A power demand response control method and device, a terminal and a storage medium

Abstract

The application provides a power demand response control method and device, a terminal and a storage medium, and the method comprises the following steps: acquiring power data of a power grid side and a user side in a previous period; inputting the power data into a trained power load prediction model to predict power load demand in a next period; based on the power load demand, determining scheduling information corresponding to each power equipment of the user side by using a particle swarm optimization algorithm, and controlling each power equipment according to the scheduling information. By inputting the power data in the previous period into the trained power load prediction model, the power load demand in the next period is predicted, the change of the power grid load can be quickly and accurately reflected, the scheduling information is determined by using the particle swarm optimization algorithm, the optimal scheduling of the power equipment is realized, the delay of the power grid regulation is avoided, and the energy utilization efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a power demand response control method, device, terminal and storage medium. Background Technology

[0002] With the rapid growth of global energy demand, the balance between power supply and demand has become an increasingly critical factor restricting the stability and sustainable development of the power grid. Traditional power systems meet this growing demand by building more power plants and grid facilities; however, this approach not only requires substantial investment but also places enormous pressure on the environment and resources. Furthermore, the widespread application of renewable energy sources (such as wind and solar power) has led to fluctuations and uncertainties in power supply, further increasing the difficulty of load regulation on the power grid.

[0003] To address these issues, demand response (DR) technology emerged. Demand response refers to adjusting user electricity consumption behavior during peak demand periods to reduce system load and balance grid supply and demand. The main strategies of demand response include load shedding, load shifting, and load augmentation. These measures effectively reduce reliance on traditional power generation methods, lower system operating costs, optimize grid load distribution, and improve energy efficiency.

[0004] Although demand response technology has made some progress, existing power demand response systems mostly rely on manual intervention or static dispatch, resulting in a relatively slow response process. They cannot quickly and accurately reflect changes in grid load, leading to delays in grid regulation and consequently poor energy utilization efficiency.

[0005] Therefore, existing technologies have shortcomings and need to be improved and developed. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a power demand response control method, device, terminal and storage medium to address the above-mentioned defects of the prior art, and to solve the problem that the power demand response system in the prior art relies on manual intervention or static scheduling, resulting in poor energy utilization efficiency of power equipment.

[0007] The technical solution adopted by this invention to solve the technical problem is as follows:

[0008] A power demand response control method, wherein the method includes:

[0009] Obtain power data from the grid side and the user side for the previous time period;

[0010] The power data is input into a trained power load prediction model to predict the power load demand for the next period.

[0011] Based on the power load demand, the scheduling information corresponding to each power device on the user side is determined by the particle swarm optimization algorithm, and the power devices are controlled according to the scheduling information.

[0012] In one embodiment of this application, the power data includes grid operation parameters on the grid side and electricity consumption data on the user side; the grid operation parameters include: voltage, current and frequency on the grid side; the electricity consumption data includes: electricity consumption on the user side and the operating status of each power device.

[0013] In one embodiment of this application, the power load prediction model is a long short-term memory network.

[0014] In one embodiment of this application, the scheduling information includes: start-up and stop times, operating modes, and operating power of power equipment.

[0015] In one embodiment of this application, the power demand response control method further includes:

[0016] Historical electricity consumption data of each user is obtained, and the historical electricity consumption data is classified based on a clustering analysis algorithm to obtain the electricity consumption characteristics of each user.

[0017] Based on the electricity consumption characteristics, a corresponding personalized scheduling strategy is generated and the personalized scheduling strategy is sent to the user side;

[0018] The personalized scheduling strategy includes the optimal usage time and energy-saving mode of power equipment.

[0019] In one embodiment of this application, after inputting the power data into a trained power load prediction model to predict the power load demand for the next time period, the method further includes:

[0020] Obtain the real-time grid load of the target virtual power plant and the power generation and energy storage capacity of the distributed energy resources in the target virtual power plant;

[0021] Based on the real-time grid load, power generation capacity, and energy storage capacity, a linear programming optimization model is used to schedule the distributed energy resources in the target virtual power plant.

[0022] In one embodiment of this application, after determining the scheduling information corresponding to each power device on the user side using a particle swarm optimization algorithm based on the power load demand, and controlling each power device according to the scheduling information, the method further includes:

[0023] Real-time power data and user feedback data from both the grid and user sides are acquired, and the power load demand is adjusted based on the real-time power data and user feedback data to obtain the current power load demand.

[0024] The current scheduling strategy is obtained based on the current power load demand, the current scheduling strategy is executed, and the execution result is monitored.

[0025] Adjust the current scheduling strategy based on the execution results to complete loop optimization.

[0026] This application also provides a power demand response control device, wherein the device includes:

[0027] The acquisition module is used to acquire power data from the grid side and the user side in the previous time period;

[0028] The prediction module inputs the power data into a trained power load prediction model to predict the power load demand for the next period.

[0029] The control module, based on the power load demand, determines the scheduling information corresponding to each power device on the user side through a particle swarm optimization algorithm, and controls each power device according to the scheduling information.

[0030] This application also provides a terminal, comprising: a memory, a processor, and a power demand response control program stored in the memory and executable on the processor, wherein the power demand response control program, when executed by the processor, implements the steps of the power demand response control method as described above.

[0031] This application also provides a computer-readable storage medium storing a computer program that can be executed to implement the steps of the power demand response control method as described above.

[0032] This invention provides a power demand response control method, device, terminal, and storage medium. The method includes: acquiring power data from the grid side and the user side in the previous time period; inputting the power data into a trained power load prediction model to predict the power load demand in the next time period; based on the power load demand, determining the scheduling information corresponding to each power device on the user side using a particle swarm optimization algorithm, and controlling each power device according to the scheduling information. This invention, by inputting power data from the previous time period into a trained power load prediction model to predict the power load demand in the next time period, can quickly and accurately reflect changes in grid load. Furthermore, by using a particle swarm optimization algorithm to determine the scheduling information, optimal scheduling of power devices is achieved, avoiding grid regulation delays and thus improving energy utilization efficiency. Attached Figure Description

[0033] Figure 1 This is a flowchart of a preferred embodiment of the power demand response control method in this invention;

[0034] Figure 2 This is a structural block diagram of the power demand response system in this invention;

[0035] Figure 3 This is a functional principle block diagram of a preferred embodiment of the power demand response control device in this invention;

[0036] Figure 4 This is a functional principle block diagram of a preferred embodiment of the terminal in this invention. Detailed Implementation

[0037] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0038] Existing demand response systems often rely heavily on manual intervention or static scheduling, resulting in delayed responses and an inability to quickly and accurately reflect changes in grid load. This leads to delayed grid regulation and increased load fluctuation risks. Furthermore, existing systems largely depend on simple load forecasting models and rule-driven control strategies, failing to achieve precise load scheduling and dynamic optimization, resulting in inefficient power dispatch and unsatisfactory response outcomes. This invention proposes a power demand response control method that addresses the problems of poor real-time performance and insufficient response accuracy in existing demand response systems. By introducing intelligent control, real-time data analysis, and user interaction mechanisms, this invention can more accurately regulate grid load, improving system stability and energy utilization efficiency.

[0039] Please see Figure 1 , Figure 1 This is a flowchart of the power demand response control method in this invention. For example... Figure 1 As shown, the power demand response control method described in this embodiment of the invention includes:

[0040] Step S100: Obtain the power data of the grid side and the user side in the previous time period.

[0041] like Figure 2As shown, the demand response system of this application can be equipped with a data acquisition module 100 to collect and store power data from the grid side and the user side in real time. In one embodiment, the data acquisition module 100 also integrates a real-time monitoring unit 101 and a data transmission unit 102; the real-time monitoring unit 101 is used to connect to smart meters, sensors, and user-side power equipment installed in the circuit on the user side via the Internet of Things to obtain power consumption data on the user side, and is also used to connect to the grid side via sensors to collect grid operating parameters. The data transmission unit 102 is used to transmit the power data composed of the power consumption data and grid operating parameters adopted by the real-time monitoring unit 101 to the load forecasting module 200.

[0042] In this embodiment of the application, the power data includes grid operation parameters on the grid side and power consumption data on the user side; the grid operation parameters include: voltage, current and frequency on the grid side; the power consumption data includes: power consumption on the user side and the operating status of each power device.

[0043] like Figure 1 As shown, the power demand response control method described in this embodiment further includes:

[0044] Step S200: Input the power data into the trained power load prediction model to predict the power load demand for the next period.

[0045] The demand response system of this application embodiment can be equipped with a load forecasting module 200. The load forecasting module 200 is used to perform power load forecasting based on the power load forecasting model and real-time collected power data, obtain the load forecasting results, and dynamically adjust the power load forecasting model in combination with the power grid demand.

[0046] The load forecasting module 200 includes a data preprocessing unit 201, a load forecasting model unit 202, and a dynamic adjustment unit 203. The data preprocessing unit 201 is used to clean, denoise, and normalize power data. The load forecasting model unit 202 is used to perform load forecasting based on the collected real-time power data and historical load data, using a power load forecasting model to calculate the power load demand within a preset period. The dynamic adjustment unit 203 is used to dynamically adjust the load forecasting model according to changes in grid demand, combined with renewable energy generation data and weather forecast information.

[0047] In this embodiment, the power load forecasting model is a long short-term memory network. Existing systems often rely on simple load forecasting models and rule-driven control strategies, which cannot achieve accurate load scheduling and dynamic optimization, resulting in inefficient power scheduling and unsatisfactory response.

[0048] The power load forecasting model uses a long short-time memory network for load forecasting, and the load forecasting calculation formula is as follows:

[0049] h t =σ(W hh h t-1 +W hx x t +b h )

[0050]

[0051] Among them, h t h represents the hidden state at time t. t-1 Let x be the hidden state at time t-1, used to update the hidden state at the current time; t The input features at the current moment represent historical electricity load data; W hh W is the hidden layer weight matrix, used to connect the hidden state from the previous time step to the hidden state at the current time step; hx The input weight matrix is ​​used to connect the input features x at the current time step. t The hidden state at the current moment; b h σ is the hidden layer bias term, used to adjust the output of the hidden layer; σ is the activation function, w hy This is the output layer weight matrix, used to weight the hidden state h. t Mapped to predicted value; b y This is the output layer bias term, used to adjust the prediction results; Let t be the predicted load value at time t, i.e., the predicted power load demand.

[0052] For example, the power data acquired in this application from both the grid and user sides for the previous time period includes users' electricity consumption over the past week, weather data (temperature, humidity, etc.), and the real-time power load of the grid. Using a Long Short-Term Memory (LSTM) model, the system can predict the power load demand for the next time period (e.g., the next 24 hours). Furthermore, based on weather changes and fluctuations in renewable energy generation, the parameters of the prediction model are dynamically adjusted to more accurately address demand fluctuations, providing accurate load forecasts and supporting power system dispatch decisions. This dynamic adjustment mechanism enhances the system's ability to respond to sudden load changes and external factors.

[0053] This application, by introducing a Long Short-Term Memory (LSTM) network, can capture the long-term dependencies in power load data and exhibits high accuracy when processing nonlinear and complex time-series data. Compared to traditional load forecasting methods, it can better handle the periodic fluctuations and sudden changes in power load. Through precise load forecasting, this invention can more accurately predict future power demand, allowing for advance grid dispatching, reducing load fluctuations, and avoiding power system overload or energy waste, thereby improving system operating efficiency and stability. In short, this application improves the accuracy of power load forecasting.

[0054] like Figure 1 As shown, the power demand response control method described in this embodiment further includes:

[0055] Step S300: Based on the power load demand, determine the scheduling information corresponding to each power device on the user side using the particle swarm optimization algorithm, and control each power device according to the scheduling information.

[0056] The demand response system of this application embodiment can be equipped with a scheduling management module 300. The scheduling management module 300 is used to schedule user-side power equipment according to load forecast results, adjust the operating mode and time period of the power equipment, and adjust peak load. The scheduling management module 300 includes: a scheduling decision unit 301, an equipment control unit 302, and a feedback adjustment unit 303. The scheduling decision unit 301 is used to schedule user-side power equipment using a particle swarm optimization algorithm based on power load demand and a power demand curve generated based on the power load demand. The equipment control unit 302 is used to execute scheduling decisions, control the switching and power regulation of power equipment, balance the grid load, and reduce peak load. The feedback adjustment unit 303 is used to adjust the equipment scheduling strategy in real time according to changes in power load, optimize equipment utilization efficiency, and avoid over-scheduling or equipment loss.

[0057] A user's electrical equipment typically includes multiple devices, such as air conditioners, water heaters, and washing machines. This application uses a particle swarm optimization (PSO) algorithm to optimize the start-up and shutdown times and operating power of these devices based on power load demand. For example, air conditioners can start during periods of low power demand (such as nighttime or off-peak hours) and adjust their power to avoid excessive grid load. Washing machines may operate during off-peak hours to avoid starting during periods of high power load, reducing pressure on the grid. Through these optimizations, the system can balance the grid load, reduce peak power load, lower power consumption, and improve the user's electricity economy.

[0058] This application's embodiments, by simulating the collective behavior of particle swarm optimization, can effectively find the optimal scheduling scheme for power equipment, thereby minimizing system operating costs or maximizing economic benefits. The scheduling decision considers different characteristics of the equipment (such as power demand, scheduling periods, etc.) and seeks the global optimal solution through iterative iteration. By optimizing the scheduling scheme for power equipment, this invention reduces overuse or idleness of equipment, reduces energy waste, lowers the peak-valley difference in the power grid, and balances power demand and supply, thereby effectively reducing electricity costs and improving the economic efficiency of the power grid. Therefore, this application optimizes power equipment scheduling and reduces costs.

[0059] In this embodiment of the application, the scheduling information includes: the start-up and shutdown time, operating mode, and operating power of the power equipment.

[0060] Specifically, when the scheduling decision unit 301 schedules user-side power equipment using the particle swarm optimization algorithm, it schedules the user-side power equipment by minimizing a preset first objective function, and calculates the optimal equipment operation mode. The formula for calculating the first objective function is as follows:

[0061]

[0062] In the formula, f(x) is the first objective function, representing the cost function to be optimized, used to minimize the deviation between peak electricity demand and equipment scheduling periods; p i Let T represent the power demand of the i-th power device, and let T represent the power consumption of the power device during dispatch. i Let T be the scheduling time of the i-th power equipment, that is, the operating period of the power equipment at a certain moment; o This refers to the operating period when the load of the i-th power device is lower than the preset load threshold, where n is the total number of power devices.

[0063] For example, if a user has an air conditioner, refrigerator, and washing machine, the personalized response module 400 first analyzes the user's historical electricity consumption data. Combining this with weather forecasts and air conditioner usage frequency, it discovers that the user uses the air conditioner frequently during periods of higher summer temperatures. Based on this analysis, the system generates a personalized scheduling plan, scheduling air conditioner use during periods of lower electricity demand and suggesting that the user perform high-power operations such as washing clothes when electricity prices are lower, thereby optimizing the user's electricity usage. Through this personalized scheduling, users can reduce electricity costs while ensuring a comfortable electricity experience, supporting the stable operation of the power grid, accurately identifying different users' electricity consumption patterns and behavioral characteristics, achieving differentiated power scheduling, optimizing the user's electricity experience, and reducing electricity costs.

[0064] In this embodiment of the application, the power demand response control method further includes: acquiring historical electricity consumption data of each user, classifying each historical electricity consumption data based on a clustering analysis algorithm to obtain the electricity consumption characteristics of each user; generating a corresponding personalized scheduling strategy based on the electricity consumption characteristics, and sending the personalized scheduling strategy to the user side; wherein, the personalized scheduling strategy includes the optimal usage time and energy-saving mode of power equipment.

[0065] Specifically, the demand response system in this embodiment can also be equipped with a personalized response module 400, used to analyze user electricity consumption characteristics based on historical electricity consumption data and behavioral data, generate personalized scheduling strategies, and dynamically optimize them in real time based on user feedback. The personalized response module 400 includes: a user behavior analysis unit 401, a personalized scheduling unit 402, and a real-time feedback unit 403. The user behavior analysis unit 401 is used to identify user electricity consumption characteristics using cluster analysis by analyzing historical electricity consumption data, behavioral data, and weather change factors. The personalized scheduling unit 402 is used to generate personalized power scheduling schemes based on user analysis results, wherein the power scheduling schemes include optimal equipment usage time and energy-saving modes. The real-time feedback unit 403 is used to dynamically optimize the user's personalized scheduling scheme by receiving user electricity consumption feedback data.

[0066] In this embodiment, the clustering analysis algorithm is the K-means clustering algorithm. Based on the clustering analysis algorithm, the historical electricity consumption data are classified to obtain the electricity consumption characteristics of each user, specifically including:

[0067] Input the user's historical electricity consumption dataset X = {x1, x2, ..., x...} i}, where x i The electricity consumption characteristics of the i-th user;

[0068] Randomly initialize k cluster centers C1, C2, ..., C k ;

[0069] For each user x i Calculate the Euclidean distance between it and each cluster center, and assign it to the nearest cluster center;

[0070] Update the cluster center positions until convergence, obtain the clustering results, i.e. the electricity consumption characteristics of each user, and finally generate a personalized scheduling strategy for each user based on the clustering results.

[0071] This application embodiment can categorize users into different groups based on their historical electricity consumption data, identifying the electricity consumption characteristics of different user groups. Based on these characteristics, a personalized power demand response scheduling plan is developed for each user, flexibly adjusting the user's electricity consumption periods or load allocation. This personalized scheduling plan optimizes power allocation based on the user's actual needs, load characteristics, and electricity consumption habits, not only improving user comfort and satisfaction but also avoiding resource waste caused by a single global scheduling strategy. Simultaneously, it enhances the flexibility and accuracy of the demand response system and reduces grid load fluctuations.

[0072] In one embodiment of this application, after step S200, the method further includes: obtaining the real-time grid load of the target virtual power plant and the power generation capacity and energy storage capacity of the distributed energy in the target virtual power plant; and scheduling the distributed energy in the target virtual power plant using a linear programming optimization model based on the real-time grid load, power generation capacity and energy storage capacity.

[0073] The demand response system in this embodiment can also include an energy management module 500. This energy management module 500 integrates distributed energy resources on the user side into a target virtual power plant, schedules the distributed energy resources within the target virtual power plant, and optimizes power dispatch strategies and adjusts response control system parameters based on real-time data feedback. The energy management module 500 includes a virtual power plant dispatch unit 501 and an energy flow optimization unit 502. The virtual power plant dispatch unit 501 integrates and dispatches distributed energy resources to construct a virtual power plant and optimize resource allocation. The energy flow optimization unit 502 uses linear programming to schedule energy resources within the virtual power plant based on real-time grid load and the generation and storage capacity of distributed energy resources to meet grid load demands.

[0074] The optimization objective of the linear programming optimization model is to minimize the total energy consumption of the system. The constraints include grid load and equipment energy storage capacity. The linear programming optimization model is expressed as follows:

[0075]

[0076] In the formula, Z is the objective function of the linear programming optimization model, representing the total energy consumption; c i Let x be the unit energy cost of the i-th distributed energy source, representing the cost per unit of energy consumed; i Let be the scheduling quantity of the i-th energy resource; n be the total number of distributed energy resources; and the constraints of the linear programming are:

[0077] A·x≤b

[0078] In the formula, A is the constraint matrix, which contains the scheduling relationship of each energy resource in different time periods and the constraints of various resources; x is the variable vector, representing the scheduling amount of all energy resources; b is the constraint vector, representing the specific values ​​of various constraints, such as maximum energy storage capacity, minimum output limit, load demand, etc.

[0079] For example, a virtual power plant in a region consists of multiple distributed energy resources, including multiple wind turbines, solar panels, and energy storage systems. The virtual power plant dispatch unit determines which energy resources should be prioritized for dispatch and how to use the energy storage system to meet grid demand based on real-time grid load data and available energy resources (such as solar and wind power generation). For instance, if wind turbines are generating more electricity than the grid load is low, the system will select the energy storage system to store the surplus energy and release it during peak demand periods, reducing reliance on traditional power generation resources.

[0080] This application uses a linear programming algorithm and an energy flow optimization unit to ensure optimal scheduling of energy resources, thereby minimizing energy consumption and guaranteeing grid load balance.

[0081] This application employs a linear programming optimization model, which can efficiently handle complex energy dispatching problems. It considers the different characteristics and constraints of multiple energy resources (such as energy storage capacity and minimum output), minimizing energy consumption while meeting grid load demands. Through precise resource allocation, the system can effectively dispatch distributed energy resources, improving the overall efficiency of the virtual power plant. This invention reduces energy consumption and improves the system's economic efficiency through the rational dispatching of energy resources. Simultaneously, the linear programming optimization model ensures the system operates under various constraints, guaranteeing grid power supply stability and reducing power outages or equipment overloads caused by sudden load fluctuations. Therefore, this application optimizes energy resource dispatching and enhances system stability.

[0082] In this embodiment of the application, after step S300, the method further includes: acquiring power data and user feedback data from the grid side and the user side in real time; adjusting the power load demand based on the real-time power data and user feedback data to obtain the current power load demand; obtaining the current scheduling strategy based on the current power load demand; executing the current scheduling strategy and monitoring the execution result; and adjusting the current scheduling strategy based on the execution result to complete the cyclic optimization.

[0083] Specifically, based on real-time power load data and user feedback data, a feedback loop-based closed-loop control method is adopted to optimize the scheduling strategy. User feedback data can include: electricity consumption data, electricity usage behavior data, electricity bill feedback, and user suggestions. Electricity usage behavior data refers to records of users' electricity consumption patterns over different time periods, including peak and off-peak electricity consumption, electricity consumption of specific devices (such as air conditioners and water heaters), and whether users have responded to the electricity demand response plan (such as reducing electricity consumption or adjusting usage time). Electricity bill feedback refers to user feedback on electricity bills, including the accuracy of the bills, changes in electricity prices, and satisfaction with the electricity pricing structure. User suggestions refer to suggestions and complaints submitted by users through customer service channels regarding electricity supply, prices, and services; this information can reveal users' expectations and needs for electricity services.

[0084] This invention achieves intelligent scheduling and optimized control of power equipment through the collaborative work of multiple modules, combining load forecasting, personalized scheduling, and energy management. The scheduling management module uses a particle swarm optimization algorithm to schedule power equipment, reducing peak grid load and minimizing energy waste. The personalized response module generates customized scheduling schemes based on historical user data and behavioral analysis, improving user electricity efficiency. The energy management module enhances the sustainability and economy of the power system through the integration and scheduling of virtual power plants and distributed energy resources.

[0085] This invention provides highly automated and intelligent decision support for power demand response systems. It can sense changes in power demand in real time and respond rapidly, thereby optimizing grid dispatching, equipment operation, and energy allocation. The system enables automated load forecasting, equipment dispatching, and resource optimization, significantly improving the intelligence level of the power system, reducing manual intervention, and increasing work efficiency. Simultaneously, the system can dynamically adjust decisions based on real-time data, enhancing the power system's adaptability to external disturbances such as load changes and unexpected events, ensuring the long-term stable operation of the power system.

[0086] The optimized scheduling of this invention not only improves the economic efficiency of the power system but also reduces energy waste and unnecessary power consumption. By precisely scheduling distributed energy resources and reducing peak-valley differences, the system can effectively reduce its dependence on traditional thermal power generation, reduce carbon emissions, promote the application of green energy, save energy, reduce environmental pollution, conform to the concept of sustainable development, and contribute to the construction of a low-carbon and environmentally friendly power system.

[0087] In one embodiment, such as Figure 3 As shown, based on the above-described power demand response control method, the present invention also provides a power demand response control device, comprising:

[0088] The acquisition module 10 is used to acquire power data from the grid side and the user side in the previous time period;

[0089] The prediction module 20 inputs the power data into the trained power load prediction model to predict the power load demand for the next period.

[0090] The control module 30 determines the scheduling information corresponding to each power device on the user side through the particle swarm optimization algorithm based on the power load demand, and controls each power device according to the scheduling information.

[0091] Figure 4 A schematic diagram of the structure of a terminal provided in an embodiment of this application. The terminal may include:

[0092] The memory 501, the processor 502, and the computer program stored on the memory 501 and capable of running on the processor 502.

[0093] When processor 502 executes the program, it implements the power demand response control method provided in the above embodiments.

[0094] Furthermore, the terminal also includes:

[0095] Communication interface 503 is used for communication between memory 501 and processor 502.

[0096] The memory 501 is used to store computer programs that can run on the processor 502.

[0097] The memory 501 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0098] If the memory 501, processor 502, and communication interface 503 are implemented independently, they can be interconnected via a bus to communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, only one line is used in the diagram, but this does not imply that there is only one bus or one type of bus.

[0099] Optionally, in a specific implementation, if the memory 501, processor 502, and communication interface 503 are integrated on a single chip, then the memory 501, processor 502, and communication interface 503 can communicate with each other through an internal interface.

[0100] Processor 502 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0101] This embodiment also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the power demand response control method described above.

[0102] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0103] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0104] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0105] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can read and execute instructions from or in conjunction with such an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by or in conjunction with an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). In addition, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically by optically scanning paper or other media, followed by editing, interpreting or otherwise processing as necessary, and then stored in computer memory.

[0106] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0107] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it includes one or a combination of the steps of the method embodiments.

[0108] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

[0109] In summary, this invention discloses a power demand response control method, device, terminal, and storage medium. The method includes: acquiring power data from the grid side and the user side in the previous time period; inputting the power data into a trained power load prediction model to predict the power load demand in the next time period; determining the scheduling information corresponding to each power device on the user side based on the power load demand using a particle swarm optimization algorithm, and controlling each power device according to the scheduling information. This invention, by inputting power data from the previous time period into a trained power load prediction model to predict the power load demand in the next time period, can quickly and accurately reflect changes in grid load. Furthermore, by using a particle swarm optimization algorithm to determine the scheduling information, it achieves optimal scheduling of power devices, avoids grid regulation delays, and thus improves energy utilization efficiency.

[0110] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A power demand response control method, characterized in that, The method includes: Obtain power data from the grid side and the user side for the previous time period; The power data is input into a trained power load prediction model to predict the power load demand for the next period. Based on the power load demand, the scheduling information corresponding to each power device on the user side is determined by the particle swarm optimization algorithm, and the power device is controlled according to the scheduling information. The power load prediction model is a long short-time memory network; The power demand response control method further includes: Historical electricity consumption data of each user is obtained, and the historical electricity consumption data is classified based on a clustering analysis algorithm to obtain the electricity consumption characteristics of each user. Based on the electricity consumption characteristics, a corresponding personalized scheduling strategy is generated and the personalized scheduling strategy is sent to the user side; The personalized scheduling strategy includes the optimal usage time and energy-saving mode of power equipment; After inputting the power data into a trained power load forecasting model to predict the power load demand for the next time period, the method further includes: Obtain the real-time grid load of the target virtual power plant and the power generation and energy storage capacity of the distributed energy resources in the target virtual power plant; Based on the real-time grid load, power generation capacity, and energy storage capacity, a linear programming optimization model is used to schedule the distributed energy resources in the target virtual power plant. Based on the aforementioned power load demand, after determining the scheduling information corresponding to each power device on the user side using a particle swarm optimization algorithm, and controlling each power device according to the scheduling information, the process further includes: Real-time power data and user feedback data from both the grid and user sides are acquired, and the power load demand is adjusted based on the real-time power data and user feedback data to obtain the current power load demand. The current scheduling strategy is obtained based on the current power load demand, the current scheduling strategy is executed, and the execution result is monitored. Adjust the current scheduling strategy based on the execution results to complete loop optimization; The formula for the power load forecasting model is as follows: in, For the first The hidden state at any given moment; For the first The hidden state at each moment is used to update the hidden state at the current moment; The input features at the current moment represent historical power load data; This is the hidden layer weight matrix, used to connect the hidden state of the previous time step to the hidden state of the current time step; The input weight matrix is ​​used to connect the input features at the current time step. The hidden state at the current moment; This is the hidden layer bias term, used to adjust the output of the hidden layer; For activation function, This is the output layer weight matrix, used to weight the hidden states. Mapped to predicted values; This is the output layer bias term, used to adjust the prediction results; For the first The power load demand at any given time; The user-side power equipment is scheduled by minimizing a preset first objective function, the formula for which the first objective function is calculated is: In the formula, Let be the first objective function, representing the cost function that needs to be optimized, used to minimize the deviation between peak electricity demand and equipment scheduling periods; For the first The power demand of an electrical device represents the power consumption of the electrical device during dispatch. For the first The scheduling time of an electrical device, that is, the operating period of an electrical device at a certain moment; For the first During operation periods when the load of a power device is lower than a preset load threshold, The total number of electrical equipment; The clustering analysis algorithm is the K-means clustering algorithm. Based on the clustering analysis algorithm, the historical electricity consumption data are classified to obtain the electricity consumption characteristics of each user. Specifically, it includes: inputting the user's historical electricity consumption dataset. ,in, For the first Electricity consumption characteristics of individual users; random initialization Cluster centers For each user Calculate the Euclidean distance between each user and each cluster center, assign the user to the nearest cluster center, update the cluster center positions until convergence, and obtain the electricity consumption characteristics of each user. The optimization objective of the linear programming optimization model is to minimize the total energy consumption of the system. The constraints include grid load and equipment energy storage capacity. The linear programming optimization model is expressed as follows: In the formula, Let be the objective function of the linear programming optimization model, representing the total energy consumption; For the first The unit energy cost of a distributed energy source represents the cost per unit of energy consumed. For the first The amount of energy resources allocated; Let be the total number of distributed energy resources; where the constraints of the linear programming are: In the formula, The constraint matrix contains the scheduling relationships of each energy resource in different time periods and the constraints on various resources; Let be a variable vector, representing the total amount of energy resources to be allocated; This is a constraint vector, representing the specific values ​​of various constraints.

2. The power demand response control method according to claim 1, characterized in that, The power data includes grid operation parameters on the grid side and electricity consumption data on the user side; The power grid operating parameters include: voltage, current and frequency on the power grid side; the electricity consumption data includes: electricity consumption on the user side and the operating status of each power device.

3. The power demand response control method according to claim 1, characterized in that, The scheduling information includes: start-up and shutdown times, operating modes, and operating power of power equipment.

4. A power demand response control device, characterized in that, The device includes: The acquisition module is used to acquire power data from the grid side and the user side in the previous time period; The prediction module inputs the power data into a trained power load prediction model to predict the power load demand for the next period. The control module, based on the power load demand, determines the scheduling information corresponding to each power device on the user side through a particle swarm optimization algorithm, and controls each power device according to the scheduling information; The power load prediction model is a long short-time memory network; The power demand response control device is used for: Historical electricity consumption data of each user is obtained, and the historical electricity consumption data is classified based on a clustering analysis algorithm to obtain the electricity consumption characteristics of each user. Based on the electricity consumption characteristics, a corresponding personalized scheduling strategy is generated and the personalized scheduling strategy is sent to the user side; The personalized scheduling strategy includes the optimal usage time and energy-saving mode of power equipment; Obtain the real-time grid load of the target virtual power plant and the power generation and energy storage capacity of the distributed energy resources in the target virtual power plant; Based on the real-time grid load, power generation capacity, and energy storage capacity, a linear programming optimization model is used to schedule the distributed energy resources in the target virtual power plant. Real-time power data and user feedback data from both the grid and user sides are acquired, and the power load demand is adjusted based on the real-time power data and user feedback data to obtain the current power load demand. The current scheduling strategy is obtained based on the current power load demand, the current scheduling strategy is executed, and the execution result is monitored. Adjust the current scheduling strategy based on the execution results to complete loop optimization; The formula for the power load forecasting model is as follows: in, For the first The hidden state at any given moment; For the first The hidden state at each moment is used to update the hidden state at the current moment; The input features at the current moment represent historical power load data; This is the hidden layer weight matrix, used to connect the hidden state of the previous time step to the hidden state of the current time step; The input weight matrix is ​​used to connect the input features at the current time step. The hidden state at the current moment; This is the hidden layer bias term, used to adjust the output of the hidden layer; For activation function, This is the output layer weight matrix, used to weight the hidden states. Mapped to predicted values; This is the output layer bias term, used to adjust the prediction results; For the first The power load demand at any given time; The user-side power equipment is scheduled by minimizing a preset first objective function, the formula for which the first objective function is calculated is: In the formula, Let be the first objective function, representing the cost function that needs to be optimized, used to minimize the deviation between peak electricity demand and equipment scheduling periods; For the first The power demand of an electrical device represents the power consumption of the electrical device during dispatch. For the first The scheduling time of an electrical device, that is, the operating period of an electrical device at a certain moment; For the first During operation periods when the load of a power device is lower than a preset load threshold, The total number of electrical equipment; The clustering analysis algorithm is the K-means clustering algorithm. Based on the clustering analysis algorithm, the historical electricity consumption data are classified to obtain the electricity consumption characteristics of each user. Specifically, it includes: inputting the user's historical electricity consumption dataset. ,in, For the first Electricity consumption characteristics of individual users; random initialization Cluster centers For each user Calculate the Euclidean distance between each user and each cluster center, assign the user to the nearest cluster center, update the cluster center positions until convergence, and obtain the electricity consumption characteristics of each user. The optimization objective of the linear programming optimization model is to minimize the total energy consumption of the system. The constraints include grid load and equipment energy storage capacity. The linear programming optimization model is expressed as follows: In the formula, Let be the objective function of the linear programming optimization model, representing the total energy consumption; For the first The unit energy cost of a distributed energy source represents the cost per unit of energy consumed. For the first The amount of energy resources allocated; Let be the total number of distributed energy resources; where the constraints of the linear programming are: In the formula, The constraint matrix contains the scheduling relationships of each energy resource in different time periods and the constraints on various resources; Let be a variable vector, representing the total amount of energy resources to be allocated; This is a constraint vector, representing the specific values ​​of various constraints.

5. A terminal, characterized in that, include: The device includes a memory, a processor, and a power demand response control program stored in the memory and executable on the processor, wherein the power demand response control program, when executed by the processor, implements the steps of the power demand response control method as described in any one of claims 1 to 3.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that can be executed to implement the steps of the power demand response control method as described in any one of claims 1 to 3.

Images