Power grid demand side response optimization method and system based on multilevel game analysis

Abstract

The invention discloses a power grid demand side response optimization method and system based on multi-level game analysis. The method comprises the following steps: firstly, constructing a hierarchical model based on a distributed multi-level game structure; analyzing real-time load data of the power grid and constructing a load prediction model in a power grid operator layer; in the power consumer layer, constructing a demand response model for the power consumers of each area unit; constructing a service provider response optimization model on the service provider layer; and finally, obtaining a distributed power grid demand side response optimization scheme. According to the method, participation subjects of demand response are divided, a dynamic interaction mechanism between layers is established through a multi-layer game model, collaborative optimization of an optimal strategy is realized in subjects in different area units in a distributed architecture in a layered game mode, the problem that a traditional single-layer optimization model cannot balance multi-subject benefits is effectively solved, and the optimization efficiency is improved. And the precision and coordination of the demand response are obviously improved under the scenes of complex load fluctuation and multi-party participation.

Description

Technical Field

[0001] The present invention relates to a demand side response optimization method, and belongs to the technical field of power grid optimization, and in particular to a power grid demand side response optimization method and system based on multi-level game analysis. Background Art

[0002] With the growing demand for power grids and the increasing proportion of renewable energy, grid demand-side response has gradually become an important technical means in grid management. By guiding power users to adjust their load usage habits, it can alleviate the imbalance between supply and demand of the power grid and reduce peak load pressure to a certain extent. However, in actual applications, there are many technical difficulties in the formulation and optimization of demand-side response strategies.

[0003] At present, most traditional demand-side response optimization methods adopt a single-level centralized optimization model, which is usually based on the grid operator's prediction of load and simple assumptions about user behavior to formulate a unified demand response strategy. The single-level optimization model has obvious limitations in dealing with the complex dynamic interactions between the grid, power users and service providers: first, the roles and interests of power users and service providers in demand-side response are different, and a single model is difficult to fully reflect the strategic interaction and interest balance between the various subjects; second, it is difficult to effectively deal with the nonlinear strategic relationship and dynamic game problems of different participating subjects, resulting in poor applicability of the optimization results. In addition, since the load changes of the power grid are highly real-time, and the traditional demand response optimization methods are usually based on static data or offline analysis, the power grid has response lags or decision-making errors when facing sudden load fluctuations, which in turn affects the operating efficiency and stability of the power system. In addition, it is difficult to meet the demand response scenarios where large-scale power users and service providers coexist in terms of computational efficiency, and lacks the ability to adapt to large-scale parallel optimization.

[0004] Although some existing solutions have attempted to introduce game theory into demand-side response optimization to improve the accuracy of demand response by simulating the strategic interaction between different subjects, the existing game analysis methods usually remain at a single level or small-scale simulation, which is difficult to adapt to the multi-level complex structure of the actual power grid, and the existing distributed architecture still has deficiencies in real-time synchronization and global coordination, which easily leads to the problem of uncoordinated response between different regions. Therefore, there is an urgent need for a method that can meet the needs of multi-agent collaborative optimization and has good regional coordination to solve the above-mentioned defects in the existing technology. Summary of the invention

[0005] The purpose of the present invention is to overcome the above-mentioned defects and problems existing in the prior art, and to provide a power grid demand side response optimization method and system based on multi-level game analysis that can meet the requirements of multi-agent collaborative optimization and has good regional coordination.

[0006] To achieve the above objectives, the technical solution of the present invention is: an optimization method for grid demand-side response based on multi-level game analysis, including:

[0007] S1. Obtain the real-time grid load data, the demand response parameters of power users, and the service capacity parameters of each regional service provider, and divide them into different regional units; construct a hierarchical model based on a distributed multi-level game structure, and perform regional hierarchical modeling on grid operators, power users, and service providers within each regional unit;

[0008] S2. At the grid operator level, analyze the real-time grid load data of each regional unit, construct a load forecasting model, and calculate the demand response optimization objective function at the grid operator level. Use the calculation result as the strategy guidance of the grid operator at this regional level for the lower-level power user layer, generate regional load management requirements, and synchronize them to the power user layer of the corresponding region;

[0009] S3. At the power user level, based on the demand response parameters of power users and the strategy guidance received from the grid operator, construct a demand response model for power users in each regional unit. Calculate the load response plan and adjustment strategy of power users according to the demand response model, use the load response plan and adjustment strategy as the demand input to the service provider layer, and synchronize them to the service provider layer of each regional unit;

[0010] S4. At the service provider level, based on the service capacity parameters of each regional service provider and the load response requirements of power users, construct a service provider response optimization model, calculate the optimal service strategy of each service provider, and feedback the optimal service strategy to the power user layer;

[0011] S5. In the hierarchical model of the multi-level game structure, based on parallel computing technology, synchronously execute the game optimization processes of the grid operator layer, power user layer, and service provider layer on the computing nodes of each regional unit, determine the optimal response strategy combination of each regional unit, and obtain a distributed grid demand-side response optimization solution.

[0012] The specific steps of step S1 include:

[0013] S11. Collect the real-time grid load data D i (t), the set of demand response parameters of power users the set of service capacity parameters of each regional service provider

[0014] S12. Based on the grid topology and geographical location, divide grid operators, power users, and service providers into different regional units, and its expression is as follows:

[0015] R = {R1, R2,..., R i ,..., R N};

[0016] Wherein: R i represents the i-th regional unit, including the corresponding grid operator O i , the set of electricity users U i and the set of service providers S i ;

[0017] S13. In each regional unit R i , based on the real-time grid load data D i (t), the set of demand response parameters of electricity users and the set of service capacity parameters of each regional service provider respectively construct a grid operator layer model based on a distributed multi-level game structure an electricity user layer model and a service provider layer model and establish a regional hierarchical game model

[0018] S14. Connect the hierarchical game models G i within each regional unit R (i) through a distributed architecture to form an overall distributed multi-level game model G = {G (1) , G (2) ,..., G (N)}.

[0019] The step S2 specifically includes:

[0020] S21. At the grid operator layer, analyze the real-time grid load data D i of each regional unit R i (t), construct a load forecasting model f i (t), and predict the load demand at a future time and the load demand within a future time period to obtain a predicted load sequence

[0021] The expression of the load demand at the future time is as follows:

[0022]

[0023] Wherein: is the predicted load of the i-th regional unit at the future time t + Δt, W i (t) is the set of weather factor parameters of the i-th regional unit, H i (t) is the set of historical load behavior parameters of the i-th regional unit;

[0024] S22. Based on the predicted load sequence Establish the demand response optimization objective function at the grid operator level, and consider the strategic interaction between the lower-level power users and service providers in combination with the multi-level game model. Its expression is as follows:

[0025]

[0026] Where: is the demand response optimization objective function, is the strategy set of the grid operator at the future time t + τ in the i-th regional unit, including demand response incentive measures and load regulation plans; is the operating cost function of the grid operator, is the penalty coefficient of the grid operator, is the load deviation penalty function;

[0027] S23. Under the condition of meeting the grid operation constraints, solve the demand response optimization objective function Obtain the optimal strategy of the grid operator in the regional unit R i

[0028] S24. Take the optimal strategy as the strategy guidance of the regional grid operator for the lower-level power user layer, generate the regional load management requirements, and synchronize them to the power user layer of the corresponding region to guide the power users to adjust the load, so as to optimize the grid demand-side response; the regional load management requirements include the demand response incentive signal I i (τ) and the load regulation plan A i (τ).

[0029] In step S23, the operation constraints include: power balance constraint, generator output constraint and network security constraint;

[0030] The expression of the power balance constraint is as follows:

[0031]

[0032] Where: is the power generation power of the g-th generator set in the i-th regional unit at time t + τ, G i is the number of generator sets, is the power input from other regions to the i-th regional unit at time t + τ, is the load reduction obtained through demand response;

[0033] The expression of the generator output constraint is as follows:

[0034]

[0035] Where:​ and are the minimum and maximum power outputs of the g-th generating unit in the i-th regional unit, respectively;

[0036] The network security constraint is: to meet the requirements of power flow and voltage stability of the power grid transmission lines.

[0037] The step S3 specifically includes:

[0038] S31. Construct a demand response model for each electricity user and, under the constraint conditions of meeting the electricity consumption demand and comfort constraint of the electricity user itself, solve the demand response model to obtain the optimal load response plan of the electricity user and the optimal adjustment strategy

[0039] The expression of the demand response model is as follows:

[0040]

[0041] Where: is the load adjustment strategy of the j-th electricity user in the i-th regional unit at time t + τ, is the utility function of the electricity user, is the revenue function obtained by the electricity user for participating in the demand response, is the cost function for the electricity user to implement the load adjustment strategy; is the demand response ability parameter of the j-th electricity user in the i-th regional unit;

[0042] The constraint conditions are as follows:

[0043]

[0044] Where: and are the minimum and maximum allowable loads of the electricity user, respectively, is the set of feasible adjustment strategies of the electricity user;

[0045] S32. Aggregate the load response demands and load adjustment strategies of all electricity users in the same regional unit to form the optimal load response plan and the optimal adjustment strategy

[0046] The step S4 specifically includes:

[0047] S41. Construct an optimization model for the response of each service provider ​Under the constraints of meeting the capabilities of service providers and the response requirements of users, solve the service provider response optimization model to obtain the optimal service strategy of the service provider

[0048] The expression of the response optimization model is as follows:

[0049]

[0050] Where: is the service provision strategy of the k-th service provider in the i-th regional unit at time t + τ, is the utility function of the service provider, is the revenue function obtained by the service provider for providing services, is the cost function of the service provider for providing services;

[0051] The constraint conditions are as follows:

[0052]

[0053] Where: is the maximum service capacity of the service provider, and are the minimum and maximum load response requirements of regional power users respectively;

[0054] S42. Aggregate the optimal service strategies of all service providers to form the optimal service strategy set of regional service providers

[0055] The specific steps of step S5 are as follows:

[0056] S51. Use the optimal strategy of the grid operator layer the optimal adjustment strategy of the power user layer and the optimal service strategy of the service provider as the optimization variables of each layer, and synchronously execute the game optimization processes of the grid operator layer, power user layer, and service provider layer on the computing nodes of each regional unit R i to solve the initial optimization strategy and determine the final optimal response strategy combination of the regional unit

[0057] S52. Aggregate the optimal response strategy combinations of each regional unit to generate a distributed grid demand-side response optimization plan G * and satisfy the global coordination conditions;

[0058] The expression of the distributed demand-side response optimization plan is as follows:

[0059] G * ={G *(1) ,G*(2) , ..., G *(N)};

[0060] Among them: The response scheme of a single regional unit is

[0061] The global coordination conditions are as follows:

[0062]

[0063] Among them: is the total power generation in the i-th regional unit, is the input power of the i-th regional unit.

[0064] A grid demand-side response optimization system based on multi-level game analysis, which is applied to the above method. The system includes:

[0065] A hierarchical modeling module, which is used to obtain real-time grid load data, demand response parameters of power users, and service capacity parameters of each regional service provider and divide them into different regional units; construct a hierarchical model based on a distributed multi-level game structure, and perform regional hierarchical modeling on grid operators, power users, and service providers within each regional unit;

[0066] A regional load management demand generation module, which is used at the grid operator level to analyze the real-time grid load data of each regional unit, construct a load forecasting model, and calculate the demand response optimization objective function at the grid operator level. The calculation result is used as the strategy guidance of the grid operator of this region for the lower-level power user layer, generate regional load management demands, and synchronize them to the power user layer of the corresponding region;

[0067] A demand response module, which is used at the power user level to construct a demand response model for power users in each regional unit based on the demand response parameters of power users and the received strategy guidance from the grid operator, calculate the load response plan and adjustment strategy of power users according to the demand response model, use the load response plan and adjustment strategy as the demand input to the service provider layer, and synchronize them to the service provider layer of each regional unit;

[0068] A response optimization module, which is used at the service provider level to construct a service provider response optimization model based on the service capacity parameters of each regional service provider and the load response requirements of power users, calculate the optimal service strategies of each service provider, and feedback the optimal service strategies to the power user layer;

[0069] The demand-side response optimization module is used to synchronously execute the game optimization processes of the grid operator layer, the power user layer, and the service provider layer on the computing nodes of each regional unit based on parallel computing technology in the hierarchical model of the multi-level game structure, determine the optimal response strategy combination of each regional unit, and obtain a distributed grid demand-side response optimization solution.

[0070] A grid demand-side response optimization device based on multi-level game analysis, the device includes a processor and a memory;

[0071] The memory is used to store computer program code and transmit the computer program code to the processor;

[0072] The processor is used to execute the above-mentioned grid demand-side response optimization method based on multi-level game analysis according to the instructions in the computer program code.

[0073] A computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed on a computer, the above-mentioned grid demand-side response optimization method based on multi-level game analysis is realized.

[0074] Compared with the prior art, the beneficial effects of the present invention are:

[0075] 1. In the grid demand-side response optimization method and system based on multi-level game analysis of the present invention, the method first constructs a hierarchical model based on a distributed multi-level game structure, and then analyzes the real-time grid load data of each regional unit in the grid operator layer to construct a load prediction model. In the power user layer, a demand response model is constructed for the power users of each regional unit by using the demand response parameters of the power users and the strategy guidance received from the grid operator. Then, in the service provider layer, a service provider response optimization model is constructed based on the service provider service capacity parameters and the load response requirements of the power users of each regional unit. Finally, a distributed grid demand-side response optimization solution is output; in the application of this design, the participating subjects of demand response are divided, a dynamic interaction mechanism between each layer is established through a multi-level game model, and the subjects in different regional units in the distributed architecture achieve collaborative optimization of the optimal strategy in a hierarchical game manner, which can effectively solve the problem that the traditional single-layer optimization model cannot balance the interests of multiple subjects, and the multi-level game model can accurately describe the load response characteristics of power users and the service adaptation ability of service providers, significantly improving the accuracy and coordination of demand response in scenarios of complex load fluctuations and multi-party participation.

[0076] 2. In the optimization method and system for grid demand response based on multi-level game analysis of the present invention, parallel computing technology is adopted to synchronously solve the optimization processes of the grid operator layer, power user layer, and service provider layer under a distributed architecture. The computing nodes of each regional unit independently solve the optimization model, and at the same time, a fast convergence of the multi-level game strategy is achieved through an iterative mechanism, enabling it to significantly improve the computing efficiency in a large-scale grid environment, significantly reduce the time complexity of demand response optimization, and achieve real-time optimization response during peak load periods. BRIEF DESCRIPTION OF THE DRAWINGS

[0077] Figure 1 is the flowchart of the method of the present invention.

[0078] Figure 2 is the schematic diagram of the hierarchical structure of the distributed multi-level game model in Embodiment 1 of the present invention.

[0079] Figure 3 is the system structure diagram of the present invention.

[0080] Figure 4 is the device structure diagram of the present invention.

[0081] In the figure: hierarchical modeling module 1, regional load management demand generation module 2, demand response module 3, response optimization module 4, demand-side response optimization module 5, processor 6, memory 7, computer program code 71. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0082] The present invention will be further described in detail below in conjunction with the drawings and specific embodiments.

[0083] Embodiment 1:

[0084] Refer to Figure 1 , an optimization method for grid demand response based on multi-level game analysis, including:

[0085] S1. Obtain the real-time grid load data, demand response parameters of power users, and service capacity parameters of each regional service provider, and divide them into different regional units; construct a hierarchical model based on a distributed multi-level game structure, and perform regional hierarchical modeling on grid operators, power users, and service providers within each regional unit; as Figure 2 shown.

[0086] Further, step S1 specifically includes:

[0087] S11. Collect the real-time grid load data D i (t), the set of demand response parameters of power users the set of service capacity parameters of each regional service provider

[0088] D i D(t) represents the real-time grid load data of the i-th regional unit at time t, where i = 1, 2,..., N, and N is the total number of regional units;

[0089]

[0090] Where: is the demand response capability parameter of the j-th power user in the i-th regional unit, where j = 1, 2,..., M i , M i is the number of power users in the i-th regional unit;

[0091]

[0092] Where: is the service capability parameter of the k-th service provider in the i-th regional unit, where k = 1, 2,..., L i , L i is the number of service providers in the i-th regional unit;

[0093] S12. Based on the grid topology and geographical location, the grid operator, power users, and service providers are divided into different regional units, and its expression is as follows:

[0094] R = {R1, R2,..., R i ,..., R N};

[0095] Where: R i represents the i-th regional unit, including the corresponding grid operator O i , the set of power users U i and the set of service providers S i ;

[0096] S13. In each regional unit R i , according to the real-time grid load data D i (t), the set of demand response parameters of power users and the set of service capability parameters of each regional service provider respectively construct a grid operator layer model a power user layer model and a service provider layer model based on a distributed multi-level game structure, and establish a regional hierarchical game model to describe the strategic interaction and revenue relationship among the participants of each layer;

[0097] S14. The hierarchical game model G within each regional unit R i (i) ​Connect through a distributed architecture to form an overall distributed multi-level game model G = {G (1) , G (2) ,..., G (N)}.

[0098] S2. At the grid operator level, analyze the real-time grid load data of each regional unit, construct a load forecasting model, and calculate the demand response optimization objective function at the grid operator level. Use the calculation result as the strategy guidance of the grid operator in this region for the lower-level power user layer, generate regional load management requirements, and synchronize them to the power user layer in the corresponding region;

[0099] Further, the step S2 specifically includes:

[0100] S21. At the grid operator level, analyze the real-time grid load data D i (t) of each regional unit R i , construct a load forecasting model f i (t), and predict the load demand at future times and the load demand within the future time period to obtain a predicted load sequence The load forecasting model f i (t) predicts the load demand at future times based on historical load data, weather factors, and user load behavior parameters;

[0101] The expression for the load demand at future times is as follows:

[0102]

[0103] Where: is the predicted load of the i-th regional unit at future time t + Δt, W i (t) is the weather factor parameter set of the i-th regional unit, H i (t) is the historical load behavior parameter set of the i-th regional unit;

[0104] S22. Based on the predicted load sequence establish the demand response optimization objective function at the grid operator level, and consider the strategic interaction of the lower-level power users and service providers in combination with the multi-level game model. Its expression is as follows:

[0105]

[0106] Where: is the demand response optimization objective function, is the strategy set of the grid operator at future time t + τ in the i-th regional unit, including demand response incentive measures and load regulation schemes; is the operating cost function of the grid operator, depending on the strategy and predicted load is the penalty coefficient of the grid operator, which is used to balance the operating cost and the load deviation penalty; is the load deviation penalty function, which considers the optimal response strategies of power users and service providers and the impact on the grid load;

[0107] S23. Under the condition of meeting the grid operation constraints, solve the demand response optimization objective function to obtain the optimal strategy of the grid operator in the regional unit R i ; The said operation constraints include: power balance constraint, generator output constraint and network security constraint;

[0108] The expression of the power balance constraint is as follows:

[0109]

[0110] Where: is the power generation power of the gth generator set in the ith regional unit at time t+τ, and G i is the number of generator sets, is the power input from other regions to the ith regional unit at time t+τ, is the load reduction obtained through demand response;

[0111] The expression of the generator output constraint is as follows:

[0112]

[0113] Where: and are respectively the minimum and maximum outputs of the gth generator set in the ith regional unit;

[0114] The network security constraint is: to meet the requirements of power flow and voltage stability of the grid transmission lines.

[0115] S24. Take the optimal strategy as the strategy guidance of the regional grid operator for the lower-layer power user layer, generate regional load management requirements, including demand response incentive signal I i (τ) and load regulation plan A i (τ); Synchronize I i (τ) and A i (τ) to the power user layer of the corresponding region to guide the power users to adjust the load, so as to optimize the grid demand-side response.

[0116] S3. At the electricity user layer, based on the demand response parameters of electricity users and the strategy guidelines received from the grid operator, a demand response model is constructed for the electricity users in each regional unit. According to the demand response model, the load response plan and adjustment strategy of the electricity users are calculated, and the load response plan and adjustment strategy are used as the demand input to the service provider layer and synchronized to the service provider layer of each regional unit;

[0117] Further, step S3 specifically includes:

[0118] S31. Construct a demand response model for each electricity user and solve the demand response model under the constraint conditions that meet the electricity user's own electricity demand and comfort constraints to obtain the optimal load response plan of the electricity user and the optimal adjustment strategy The demand response model is based on the demand response ability parameters of electricity users demand response incentive signal I i (τ) and load regulation plan A i (τ), which are used to describe the load adjustment strategy and response behavior of electricity users;

[0119] The expression of the demand response model is as follows:

[0120]

[0121] Where: is the load adjustment strategy of the jth electricity user in the ith regional unit at time t+τ; is the utility function of the electricity user, representing the difference between revenue and cost; is the revenue function obtained by the electricity user for participating in demand response; is the cost function for the electricity user to implement the load adjustment strategy, which depends on the load adjustment strategy and the demand response ability parameter

[0122] The constraint conditions are as follows:

[0123]

[0124] Where: and are the minimum and maximum allowable loads of the electricity user respectively, is the set of feasible adjustment strategies for the electricity user;

[0125] S32. Aggregate the load response demands and load adjustment strategies of all electricity users in the same regional unit to form the optimal load response plan of the regional electricity users and the optimal adjustment strategy

[0126] S4. At the service provider level, based on the service capacity parameters of each regional service provider and the load response requirements of electricity users, construct a service provider response optimization model, calculate the optimal service strategy for each service provider, and feedback the optimal service strategy to the electricity user level;

[0127] Further, the step S4 specifically includes:

[0128] S41. For each service provider Construct a service provider response optimization model And under the constraint conditions of meeting the service provider's capabilities and user response requirements, solve the service provider response optimization model to obtain the optimal service strategy of the service provider

[0129] The expression of the response optimization model is as follows:

[0130]

[0131] Where: is the service provision strategy of the kth service provider in the ith regional unit at time t + τ; is the utility function of the service provider; is the revenue function obtained by the service provider for providing services, which depends on the service provision strategy and the load response requirements of electricity users is the cost function of the service provider for providing services, which depends on the service provision strategy and the service capacity parameters

[0132] The constraint conditions are as follows:

[0133]

[0134] Where: is the maximum service capacity of the service provider, and are the minimum and maximum load response requirements of regional electricity users, respectively;

[0135] S42. Aggregate the optimal service strategies of all service providers to form the optimal service strategy set of regional service providers

[0136] S5. In the hierarchical model of the multi-level game structure, based on parallel computing technology, the game optimization processes of the grid operator layer, the power user layer, and the service provider layer are synchronously executed on the computing nodes of each regional unit to determine the optimal response strategy combination of each regional unit and obtain a distributed grid demand-side response optimization solution.

[0137] Further, step S5 specifically includes:

[0138] S51. Using the optimal strategy of the grid operator layer the optimal adjustment strategy of the power user layer and the optimal service strategy of the service provider as the optimization variables for each layer, the game optimization processes of the grid operator layer, the power user layer, and the service provider layer are synchronously executed on the computing nodes of each regional unit R i to solve the initial optimization strategy and determine the final optimal response strategy combination of the regional unit.

[0139] During this process, a parallel computing model is constructed within each regional unit R i . The optimization models of the grid operator layer, the power user layer, and the service provider layer are deployed in parallel to each computing node through a distributed architecture. At the same time, the initial optimization strategy is independently solved using the load prediction values, user response parameters, and service capacity parameters of each layer model. And in each iteration, the grid operator layer interacts with the response strategies of the power user layer and the service provider layer and by issuing strategy guidelines. The service provider layer and the power user layer respectively update their response strategies according to the model calculations and feedback them to the grid operator layer. Through multiple iterative calculations, the strategy combination of the grid operator layer, the power user layer, and the service provider layer gradually converges to the Nash equilibrium point of each layer using the equilibrium solution algorithm of the game model to determine the final optimal response strategy combination of the regional unit.

[0140] S52. Aggregate the optimal response strategy combinations of each regional unit to generate a distributed grid demand-side response optimization solution G * that satisfies the global coordination condition;

[0141] The expression of the distributed demand-side response optimization solution is as follows:

[0142] G * ={G *(1) , G *(2) ,..., G *(N)};

[0143] where: the response solution of a single regional unit is

[0144] The global coordination conditions are as follows:

[0145]

[0146] Where: is the total power generation in the i-th regional unit, is the input power of the i-th regional unit.

[0147] Furthermore, after obtaining the optimal response strategy combination, real-time monitoring of each regional unit can be continued, real-time grid load change data can be collected, and the demand response optimization plan can be dynamically adjusted according to the load fluctuation conditions in each regional unit and the feedback of multi-level game optimization, and the optimal response strategy combination of each regional unit can be updated.

[0148] Based on the demand-side response optimization plan calculated in the previous stage, combined with the real-time grid load change data, dynamically track the load fluctuation conditions of each regional unit, detect the applicability of the current response strategy in actual operation through dynamic analysis of the collected data, and the process depends on the feedback mechanism in the distributed multi-level game structure in the previous steps, compare the actual execution status of each layer with the theoretical optimization goal, and identify the strategy deviation existing in the operation.

[0149] On this basis, through the distributed computing node to synchronously adjust the strategy, dynamically correct the load fluctuation conditions of each unit with the optimal response strategy combination calculated in step S5, consider the strategy interaction relationship between the grid operator layer, the power user layer and the service provider layer during the optimization process, and dynamically update the weights and strategy sets of the game variables of each layer according to the amplitude and frequency characteristics of the grid load fluctuation, so as to ensure that the adjusted strategy can optimize the load response effect while meeting the power balance and network security constraints. Specifically, the grid operator adjusts the load management demand signal according to the real-time feedback, the power user makes precise responses according to the adjusted load plan, and the service provider adapts the service strategy according to the latest optimization results.

[0150] At the same time, to ensure the consistency of the strategy update of each regional unit, utilize the parallel computing characteristics of the distributed multi-level game model to synchronously adjust the strategy process among units in real time. Finally, by integrating the optimization results in each regional unit, dynamically update the optimal response strategy combination of each region to ensure the overall coordination and real-time nature of the grid demand-side response.

[0151] In this embodiment, the solution is simulated through case analysis. The power grid of a certain city is facing huge pressure brought by the summer electricity peak. Especially from 1 pm to 3 pm on weekdays, the centralized use of residential air conditioners, commercial office equipment and industrial production equipment leads to a rapid increase in the grid load. The peak-valley difference is as high as 35%. Traditional grid load management methods are difficult to quickly adapt to such real-time load changes and usually rely on simple load forecasting and offline strategies, resulting in frequent problems of voltage instability and local power supply shortages in some areas.

[0152] To improve this situation, the power dispatching center of a certain city applies this solution, optimizes the demand response strategy through a distributed multi-level game analysis model, coordinates grid operators, power users and service providers, dynamically manages load fluctuations, and effectively cuts peaks and fills valleys during peak electricity consumption periods.

[0153] The power grid of this city is divided into 10 regional units. Each unit includes a grid operator node, 50 - 200 power users (including residential, commercial and industrial users) and 5 - 10 service providers. Each regional unit runs a distributed multi-level game model through an independent computing node.

[0154] The system starts running at 9 am, and collects the grid load data of each region, the demand response parameters of power users and the service capacity parameters of service providers in real time. The real-time load data of Region 1 is 52 MW, Region 2 is 68 MW, and the total load demand is 620 MW. The demand response parameters of power users cover the adjustable load of users and comfort limits. Among the 50 users in Region 1, 35% of the users have an adjustable load of more than 2 kW. The service capacity data of service providers shows that the total adjustment capacity of service providers in Region 1 is 10 MW.

[0155] Based on the collected data, the system constructs a multi-level game model in each region. The grid operator layer predicts the load peak from 1 pm to 3 pm according to the load forecasting model. The peak in Region 1 is 72 MW, Region 2 is 88 MW, and the total peak prediction is 820 MW. The grid operator layer sets the peak shaving target at 20% and issues the load reduction demand to the power user layer. The power user layer combines the demand response incentive signal and its own response capacity parameters to calculate the optimal adjustment strategy. The total user response in Region 1 is 12 MW, of which 5 MW is contributed by industrial users, 4 MW by commercial users, and 3 MW by residential users. The service provider layer formulates the optimal service strategy according to the response needs of users to ensure that the load adjustment of power users is supported.

[0156] Through parallel computing technology, the system synchronously runs the game optimization process in each regional unit. The computing nodes in Region 1 iterate with a time step of 30 seconds. Eventually, after 20 iterations, the strategy combinations of each layer converge to the Nash equilibrium point. The final load reduction strategy for Region 1 is as follows: industrial users reduce 5 MW, commercial users reduce 3 MW, residential users reduce 2 MW, and the service provider provides 4 MW of auxiliary regulation.

[0157] From 1 pm to 3 pm, the system executes the response strategy according to the optimization plan. The real-time load in Region 1 drops from 72 MW to 60 MW after the implementation of the response, in Region 2 from 88 MW to 74 MW, and the total load drops from 820 MW to 656 MW, successfully achieving the peak shaving goal. The real-time load data is fed back to the system through sensors without any serious deviation.

[0158] To verify the effectiveness of this solution, a comparative experiment was conducted between this solution and traditional methods. The comparison data is shown in the following table:

[0159]

[0160] In the above comparative experiment, the traditional optimization method requires 2 minutes of response time to complete the load reduction goal and the reduction ratio is only 10%. This is mainly because the traditional method cannot adjust the strategy in real time, and the user participation rate and the service provider's response efficiency are relatively low. While this solution can complete the optimization and implement the response within 30 seconds, with a significant peak shaving effect, and the participation of users and service providers has been greatly improved. In addition, the distributed parallel computing architecture of the system reduces the computing time from 480 seconds to 120 seconds, showing significant advantages in terms of real-time performance and efficiency.

[0161] This solution uses a dynamic optimization algorithm to combine the real-time load data of the power grid, the behavior characteristics of power users, and the ability parameters of service providers, and dynamically adjusts the optimization objectives and constraint conditions in the multi-level game model. When the gap between the load forecast and the actual response is large, the optimization plan is adjusted in real time to enable the demand-side response of the power grid to adapt to load fluctuations. The above examples prove that this solution has obvious advantages in multi-level collaborative optimization, real-time response, and large-scale computing efficiency, and can provide reliable technical support for the demand-side response of the power grid.

[0162] The experimental results show that in scenarios with high-concurrency load demands, the computing time of this solution is reduced by about 35% compared with the traditional optimization method. And compared with the traditional optimization method based on static data, this solution can generate adjustment strategies in a timely manner under load mutations to effectively relieve the load pressure of the power grid. The experiment shows that the dynamic optimization algorithm of this solution improves the load forecast accuracy of demand response by 12% and the peak shaving ability during peak load periods by 15%.

[0163] Example 2:

[0164] See Figure 3 , a power grid demand response optimization system based on multi-level game analysis, which is applied to the method described in Embodiment 1. The system includes:

[0165] A hierarchical modeling module 1, configured to obtain real-time grid load data, demand response parameters of power users, and service capacity parameters of each regional service provider, and divide them into different regional units; construct a hierarchical model based on a distributed multi-level game structure, and perform regional hierarchical modeling on grid operators, power users, and service providers within each regional unit;

[0166] Furthermore, the hierarchical modeling module 1 is configured to perform hierarchical modeling according to the following steps:

[0167] S11. Collect real-time grid load data D i (t), the set of demand response parameters of power users The set of service capacity parameters of each regional service provider

[0168] S12. Based on the grid topology and geographical location, divide grid operators, power users, and service providers into different regional units, and its expression is as follows:

[0169] R = {R1, R2,..., R i ,..., R N};

[0170] Where: R i Represents the i-th regional unit, including the corresponding grid operator O i , the set of power users U i And the set of service providers S i ;

[0171] S13. Within each regional unit R i , according to the real-time grid load data D i (t), the set of demand response parameters of power users The set of service capacity parameters of each regional service provider Respectively construct a grid operator layer model based on a distributed multi-level game structure Power user layer model And service provider layer model And establish a regional hierarchical game model

[0172] S14. Connect the hierarchical game models G i Within each regional unit R through a distributed architecture to form an overall distributed multi-level game model G = {G (i) , G (1) , G(2) , ..., G (N)}。

[0173] The regional load management demand generation module 2 is used to analyze the real-time grid load data of each regional unit at the grid operator layer, construct a load forecasting model, calculate the demand response optimization objective function at the grid operator layer, and use the calculation result as the strategy guidance of the grid operator of this region for the lower-level power user layer, generate regional load management demands and synchronize them to the power user layer of the corresponding region;

[0174] Furthermore, the regional load management demand generation module 2 is used to perform load forecasting according to the following steps:

[0175] S21. At the grid operator layer, analyze the real-time grid load data D i (t) of each regional unit R i to construct a load forecasting model f i (t), and predict the load demand at future moments and the load demand within a future time period to obtain a predicted load sequence

[0176] The expression of the load demand at the future moment is as follows:

[0177]

[0178] Where: is the predicted load of the i-th regional unit at the future moment t + Δt, W i (t) is the weather factor parameter set of the i-th regional unit, H i (t) is the historical load behavior parameter set of the i-th regional unit;

[0179] S22. Based on the predicted load sequence establish the demand response optimization objective function at the grid operator layer, and consider the strategic interaction of lower-level power users and service providers in combination with the multi-level game model. Its expression is as follows:

[0180]

[0181] Where: is the demand response optimization objective function, is the strategy set of the grid operator at the future moment t + τ in the i-th regional unit, including demand response incentive measures and load regulation schemes; is the operating cost function of the grid operator, is the penalty coefficient of the grid operator, is the load deviation penalty function;

[0182] S23. Solve the demand response optimization objective function under the premise of meeting the grid operation constraints. Obtain the optimal strategy of the grid operator in the regional unit R i

[0183] The operation constraints include: power balance constraint, generator output constraint, and network security constraint;

[0184] The expression of the power balance constraint is as follows:

[0185]

[0186] Where: is the power generation power of the gth generator set in the ith regional unit at time t+τ, and G i is the number of generator sets, is the power input from other regions to the ith regional unit at time t+τ, is the load reduction obtained through demand response;

[0187] The expression of the generator output constraint is as follows:

[0188]

[0189] Where: and are the minimum and maximum outputs of the gth generator set in the ith regional unit respectively;

[0190] The network security constraint is: meeting the requirements of power flow and voltage stability of the grid transmission line;

[0191] S24. Use the optimal strategy as the strategy guidance of the regional grid operator for the lower-layer power user layer, generate regional load management requirements, and synchronize them to the power user layer of the corresponding region to guide power users to adjust the load, so as to optimize the grid demand-side response; the regional load management requirements include the demand response incentive signal I i (τ) and the load regulation plan A i (τ).

[0192] The demand response module 3 is used to construct a demand response model for the power users of each regional unit in the power user layer based on the demand response parameters of the power users and the strategy guidance received from the grid operator, calculate the load response plan and adjustment strategy of the power users according to the demand response model, use the load response plan and adjustment strategy as the demand input to the service provider layer, and synchronize them to the service provider layer of each regional unit;

[0193] Further, the demand response module 3 is used to optimize demand response according to the following steps:

[0194] S31. Build a demand response model for each power user and, under the constraint conditions that satisfy the power user's own electricity demand and comfort constraints, solve the demand response model to obtain the optimal load response plan and optimal adjustment strategy of the power user and optimal adjustment strategy

[0195] The expression of the demand response model is as follows:

[0196]

[0197] Where: is the load adjustment strategy of the j-th power user in the i-th regional unit at time t+τ, is the utility function of the power user, is the revenue function obtained by the power user for participating in demand response, is the cost function for the power user to implement the load adjustment strategy; is the demand response ability parameter of the j-th power user in the i-th regional unit;

[0198] The constraint conditions are as follows:

[0199]

[0200] Where: and are the minimum and maximum allowable loads of the power user respectively, is the set of feasible adjustment strategies of the power user;

[0201] S32. Aggregate the load response demands and load adjustment strategies of all power users in the same regional unit to form the optimal load response plan and optimal adjustment strategy of the regional power users and optimal adjustment strategy

[0202] The response optimization module 4 is used to build a service provider response optimization model at the service provider layer based on the service ability parameters of each regional service provider and the load response demands of power users, calculate the optimal service strategies of each service provider, and feedback the optimal service strategies to the power user layer;

[0203] Further, the response optimization module 4 is used to optimize the service strategy according to the following steps:

[0204] S41. Build a service provider response optimization model for each service provider Build a service provider response optimization model Under the constraints of meeting the capabilities of service providers and the response requirements of users, solve the service provider response optimization model to obtain the optimal service strategy of the service provider

[0205] The expression of the response optimization model is as follows:

[0206]

[0207] Where: is the service provision strategy of the Kth service provider in the ith regional unit at time t + τ, is the utility function of the service provider, is the revenue function obtained by the service provider for providing services, is the cost function of the service provider for providing services;

[0208] The constraint conditions are as follows:

[0209]

[0210] Where: is the maximum service capacity of the service provider, and are the minimum and maximum load response requirements of regional power users, respectively;

[0211] S42. Aggregate the optimal service strategies of all service providers to form the optimal service strategy set of regional service providers

[0212] The demand-side response optimization module 5 is used to synchronously execute the game optimization processes of the grid operator layer, power user layer, and service provider layer on the computing nodes of each regional unit based on parallel computing technology in the hierarchical model of the multi-level game structure, determine the optimal response strategy combination of each regional unit, and obtain a distributed grid demand-side response optimization solution.

[0213] Furthermore, the demand-side response optimization module 5 is used to determine the optimal response strategy according to the following steps:

[0214] S51. Using the optimal strategy of the grid operator layer the optimal adjustment strategy of the power user layer and the optimal service strategy of the service provider as the optimization variables of each layer, synchronously execute the game optimization processes of the grid operator layer, power user layer, and service provider layer on the computing nodes of each regional unit R i to solve the initial optimization strategy and determine the final optimal response strategy combination of the regional unit

[0215] S52. Aggregate the optimal response strategy combinations of each regional unit to generate a distributed grid demand-side response optimization plan G * , and satisfy the global coordination condition;

[0216] The expression of the distributed demand-side response optimization plan is as follows:

[0217] G * ={G *(1) , G *(2) ,..., G *(N)};

[0218] Wherein: The response plan of a single regional unit is

[0219] The global coordination condition is as follows:

[0220]

[0221] Wherein: is the total power generation within the I-th regional unit, is the input power of the i-th regional unit.

[0222] Example 3:

[0223] Refer to Figure 4 , a grid demand-side response optimization device based on multi-level game analysis, the device includes a processor 6 and a memory 7;

[0224] The memory 7 is used to store the computer program code 71 and transmit the computer program code 71 to the processor 6;

[0225] The processor 6 is used to execute the grid demand-side response optimization method described in Example 1 according to the instructions in the computer program code 71.

[0226] In this embodiment, there is also a computer-readable storage medium, and computer-executable instructions are stored in the computer-readable storage medium. When the computer-executable instructions are executed on a computer, the grid demand-side response optimization method described in Example 1 is implemented.

[0227] Generally speaking, the computer instructions for implementing the method of the present invention can be carried by any combination of one or more computer-readable storage media. A non-temporary computer-readable storage medium can include any computer-readable medium except for the signal propagating temporarily itself.

[0228] A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the computer-readable storage medium include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, the computer-readable storage medium can be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0229] Computer program code for performing the operations of the present invention can be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. In particular, the Python language suitable for neural network computing and platform frameworks based on TensorFlow, PyTorch, etc. can be used. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or connected to an external computer (for example, by connecting through an Internet service provider via the Internet).

[0230] For the above-mentioned devices and non-transitory computer-readable storage media, reference can be made to the specific description of a method for optimizing grid demand-side response based on multi-level game analysis and its beneficial effects, which will not be elaborated here.

[0231] Although the embodiments of the present invention have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. An optimization method for power grid demand-side response based on multi-level game analysis, characterized in that, Including: S1. Obtain the real-time grid load data, demand response parameters of power users, and service capacity parameters of regional service providers, and divide them into different regional units; Construct a hierarchical model based on a distributed multi-level game structure, and perform regional hierarchical modeling on grid operators, power users, and service providers within each regional unit; S2. At the grid operator layer, analyze the real-time grid load data of each regional unit, construct a load forecasting model, and calculate the demand response optimization objective function of the grid operator layer. Use the calculation result as the strategy guidance of the grid operator in this region for the lower-level power user layer, generate regional load management requirements, and synchronize them to the power user layer of the corresponding region; S3. At the power user layer, based on the demand response parameters of power users and the received strategy guidance from the grid operator, construct a demand response model for power users in each regional unit. Calculate the load response plan and adjustment strategy of power users according to the demand response model, use the load response plan and adjustment strategy as the demand input to the service provider layer, and synchronize them to the service provider layer of each regional unit; S4. At the service provider layer, construct a service provider response optimization model based on the service capacity parameters of regional service providers and the load response requirements of power users, calculate the optimal service strategy of each service provider, and feedback the optimal service strategy to the power user layer; S5. In the hierarchical model of the multi-level game structure, based on parallel computing technology, synchronously execute the game optimization processes of the grid operator layer, power user layer, and service provider layer on the computing nodes of each regional unit, determine the optimal response strategy combination of each regional unit, and obtain a distributed grid demand-side response optimization solution.

2. The grid demand-side response optimization method based on multi-level game analysis according to claim 1, characterized in that: The step S1 specifically includes: S11. Collect the real-time load data D of the power grid i (t), the set of demand response parameters of power users The set of service capacity parameters of each regional service provider S12. Based on the grid topology structure and geographical location, divide grid operators, power users, and service providers into different regional units, and its expression is as follows: R = {R1R2,..., R i ,..., R N} Where: R i represents the i-th regional unit, including the corresponding grid operator O i , the set of electricity users U i and the set of service providers S i ; S13. In each regional unit R i according to the real-time grid load data D i (t), the set of demand response parameters of electricity users and the set of service capacity parameters of each regional service provider respectively construct the grid operator layer model based on the distributed multi-level game structure the electricity user layer model and the service provider layer model and establish a regional hierarchical game model S14. Connect the hierarchical game models G i within each regional unit R (i) through a distributed architecture to form an overall distributed multi-level game model G = {G (1) , G (2) ,..., G (N)}.

3. The grid demand-side response optimization method based on multi-level game analysis according to claim 2, characterized in that: The step S2 specifically includes: S21. At the grid operator level, analyze the real-time grid load data D i of each regional unit R i (t), construct a load forecasting model f i (t), and predict the load demand at future times and the load demand within a future time period to obtain a predicted load sequence The expression of the load demand at the future moment is as follows: Wherein: is the predicted load of the i-th regional unit at the future time t+Δt, W i (t) is the weather factor parameter set of the i-th regional unit, H i (t) is the historical load behavior parameter set of the i-th regional unit; S22. Based on the predicted load sequence Establish the demand response optimization objective function at the grid operator level, and consider the strategic interaction between downstream electricity users and service providers in combination with the multi-level game model. Its expression is as follows: Wherein: is the objective function for demand response optimization, is the strategy set of the grid operator at the future time t + τ in the i-th regional unit, including demand response incentive measures and load regulation schemes; is the operating cost function of the grid operator, is the penalty coefficient of the grid operator, is the load deviation penalty function; S23. Solve the demand response optimization objective function under the condition of meeting the power grid operation constraints Obtain the optimal strategy of the grid operator in the regional unit R i ​ S24. Use the optimal strategy as the strategy guidance for the regional grid operator to the lower-level power user layer, generate regional load management requirements, and synchronize them to the power user layer of the corresponding region to guide power users to adjust their loads, so as to optimize the demand-side response of the power grid; the regional load management requirements include the demand response incentive signal I i (τ) and the load regulation plan A i (τ).

4. The grid demand-side response optimization method based on multi-level game analysis according to claim 3, characterized in that: In the step S23, the operation constraint conditions include: power balance constraint, generator output constraint, and network security constraint; The expression of the power balance constraint is as follows: Wherein: is the power generation power of the g-th generator set in the i-th regional unit at time t+τ, G i is the number of generator sets, is the power input from other regions by the i-th regional unit at time t+τ, is the load reduction obtained through demand response; The expression of the generator output constraint is as follows: Wherein: and are respectively the minimum and maximum power outputs of the g-th generating unit in the i-th regional unit; The network security constraint is: meeting the requirements of power flow and voltage stability of the grid transmission line.

5. The grid demand-side response optimization method based on multi-level game analysis according to claim 3, characterized in that: The step S3 specifically includes: S31. Build a demand response model for each electricity user and solve the demand response model under the constraint conditions that satisfy the electricity users' own electricity consumption needs and comfort constraints, so as to obtain the optimal load response plan of the electricity users and the optimal adjustment strategy The expression of the demand response model is as follows: wherein: is the load adjustment strategy of the j-th power user in the i-th regional unit at time t + τ, is the utility function of the power user, is the revenue function obtained by the power user for participating in demand response, is the cost function for the power user to implement the load adjustment strategy; is the demand response capacity parameter of the j-th power user in the i-th regional unit; The constraint conditions are as follows: Wherein: and are the minimum and maximum allowable loads of the electricity user respectively, is the set of feasible adjustment strategies for the electricity user; S32. Aggregate the load response requirements of all electricity users within the same regional unit and the load adjustment strategies to form the optimal load response plan and the optimal adjustment strategy of regional electricity users and the optimal adjustment strategy 6. The grid demand-side response optimization method based on multi-level game analysis according to claim 5, characterized in that: The step S4 specifically includes: S41. For each service provider Construct an optimization model for service provider response And under the constraints of service provider capabilities and user response requirements, solve the optimization model for service provider response to obtain the optimal service strategy of the service provider The expression of the response optimization model is as follows: Wherein: is the service provision strategy of the k-th service provider in the i-th regional unit at time t + τ, is the utility function of the service provider, is the revenue function obtained by the service provider for providing services, is the cost function of the service provider for providing services; The constraint conditions are as follows: Wherein: is the maximum service capacity of the service provider, and are the minimum and maximum load response demands of regional electricity users, respectively; S42. Aggregate the optimal service strategies of all service providers to form the set of optimal service strategies of regional service providers 7. The method for optimizing the power grid demand-side response based on multi-level game analysis according to claim 6, characterized in that: The step S5 specifically includes: S51. The optimal strategy of the grid operator layer The optimal adjustment strategy of the electricity user layer And the optimal service strategy of the service provider As the optimization variables of each layer, on the computing nodes of each regional unit R i Synchronously execute the game optimization processes of the grid operator layer, the electricity user layer, and the service provider layer, solve the initial optimization strategy, and determine the final optimal response strategy combination of the regional unit S52. Aggregate the optimal response strategy combinations of each regional unit to generate a distributed grid demand-side response optimization plan G * , and satisfy the global coordination condition; The expression of the distributed demand-side response optimization scheme is as follows: G * = {G *(1) , G *⑵) ,..., G *(N)}; Among them: The response scheme of a single regional unit is The global coordination conditions are as follows: Wherein: is the total power generation within the i-th regional unit, is the input power of the i-th regional unit.

8. An optimized system for grid demand-side response based on multi-level game analysis, characterized in that This system is applied to the method described in any one of claims 1-7, and the system includes: A hierarchical modeling module (1), configured to obtain real-time power grid load data, demand response parameters of power users, and service capacity parameters of each regional service provider and divide them into different regional units; construct a hierarchical model based on a distributed multi-level game structure, and perform regional hierarchical modeling on power grid operators, power users, and service providers within each regional unit; A regional load management demand generation module (2), configured to analyze the real-time power grid load data of each regional unit at the power grid operator layer, construct a load forecasting model, and calculate the demand response optimization objective function at the power grid operator layer, use the calculation result as the strategy guidance of the regional power grid operator for the lower-layer power user layer, generate regional load management demands, and synchronize them to the power user layer of the corresponding region; A demand response module (3), configured to construct a demand response model for power users in each regional unit at the power user layer based on the demand response parameters of power users and the received strategy guidance of the power grid operator, calculate the load response plan and adjustment strategy of power users according to the demand response model, use the load response plan and adjustment strategy as the demand input to the service provider layer, and synchronize them to the service provider layer of each regional unit; A response optimization module (4), configured to construct a service provider response optimization model based on the service capacity parameters of each regional service provider and the load response requirements of power users at the service provider layer, calculate the optimal service strategy of each service provider, and feedback the optimal service strategy to the power user layer; A demand-side response optimization module (5), configured to, in the hierarchical model of the multi-level game structure, synchronously execute the game optimization processes of the power grid operator layer, the power user layer, and the service provider layer on the computing nodes of each regional unit based on parallel computing technology, determine the optimal response strategy combination of each regional unit, and obtain a distributed power grid demand-side response optimization scheme.

9. An optimized device for power grid demand-side response based on multi-level game analysis, characterized in that: The device includes a processor (6) and a memory (7); The memory (7) is used to store computer program code (71) and transmit the computer program code (71) to the processor (6); The processor (6) is used to execute the method for optimizing the power grid demand-side response based on multi-level game analysis described in any one of claims 1-7 according to the instructions in the computer program code (71).

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed on a computer, the method for optimizing the power grid demand-side response based on multi-level game analysis described in any one of claims 1-7 is implemented.

Images