Control method, control device and equipment of power system and storage medium

Abstract

The invention discloses a control method of a power system, a control device of the power system, control equipment of the power system, computer equipment and a computer storage medium, and the method comprises the steps: monitoring the environment quality information and operation information of the power system, and predicting a power generation state of the distributed energy equipment and a power utilization state of the power utilization side equipment according to the operation information, determining an operation target according to a first prediction result of the power generation state and a second prediction result of the power utilization state, and controlling power generation of the distributed energy equipment and power utilization of the power utilization side equipment according to the environment quality information, the operation information and the operation target. Therefore, the operation target of the electric power system can be matched with the power utilization state and the power generation state, the environment quality of the position where the electric power system is located can be considered when the electric power system operates, the influence of the operation environment quality of the electric power system is reduced, and the wind and light abandoning phenomenon is avoided.

Description

Technical Field

[0001] The present application relates to the technical field of power systems, and in particular to a control method for a power system, a control device for a power system, a control equipment for a power system, a computer equipment and a computer storage medium. Background Art

[0002] With the increase in energy demand and the improvement of environmental protection awareness, the use of new energy such as wind energy and solar energy has gradually become a hot project. However, due to the intermittent and volatile nature of clean energy, there are many challenges in the access and consumption of new energy in the power grid, resulting in low energy utilization efficiency. In addition, the use of new energy requires the installation of corresponding power generation facilities, and the installation and operation and maintenance of power generation facilities will affect the local ecological environment, ultimately leading to serious wind and solar power abandonment. Summary of the invention

[0003] The present application provides a control method for an electric power system, a control device for an electric power system, a control equipment for an electric power system, a computer equipment and a computer storage medium.

[0004] The present application provides a control method for a power system, wherein the power system includes a distributed energy device and a power consumption device, and the method includes:

[0005] Monitoring environmental quality information and operation information of the power system, wherein the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is used to indicate the status of the distributed energy device and the status of the power consumption side device;

[0006] Predicting the power generation status of the distributed energy device and the power consumption status of the power consumption side device according to the operation information;

[0007] Determining an operation target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state;

[0008] The power generation of the distributed energy device and the power consumption of the power-consuming device are controlled according to the environmental quality information, the operating information and the operating target.

[0009] In the control method of the power system provided in the embodiment of the present application, the environmental quality information and operation information of the power system can be monitored, and the power generation status of the distributed energy equipment and the power consumption status of the power-consuming side equipment can be predicted based on the operation information, and the operation target of the power system can be determined based on the first prediction result of the power generation status and the second prediction result of the power consumption status, and then the power generation of the distributed energy equipment and the power consumption of the power-consuming side equipment can be controlled based on the environmental quality information, the operation information and the operation target.

[0010] Thus, in the implementation mode of the present application, the power generation state of the distributed energy equipment and the power consumption state of the power consumption side equipment can be predicted according to the operation information, and the operation target of the power system can be determined based on the power generation state prediction result and the power consumption state prediction result, so that the operation target of the power system can be matched with the power consumption state of the distributed energy equipment and the power generation state of the power generation side equipment to a certain extent, ensuring the reliable and long-term operation of the power system, thereby avoiding the occurrence of wind and solar power abandonment to a certain extent. In addition, the implementation mode of the present application can control the operation of the distributed energy equipment and the power consumption side equipment according to the environmental quality information, operation information and the operation target, so that when the power system operates according to the operation target, the environmental quality of its own location can be taken into account, thereby reducing the impact of the power system operation on the environmental quality of the location of the system to a certain extent, thereby further avoiding the occurrence of wind and solar power abandonment.

[0011] In certain embodiments of the present application, the operation information includes multiple types, each type of the operation information corresponds to a preset time scale, and the power generation state of the distributed energy device and the power consumption state of the power consumption side device are predicted according to the operation information, including:

[0012] According to the operation information corresponding to the preset time scale, predicting the power generation state and the power consumption state corresponding to the preset time scale;

[0013] Determining the operation target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state includes:

[0014] The operation target corresponding to the preset time scale is determined according to the first prediction result and the second prediction result corresponding to the preset time scale.

[0015] In this way, in the implementation mode of the present application, the power system can predict the power generation status and power consumption status corresponding to each time scale based on the operating information of multiple time scales, thereby determining the operating target corresponding to each time scale, and then can control the distributed energy equipment and the power consumption side equipment according to the operating target corresponding to each time scale, thereby realizing the reasonable control of the distributed energy equipment and the power consumption side equipment to a certain extent, and ensuring the long-term operation of the power system to a certain extent.

[0016] In certain embodiments of the present application, predicting the power generation state and the power consumption state corresponding to the preset time scale according to the operation information corresponding to the preset time scale includes:

[0017] The power generation state and the power consumption state corresponding to the preset time scale are predicted based on a pre-trained neural network model corresponding to the preset time scale and the operation information corresponding to the preset time scale.

[0018] Thus, in the implementation mode of the present application, the power generation state and the power consumption state can be predicted based on the pre-trained neural network model, so the credibility of the prediction results of the power generation state and the power consumption state can be guaranteed.

[0019] In certain embodiments of the present application, the preset time scale includes a minute-level time scale, an hour-level time scale, and a day-level time scale.

[0020] In certain embodiments of the present application, the neural network model includes a long short-term memory network model corresponding to the minute time scale, a gated recurrent unit network model corresponding to the hour time scale, and a transformer model corresponding to the day time scale.

[0021] In certain embodiments of the present application, the operation target includes a first target, the first prediction result includes a power generation prediction, and the second prediction result includes a power consumption prediction, and determining the operation target corresponding to the preset time scale according to the first prediction result and the second prediction result corresponding to the preset time scale includes:

[0022] Based on the power generation forecast and the power consumption forecast corresponding to the minute time scale, the power generation expectation of the distributed energy device is determined to obtain the first target.

[0023] Thus, in the implementation mode of the present application, the expected power generation of the distributed energy equipment can be determined through power generation prediction and power consumption prediction, thereby obtaining the first goal.

[0024] In certain embodiments of the present application, the distributed energy device includes an electric energy storage module, the operation target includes a second target, and determining the operation target corresponding to the preset time scale according to the first prediction result and the second prediction result corresponding to the preset time scale includes:

[0025] According to the power generation forecast and the power consumption forecast corresponding to the hourly time scale, the expected charging amount and the expected discharging amount of the energy storage module are determined to obtain the second target.

[0026] Thus, in the implementation mode of the present application, the expected charging amount and the expected discharging amount of the energy storage module can be determined by predicting the power generation amount and the power consumption amount, thereby obtaining the second target.

[0027] In certain embodiments of the present application, the operation target includes a third target, and determining the operation target corresponding to the preset time scale according to the first prediction result and the second prediction result corresponding to the preset time scale includes:

[0028] Based on the power generation forecast and the power consumption forecast corresponding to the daily time scale, the expected operating state of the power system is determined to obtain the third goal.

[0029] Thus, in the implementation mode of the present application, the expected working state of the power system can be determined through power generation forecast and power consumption forecast, thereby obtaining the third goal.

[0030] In certain embodiments of the present application, controlling the power generation of the distributed energy device and the power consumption of the power-consuming device according to the environmental quality information, the operating information and the operating target includes:

[0031] constructing a state vector according to the environmental quality information and the operation information;

[0032] constructing a reward function based on the operational objectives and the predetermined environmental quality objectives;

[0033] Determine, according to the state vector, the reward function and the pre-trained agent model, a first action parameter corresponding to the distributed energy device and a second action parameter of the power-side device;

[0034] The power generation of the distributed energy device is controlled according to the first action parameter, and the power consumption of the power-consuming device is controlled according to the second action parameter.

[0035] Thus, in the implementation mode of the present application, a state vector can be constructed based on environmental quality information and operating information, and a reward function can be constructed based on operating objectives and predetermined environmental quality objectives, and based on the state vector, the reward function and the pre-trained intelligent agent model, the first action parameters corresponding to the distributed energy equipment and the second action parameters of the power-consuming side equipment can be determined, thereby controlling the operation of the distributed energy equipment through the first action parameters and controlling the operation of the power-consuming side equipment through the second action parameters.

[0036] In certain embodiments of the present application, the environmental quality target includes at least one of greenhouse gas emission reduction, air quality improvement, and comprehensive environmental benefit assessment parameters.

[0037] In certain embodiments of the present application, the distributed energy device includes a plurality of power station modules, the operation information includes module operation sub-information of each of the power station modules, the operation target includes module operation target of each of the power station modules, and the power generation of the distributed energy device and the power consumption of the power consumption side device are controlled according to the environmental quality information, the operation information and the operation target, including:

[0038] The power generation of the power plant module and the power consumption of the power consumption side equipment are controlled according to the environmental quality information, the module operation sub-information of the power plant module and the module operation target.

[0039] Thus, in the embodiment of the present application, the power plant modules and the power consumption side equipment can be controlled according to the environmental quality information, as well as the module operation sub-information and module operation target of the power plant modules.

[0040] In certain embodiments of the present application, the operation target includes the total operation target of the plurality of power station modules, and the controlling the power generation of the distributed energy device and the power consumption of the power-consuming device according to the environmental quality information, the operation information and the operation target includes:

[0041] The power generation and the power consumption of the power-consuming side equipment are controlled according to the environmental quality information, the overall operation target and the module operation sub-information of each power plant module.

[0042] Thus, in the embodiment of the present application, multiple power plant modules and power-consuming side equipment can be controlled according to environmental quality information, the total operating target corresponding to multiple power plant modules, and the module operating sub-information of each power plant module.

[0043] The present application provides a control device for an electric power system, wherein the electric power system includes a distributed energy device and a power consumption side device, and the device includes:

[0044] A monitoring module configured to monitor environmental quality information and operation information of the power system, wherein the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is configured to indicate the status of the distributed energy device and the status of the power-side device;

[0045] A prediction module, configured to predict the power generation state of the distributed energy device and the power consumption state of the power consumption side device according to the operation information;

[0046] a determination module configured to determine an operation target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state;

[0047] The control module is configured to control the power generation of the distributed energy device and the power consumption of the power-consuming device according to the environmental quality information, the operating information and the operating target.

[0048] The present application embodiment provides a control device for an electric power system, wherein the electric power system includes a distributed energy device and a power consumption side device, and the device includes:

[0049] An Internet of Things and edge computing real-time monitoring module, configured to monitor environmental quality information and operation information of the power system, wherein the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is used to indicate the status of the distributed energy device and the status of the power-side device;

[0050] A multi-level prediction module, configured to predict the power generation state of the distributed energy device and the power consumption state of the power consumption side device according to the operation information;

[0051] The distributed artificial intelligence coordination optimization control module is configured to determine the operating target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state, and to determine the operating target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state, and to control the power generation of the distributed energy equipment and the power consumption of the power consumption side equipment according to the environmental quality information, the operating information and the operating target.

[0052] An embodiment of the present application provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the control method of the power system described above is implemented.

[0053] An embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program. When the computer program is executed by one or more processors, the control method of the power system described above is implemented.

[0054] The control device of the power system, the control equipment of the power system, the computer equipment and the computer-readable storage medium provided in the embodiments of the present application can predict the power generation state of the distributed energy equipment and the power consumption state of the power consumption side equipment according to the operation information, and determine the operation target of the power system based on the power generation state prediction result and the power consumption state prediction result, so that the operation target of the power system can be matched with the power consumption state of the distributed energy equipment and the power generation state of the power generation side equipment to a certain extent, ensuring the reliable and long-term operation of the power system, thereby avoiding the occurrence of wind and solar power abandonment to a certain extent. In addition, the embodiments of the present application can control the operation of the distributed energy equipment and the power consumption side equipment according to the environmental quality information, the operation information and the operation target, so that when the power system operates at the operation target, the environmental quality of its own location can be taken into account, thereby reducing the impact of the power system operation on the environmental quality of the location of the system to a certain extent, thereby further avoiding the occurrence of wind and solar power abandonment.

[0055] Additional aspects and advantages of the embodiments of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the embodiments of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0057] Figure 1 A schematic flow chart of a control method for a power system in certain embodiments of the present application;

[0058] Figure 2 A schematic diagram of a control device for a power system in certain embodiments of the present application;

[0059] Figure 3 A schematic diagram of a control device for a power system in certain embodiments of the present application;

[0060] Figure 4 A schematic flow chart of a control method for a power system in certain embodiments of the present application;

[0061] Figure 5 A schematic flow chart of a control method for a power system in certain embodiments of the present application;

[0062] Figure 6 It is a flowchart of a control method of a power system in certain embodiments of the present application. DETAILED DESCRIPTION

[0063] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions from beginning to end. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the embodiments of the present application, and cannot be understood as limiting the embodiments of the present application.

[0064] With the continuous increase in energy demand and the rapid development of new energy technologies, distributed power generation facilities based on new energy such as solar energy and wind energy have gradually become a current research hotspot.

[0065] Among them, the distributed generation facilities corresponding to solar energy can be understood as photovoltaic power generation facilities built near the user's residence, with the operation mode of self-generation and self-use on the user side, surplus electricity connected to the grid, and balanced regulation in the distribution system. The distributed generation facilities corresponding to wind energy can be understood as distributed generation systems that use wind turbines as distributed power sources to convert wind energy into electrical energy.

[0066] However, new energy sources such as solar energy and wind energy are intermittent and volatile, which makes their coordination with traditional power grids complicated. In addition, the existing power system mainly relies on centralized power production and distribution, which makes it difficult to fully utilize distributed new energy sources, resulting in low utilization efficiency of new energy, which in turn causes serious wind and solar power abandonment.

[0067] Based on the above problems that may be encountered, please refer to Figure 1 The present application provides a control method for a power system, wherein the power system includes a distributed energy device and a power consumption device, and the method includes:

[0068] 01: Monitor the environmental quality information and operation information of the power system, wherein the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is used to indicate the status of distributed energy equipment and the status of power consumption side equipment;

[0069] 02: Based on the operation information, predict the power generation status of distributed energy equipment and the power consumption status of power-consuming equipment;

[0070] 03: Determine the operation target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state;

[0071] 04: Control the power generation of distributed energy equipment and the power consumption of power-consuming equipment based on environmental quality information, operation information and operation targets.

[0072] See also Figure 2, the embodiment of the present application provides a control device 200 of an electric power system, and the electric power system includes distributed energy equipment and power-side equipment. The control method of the embodiment of the present application can be implemented by the control device 200 of the embodiment of the present application. Specifically, the control device 200 includes a monitoring module 210, a prediction module 220, a determination module 230 and a control module 240. Among them, the monitoring module 210 is configured to monitor the environmental quality information and operation information of the power system, and the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is used to indicate the state of the distributed energy equipment and the state of the power-side equipment. The prediction module 220 is configured to predict the power generation state of the distributed energy equipment and the power consumption state of the power-side equipment according to the operation information. The determination module 230 is configured to determine the operation target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state. The control module 240 is configured to control the power generation of the distributed energy equipment and the power consumption of the power-side equipment according to the environmental quality information, the operation information and the operation target.

[0073] See also Figure 3 , the embodiment of the present application provides a control device 300 of an electric power system, and the electric power system includes distributed energy equipment and power-consuming side equipment. The control method of the embodiment of the present application can be implemented by the control device 300 of the embodiment of the present application. Specifically, the control device 300 includes an Internet of Things and edge computing real-time monitoring module 310, a multi-level prediction module 320, and a distributed artificial intelligence coordination optimization control module 330. Among them, the Internet of Things and edge computing real-time monitoring module 310 is configured to monitor the environmental quality information and operation information of the power system, and the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is used to indicate the status of the distributed energy equipment and the status of the power-consuming side equipment. The multi-level prediction module 320 is configured to predict the power generation status of the distributed energy equipment and the power consumption status of the power-consuming side equipment based on the operation information. The distributed artificial intelligence coordination optimization control module 330 is configured to determine the operating objectives of the power system based on the first prediction result of the power generation state and the second prediction result of the power consumption state, and to determine the operating objectives of the power system based on the first prediction result of the power generation state and the second prediction result of the power consumption state, and to control the power generation of distributed energy equipment and the power consumption of power-consuming equipment based on environmental quality information, operating information and operating objectives.

[0074] The embodiment of the present application also provides a computer device, which includes a memory and a processor. The location determination method of the computer device of the embodiment of the present application can be implemented by the computer device of the embodiment of the present application. Specifically, a computer program is stored in the memory, and the processor is used to monitor the environmental quality information and operation information of the power system, and is used to predict the power generation status of the distributed energy equipment and the power consumption status of the power-consuming side equipment according to the operation information, and is used to determine the operation target of the power system according to the first prediction result of the power generation status and the second prediction result of the power consumption status, and is used to control the power generation of the distributed energy equipment and the power consumption of the power-consuming side equipment according to the environmental quality information, the operation information and the operation target. The environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is used to indicate the status of the distributed energy equipment and the status of the power-consuming side equipment.

[0075] Specifically, in the implementation mode of the present application, the power system can obtain environmental quality information and operating information, and based on the operating information, predict the power generation status of the distributed energy equipment in the future and the power consumption status of the power-consuming equipment in the future, and determine the operating target based on the power generation status prediction results and the power consumption status prediction results, and then control the power generation of the distributed energy equipment and the power consumption of the power-consuming equipment based on the environmental quality information, the operating information and the operating target, so as to make the power system approach the operating target while taking into account the environmental quality of the location.

[0076] Thus, in the implementation mode of the present application, the power generation state of the distributed energy equipment and the power consumption state of the power consumption side equipment can be predicted according to the operation information, and the operation target of the power system can be determined based on the power generation state prediction result and the power consumption state prediction result, so that the operation target of the power system can be matched with the power consumption state of the distributed energy equipment and the power generation state of the power generation side equipment to a certain extent, ensuring the reliable and long-term operation of the power system, thereby avoiding the occurrence of wind and solar power abandonment to a certain extent. In addition, the implementation mode of the present application can control the operation of the distributed energy equipment and the power consumption side equipment according to the environmental quality information, operation information and the operation target, so that when the power system operates according to the operation target, the environmental quality of its own location can be taken into account, thereby reducing the impact of the power system operation on the environmental quality of the location of the system to a certain extent, thereby further avoiding the occurrence of wind and solar power abandonment.

[0077] It is understandable that in the implementation mode of the present application, the power generation side of the power system can be set according to the actual situation. For example, in one example, the power generation side of the power system is composed of traditional power generation equipment and the above-mentioned distributed energy equipment. In one example, traditional power generation equipment may include thermal power stations and hydropower stations, and new energy power generation facilities may include solar panels, wind turbines, and battery energy storage systems.

[0078] In another example, the distributed energy equipment includes solar panels, battery energy storage systems connected to the solar panels, wind turbines, battery energy storage systems connected to the wind turbines, on-site facilities of thermal power stations (such as boilers and steam turbines), on-site facilities of hydroelectric power stations (such as main valves), etc.

[0079] It can also be understood that in the implementation manner of the present application, the power-consuming side equipment of the power system can be understood as residential power equipment, commercial power equipment, industrial power equipment, etc., and can be specifically set according to actual conditions.

[0080] It should be noted that in the implementation of the present application, the operation information of the power system can be set according to the actual situation. For example, in one example, the operation information of the power system includes power generation side information and power consumption side information, wherein the power generation side information includes the real-time power generation power, voltage, current, etc. of the distributed energy equipment, and the power consumption side information includes the voltage, current and power consumption of the power consumption side equipment.

[0081] It should also be noted that in the implementation mode of the present application, the environmental quality information of the power system can also be set according to actual conditions. For example, in one example, the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data. Among them, the air quality monitoring data includes PM2.5, PM10 concentration, etc., the greenhouse gas emission data may include carbon dioxide emissions, methane emissions, etc., and the meteorological data may include temperature, humidity, wind speed, etc.

[0082] It can be understood that in the embodiments of the present application, the power system can determine the operating target based on the operating information, or in other words, can determine the operating target for a period of time in the future based on the operating information of a period of time in the past.

[0083] For instance, in one example, the power system can predict the power generation status (or power generation) of the distributed energy equipment in July this year based on the power generation of the distributed energy equipment in July last year and the power consumption of the power-consuming equipment in July last year (i.e., operating information), thereby obtaining the operating target for July this year.

[0084] In another example, the power system can predict the power generation demand of the distributed energy equipment in the next 10 minutes based on the current power generation of the distributed energy equipment and the current power generation of the power-consuming equipment (i.e., operating information) as the power generation status prediction result, thereby obtaining the operating target for the next 10 minutes.

[0085] It can also be understood that in the embodiments of the present application, the power system can control distributed energy equipment and power-consuming side equipment based on operating information, environmental quality information and operating objectives when the operating objectives are obtained.

[0086] For example, let the first prediction result include the power generation in the next few minutes to hours, and the second prediction result include the power consumption in the next few minutes to hours, then:

[0087] If it is determined based on the first prediction result and the second prediction result that the power generation in the next few minutes to hours will increase sharply and be greater than the power consumption in the next few minutes to hours, the operation target can be determined as "in the next few minutes to hours, on the premise of meeting the power demand, increase the charging amount of the energy storage battery on the power generation side and reduce the discharging amount of the energy storage battery on the power generation side.

[0088] On the contrary, if it is determined based on the first prediction result and the second prediction result that the power generation in the next few minutes to hours will drop sharply and will be greater than the power consumption in the next few minutes to hours, the operation target can be determined as "in the next few minutes to hours, reduce the charging amount of the energy storage battery on the power generation side and increase the discharging amount of the energy storage battery on the power generation side to meet the power consumption demand."

[0089] In addition, it can be understood that the prediction operation of the power generation status and the power consumption status can be set according to the actual situation. For example, in one example, the prediction results of the power generation status and the power consumption status can be determined based on the statistical results of historical operation information, such as predicting the power generation in July this year based on the power generation in July last year.

[0090] In addition, it should be noted that in the implementation of the present application, the specific monitoring process of the operation information and environmental quality information can be set according to the actual situation. For example, in one example, in the control device of the power system, the Internet of Things and edge computing real-time monitoring module 310 can monitor the operation information of distributed energy equipment (such as solar panels, wind turbines) and electrical equipment in real time through the Internet of Things sensor network, and monitor the environmental quality information of the entire power system, including but not limited to the collection of data such as power generation, energy consumption and equipment failure.

[0091] Furthermore, in the IoT and edge computing real-time monitoring module 310, current sensors, voltage sensors, temperature sensors, humidity sensors, light sensors and other sensors can be used to monitor the operation information of distributed energy generation equipment (such as solar panels, wind turbines) and electrical equipment, as well as the environmental quality information of the entire power system. The data collected by the sensors are transmitted to the edge computing node through wireless communication technology (such as Zigbee, LoRa, Wi-Fi, etc.).

[0092] Furthermore, edge computing nodes can be composed of NPU (Neural Processing Unit) and other high-performance embedded processors, with strong data computing and data storage capabilities. At the same time, after receiving sensor data, edge computing nodes can perform real-time data processing on sensor data, such as data cleaning, filtering, feature extraction and preliminary analysis, thereby completing the collection and preprocessing of operation information and environmental quality information.

[0093] Data cleaning can include data denoising, data calibration, data feature extraction, data aggregation, and abnormal data detection. Data denoising can remove high-frequency noise through a low-pass filter to ensure data accuracy. The transfer function of the filter can be expressed as follows:

[0094]

[0095] In the formula, ω c is the cutoff frequency of the filter.

[0096] Data calibration refers to calibrating the data collected by the sensor to eliminate the data collection error of the sensor. The calibration process can be referred to as follows:

[0097]

[0098] Where D calibrated is the calibrated data, D raw is the raw data collected by the sensor, D offset is the sensor offset, D scale is the scale factor of the sensor.

[0099] In addition, it can be understood that feature extraction can be understood as the extraction of key data features, such as current (I), voltage (V), power (P), etc. Data aggregation can be understood as aggregating data from different sensors to form comprehensive indicators, such as average power, peak current, etc. Data anomaly monitoring can be understood as anomaly detection through machine learning algorithms (such as k-mean clustering) to identify abnormal conditions in equipment operation.

[0100] In certain embodiments of the present application, the operation information includes multiple types, each type of operation information corresponds to a preset time scale, and then, the above step 02 includes:

[0101] According to the operation information corresponding to the preset time scale, the power generation state and the power consumption state corresponding to the preset time scale are predicted.

[0102] Meanwhile, the above step 03 includes:

[0103] An operation target corresponding to the preset time scale is determined according to the first prediction result and the second prediction result corresponding to the preset time scale.

[0104] The prediction module of the embodiment of the present application is configured to predict the power generation state and power consumption state corresponding to the preset time scale based on the operation information corresponding to the preset time scale. The determination module of the embodiment of the present application is configured to determine the operation target corresponding to the preset time scale based on the first prediction result and the second prediction result corresponding to the preset time scale.

[0105] The multi-level prediction module of the embodiment of the present application is configured to predict the power generation state and power consumption state corresponding to the preset time scale based on the operation information corresponding to the preset time scale. The distributed artificial intelligence coordination optimization control module of the embodiment of the present application is configured to determine the operation target corresponding to the preset time scale based on the first prediction result and the second prediction result corresponding to the preset time scale.

[0106] The processor of the embodiment of the present application is also used to predict the power generation status and power consumption status corresponding to the preset time scale based on the operating information corresponding to the preset time scale, and to determine the operating target corresponding to the preset time scale based on the first prediction result and the second prediction result corresponding to the preset time scale.

[0107] Specifically, in order to ensure the rationality and reliability of the operating objectives, in the implementation mode of the present application, the power system can formulate operating objectives at different time scales based on the operating information at different time scales.

[0108] In one example, the preset time scale includes a minute time scale, an hour time scale, and a day time scale. Therefore, the operation target in the implementation mode of the present application may include a short-term target (or first target) corresponding to the minute time scale, a medium-term target (or second target) corresponding to the hour time scale, and a long-term target (or third target) corresponding to the day time scale.

[0109] For example, taking the minute-level time scale as an example, the operating information of the past few minutes to hours can be used to predict the power generation of distributed energy equipment in the next few minutes to hours, and the power consumption of power-consuming equipment in the next few minutes to hours, and determine the power generation target of the power generation equipment on the power generation side in the next few minutes to hours to ensure that the electricity demand on the power consumption side can be met.

[0110] For example, taking the hourly time scale as an example, the operating information of the past few hours to days can be used to predict the power generation of distributed energy equipment in the next few hours to days, the power consumption of power-consuming equipment in the next few hours to days, and the charging and discharging targets of the energy storage equipment on the power generation side in the next few minutes to hours, so as to store excess energy during peak power generation periods and release the stored energy during low power generation periods.

[0111] For example, taking the day-level time scale as an example, the operating information from the past few days, weeks or even months can be used to predict the power generation of distributed energy equipment in the next few days, weeks or even months, and the power consumption of power-consuming equipment in the next few days, weeks or even months, and determine the development goals of the entire power system.

[0112] In this way, in the implementation mode of the present application, the power system can predict the power generation status and power consumption status corresponding to each time scale based on the operating information of multiple time scales, thereby determining the operating target corresponding to each time scale, and then can control the distributed energy equipment and the power consumption side equipment according to the operating target corresponding to each time scale, thereby realizing the reasonable control of the distributed energy equipment and the power consumption side equipment to a certain extent, and ensuring the long-term operation of the power system to a certain extent.

[0113] In certain embodiments of the present application, the above-mentioned predicting the power generation state and the power consumption state corresponding to the preset time scale according to the operation information corresponding to the preset time scale includes:

[0114] According to a pre-trained neural network model corresponding to a preset time scale and operation information corresponding to the preset time scale, the power generation state and power consumption state corresponding to the preset time scale are predicted.

[0115] The prediction module of the implementation mode of the present application is configured to predict the power generation state and power consumption state corresponding to the preset time scale based on a pre-trained neural network model corresponding to the preset time scale and the operation information corresponding to the preset time scale.

[0116] The multi-level prediction module in the implementation manner of the present application is configured to predict the power generation state and power consumption state corresponding to the preset time scale based on a pre-trained neural network model corresponding to the preset time scale and the operation information corresponding to the preset time scale.

[0117] The processor of the embodiment of the present application is also used to predict the power generation state and power consumption state corresponding to the preset time scale based on a pre-trained neural network model corresponding to the preset time scale, and the operating information corresponding to the preset time scale.

[0118] Specifically, in order to ensure the accuracy of the power generation status prediction results and the power consumption status prediction results corresponding to each time scale, in the implementation mode of the present application, a pre-trained neural network model corresponding to each time scale can be called to complete the corresponding prediction work.

[0119] In one example, multiple preset time scales include a minute time scale, an hour time scale, and a day time scale. Furthermore, the neural network model in the implementation mode of the present application includes a long short-term memory network model (Long-Short Term Memory, LSTM) corresponding to the minute time scale, a gated recurrent unit network model (Gated Recurrent Unit, GRU) corresponding to the hour time scale, and a transformer model (Transformer) corresponding to the day time scale.

[0120] For example, taking the minute-level time scale as an example, the operating information of the past few minutes to hours can be input into the LSTM terminal to obtain the LSTM-predicted power generation of distributed energy equipment in the next few minutes to hours and the power consumption of power-consuming equipment in the next few minutes to hours.

[0121] For example, taking the hourly time scale as an example, the operating information of the past few hours to days can be input into the GRU to obtain the GRU's prediction of the power generation of distributed energy equipment in the next few hours to days and the power consumption of power-consuming equipment in the next few hours to days.

[0122] For example, taking the day-level time scale as an example, the operating information from the past few days to weeks or even months can be input into Transformer to obtain the Transformer's prediction of the power generation of distributed energy equipment in the next few days to weeks or even months, and the power consumption of power-consuming equipment in the next few days to weeks or even months.

[0123] Thus, in the implementation mode of the present application, the power generation state and the power consumption state can be predicted based on the pre-trained neural network model, so the credibility of the prediction results of the power generation state and the power consumption state can be guaranteed.

[0124] In addition, it can be understood that the training process of the above-mentioned LSTM, GRU and Transformer can be set according to actual conditions. For example, in one example, the training process of LSTM includes: 1) collecting information on the distributed energy equipment and the power consumption side equipment on a minute scale. The collected information may include parameters such as power, current, voltage, air pressure, temperature, humidity, etc., to obtain operation information corresponding to the minute scale. 2) Normalizing the collected operation information to eliminate the dimensional differences between various types of information and ensure that the information input to LSTM is at the same level. 3) Constructing time series data based on the normalized operation information, and inputting the time series data into LSTM to obtain the prediction results of LSTM. 4) Calculating the mean square error loss value based on the prediction results. 5) Using the Adam optimization algorithm and the mean square error loss value, update LSTM. 6) Evaluating the prediction performance of LSTM based on the cross-validation method. 7) When the prediction performance evaluation results meet the preset conditions, end the training.

[0125] In another example, the training process of GRU includes: 1) Collecting information on distributed energy equipment and electricity-consuming equipment on an hourly scale. The collected information may include parameters such as power, current, voltage, air pressure, temperature, humidity, etc., to obtain operating information corresponding to the hourly scale. 2) De-noising and normalizing the collected operating information. 3) Constructing time series data based on the processed operating information, and inputting the time series data into GRU to obtain the prediction results of LSTM. 4) Calculating the mean square error loss value based on the prediction results. 5) Using the GRU optimization algorithm and the mean square error loss value, update the GRU. 6) Evaluating the prediction performance of GRU based on the rolling prediction method. 7) When the prediction performance evaluation results meet the preset conditions, end the training.

[0126] In another example, the training process of Transformer includes: 1) Collecting information on distributed energy equipment and electricity-consuming equipment on a daily scale. The collected information may include parameters such as power, current, voltage, air pressure, temperature, humidity, etc., to obtain operating information corresponding to the daily scale. 2) De-noising and feature engineering the collected operating information. 3) Constructing time series data based on the processed operating information, and inputting the time series data into Transformer to obtain the prediction results of Transformer. 4) Calculating the mean square error loss value based on the prediction results. 5) Using the Adam optimization algorithm and the mean square error loss value, update Transformer. 6) Evaluating the prediction performance of Transformer based on the method verified by the time series tutorial. 7) When the prediction performance evaluation results meet the preset conditions, end the training.

[0127] In certain embodiments of the present application, the operation target includes a first target, the first prediction result includes a power generation prediction, and the second prediction result includes a power consumption prediction. Furthermore, the step of determining the operation target corresponding to the preset time scale according to the first prediction result and the second prediction result corresponding to the preset time scale includes:

[0128] Based on the power generation forecast and power consumption forecast corresponding to the minute time scale, the power generation expectation of the distributed energy equipment is determined to obtain the first target.

[0129] The determination module of the implementation mode of the present application is configured to determine the power generation expectation of the distributed energy equipment based on the power generation forecast and power consumption forecast corresponding to the minute time scale to obtain the first target.

[0130] The distributed artificial intelligence coordination optimization control module of the implementation mode of the present application is configured to determine the power generation expectation of the distributed energy equipment based on the power generation forecast and power consumption forecast corresponding to the minute time scale to obtain the first target.

[0131] The processor of the embodiment of the present application is also used to determine the power generation expectation of the distributed energy equipment based on the power generation forecast and power consumption forecast corresponding to the minute time scale to obtain the first target.

[0132] Specifically, in the implementation mode of the present application, the power system can obtain power generation forecasts and power consumption forecasts on a minute scale based on the operating information collected on a minute basis, and determine the expected power generation of distributed energy equipment in the next few minutes to hours through the power generation forecasts and power consumption forecasts, thereby achieving the first goal.

[0133] Specifically, when obtaining a power generation forecast and a power consumption forecast corresponding to a minute time scale, or in other words, when predicting the power generation of distributed energy devices per minute and the power consumption of power-consuming devices per minute in the next few minutes to hours based on the operating information collected on a minute-by-minute basis in the past few minutes to hours, the implementation mode of the present application can determine the expected power generation of distributed energy devices in the next few minutes to hours based on the power generation forecast of distributed energy devices per minute and the power consumption forecast of power-consuming devices per minute in the next few minutes to hours.

[0134] For example, in one example, the power generation of distributed energy devices in the next few hours is predicted to be P per minute. g , the power generation of the power consumption equipment in the next few minutes to hours is predicted to be P u , then: In P g Less than or equal to P u In the case of P, it can be determined that the power generation of distributed energy equipment is expected to be higher than Pu The first goal is to "distributed energy equipment with a value higher than P u The value of 'generates electricity', thereby meeting the electricity demand on the electricity consumption side.

[0135] Thus, in the implementation mode of the present application, the expected power generation of the distributed energy equipment can be determined through power generation prediction and power consumption prediction, thereby obtaining the first goal.

[0136] In certain embodiments of the present application, the distributed energy device includes an electric energy storage module, and the operation target includes a second target. Furthermore, the step of determining the operation target corresponding to the preset time scale according to the first prediction result and the second prediction result corresponding to the preset time scale includes:

[0137] According to the power generation forecast and power consumption forecast corresponding to the hourly time scale, the expected charging amount and discharging amount of the energy storage module are determined to obtain the second target.

[0138] The determination module of the implementation mode of the present application is configured to determine the expected charging amount and discharging amount of the energy storage module based on the power generation forecast and power consumption forecast corresponding to the hourly time scale to obtain the second target.

[0139] The distributed artificial intelligence coordination optimization control module of the implementation mode of the present application is configured to determine the expected charging amount and discharging amount of the energy storage module based on the power generation forecast and power consumption forecast corresponding to the hourly time scale to obtain the second target.

[0140] The processor of the embodiment of the present application is also used to determine the expected charging amount and discharging amount of the energy storage module based on the power generation forecast and power consumption forecast corresponding to the hourly time scale to obtain the second target.

[0141] Specifically, in the implementation mode of the present application, the power system can obtain power generation forecasts and power consumption forecasts on an hourly scale based on the operating information collected on an hourly basis, and determine the expected charging and discharging amounts of the energy storage module through the power generation forecasts and power consumption forecasts, thereby obtaining the second goal.

[0142] Specifically, when obtaining a power generation forecast and a power consumption forecast corresponding to an hourly time scale, or in other words, when predicting the hourly power generation of distributed energy devices and the hourly power consumption of power-consuming devices in the next few hours to days based on the operating information collected hourly in the past few hours to days, the implementation mode of the present application can determine the expected charging amount and discharging amount of the energy storage module in the next few hours to days based on the hourly power generation forecast of distributed energy devices and the hourly power consumption forecast of power-consuming devices in the next few hours to days.

[0143] For example, in one example, the hourly power generation forecast of distributed energy devices in the next few hours is P g , P g Equal to P g1 +P g2 , P g1 For the prediction of the power generation of solar panels every hour in the next few hours to several hours, P g2 is the hourly power generation forecast of the thermal power station in the next few hours to several hours (known quantity), and the hourly power generation forecast of the power consumption equipment in the next few hours to several hours is P u The energy storage module is charged with C per hour. in The energy storage module releases C out , then: In P g1 is less than or equal to a preset reference value, and P g Less than or equal to P u In the case of lowering C in And improve C out In P g1 Greater than the above reference value, and P g Greater than P u In the case of lowering C out And improve C in This ensures that excess energy is stored during peak periods of high power generation and that the stored energy is released during trough periods of low power generation.

[0144] Thus, in the implementation mode of the present application, the expected charging amount and the expected discharging amount of the energy storage module can be determined by predicting the power generation amount and the power consumption amount, thereby obtaining the second target.

[0145] In certain embodiments of the present application, the operation target includes a third target, and the step of determining the operation target corresponding to the preset time scale according to the first prediction result and the second prediction result corresponding to the preset time scale includes:

[0146] Based on the power generation forecast and power consumption forecast corresponding to the daily time scale, the expected working state of the power system is determined to obtain the third goal.

[0147] The determination module of the implementation mode of the present application is configured to determine the expected working state of the power system based on the power generation forecast and power consumption forecast corresponding to the daily time scale to obtain the third goal.

[0148] The distributed artificial intelligence coordination optimization control module of the implementation mode of the present application is configured as the determination module of the implementation mode of the present application is configured to determine the expected working state of the power system based on the power generation forecast and power consumption forecast corresponding to the day-level time scale to obtain the third goal.

[0149] The processor of the implementation mode of the present application is also used to determine the expected working state of the power system based on the power generation forecast and power consumption forecast corresponding to the daily time scale to obtain the third goal.

[0150] Specifically, in the implementation mode of the present application, the power system can obtain a daily power generation forecast and a power consumption forecast based on the operating information collected on a daily basis, and determine the expected future working status of the power system through the power generation forecast and the power consumption forecast, thereby obtaining the third goal.

[0151] Specifically, when obtaining a power generation forecast and a power consumption forecast corresponding to a daily time scale, or in other words, when predicting the daily power generation of distributed energy equipment and the daily power consumption of power-consuming equipment in the next few days to weeks or even months based on the operating information collected on a daily basis in the next few days to weeks or even months, the implementation mode of the present application can determine the expected overall working status of the power system in the next few days to weeks or even months based on the daily power generation forecast of distributed energy equipment and the daily power consumption forecast of power-consuming equipment in the next few days to weeks or even months, such as the waste of power generation of distributed energy equipment and the failure to meet the power consumption demand of power-consuming equipment.

[0152] It should be noted that, in the implementation mode of the present application, the third goal can be understood as advisory content, which can be provided to the staff of the power system for long-term planning and strategic decision-making to ensure the sustainable development and expansion of the system.

[0153] It should also be noted that in the implementation of the present application, the specific content of the third goal can be set according to the actual situation. In one example, the third goal includes optimizing the energy structure, such as adjusting the quantity ratio between new energy power generation equipment (such as photovoltaic power generation facilities, wind power facilities, etc.) and traditional power generation equipment (such as facilities in thermal power plants, etc.). In another example, the third goal includes a maintenance and repair plan.

[0154] Thus, in the implementation mode of the present application, the expected working state of the power system can be determined through power generation forecast and power consumption forecast, thereby obtaining the third goal.

[0155] In addition, in one example, the operation target in the embodiment of the present application includes the first target, the second target and the third target. Accordingly, the specific process of the power system acquiring the first target, the second target and the third target, and controlling the power generation side and the power consumption side according to the first target, the second target and the third target may include:

[0156] First, real-time data analysis is performed on edge computing nodes to reduce data transmission latency and network bandwidth requirements.

[0157] Then, for each edge computing node, the edge computing node triggers the corresponding processing logic according to the predefined time rules.

[0158] Next, the edge computing node stores critical data locally for data recovery in case of network instability or disconnection.

[0159] Next, the edge computing nodes regularly synchronize data with the central server to ensure data consistency and integrity.

[0160] After that, the LSTM for short-term prediction, the GRU for medium-term prediction, and the Transformer for long-term prediction are trained, and the trained LSTM, GRU, and Transformer are deployed in the edge computing nodes. It can be understood that the training process of LSTM, GRU, and Transformer can refer to the above content, and to avoid repetition, it will not be repeated here.

[0161] Short-term prediction (LSTM) is targeted at a few minutes to a few hours and is used to ensure the efficient operation and stability of the power system in a short period of time. Specifically, based on the operating information (such as power generation) of distributed energy equipment (such as solar panels, wind turbines, battery energy storage systems and other distributed energy equipment) and the operating information (such as power consumption) of power-consuming equipment (such as residential, commercial and industrial power equipment), the power generation in the next few minutes to a few hours and the power demand in the next few minutes to a few hours can be predicted. Furthermore, based on the power generation in the next few minutes to a few hours and the power demand in the next few minutes to a few hours, the power generation power target of the distributed energy equipment can be formulated, so as to achieve a rapid response to changes in power demand and adjust the operating status of the power-consuming equipment.

[0162] The medium-term forecast (GRU) is for the next few hours to days, and is used for medium-term energy storage and grid dispatching to improve the overall efficiency of the system. Specifically, the power generation in the next few hours to days and the power demand in the next few hours to days can be predicted based on the operating information (such as power generation) of distributed energy equipment (such as solar panels, wind turbines, battery energy storage systems and other distributed energy equipment) and the operating information (such as power consumption) of power-consuming equipment (such as residential, commercial and industrial power equipment). Furthermore, based on the power generation in the next few hours to days and the power demand in the next few hours to days, the charging and discharging strategy of the battery energy storage system (including charging expectation, discharge expectation, charging period prediction, discharge time prediction, etc.) can be formulated to store excess energy during the peak period of high power generation and release the stored energy during the trough period of power generation. The grid dispatching strategy can also be set to balance supply and demand and ensure the stable operation of the power grid.

[0163] Long-term forecasting (Transformer) is aimed at a few days to a few weeks or even a few months, and is used to ensure the long-term planning and strategic decision-making of the power system, and provide advice to decision makers to ensure the sustainable development and expansion of the system. Specifically, based on the operating information (such as power generation) of distributed energy equipment (such as distributed energy equipment such as solar panels, wind turbines, battery energy storage systems, and traditional power stations such as thermal power stations and hydropower stations), and the operating information (such as power consumption) of power-consuming equipment (such as residential, commercial and industrial power equipment), the power generation in the next few days to weeks or even months and the power demand in the next few days to weeks or even months can be predicted. Furthermore, based on the power generation in the next few days to weeks or even months, and the power demand in the next few minutes to hours, the power generation plan of new energy and traditional energy can be formulated, and the energy structure can be optimized, such as grid planning and expansion to ensure that the grid can adapt to future changes in supply and demand, and arrange long-term equipment maintenance and upgrade plans to ensure the reliability and efficiency of the system.

[0164] It is understandable that short-term forecast results are used for real-time optimization and exception handling to ensure that the system can respond quickly to load changes and emergencies. Medium-term forecast results are used to formulate medium-term energy dispatch and energy storage strategies to balance supply and demand and improve the overall efficiency of the system. Long-term forecast results are used to formulate long-term energy production and supply strategies, plan grid expansion and upgrades, and arrange equipment maintenance and upgrade plans to ensure the sustainable development and expansion of the system. By combining the results of short-term, medium-term and long-term forecasts, the power system can achieve comprehensive dynamic optimization and adaptive adjustment, thereby ensuring the efficient and stable operation of the intelligent management system for local consumption of new energy within different time frames.

[0165] See also Figure 4 In certain embodiments of the present application, step 04 includes:

[0166] 040: Construct a state vector based on environmental quality information and operation information;

[0167] 041: Construct a reward function based on operational goals and predetermined environmental quality goals;

[0168] 042: Determine the first action parameter corresponding to the distributed energy device and the second action parameter of the power-side device according to the state vector, the reward function and the pre-trained intelligent agent model;

[0169] 043: Controlling the power generation of the distributed energy device according to the first action parameter, and controlling the power consumption of the power-consuming side device according to the second action parameter.

[0170] The control module of the implementation mode of the present application is configured to construct a state vector based on environmental quality information and operating information, and to construct a reward function based on operating goals and predetermined environmental quality goals, and to determine the first action parameters corresponding to the distributed energy device and the second action parameters of the power-consuming side device based on the state vector, the reward function and a pre-trained intelligent agent model, and to control the power generation of the distributed energy device according to the first action parameters, and to control the power consumption of the power-consuming side device according to the second action parameters.

[0171] The distributed artificial intelligence coordination optimization control module of the implementation mode of the present application is configured to construct a state vector based on environmental quality information and operating information, and to construct a reward function based on operating goals and predetermined environmental quality goals, and to determine the first action parameters corresponding to the distributed energy equipment and the second action parameters of the power-consuming side equipment based on the state vector, the reward function and the pre-trained intelligent agent model, and to control the power generation of the distributed energy equipment according to the first action parameters, and to control the power consumption of the power-consuming side equipment according to the second action parameters.

[0172] The processor of the embodiment of the present application is also used to construct a state vector based on environmental quality information and operating information, and to construct a reward function based on operating objectives and predetermined environmental quality objectives, and to determine the first action parameters corresponding to the distributed energy device and the second action parameters of the power-consuming device based on the state vector, the reward function and a pre-trained intelligent agent model, and to control the power generation of the distributed energy device according to the first action parameters, and to control the power consumption of the power-consuming device according to the second action parameters.

[0173] Specifically, in the implementation mode of the present application, in order to ensure the robust control of distributed energy equipment and power-consuming equipment, the implementation mode of the present application can determine the action parameters of the distributed energy equipment and the power-consuming equipment through a pre-trained intelligent agent model (or reinforcement learning model), and control the distributed energy equipment and the power-consuming equipment according to the action parameters.

[0174] In an example, the operating information includes the current power generation and fault status of the distributed energy equipment, and the current power consumption of the power-consuming equipment. The current power generation of the distributed energy equipment, the fault status of the distributed energy equipment, the current power consumption of the power-consuming equipment, and the environmental quality information of the location of the power system can be used as the state vector St.

[0175] Next, a reward function rt is constructed based on the operating objectives and predetermined environmental quality goals (such as carbon emission reduction, air quality improvement, etc.).

[0176] Next, based on the current state St and strategy π, the optimal action At is selected, where the action can be operations such as adjusting power generation, load scheduling, and energy storage.

[0177] Finally, according to algorithms such as Q-learning or Deep Q-Network (DQN), the strategy is updated according to actual feedback to gradually improve the quality of decision making.

[0178] In another example, the operation information includes the real-time power generation, operation status and fault information of distributed energy (such as solar panels, wind turbines) and traditional power stations, as well as the real-time power consumption of the power-consuming equipment in the power system, and the power generation and power consumption in the past period of time. The environmental quality information at the location of the power system includes meteorological data such as temperature, humidity, wind speed, etc. from the past period of time to the present, and also includes air quality data such as PM2.5 and PM10 concentrations from the past period of time to the present.

[0179] Furthermore, the above-mentioned real-time power generation, operating status, fault information, real-time power consumption, power generation and power consumption in the past period of time, and environmental quality information are input into the intelligent agent model as state vectors. The output results of the intelligent agent model may include: power generation adjustment of distributed energy equipment and traditional power stations, load adjustment of controllable power equipment, start-stop state changes of controllable power equipment, charging and discharging strategies of energy storage batteries on the power generation side, expected carbon emission reduction, and expected air quality improvement.

[0180] Thus, in the implementation mode of the present application, a state vector can be constructed based on environmental quality information and operating information, and a reward function can be constructed based on operating objectives and predetermined environmental quality objectives, and based on the state vector, the reward function and the pre-trained intelligent agent model, the first action parameters corresponding to the distributed energy equipment and the second action parameters of the power-consuming side equipment can be determined, thereby controlling the operation of the distributed energy equipment through the first action parameters and controlling the operation of the power-consuming side equipment through the second action parameters.

[0181] In addition, it can be understood that in the implementation mode of the present application, the specific setting method of the environmental quality target can be set according to actual conditions. For example, in one example, the environmental quality target includes at least one of the greenhouse gas emission reduction, air quality improvement and comprehensive environmental benefit assessment parameters.

[0182] Furthermore, the three environmental quality targets of greenhouse gas emission reduction, air quality improvement and comprehensive environmental benefit assessment parameters can be calculated by the following two formulas, namely:

[0183] ΔCO2=P renewable ×T×EF

[0184] In the formula, ΔCO2 is the amount of CO2 reduced, P renewableis the power generation of distributed energy, T is time, and EF is the emission reduction factor of traditional fossil energy.

[0185]

[0186] In the formula, AQI improve Air Quality Improvement Index, AQI before AQI is the air quality index before the use of distributed energy. after It is the air quality index after adopting distributed energy.

[0187] E total =ω1×ΔCO2+ω2×AQI improve +ω3×E other

[0188] In the formula, E total is the comprehensive environmental benefit assessment parameter, ω1, ω2 and ω3 are the weights of different indicators, E other Scores for other environmental benefit indicators (such as energy saving, noise reduction, etc.).

[0189] In addition, in Figure 3 In an example shown, three environmental quality targets, namely, greenhouse gas emission reduction, air quality improvement, and comprehensive environmental benefit assessment parameters, are determined by a comprehensive environmental benefit assessment module 350 in the control device 300 .

[0190] Specifically, the comprehensive environmental benefit assessment module 350 can provide decision support based on the environmental benefits of the real-time power system location based on the data collected by each sensor. For example, the real-time power generation data of distributed energy power generation equipment, the real-time power consumption data of power-consuming equipment, air quality monitoring data (such as PM2.5, PM10 concentration), greenhouse gas emission data (such as CO2, CH4 emissions), meteorological data (such as temperature, humidity, wind speed, etc.), etc. can be collected through the aforementioned sensors, and then the collected data are preprocessed accordingly. For example, in one example, the preprocessing can be to first clean the collected data to remove noise and outliers to ensure the accuracy of the data, and to fuse data from different sources to generate a comprehensive indicator data set, and then normalize the data to ensure that the dimensional differences between different types of data are eliminated.

[0191] Furthermore, the comprehensive environmental benefit assessment module 350 can calculate corresponding environmental benefit indicators based on the preprocessed data to provide data support for the distributed artificial intelligence coordination optimization control module 330.

[0192] In certain embodiments of the present application, the distributed energy device includes a plurality of power station modules, the operation information includes module operation sub-information of each power station module, and the operation target includes the module operation target of each power station module. Then, step 04 includes:

[0193] The power generation of the power station module and the power consumption of the power consumption side equipment are controlled according to the environmental quality information, the module operation sub-information and the module operation target of the power station module.

[0194] The control module of the embodiment of the present application is also used to control the power generation of the power plant module and the power consumption of the power-consuming side equipment according to the environmental quality information, as well as the module operation sub-information and module operation target of the power plant module.

[0195] The distributed artificial intelligence coordination optimization control module of the implementation mode of the present application is configured to control the power generation of the power station module and the power consumption of the power-consuming side equipment based on environmental quality information, as well as the module operation sub-information and module operation goals of the power station module.

[0196] The processor of the embodiment of the present application is also used to control the power generation of the power plant module and the power consumption of the power-consuming side equipment according to the environmental quality information, as well as the module operation sub-information and module operation target of the power plant module.

[0197] Specifically, in an electric power system, the power generation side may include multiple power stations, such as wind power stations and solar power stations, and the power generation of wind power stations and solar power stations are both affected by local weather. For example, when the wind is high at night, the power generation of the wind power station may be high while the power generation of the solar power station may be low.

[0198] Therefore, considering that the power generation of different power plants may be different, when the same control method is adopted for each power plant, the power generation effect achieved by all power plants may not reach the expected situation. Therefore, in the implementation of the present application, the power plant module and the power-consuming side equipment can be controlled according to the environmental quality information, the module operation sub-information and module operation target of the power plant module. That is, the implementation of the present application can control the power plant module separately for the module operation sub-information and module operation target of a certain power plant module to ensure the stable operation of the power plant module.

[0199] For instance, in one example, the distributed energy equipment includes a wind power station and a solar power station, and then: based on the current power generation of the wind power station, the historical power generation of the wind power station in the past period of time (such as the past 7 days), the current power consumption of the power-consuming side equipment and the historical power consumption of the power-consuming side equipment in the past period of time (such as the past 7 days), predict the power generation status of the wind power station in the future period of time (such as the next 7 days), and the power consumption status of the power-consuming side equipment in the future period of time (such as the next 7 days), and determine the operation target of the wind power station based on the power generation status prediction results and the power consumption status prediction results, and then based on the environmental quality information of the location of the power system, combined with the operation target of the wind power station, the current power generation of the wind power station, the historical power generation of the wind power station in the past period of time (such as the past 7 days), the current power consumption of the power-consuming side equipment and the historical power consumption of the power-consuming side equipment in the past period of time (such as the past 7 days), control the operation of the facilities in the wind power station.

[0200] Similarly, for solar power stations, the following are: predicting the power generation status of the wind power station in the future (such as the next 7 days) and the power consumption status of the power-consuming equipment in the future (such as the next 7 days) based on the current power generation of the solar power station, the historical power generation of the solar power station in the past (such as the past 7 days), the current power consumption of the power-consuming equipment and the historical power consumption of the power-consuming equipment in the past (such as the past 7 days), and determining the operation target of the solar power station based on the power generation status prediction results and the power consumption status prediction results, and then controlling the operation of the facilities in the solar power station based on the environmental quality information of the location of the power system and the operation target of the solar power station, the current power generation of the solar power station, the historical power generation of the solar power station in the past (such as the past 7 days), the current power consumption of the power-consuming equipment and the historical power consumption of the power-consuming equipment in the past (such as the past 7 days).

[0201] Optionally, in certain embodiments of the present application, a corresponding intelligent agent model may be configured for each power station in the power system to determine the control parameters of the power station.

[0202] Optionally, in certain embodiments of the present application, power stations within the power system are set up in a distributed manner, and thus, the power station and power-consuming equipment that has an electricity supply relationship with the power station can be controlled based on environmental quality information combined with module operation sub-information and module operation targets of a power station (or a power station module).

[0203] Thus, in the embodiment of the present application, the power plant modules and the power consumption side equipment can be controlled according to the environmental quality information, as well as the module operation sub-information and module operation target of the power plant modules.

[0204] In certain embodiments of the present application, the operation target includes the total operation target of multiple power plant modules, and then, step 04 includes:

[0205] The power generation and power consumption of the power-consuming side equipment are controlled according to the environmental quality information, the overall operation target and the module operation sub-information of each power station module.

[0206] The control module of the implementation mode of the present application is also used to control the power generation and power consumption of the power-consuming side equipment according to the environmental quality information, the overall operation target and the module operation sub-information of each power station module.

[0207] The distributed artificial intelligence coordination optimization control module of the implementation mode of the present application is configured to control the power generation and power consumption of the power-consuming side equipment according to the environmental quality information, the overall operation target and the module operation sub-information of each power station module.

[0208] The processor of the embodiment of the present application is also used to control the power generation and power consumption of the power-consuming side equipment according to the environmental quality information, the overall operation target and the module operation sub-information of each power plant module.

[0209] Specifically, in the implementation mode of the present application, in order to ensure the balance between the power generation side and the power consumption side in the power system, in the implementation mode of the present application, all power station modules and power consumption side equipment can be controlled according to the overall operating target corresponding to all power station modules on the power generation side and the module operation sub-information of each power station module.

[0210] For instance, in one example, the power generation side includes a wind power station and a solar power station, then: based on the current power generation of the wind power station, the historical power generation of the wind power station in the past period of time (such as the past 7 days), the current power generation of the solar power station, the historical power generation of the solar power station in the past period of time (such as the past 7 days), the current power consumption of the power-consuming side equipment and the historical power consumption of the power-consuming side equipment in the past period of time (such as the past 7 days), predict the power generation status of the wind power station and the solar power station in the future period of time (such as the next 7 days), and predict the power consumption status of the power-consuming side equipment in the future period of time (such as the next 7 days), and Based on the power generation status prediction results and the power consumption status prediction results, the total operating target corresponding to the wind power station and the solar power station is determined, and then based on the environmental quality information of the location of the power system, combined with the total operating target, the current power generation of the wind power station, the historical power generation of the wind power station in the past period of time (such as the past 7 days), the current power generation of the solar power station, the historical power generation of the solar power station in the past period of time (such as the past 7 days), the current power consumption of the power-consuming side equipment and the historical power consumption of the power-consuming side equipment in the past period of time (such as the past 7 days), the operation of the facilities in the wind power station, the facilities in the solar power station and the power-consuming side equipment is controlled.

[0211] Thus, in the embodiment of the present application, multiple power plant modules and power-consuming side equipment can be controlled according to environmental quality information, the total operating target corresponding to multiple power plant modules, and the module operating sub-information of each power plant module.

[0212] In addition, to more clearly illustrate the implementation methods of this application, please refer to Figure 3 , and see Figure 5 and Figure 6 , Figure 5 and Figure 6 All of them are flowchart diagrams of control methods of power systems in certain implementation modes of the present application.

[0213] like Figure 5 and Figure 6 As shown, the distributed artificial intelligence coordination optimization control module 330 is used to integrate the information of each module, optimize the energy distribution and storage strategy in real time, realize the maximum local consumption of new energy, and ensure the stability and efficient operation of the power grid. Furthermore, the distributed artificial intelligence coordination optimization control module 330 can be used for local optimization of edge computing nodes and global optimization of central control nodes.

[0214] Among them, the edge computing node can be equipped with a reinforcement learning (RL) model, and based on the reinforcement learning model, according to local data, such as the real-time power generation of distributed energy (such as solar panels, wind turbines) and traditional power stations, the real-time power consumption of all electrical equipment in the power system, meteorological data (such as temperature, humidity, wind speed, etc.), air quality data (such as PM2.5, PM10 concentration), the operating status of distributed energy equipment and electrical equipment, the fault information of distributed energy equipment and electrical equipment, etc., combined with the power generation of distributed energy (such as solar panels, wind turbines) and traditional power stations in the past period of time, the power consumption of all electrical equipment in the power system, environmental data and other time series data, the "power generation target formulated through short-term forecast results, and the charging and discharging strategy formulated through medium-term forecast results" are optimized to output local power generation, local power consumption, local charging and discharging status, local environmental benefit estimation, etc.

[0215] The central computing node can be equipped with a multi-agent reinforcement learning (MARL) model, and based on the multi-agent reinforcement learning model, according to local data, such as the real-time total power generation of distributed energy (such as solar panels, wind turbines) and traditional power stations, the real-time total power consumption of all electrical equipment in the power system, meteorological data (such as temperature, humidity, wind speed, etc.), air quality data (such as PM2.5, PM10 concentration), the operating status of distributed energy equipment and electrical equipment, fault information of distributed energy equipment and electrical equipment, local data of each edge computing node (including local power generation, power consumption, equipment status), etc., combined with the power generation of distributed energy (such as solar panels, wind turbines) and traditional power stations in the past period of time, the power consumption of all electrical equipment in the power system, environmental data and other time series data, the "power generation target formulated through short-term forecast results, and the charging and discharging strategy formulated through medium-term forecast results" are optimized to output the overall power generation, overall power consumption, overall charging and discharging status, overall environmental benefit estimation, etc.

[0216] It is understandable that the reinforcement learning model (or single agent) carried by the edge computing node is used to output the optimization strategy of a single agent, such as power generation scheduling, load scheduling, energy storage strategy, etc. The central computing node can be equipped with a multi-agent reinforcement learning model to output the collaborative optimization strategy of multiple agents, including power generation scheduling, load scheduling, energy storage strategy, etc. of each distributed energy node or equipment.

[0217] It can also be understood that in the edge computing node, the operation process of the reinforcement learning model may include: a single agent evaluates the current state of the environment, selects the optimal action based on the current state and strategy, obtains rewards based on the results of the action, updates the strategy, and then adjusts the strategy based on feedback for continuous optimization.

[0218] In the central computing node, the operation process of the multi-agent reinforcement learning model may include: multiple agents evaluate their own local states and overall global states respectively, and then each agent selects the optimal action based on its own local state and global strategy. Then, each agent is rewarded according to its own performance and the performance of the overall system. Then, multiple agents share information and jointly optimize the overall system performance through collaboration or competition. Finally, the central control node performs global coordination and dynamic adjustments based on the global state and the decisions of the agents.

[0219] In addition, it is also understandable that the reinforcement learning model of edge computing nodes is suitable for a single control entity or a relatively independent system. For example, the optimal control of a single wind turbine generator set maximizes power generation efficiency by adjusting its angle and speed.

[0220] Correspondingly, the multi-agent reinforcement learning model in the central computing node is suitable for complex systems with multiple interrelated control entities that need to work together. For example, a large distributed energy network includes multiple wind turbines, solar panels, energy storage systems, and load centers, where the agents need to coordinate their work to optimize overall energy management and scheduling.

[0221] In addition, in Figure 3 In one example shown, the control device also includes an energy trading reward mechanism 340. The energy trading reward mechanism 340 can provide a corresponding reward mechanism based on the user's new energy consumption and energy-saving performance to encourage the user to actively participate in new energy management. Specifically, a corresponding new energy trading platform can be set up to record and evaluate the user's energy use and production, and provide incentives based on the energy use and production to promote users to use new energy more efficiently.

[0222] The embodiment of the present application further provides a computer-readable storage medium storing a computer program, which implements the above-mentioned power system control method when the computer program is executed by one or more processors.

[0223] In the description of this specification, the descriptions with reference to the terms "specifically", "further", "particularly", "understandably", etc. are intended to mean that the specific features, structures, materials or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms are not intended to refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0224] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by technicians in the technical field to which the embodiments of the present application belong.

[0225] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A control method for an electric power system, characterized in that: The power system includes distributed energy equipment and power consumption equipment, and the method includes: Monitoring environmental quality information and operation information of the power system, wherein the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is used to indicate the status of the distributed energy device and the status of the power consumption side device; Predicting the power generation status of the distributed energy device and the power consumption status of the power consumption side device according to the operation information; Determining an operation target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state; The power generation of the distributed energy device and the power consumption of the power-consuming device are controlled according to the environmental quality information, the operating information and the operating target.

2. The method according to claim 1, characterized in that: The operation information includes multiple types, each type of the operation information corresponds to a preset time scale, and the power generation state of the distributed energy device and the power consumption state of the power consumption side device are predicted according to the operation information, including: According to the operation information corresponding to the preset time scale, predicting the power generation state and the power consumption state corresponding to the preset time scale; Determining the operation target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state includes: The operation target corresponding to the preset time scale is determined according to the first prediction result and the second prediction result corresponding to the preset time scale.

3. The method according to claim 2, characterized in that The predicting, based on the operation information corresponding to the preset time scale, the power generation state and the power consumption state corresponding to the preset time scale comprises: The power generation state and the power consumption state corresponding to the preset time scale are predicted based on a pre-trained neural network model corresponding to the preset time scale and the operation information corresponding to the preset time scale.

4. The method according to claim 3, characterized in that The preset time scale includes a minute-level time scale, an hour-level time scale, and a day-level time scale.

5. The method according to claim 4, characterized in that The neural network model includes a long short-term memory network model corresponding to the minute time scale, a gated recurrent unit network model corresponding to the hour time scale, and a transformer model corresponding to the day time scale.

6. The method according to claim 4, characterized in that The operation target includes a first target, the first prediction result includes a power generation prediction, and the second prediction result includes a power consumption prediction, and determining the operation target corresponding to the preset time scale according to the first prediction result and the second prediction result corresponding to the preset time scale includes: Based on the power generation forecast and the power consumption forecast corresponding to the minute time scale, the power generation expectation of the distributed energy device is determined to obtain the first target.

7. The method according to claim 6, characterized in that The distributed energy device includes an electric energy storage module, the operation target includes a second target, and determining the operation target corresponding to the preset time scale according to the first prediction result and the second prediction result corresponding to the preset time scale includes: According to the power generation forecast and the power consumption forecast corresponding to the hourly time scale, the expected charging amount and the expected discharging amount of the energy storage module are determined to obtain the second target.

8. The method according to claim 6, characterized in that The operation target includes a third target, and determining the operation target corresponding to the preset time scale according to the first prediction result and the second prediction result corresponding to the preset time scale includes: Based on the power generation forecast and the power consumption forecast corresponding to the daily time scale, the expected working state of the power system is determined to obtain the third goal.

9. The method according to claim 1, characterized in that: The controlling the power generation of the distributed energy device and the power consumption of the power-consuming device according to the environmental quality information, the operation information and the operation target includes: constructing a state vector according to the environmental quality information and the operation information; constructing a reward function based on the operational objectives and the predetermined environmental quality objectives; Determine, according to the state vector, the reward function and the pre-trained agent model, a first action parameter corresponding to the distributed energy device and a second action parameter of the power-side device; The power generation of the distributed energy device is controlled according to the first action parameter, and the power consumption of the power-consuming device is controlled according to the second action parameter.

10. The method according to claim 9, characterized in that The environmental quality target includes at least one of greenhouse gas emission reduction, air quality improvement and comprehensive environmental benefit assessment parameters.

11. The method according to claim 1, characterized in that The distributed energy device includes a plurality of power station modules, the operation information includes module operation sub-information of each of the power station modules, the operation target includes module operation target of each of the power station modules, and the controlling of power generation of the distributed energy device and power consumption of the power consumption side device according to the environmental quality information, the operation information and the operation target includes: The power generation of the power plant module and the power consumption of the power consumption side equipment are controlled according to the environmental quality information, the module operation sub-information of the power plant module and the module operation target.

12. The method according to claim 11, characterized in that The operation target includes the total operation target of the plurality of power station modules, and the controlling of the power generation of the distributed energy device and the power consumption of the power-consuming device according to the environmental quality information, the operation information and the operation target includes: The power generation and power consumption of the power-consuming devices are controlled according to the environmental quality information, the overall operation target, and the module operation sub-information of each power plant module.

13. A control device for an electric power system, characterized in that: The power system includes distributed energy equipment and power consumption equipment, and the device includes: A monitoring module configured to monitor environmental quality information and operation information of the power system, wherein the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is used to indicate the status of the distributed energy device and the status of the power-side device; A prediction module, configured to predict the power generation state of the distributed energy device and the power consumption state of the power consumption side device according to the operation information; a determination module configured to determine an operation target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state; The control module is configured to control the power generation of the distributed energy device and the power consumption of the power-consuming device according to the environmental quality information, the operating information and the operating target.

14. A control device for an electric power system, characterized in that: The power system includes distributed energy equipment and power consumption equipment, and the equipment includes: An Internet of Things and edge computing real-time monitoring module, configured to monitor environmental quality information and operation information of the power system, wherein the environmental quality information includes at least one of air quality monitoring data, greenhouse gas emission data and meteorological data, and the operation information is used to indicate the status of the distributed energy device and the status of the power-side device; A multi-level prediction module, configured to predict the power generation state of the distributed energy device and the power consumption state of the power consumption side device according to the operation information; The distributed artificial intelligence coordination optimization control module is configured to determine the operating target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state, and to determine the operating target of the power system according to the first prediction result of the power generation state and the second prediction result of the power consumption state, and to control the power generation of the distributed energy equipment and the power consumption of the power consumption side equipment according to the environmental quality information, the operating information and the operating target.

15. A computer device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the method according to any one of claims 1 to 12 is implemented.

16. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by one or more processors, the method according to any one of claims 1 to 12 is implemented.

Images