photovoltaic power generation prediction model and the power consumption prediction model can be put into use.

[0087] The data prediction module 300 is used to obtain real-time environmental conditions and output photovoltaic power generation prediction data and power consumption prediction data based on the trained photovoltaic power generation prediction model and power consumption prediction model, respectively. The photovoltaic power generation prediction data includes the power generation data of each photovoltaic power generation group in each time period, and the power consumption prediction data includes the power consumption data of each power consumption equipment group in each time period.

[0088] In this system, the data prediction module 300 obtains real-time environmental conditions. For photovoltaic power generation, the main factors affecting the power generation are temperature and light intensity. For humidity, its influence is not direct, but indirectly affects the power generation by affecting the dirt layer and insulation on the surface of the photovoltaic panel. When predicting the power generation data of each photovoltaic power generation device, an online query is performed according to the current time and the predicted time to be predicted, and the various environmental conditions within the corresponding time are determined, and imported into the photovoltaic power generation prediction model. By using the photovoltaic power generation prediction model for prediction, the power generation of each photovoltaic power generation device in each time period can be obtained, and the power generation can be determined by integration. Similarly, the environmental factors that affect the power consumption are collected and imported into the power consumption prediction model. Based on the power consumption prediction model, the predicted power consumption at each moment is determined to obtain the power consumption prediction data.

[0089] The dynamic adjustment module 400 is used to extract the total power consumption in each time period, combine power generation groups based on the total power consumption, and dynamically adjust the photovoltaic power generation equipment based on the result of the power generation group combination.

[0090] In this system, the dynamic adjustment module 400 extracts the total power consumption in each time period, divides the power consumption forecast data and the photovoltaic power generation forecast data according to a preset gradient, and divides them into data corresponding to multiple time periods. For example, if one hour is a time period, the power consumption forecast data and the photovoltaic power generation forecast data in each hour are generated accordingly. For power consumption and power generation, if the two can maintain a balanced state for a long time, there is no need to spend a lot of cost to add power storage equipment. However, if each photovoltaic power generation equipment is frequently started and stopped for a long time, on the one hand, it will cause unnecessary energy loss, and on the other hand, it will increase the aging speed of the photovoltaic power generation equipment. In order to solve this problem, in the present invention, the power generation equipment in each time period is analyzed, and the photovoltaic power generation equipment is analyzed. If four photovoltaic power generation equipment are combined, for example, if four photovoltaic power generation equipment work at the same time, their power generation can just meet the electricity consumption in that time period, but this is only the optimal solution within that time period. When the time is further extended, such as including multiple time periods in a day, then at the critical moment between the time periods, there is a high probability that the power generation equipment will be switched to work. Therefore, it is necessary to conduct a unified analysis of the global power generation equipment to minimize the start and stop times of photovoltaic power generation equipment. According to the global analysis results, some photovoltaic power generation equipment will be used as fixed power generation equipment, and other photovoltaic power generation equipment will be used as dynamic power generation equipment. Fixed power generation equipment will work continuously, and dynamic power generation equipment will only work when there is a power consumption gap, which will greatly reduce the start and stop times of photovoltaic power generation equipment and ensure the overall supply and demand balance.

[0091] like Figure 6 As shown, as a preferred embodiment of the present invention, the model training module 200 includes:

[0092] The model building unit 201 is used to build a photovoltaic power generation prediction model and a power consumption prediction model. Both the photovoltaic power generation prediction model and the power consumption prediction model are neural network models.

[0093] In this embodiment, the model construction unit 201 constructs a photovoltaic power generation prediction model and a power consumption prediction model, and uses the Keras library to construct a multi-layer perceptron (MLP) model. The model includes an input layer, three hidden layers (the number of neurons in each layer decreases) and an output layer. The hidden layer uses a ReLU activation function, and the output layer uses a linear activation function. The loss function of the model is set to mean square error (suitable for regression problems), the optimizer is Adam (adaptive learning rate optimizer), and the evaluation indicator is specified as mean absolute error (MAE).

[0094] The training data extraction unit 202 is used to extract corresponding training data from the photovoltaic power generation data, the historical power consumption data and the environmental condition data based on the training data type required by the model.

[0095] The model testing unit 203 is used to divide the training data into a training set and a test set, and train the photovoltaic power generation prediction model and the power consumption prediction model respectively to obtain the trained photovoltaic power generation prediction model and the power consumption prediction model.

[0096] In this embodiment, training data is extracted, batch size (32), maximum number of training rounds (1000 rounds) and validation set ratio (10%) are defined, early stopping strategy is enabled (training is stopped when validation set loss does not decrease for 50 consecutive rounds, and optimal model weights are retained) to prevent overfitting, and the model is trained using training set data. During the training process, the model updates weights according to the loss function and optimizer, and at the end of each training round, the evaluation index on the validation set is calculated. The trained model is used to predict the training set and test set respectively to obtain the predicted power generation or power consumption, and the mean absolute error (MAE) between the predicted results and the actual power generation on the training set and test set is calculated as the performance index of the model on the training set and test set. Based on the same method, the training of the power consumption prediction model is completed.

[0097] like Figure 7 As shown, as a preferred embodiment of the present invention, the data prediction module 300 includes:

[0098] The condition query unit 301 is used to query the network information based on the location of each photovoltaic power generation equipment and power consumption equipment to obtain real-time environmental conditions.

[0099] In this step, the conditional query unit 301 obtains the location information of each photovoltaic power generation equipment and power-consuming equipment, and performs an online query based on the above location information to query the environmental condition information corresponding to the location, which at least includes environmental condition data within the predicted time, such as temperature data, humidity data and light intensity data.

[0100] The power generation prediction unit 302 is used to extract power generation influencing factor data from real-time environmental conditions, import the data into the photovoltaic power generation prediction model, and generate photovoltaic power generation prediction data.

[0101] In this step, the power generation prediction unit 302 extracts power generation influencing factor data from the real-time environmental conditions. For photovoltaic power generation equipment, temperature, humidity and light intensity all affect power generation. Therefore, it is necessary to extract the corresponding temperature, humidity and light intensity from the real-time environmental conditions. The above data are all continuous data. For example, the temperature, humidity and light intensity at any time within a day are obtained and imported into the photovoltaic power generation prediction model to generate photovoltaic power generation prediction data.

[0102] The power consumption prediction unit 303 is used to extract power consumption influencing factor data from real-time environmental conditions, import the data into the power consumption prediction model, and generate power consumption prediction data.

[0103] In this step, the power consumption prediction unit 303 extracts power consumption influencing factor data from the real-time environmental conditions. Similarly, temperature, humidity and light intensity can be used as influencing factors affecting power consumption. The above-mentioned influencing factors can also be increased or decreased and imported into the power consumption prediction model to generate power consumption prediction data.

[0104] like Figure 8 As shown, as a preferred embodiment of the present invention, the dynamic adjustment module 400 includes:

[0105] The prediction data segmentation unit 401 is used to segment the power consumption prediction data and the photovoltaic power generation prediction data according to a preset time gradient, determine the total power consumption within each time gradient and the corresponding stage power generation data, and the stage power generation data records the stage power generation of each photovoltaic power generation equipment within the time period.

[0106] In this module, the prediction data segmentation unit 401 segments the electricity consumption prediction data and the photovoltaic power generation prediction data according to a preset time gradient. The time gradient is determined according to the total prediction time. For example, if the prediction time is 24 hours, the time gradient can be set to ten minutes or thirty minutes. If the prediction time is one week, the time gradient can be set to two hours. The data is segmented according to the time gradient to obtain the total electricity consumption in each time period and the stage power generation in the corresponding time period.

[0107] The equipment combination unit 402 is used to combine the stage power generation based on the total power consumption within the time gradient, and determine the power generation combination mode in which the total power generation can meet the preset power generation conditions.

[0108] In this module, the equipment combination unit 402 combines the stage power generation based on the total power consumption within the time gradient. The total power consumption within a certain time period can be met through a variety of combinations. For example, there are 10 groups of photovoltaic power generation equipment with a total power consumption of 5 units. The power generation of the 10 groups of photovoltaic power generation equipment within the time period is 1 unit. Then any five photovoltaic power generation equipment combined together can meet the power demand within the time period. Then, each photovoltaic power generation equipment is combined so that the total power generation of the combined photovoltaic power generation equipment within the corresponding time period is greater than the power consumption within the time period. If the total power consumption is P, then in order to ensure the stability of power supply, the total power generation of the combined photovoltaic power generation equipment is at least 1.1P, and the power generation combination mode within each time period is determined.

[0109] The equipment scheduling unit 403 is used to globally combine all power generation combinations within each time gradient, collect statistics on the results of the global combination, select the global combination with the least number of equipment switching times, and determine the dynamic control plan within each stage.

[0110] In this module, the equipment scheduling unit 403 globally combines all power generation combinations within each time gradient. If there are multiple combinations of photovoltaic power generation equipment within a time period, then there are multiple time periods within the prediction time, and there is a combination relationship between the time periods. Then the combination methods within the entire prediction time are globally combined. For example, there are a total of ten time periods within the prediction time, and the numbers of combinations within the ten time periods are 4, 5, 2, 7, 8, 2, 4, 5, 6 and 8 respectively. Then there are 4*5*2*7*8*2*4*5*6*8 global combinations. In order to reduce the number of equipment switching times within the prediction time, each global combination is analyzed and the number of equipment switching times under each global combination is counted. The definition of equipment switching is as follows: from the first time period to the next In a time period, turning off and on photovoltaic power generation equipment is counted as an independent equipment switch. For example, the photovoltaic power generation equipment turned on in the first time period are A, C and D, and the photovoltaic power generation equipment turned on in the second time period are A, D and E. Then A and D photovoltaic power generation equipment are continuously started, C photovoltaic power generation equipment is turned off, and E photovoltaic power generation equipment is started. The number of equipment switches is 2. The global combination with the least number of equipment switches is selected. Then the photovoltaic power generation equipment working in each time period is determined. In each time period, the selected photovoltaic power generation equipment is used as a fixed power generation equipment, and the unselected photovoltaic power generation equipment is used as a dynamic power generation equipment. The real-time electricity demand is monitored. When the real-time electricity demand reaches the preset proportion of the real-time power generation, the dynamic power generation equipment is started to generate electricity.

[0111] It should be understood that, although each step in the flow chart of each embodiment of the present invention is shown in sequence according to the indication of the arrow, these steps are not necessarily performed in sequence according to the order indicated by the arrow. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0112] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0113] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0114] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

[0115] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for controlling a group of photovoltaic power generation equipment, characterized in that: The method comprises: Collecting photovoltaic power generation data, historical power consumption data and environmental condition data, wherein the environmental condition data at least includes temperature data, humidity data and light intensity data; Build photovoltaic power generation prediction models and electricity consumption prediction models, build training data based on photovoltaic power generation data, historical electricity consumption data, and environmental condition data, and conduct model training; Acquire real-time environmental conditions, and output photovoltaic power generation prediction data and power consumption prediction data based on the trained photovoltaic power generation prediction model and power consumption prediction model, respectively. The photovoltaic power generation prediction data includes power generation data of each photovoltaic power generation group in each time period, and the power consumption prediction data includes power consumption data of each power consumption equipment group in each time period; The total electricity consumption in each time period is extracted, the power generation groups are combined based on the total electricity consumption, and the photovoltaic power generation equipment is dynamically adjusted based on the results of the power generation group combination.

2. The photovoltaic power generation equipment group control method according to claim 1, characterized in that: The steps of constructing a photovoltaic power generation prediction model and a power consumption prediction model, constructing training data based on photovoltaic power generation data, historical power consumption data and environmental condition data, and performing model training specifically include: Constructing photovoltaic power generation prediction model and electricity consumption prediction model, both of which are neural network models; Based on the type of training data required by the model, extract corresponding training data from photovoltaic power generation data, historical electricity consumption data, and environmental condition data; The training data is divided into a training set and a test set, and the photovoltaic power generation prediction model and the electricity consumption prediction model are trained respectively to obtain the trained photovoltaic power generation prediction model and the electricity consumption prediction model.

3. The photovoltaic power generation equipment group control method according to claim 1, characterized in that: The step of obtaining real-time environmental conditions and outputting photovoltaic power generation prediction data and electricity consumption prediction data based on the trained photovoltaic power generation prediction model and electricity consumption prediction model respectively specifically includes: Based on the location of each photovoltaic power generation equipment and power-consuming equipment, network information query is carried out to obtain real-time environmental conditions; Extract power generation influencing factor data from real-time environmental conditions, import them into the photovoltaic power generation prediction model, and generate photovoltaic power generation prediction data; The data of factors affecting electricity consumption are extracted from real-time environmental conditions, imported into the electricity consumption prediction model, and the electricity consumption prediction data is generated.

4. The photovoltaic power generation equipment group control method according to claim 1, characterized in that: The steps of extracting the total power consumption in each time period, combining power generation groups based on the total power consumption, and dynamically adjusting the photovoltaic power generation equipment based on the result of the power generation group combination specifically include: The power consumption forecast data and the photovoltaic power generation forecast data are divided according to a preset time gradient to determine the total power consumption in each time gradient and the corresponding stage power generation data, wherein the stage power generation data records the stage power generation of each photovoltaic power generation device in the time period; Combine the stage power generation based on the total power consumption within the time gradient to determine the power generation combination mode in which the total power generation can meet the preset power generation conditions; All power generation combinations within each time gradient are globally combined, the results of the global combination are statistically analyzed, the global combination with the least number of equipment switching times is selected, and the dynamic control plan within each stage is determined.

5. The photovoltaic power generation equipment group control method according to claim 4, characterized in that: In the dynamic control plan, the selected photovoltaic power generation equipment is used as fixed power generation equipment, and the unselected photovoltaic power generation equipment is used as dynamic power generation equipment to monitor the real-time electricity demand. When the real-time electricity demand reaches a preset proportion of the real-time power generation, the dynamic power generation equipment is started to generate electricity.

6. A photovoltaic power generation equipment group control system, characterized in that: The system comprises: A data acquisition module, used to collect photovoltaic power generation data, historical power consumption data and environmental condition data, wherein the environmental condition data at least includes temperature data, humidity data and light intensity data; Model training module, used to build photovoltaic power generation prediction model and electricity consumption prediction model, build training data based on photovoltaic power generation data, historical electricity consumption data and environmental condition data, and perform model training; A data prediction module is used to obtain real-time environmental conditions and output photovoltaic power generation prediction data and power consumption prediction data based on the trained photovoltaic power generation prediction model and power consumption prediction model, respectively. The photovoltaic power generation prediction data includes power generation data of each photovoltaic power generation group in each time period, and the power consumption prediction data includes power consumption data of each power consumption equipment group in each time period; The dynamic adjustment module is used to extract the total power consumption in each time period, combine the power generation groups based on the total power consumption, and dynamically adjust the photovoltaic power generation equipment based on the results of the power generation group combination.

7. The photovoltaic power generation equipment group control system according to claim 6, characterized in that: The model training module includes: A model building unit, used to build a photovoltaic power generation prediction model and a power consumption prediction model, both of which are neural network models; A training data extraction unit, used to extract corresponding training data from photovoltaic power generation data, historical power consumption data and environmental condition data based on the type of training data required by the model; The model testing unit is used to divide the training data into a training set and a test set, train the photovoltaic power generation prediction model and the electricity consumption prediction model respectively, and obtain the trained photovoltaic power generation prediction model and the electricity consumption prediction model.

8. The photovoltaic power generation equipment group control system according to claim 6, characterized in that: The data prediction module comprises: A condition query unit is used to query network information based on the location of each photovoltaic power generation equipment and power consumption equipment to obtain real-time environmental conditions; A power generation prediction unit is used to extract power generation influencing factor data from real-time environmental conditions, import the data into a photovoltaic power generation prediction model, and generate photovoltaic power generation prediction data; The power consumption prediction unit is used to extract power consumption influencing factor data from real-time environmental conditions, import it into the power consumption prediction model, and generate power consumption prediction data.

9. The photovoltaic power generation equipment group control system according to claim 6, characterized in that: The dynamic adjustment module includes: The prediction data segmentation unit is used to segment the power consumption prediction data and the photovoltaic power generation prediction data according to a preset time gradient, and determine the total power consumption in each time gradient and the corresponding stage power generation data, wherein the stage power generation data records the stage power generation of each photovoltaic power generation device in the time period; An equipment combination unit, used to combine the stage power generation based on the total power consumption within the time gradient, and determine a power generation combination mode in which the total power generation can meet the preset power generation conditions; The equipment dispatching unit is used to globally combine all power generation combinations within each time gradient, collect statistics on the results of the global combination, select the global combination with the least number of equipment switching times, and determine the dynamic control plan within each stage.

10. The photovoltaic power generation equipment group control system according to claim 9, characterized in that: In the dynamic control plan, the selected photovoltaic power generation equipment is used as fixed power generation equipment, and the unselected photovoltaic power generation equipment is used as dynamic power generation equipment to monitor the real-time electricity demand. When the real-time electricity demand reaches a preset proportion of the real-time power generation, the dynamic power generation equipment is started to generate electricity.

Images