Micro-grid group collaborative regulation method and system considering demand response

Abstract

The application discloses a micro-grid group collaborative regulation method and system considering demand response, and belongs to the technical field of micro-grid group collaborative regulation. The regulation method is specifically as follows: a demand response model is constructed based on a user interaction willingness factor and an electricity price elasticity coefficient to obtain demand response costs of each sub-micro-grid; equipment operation constraints are set based on equipment operation parameters of each sub-micro-grid in the micro-grid group, a total operation cost of the micro-grid group including a new energy waste penalty and a carbon emission cost is minimized as an upper optimization target, local operation costs of each sub-micro-grid including corresponding demand response costs are minimized as lower optimization targets, a double-layer game scheduling model is constructed, and distributed iteration is used for solving to obtain scheduling schemes and power interaction strategies of each sub-micro-grid in the micro-grid group. The double-layer game scheduling model is established, distributed iteration is used for solving, and the demand response resource is introduced to improve the new energy consumption rate, and the solving efficiency is considered.

Description

Technical Field

[0001] This invention relates to the field of microgrid group coordinated control technology, and in particular to a microgrid group coordinated control method and system that takes demand response into account. Background Technology

[0002] With the continuous increase in the penetration rate of distributed renewable energy in the distribution network, individual microgrids, limited by their regulation capacity and resource types, find it difficult to effectively mitigate the volatility of renewable energy output. Therefore, current methods involve interconnecting multiple microgrids through tie lines to form microgrid clusters, utilizing power interaction between sub-microgrids to achieve energy complementarity, thereby improving the absorption of new energy and the economic efficiency of grid operation.

[0003] Currently, the optimal scheduling of microgrid groups typically adopts a centralized solution method. This method uses the power generation output and energy storage charging and discharging power of each sub-microgrid as optimization variables. The microgrid group coordination center collects the operating parameters and load data of each sub-microgrid, constructs a global optimization model for overall solution, and adjusts the power interaction and coordination between sub-microgrids by adjusting the power generation output and energy storage charging and discharging power, thereby smoothing out the fluctuations in new energy sources.

[0004] However, in actual operation, this method of adjusting power generation output and energy storage charging and discharging power is limited by the unit ramp-up rate and the energy storage charging and discharging depth, making it difficult to fully match the rapid fluctuations in renewable energy output, and the renewable energy absorption rate remains low. Demand response, on the other hand, guides users to adjust their electricity consumption behavior through price signals or incentive compensation, providing additional flexibility without increasing equipment investment. Therefore, demand response resources can be incorporated into the optimal scheduling of microgrid clusters to provide more adjustment space for renewable energy absorption by utilizing the spatiotemporal transferability and power reduction characteristics of user-side loads, reducing the adjustment pressure on units and energy storage, and improving the renewable energy absorption level. However, if demand response resources are incorporated into the optimal scheduling of microgrid clusters, the relevant information on demand response resources of each sub-microgrid needs to be uniformly uploaded to the microgrid cluster coordination center, and the demand response call volume needs to be added as an optimization variable to the global optimization model for centralized and unified solution. The microgrid cluster coordination center may experience a decrease in optimization scheduling solution efficiency due to excessive computational load. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of current centralized solution methods in microgrid group optimization scheduling, which cannot simultaneously consider the renewable energy absorption rate and solution efficiency. It provides a microgrid group collaborative control method and system that considers demand response. By constructing a demand response model to obtain the demand response cost of each sub-microgrid, and incorporating it into the lower-level optimization objective of a two-layer game scheduling model, the demand response variable participates in distributed iterative solution as a local variable of each sub-microgrid. Each sub-microgrid only needs to share boundary interaction power for global coordination. Thus, while introducing demand response resources to improve renewable energy absorption rate, it avoids the communication and computational burden of centralized solutions, thus balancing solution efficiency.

[0006] The objective of this invention is achieved through the following technical solution: Microgrid group coordinated control methods considering demand response include: A demand response model is constructed based on the user interaction willingness factor and the electricity price elasticity coefficient, and the demand response cost of each sub-microgrid is obtained based on the demand response model. Based on the equipment operation parameters of each sub-microgrid in the microgrid group, the equipment operation constraints are set. The upper-level optimization objective is to minimize the total operating cost of the microgrid group, including the penalty for waste of new energy and the cost of carbon emissions. The lower-level optimization objective is to minimize the local operating cost of each sub-microgrid, including the corresponding demand response cost. A two-level game scheduling model is constructed. The two-level game scheduling model is solved based on distributed iteration to obtain the scheduling scheme and power interaction strategy of each sub-microgrid in the microgrid group.

[0007] Furthermore, the construction of the demand response model based on the user interaction willingness factor and the electricity price elasticity coefficient includes: Based on the basic electricity consumption data of the transferable load in each time period within each sub-microgrid and the electricity price elasticity coefficient in each time period, a transferable load response sub-model is established, with the constraints being that the total electricity consumption of the transferable load is conserved within the preset scheduling cycle and that the transfer amount does not exceed the preset upper and lower limits in each time period. Based on the original load data of each level of load that can be reduced within each sub-microgrid and the preset maximum reduction ratio of each level, a load reduction response sub-model is established. A demand response model is formed by combining the shiftable load response sub-model and the reduceable load response sub-model.

[0008] Furthermore, obtaining the demand response cost of each sub-micronet based on the demand response model includes: The response of the shiftable load in each time period under the influence of electricity price is obtained based on the shiftable load response sub-model; Obtain the maximum load call limit for each level of load reduction, and use the corresponding maximum load call limit as the response amount for each level of load reduction; Based on the user interaction willingness factor of each user in each sub-micronet, set the unit demand response cost parameter for each sub-micronet; Based on the response volume of the shiftable load and the response volume of the load that can be reduced at each level in each time period, and combined with the unit demand response cost parameter, the demand response cost of each sub-microgrid in each time period is calculated. Among them, the response volume of the shiftable load and the response volume of the load that can be reduced at each level in each time period are the demand response call volumes to be optimized. Accumulate the demand response costs for each time period within the preset scheduling period to obtain the total demand response cost for each sub-micronet within the preset scheduling period.

[0009] Furthermore, the two-layer game scheduling model is constructed with the upper-layer optimization objective of minimizing the total operating cost of the microgrid group, including penalties for renewable energy waste and carbon emission costs, and the lower-layer optimization objective of minimizing the local operating cost of each sub-microgrid, including corresponding demand response costs. This model includes: Using the power generation output, energy storage charging and discharging power, power purchase and sale with the distribution network, inter-group interaction power, and demand response call volume of each sub-microgrid as optimization variables, and combining equipment operating parameters, fuel cost coefficient, power purchase and sale price and inter-group interaction price coefficient, respectively establish the power generation fuel cost item, energy storage aging cost item, distribution network transaction cost item, inter-group interaction cost item and demand response cost item for each sub-microgrid, and obtain the local operating cost of each sub-microgrid after summing them; To minimize the corresponding local operating cost, the lower-level optimization objectives for each sub-micronet are determined. Based on the local operating cost and inter-group interaction power of each sub-microgrid corresponding to the lower-level optimization objective, the total operating cost item of the sub-microgrid and the inter-group carbon emission cost item are established respectively. Using the curtailment of wind and solar power in each sub-microgrid at each time period as the optimization variable, and combining the predicted power output of new energy sources with the preset curtailment penalty coefficient, a new energy waste penalty item is established. Based on the total operating cost of the sub-microgrid, the cost of renewable energy waste, and the cost of carbon emissions between groups, the total operating cost of the microgrid group is obtained, and minimizing the total operating cost of the microgrid group is the upper-level optimization objective. Based on the upper-level optimization objective and the lower-level optimization objective of each sub-microgrid, a two-level game scheduling model is constructed, with the microgrid group as the upper-level decision-maker and each sub-microgrid as the lower-level decision-maker.

[0010] Furthermore, the step of solving the two-layer game scheduling model based on distributed iteration to obtain the scheduling scheme and power interaction strategy of each sub-microgrid in the microgrid group includes: The power generation output, energy storage charging and discharging power, power purchase and sale with the distribution network, and demand response call volume of each sub-microgrid are used as local variables, and the corresponding inter-group interaction power is used as a globally shared variable. Each sub-microgrid initializes its corresponding local variables and global shared variables, and uploads the initialized global shared variables to the microgrid group; The microgrid group solves the upper-level optimization objective based on all received global shared variables, and obtains the updated inter-group interaction price coefficient based on the solution results; The microgrid cluster distributes the updated inter-group interaction price coefficients to each sub-microgrid. Each sub-microgrid solves the corresponding lower-level optimization objective based on the updated inter-group interaction price coefficients, and obtains the updated local variables and globally shared variables based on the solution results. Each sub-microgrid uploads the updated global shared variables to the microgrid cluster, and repeatedly executes the upper-level solution of the microgrid cluster and the local solution of each sub-microgrid until the solution results of the upper-level optimization objective and the solution results of the lower-level optimization objective of each sub-microgrid satisfy the corresponding convergence conditions. Based on the local variables and globally shared variables of each sub-microgrid, the scheduling scheme and power interaction strategy of each sub-microgrid in the microgrid group are obtained.

[0011] Furthermore, the convergence condition for the solution of the upper-level optimization objective includes that the original residuals and dual residuals of the globally shared variables are both less than the corresponding preset thresholds, and the convergence condition for the solution of the lower-level optimization objective includes that the change in each local variable is less than the corresponding preset change threshold.

[0012] Furthermore, after completing a local solution, each sub-micronet also performs the following: Based on the demand response call volume of the previous local solution and the current local solution, determine the change in demand response call volume of each sub-micronet; When the change in demand response call is less than the corresponding preset change threshold, it is determined that the demand response scheduling quantity meets the corresponding convergence condition, and the value of the current demand response scheduling quantity is fixed in the local variable.

[0013] Based on the user interaction willingness factor of each sub-micronet, a corresponding preset change threshold is set for the change in the corresponding demand response call. The larger the user interaction willingness factor, the larger the corresponding preset change threshold.

[0014] Furthermore, the equipment operation constraints include at least the upper limit constraints on the renewable energy output of each sub-microgrid in the microgrid group at each time period, as well as the charging and discharging power constraints and state of charge constraints of the energy storage devices of each sub-microgrid in the microgrid group at each time period.

[0015] Considering demand response in microgrid group coordinated control systems, the microgrid group coordinated control method for any of the above includes: The data acquisition module is used to acquire equipment operating parameters, new energy output forecast data, load data, electricity price information, and user interaction willingness factors for each sub-microgrid. The model building module is used to build a demand response model based on user interaction willingness factors and electricity price elasticity coefficients, obtain the demand response cost of each sub-microgrid, and set equipment operation constraints based on the equipment operation parameters of each sub-microgrid, so as to build a two-layer game scheduling model with minimizing the total operating cost of the microgrid group as the upper-layer optimization objective and minimizing the local operating cost of each sub-microgrid as the lower-layer optimization objective. The optimization solution module is used to solve the two-layer game scheduling model through distributed iteration to obtain the scheduling scheme and power interaction strategy of each sub-micronet.

[0016] The beneficial effects of this invention are: By constructing a demand response model, demand response costs are incorporated into the lower-level optimization objective of the two-layer game scheduling model. This allows demand response variables to participate in the solution as local variables for each sub-microgrid, thereby adding demand response resources to the microgrid group scheduling. This provides more adjustment space for renewable energy consumption and improves renewable energy utilization. Simultaneously, the two-layer game scheduling model decomposes the global optimization problem of the microgrid group into upper-level global coordination and lower-level local optimization. Each sub-microgrid only needs to upload boundary interaction power as a shared variable, while internal equipment parameters and demand response intention information remain locally. This achieves global collaborative optimization while protecting the information privacy of each sub-microgrid and reducing communication burden. Furthermore, the two-layer game scheduling model is solved based on distributed iteration. Each sub-microgrid's local subproblem includes its own demand response variables and is solved independently. This avoids the excessive computational burden caused by centralizing all variables in the microgrid group coordination center for unified solution, ensuring solution efficiency while introducing demand response resources to improve renewable energy consumption. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a process of the present invention; Figure 2 This is a schematic diagram of a distributed iterative solution process according to an embodiment of the present invention. Detailed Implementation

[0018] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0019] Example: Consideration should be given to microgrid group coordinated control methods based on demand response, such as Figure 1 As shown, it includes: A demand response model is constructed based on the user interaction willingness factor and the electricity price elasticity coefficient, and the demand response cost of each sub-microgrid is obtained based on the demand response model. Based on the equipment operation parameters of each sub-microgrid in the microgrid group, the equipment operation constraints are set. The upper-level optimization objective is to minimize the total operating cost of the microgrid group, including the penalty for waste of new energy and the cost of carbon emissions. The lower-level optimization objective is to minimize the local operating cost of each sub-microgrid, including the corresponding demand response cost. A two-level game scheduling model is constructed. The two-level game scheduling model is solved based on distributed iteration to obtain the scheduling scheme and power interaction strategy of each sub-microgrid in the microgrid group.

[0020] By constructing a demand response model to obtain the demand response cost of each sub-microgrid, the demand response resources are incorporated into the microgrid group optimization scheduling solution as a cost function, providing load-side adjustment space for renewable energy consumption.

[0021] Specifically, a two-layer game scheduling model is constructed, with the upper-layer objective of minimizing the total operating cost of the microgrid group and the lower-layer objective of minimizing the local operating cost of each sub-microgrid, including demand response cost. Demand response resources are used as local optimization variables of each sub-microgrid in the lower-layer solution. By constructing a demand response model to obtain the demand response cost of each sub-microgrid, and incorporating the demand response cost into the local optimization objective of each sub-microgrid, the demand response variables can be solved locally in each sub-microgrid without having to be uploaded to the microgrid group coordination center for global optimization. This provides load-side adjustment space for renewable energy consumption without increasing the dimensionality of global optimization variables.

[0022] Furthermore, by constructing a two-layer game scheduling model with the goal of minimizing the total operating cost of the microgrid group as the upper-layer objective and minimizing the local operating cost of each sub-microgrid, including demand response cost, as the lower-layer objective, the hierarchical structure of the two-layer game scheduling model ensures that the internal variables of each sub-microgrid are kept locally, reducing the risk of privacy leakage of demand response resources within each sub-microgrid. Moreover, global optimization only requires coordinating boundary interaction power, reducing the variable dimensionality of the global optimization problem.

[0023] Finally, the two-layer game scheduling model is solved through distributed iterative solution. Each sub-microgrid completes the solution of sub-problems, including demand response variables, locally. This avoids the computational burden and communication overhead caused by the increase in variables in centralized solution, thereby improving the renewable energy consumption rate while introducing demand response resources.

[0024] Demand response resources mainly manifest as the shifting and reduction of user-side load, which are variables at the level of electricity consumption behavior. Unlike physical variables at the equipment level, such as power generation output and energy storage charging / discharging power, these variables cannot be directly used as optimization variables in the scheduling model. Furthermore, the shifting and reduction of user-side load are primarily driven by electricity price signals. Changes in electricity prices affect users' willingness and extent to adjust their electricity consumption behavior. Moreover, changes in electricity prices in the current time period not only affect the load in that period but also the load in adjacent time periods, indicating a cross-influence across multiple time periods.

[0025] Therefore, a demand response model is constructed, in which demand response variables are defined, and the quantitative relationship between load response and multi-period electricity price changes is quantified by the electricity price elasticity coefficient. The differentiated costs of each sub-microgrid participating in demand response are quantified by the user interaction willingness factor, thereby transforming demand response resources from electricity consumption behavior variables into quantifiable optimization variables that can be used for scheduling.

[0026] Specifically, the construction of the demand response model based on user interaction willingness factor and electricity price elasticity coefficient includes: Based on the basic electricity consumption data of the transferable load in each time period within each sub-microgrid and the electricity price elasticity coefficient in each time period, a transferable load response sub-model is established, with the constraints being that the total electricity consumption of the transferable load is conserved within the preset scheduling cycle and that the transfer amount does not exceed the preset upper and lower limits in each time period. Based on the original load data of each level of load that can be reduced within each sub-microgrid and the preset maximum reduction ratio of each level, a load reduction response sub-model is established. A demand response model is formed by combining the shiftable load response sub-model and the reduceable load response sub-model.

[0027] Obtain the basic electricity consumption data of the load that can be transferred in each time period within the scheduling cycle for each sub-microgrid, and obtain the electricity price elasticity coefficient for each time period. The electricity price elasticity coefficient includes the self-elasticity coefficient and the cross-elasticity coefficient. The self-elasticity coefficient represents the impact of the electricity price change in the current time period on the load in the current time period, and the cross-elasticity coefficient represents the impact of the electricity price change in other time periods on the load in the current time period. The two together constitute the electricity price elasticity coefficient matrix.

[0028] Using the response of the load that can be shifted in each time period as the variable to be solved, and letting the corresponding response equal to the product of the electricity price elasticity coefficient matrix and the electricity price change rate in the corresponding time period, a quantitative relationship between the load response and the change in electricity price is established, forming a shiftable load response sub-model.

[0029] Meanwhile, constraints are set to ensure that the shift amount in each time period does not exceed the preset upper limit of shift amount and is not lower than the preset lower limit of shift amount, in order to prevent excessive load concentration or excessive transfer from causing local overload or voltage exceeding the limit in the distribution network. In addition, a constraint is set to ensure that the algebraic sum of the shift out and shift in each time period within the dispatch cycle is zero, that is, the total electricity consumption of users remains unchanged within the dispatch cycle, so as to ensure that the basic electricity needs of users are not affected.

[0030] The expression for the established shiftable load response sub-model is as follows: ; ; ; in, For time period Basic electricity consumption For time period The change in load, For time period The change in electricity price For time period Electricity price, Let be the electricity price elasticity coefficient matrix, when When is the self-elasticity coefficient, when When is the cross elasticity coefficient, For time period The response of the transferable load, For the scheduling period, For time period The corresponding preset lower limit of translation amount, For time period The corresponding preset upper limit of translation amount.

[0031] Obtain the load reduction level classification data within each sub-microgrid, including the original load amount corresponding to each level and its power value in each time period throughout the entire scheduling cycle, and obtain the preset maximum reduction ratio for each level. The maximum reduction ratio represents the upper limit of the maximum reduction ratio that the load of the corresponding level can be reduced based on the original load amount.

[0032] Using the reduction amount of each level of reducible load in each time period within the scheduling cycle as the variable to be solved, and letting the relationship between the corresponding reduction amount and the original load of the corresponding level and the preset maximum reduction ratio be such that the reduction amount of each level is not less than 0 and does not exceed the product of the original load of that level and the corresponding maximum reduction ratio, a reducible load response sub-model is formed, the expression of which is: ; ; in, For time period The total amount of load reduction that can be reduced. For time period The The amount of load reduction that can be achieved at each level. The total number of load levels. For time period The The preset maximum reduction ratio for the load that can be reduced.

[0033] The shiftable load response sub-model and the load reduction sub-model are combined to form a demand response model. In the formed demand response model, the shiftable load response amount and the load reduction amount together constitute the demand response call amount of each sub-microgrid in each time period.

[0034] Demand response resources are typically used in scheduling with a fixed maximum available quantity. This means that given a maximum available quantity for each load level in each time period, the scheduling model allocates load within this limit, and the allocation itself does not incur costs. However, this approach ignores the issue of users needing economic compensation for participating in demand response. Users reducing or shifting load means their current electricity consumption behavior will be restricted. If this is simply used as a fixed parameter in scheduling planning, the scheduling model will tend to allocate as much demand response available quantity as possible to reduce generation or purchase costs. This can lead to calculated reductions far exceeding what users are actually willing to accept, making it difficult to guarantee the reliability of the optimization solution.

[0035] Therefore, after constructing the demand response model, the demand response cost of each sub-micronet is further calculated based on the user interaction willingness factor, thereby linking the amount of demand response resources called up with the corresponding economic cost, thus ensuring that the optimization scheme can closely reflect reality and improve its accuracy and reliability.

[0036] Specifically, obtaining the demand response cost of each sub-micronet based on the demand response model includes: The response of the shiftable load in each time period under the influence of electricity price is obtained based on the shiftable load response sub-model; Obtain the maximum load call limit for each level of load reduction, and use the corresponding maximum load call limit as the response amount for each level of load reduction; Based on the user interaction willingness factor of each user in each sub-micronet, set the unit demand response cost parameter for each sub-micronet; Based on the response volume of the shiftable load and the response volume of the load that can be reduced at each level in each time period, and combined with the unit demand response cost parameter, the demand response cost of each sub-microgrid in each time period is calculated. Among them, the response volume of the shiftable load and the response volume of the load that can be reduced at each level in each time period are the demand response call volumes to be optimized. Accumulate the demand response costs for each time period within the preset scheduling period to obtain the total demand response cost for each sub-micronet within the preset scheduling period.

[0037] Based on the product relationship between the electricity price elasticity coefficient matrix and the electricity price change rate in the shiftable load response sub-model, the shiftable load response quantity is defined as an optimization variable jointly determined by the electricity price change rate and the corresponding electricity price elasticity coefficient for each time period. The shiftable load response quantity reflects the amount of load that users move out of or into their original electricity consumption plan in each time period under the influence of electricity price changes, and its value range is constrained by the upper and lower limits of the shift amount and the total electricity consumption conservation constraint in the shiftable load response sub-model.

[0038] Based on the load reduction response sub-model, the product of the original load amount at each level and the corresponding maximum reduction ratio is used as the upper limit of load call for each level of load reduction. The upper limit of load call is defined as the maximum value of the load reduction response amount at the corresponding level. The load reduction response amount at each level is used as the optimization variable to be solved, and the optimization variable is used as the demand response call amount of the load reduction.

[0039] Obtain the user interaction willingness factor for each user within each sub-micronet. The user interaction willingness factor is a value between 0 and 1. Determine the unit demand response cost parameter for each sub-micronet based on the user interaction willingness factor. The unit demand response cost parameter is set as the difference between 1 and the user interaction willingness factor. The closer the user interaction willingness factor is to 1, the closer the unit demand response cost parameter is to 0, indicating that users are more willing to participate in demand response and the cost required to pay per unit of response is lower.

[0040] Finally, the shiftable load response amount for each time period is multiplied by the corresponding unit demand response cost parameter to obtain the demand response cost of the shiftable load for that time period. The load reduction response amount for each level is multiplied by the corresponding unit demand response cost parameter to obtain the demand response cost of the load reduction for that level.

[0041] The demand response costs of loads that can be shifted and loads that can be reduced at each level are then summed to obtain the demand response cost of each sub-microgrid in each time period. It should be noted that the demand response cost is a function of the demand response call volume, and its value changes dynamically according to the different demand response call volumes during the solution process.

[0042] The expression for the demand response cost is: ; in, For the first Individual microgrids call demand response load The subsequent demand response cost, For the first The cost coefficient of a microgrid This is a factor related to user engagement.

[0043] The demand response costs of each sub-micronet are summed up during each time period within the scheduling cycle to obtain the total demand response cost of each sub-micronet within the scheduling cycle. This total demand response cost is then incorporated into subsequent scheduling plans as a demand response cost item to enable the application of demand response resources.

[0044] However, considering that current microgrid group optimization and scheduling typically employs a single-layer centralized model, which unifies the power generation output, energy storage charging and discharging power, and inter-group interaction power of all sub-microgrids as optimization variables and constructs a global objective function for unified solution, the shiftable and slashable loads of each sub-microgrid are added as new local variables to the model after the introduction of demand response resources. The single-layer centralized model requires solving all variables within the same optimization problem. With the addition of demand response variables, the problem size increases, the computational load per solution increases, and the solution efficiency decreases significantly.

[0045] Therefore, the single-layer centralized model is reconstructed into a two-layer game scheduling model with upper and lower layers. The upper layer only handles the coordination of inter-group interaction power, while the lower layer allows each sub-micronet to independently handle its own local variables. This decouples the updating of coupled variables from the optimization of local variables during the iteration process. In each iteration, the upper layer only needs to handle a few coupled variables, and the scale of the local variables of each sub-micronet in the lower layer remains within the scope of its local problem and does not enter the upper layer to participate in the solution, thereby improving the overall solution efficiency.

[0046] Specifically, the two-layer game scheduling model is constructed with the upper-level optimization objective of minimizing the total operating cost of the microgrid group, including penalties for renewable energy waste and carbon emission costs, and the lower-level optimization objective of minimizing the local operating cost of each sub-microgrid, including corresponding demand response costs. This model includes: Using the power generation output, energy storage charging and discharging power, power purchase and sale with the distribution network, inter-group interaction power, and demand response call volume of each sub-microgrid as optimization variables, and combining equipment operating parameters, fuel cost coefficient, power purchase and sale price and inter-group interaction price coefficient, respectively establish the power generation fuel cost item, energy storage aging cost item, distribution network transaction cost item, inter-group interaction cost item and demand response cost item for each sub-microgrid, and obtain the local operating cost of each sub-microgrid after summing them; To minimize the corresponding local operating cost, the lower-level optimization objectives for each sub-micronet are determined. Based on the local operating cost and inter-group interaction power of each sub-microgrid corresponding to the lower-level optimization objective, the total operating cost item of the sub-microgrid and the inter-group carbon emission cost item are established respectively. Using the curtailment of wind and solar power in each sub-microgrid at each time period as the optimization variable, and combining the predicted power output of new energy sources with the preset curtailment penalty coefficient, a new energy waste penalty item is established. Based on the total operating cost of the sub-microgrid, the cost of renewable energy waste, and the cost of carbon emissions between groups, the total operating cost of the microgrid group is obtained, and minimizing the total operating cost of the microgrid group is the upper-level optimization objective. Based on the upper-level optimization objective and the lower-level optimization objective of each sub-microgrid, a two-level game scheduling model is constructed, with the microgrid group as the upper-level decision-maker and each sub-microgrid as the lower-level decision-maker.

[0047] The fuel cost coefficient of each generator unit in the sub-microgrid is obtained. Based on the quadratic function relationship between the fuel cost coefficient and the power generation output, a power generation fuel cost item is established. The power generation fuel cost item increases quadratically with the increase of power generation output.

[0048] Obtain the charge-discharge cycle life conversion factor for each energy storage device in the sub-microgrid. Based on the product of the charge-discharge cycle life conversion factor and the cumulative value of the energy storage charge-discharge power, establish the energy storage aging cost item, which reflects the life loss of energy storage devices due to charge-discharge cycles.

[0049] Obtain the electricity purchase price and electricity sales price between each sub-microgrid and the upper-level distribution network at each time period. Based on the product of the electricity purchase price and the electricity purchase capacity, and the product of the electricity sales price and the electricity sales capacity, establish the distribution network transaction cost item. This item is positive when purchasing electricity and negative when selling electricity.

[0050] Obtain the inter-group interaction power guidance price for each time period issued by the microgrid group coordination center. Based on the product of the inter-group interaction power guidance price and the actual interaction power, establish the inter-group interaction cost item. This item is positive when purchasing electricity from other sub-microgrids and negative when selling electricity to other sub-microgrids.

[0051] Next, obtain the demand response cost of each sub-micronet in each time period based on the demand response model, multiply the demand response call volume by the corresponding unit demand response cost parameter, and establish a demand response cost item. This item increases as the demand response call volume increases.

[0052] The above-mentioned power generation fuel cost item, energy storage aging cost item, distribution network transaction cost item, inter-group interaction cost item, and demand response cost item are summed to obtain the local operating cost function of each sub-microgrid. The lower-level optimization objective of each sub-microgrid is determined by minimizing the local operating cost function of each sub-microgrid.

[0053] Taking one of the sub-micronets as an example, the expression for its lower-level optimization objective is: ; in, For the first The local operating cost of a microgrid For the first Individual micro-network during time period The cost of fuel for power generation, For the first Individual micro-network during time period The aging cost of energy storage For the first Individual micro-network during time period The cost of purchasing and selling electricity in the power distribution network For the first Individual micro-network during time period Inter-group interaction costs For the first Individual micro-network during time period Demand response compensation costs.

[0054] Obtain the lower-level optimization target values ​​of each sub-micronet, and sum them to establish the total operating cost item of the sub-micronet.

[0055] Obtain the inter-group interaction power between each sub-microgrid, and establish the inter-group carbon emission cost item based on the product of the inter-group interaction power and the preset carbon trading price coefficient.

[0056] Obtain the predicted power output of each sub-microgrid in each time period, determine the curtailed wind and solar power based on the difference between the predicted power output and the actual power output, use the curtailed wind and solar power as the optimization variable, and multiply it by the preset curtailment penalty coefficient as the curtailment penalty cost item.

[0057] The total operating cost of the sub-microgrid, the inter-group carbon emission cost, and the energy curtailment penalty cost are summed to obtain the total operating cost function of the microgrid group, and minimizing the total operating cost of the microgrid group is taken as the upper-level optimization objective.

[0058] The expression for the upper-level optimization objective is: ; in, The total operating cost of the microgrid group For the first The operating cost of a microgrid For the first The amount of wind and solar power curtailed by individual microgrids For the first Individual micro-networks and the first Interaction power between individual microgrids and These are the preset energy curtailment penalty coefficient and the preset carbon trading price coefficient, respectively.

[0059] When establishing the two-level game scheduling model, a mathematical model of microgrid group equipment, including new energy power generation equipment and energy storage equipment, is also established to provide corresponding equipment constraints for the two-level game scheduling model and ensure the feasibility of the scheduling scheme obtained by solving it.

[0060] The mathematical models for new energy equipment include output models for photovoltaic power generation equipment and wind power generation equipment. The expression for the output model of photovoltaic power generation equipment is as follows: ; in, For time period The output of photovoltaic power generation equipment, For photovoltaic array conversion efficiency, For time period The area of ​​the photovoltaic array in operation. For time period The amount of solar radiation.

[0061] The expression for the output model of wind power generation equipment is: ; ; in, For wind power generation equipment during the time period Wind speed at wheel hub height, For the time period Wind speed at reference height and These are the wheel hub height and reference height, respectively. The power-law exponent, For the time period The efficiency of the wind turbine air density, For power coefficient, The swept area of ​​the wind turbine in a wind power generation device.

[0062] The mathematical model of energy storage equipment includes the charging and discharging model of electric vehicle charging piles, and its expression is as follows: ; in, and The first Each charging station during the time period The charging power and discharging power, For time period Battery energy state, For time period Battery energy state, and These are the charging efficiency and the discharging efficiency, respectively.

[0063] In the mathematical model of photovoltaic (PV) power generation equipment, the conversion efficiency and PV array area determine the upper limit of output under different solar radiation levels. This upper limit is limited by the rated capacity of the PV inverter, thus forming the upper limit constraint of PV output for each time period. Similarly, in the mathematical model of wind power generation equipment, the turbine efficiency, swept area, and power coefficient determine the upper limit of output under different wind speeds. This upper limit is limited by the rated capacity of the wind turbine generator set, thus forming the upper limit constraint of wind power output for each time period.

[0064] In the mathematical model of energy storage equipment, the battery's state of energy is limited by the upper and lower limits of the battery capacity. The state of charge at each time period must not be lower than the lower limit or higher than the upper limit, thus forming a state of charge constraint. The charging and discharging power is limited by the rated power of the charging pile. The actual charging and discharging power at each time period does not exceed the rated value, thus forming a charging and discharging power constraint.

[0065] Based on this, the equipment operation constraints of the two-layer game scheduling model are obtained. The equipment operation constraints include at least the upper limit constraints of the new energy output of each sub-microgrid in the microgrid group at each time period, as well as the charging and discharging power constraints and state of charge constraints of the energy storage devices of each sub-microgrid in the microgrid group at each time period.

[0066] After the two-level game scheduling model is constructed, it includes local variables such as the power generation output, energy storage charging and discharging power, power purchase and sale with the distribution network, and demand response call volume of each sub-microgrid, as well as the global coupling variable of inter-group interaction power among the sub-microgrids. If a centralized solution is adopted, all local variables and global coupling variables need to be uploaded to the microgrid group coordination center, which then constructs a complete global optimization problem for a one-time overall solution. This results in a large computational load, a heavy communication burden, and low solution efficiency. However, if a completely decentralized solution is adopted, each sub-microgrid solves its own local problem independently, determining the inter-group interaction power based solely on local information. The independent decisions of each sub-microgrid may not reach a consensus, leading to mismatch in inter-group interaction power and making it difficult to guarantee the accuracy of the solution results.

[0067] Therefore, a distributed iterative approach is adopted to solve the two-layer game scheduling model. Each sub-microgrid independently solves the sub-problem containing local variables and inter-group interaction power locally. Only the inter-group interaction power is uploaded as a globally shared variable to the microgrid group coordination center for consistency coordination. The coordination center updates the global variables based on the interaction power uploaded by each sub-microgrid and then distributes them. Each sub-microgrid re-solves its local sub-problem based on the updated global variables, repeating the iteration until convergence. This allows the solution scale of local variables to be distributed locally to each sub-microgrid. The microgrid group coordination center only needs to handle the coordination of a few variables, namely inter-group interaction power. The amount of communication data is limited to the boundary interaction power, avoiding the communication burden and computational load problems of centralized solutions. At the same time, global consistency coordination can ensure the feasibility of independent decisions by each sub-microgrid at the global level.

[0068] Specifically, this embodiment employs the alternating direction multiplier method to achieve a distributed iterative solution for the two-level game scheduling model, wherein, as... Figure 2 As shown, the solution to the two-layer game scheduling model based on distributed iteration to obtain the scheduling scheme and power interaction strategy of each sub-microgrid in the microgrid group includes: The power generation output, energy storage charging and discharging power, power purchase and sale with the distribution network, and demand response call volume of each sub-microgrid are used as local variables, and the corresponding inter-group interaction power is used as a globally shared variable. Each sub-microgrid initializes its corresponding local variables and global shared variables, and uploads the initialized global shared variables to the microgrid group; The microgrid group solves the upper-level optimization objective based on all received global shared variables, and obtains the updated inter-group interaction price coefficient based on the solution results; The microgrid cluster distributes the updated inter-group interaction price coefficients to each sub-microgrid. Each sub-microgrid solves the corresponding lower-level optimization objective based on the updated inter-group interaction price coefficients, and obtains the updated local variables and globally shared variables based on the solution results. Each sub-microgrid uploads the updated global shared variables to the microgrid cluster, and repeatedly executes the upper-level solution of the microgrid cluster and the local solution of each sub-microgrid until the solution results of the upper-level optimization objective and the solution results of the lower-level optimization objective of each sub-microgrid satisfy the corresponding convergence conditions. Based on the local variables and globally shared variables of each sub-microgrid, the scheduling scheme and power interaction strategy of each sub-microgrid in the microgrid group are obtained.

[0069] The power generation output, energy storage charging and discharging power, power purchase and sale with the distribution network, and demand response call volume of each sub-microgrid are defined as local variables. These local variables relate to the equipment operating status and power consumption arrangements within each sub-microgrid, and are only stored locally for optimization, not shared externally, effectively reducing the risk of privacy leaks. Furthermore, the inter-group interaction power between sub-microgrids is defined as a globally shared variable. This variable relates to the power interaction of multiple sub-microgrids and requires unified coordination by the microgrid group coordination center to ensure consistency in power magnitude and direction among the sub-microgrids.

[0070] At the start of the scheduling cycle, each sub-micronet assigns initial values ​​to its local variables and global shared variables. The initial values ​​of the local variables are determined based on the actual operating status of each sub-micronet in the current time period, while the initial values ​​of the global shared variables are determined based on the initial power plan for interaction between each sub-micronet and its neighboring sub-micronets.

[0071] Each sub-microgrid uploads its initialized global shared variables to the microgrid group coordination center as input data for the first round of iterations of the microgrid group coordination center's upper-level solution.

[0072] The microgrid group coordination center receives globally shared variables uploaded by each sub-microgrid. It substitutes all globally shared variables for the current round into the upper-level optimization objective function, solving the upper-level optimization problem with the goal of minimizing the total operating cost of the microgrid group. This yields the inter-group interaction price coefficients between sub-microgrids at each time period. These inter-group interaction price coefficients are the optimal interaction price signals calculated by the microgrid group coordination center based on the current globally shared variables. They are used to guide each sub-microgrid to adjust its inter-group interaction power and local variables in the next round of local solution.

[0073] The microgrid group coordination center distributes the updated inter-group interaction price coefficients to each sub-microgrid. Upon receiving the inter-group interaction price coefficients, each sub-microgrid uses them as the settlement price in the current round's inter-group interaction cost item, substitutes them into its respective lower-level optimization objective function, and solves the lower-level optimization problem with the goal of minimizing local operating costs, obtaining updated local variables and globally shared variables. It should be noted that the upper-level and lower-level optimization objectives can be solved using existing optimization solvers.

[0074] Each sub-microgrid uploads its updated global shared variables to the microgrid group coordination center, entering the next iteration. The microgrid group coordination center re-solves the upper-level optimization problem based on the new global shared variables, updates the inter-group interactive power guidance price, and redistributes it. Each sub-microgrid receives the new guidance price and re-solves the lower-level optimization problem, updating its local variables and global shared variables. This solution and update process is repeated until the solution results for both the upper-level optimization objective and the lower-level optimization objectives of each sub-microgrid satisfy the corresponding convergence conditions. At this point, the model is considered converged, and the iteration terminates.

[0075] The convergence condition for the solution of the upper-level optimization objective includes that the original residuals and dual residuals of the globally shared variables are both less than the corresponding preset thresholds, and the convergence condition for the solution of the lower-level optimization objective includes that the change in each local variable is less than the corresponding preset change threshold.

[0076] The original residual of the globally shared variables refers to the deviation between the globally shared variables uploaded by each sub-microgrid and the globally consistent target value determined by the microgrid group coordination center. In each iteration, the microgrid group coordination center takes the arithmetic mean of the globally shared variables uploaded by each sub-microgrid as the globally consistent target value, then calculates the deviation between the globally shared variables uploaded by each sub-microgrid and this target value, and takes the L2 norm of the deviations of all sub-microgrids as the original residual. The original residual reflects the degree of divergence among the sub-microgrids in the values ​​of inter-group interaction power; the larger the value, the more inconsistent the negotiation results of the interaction power among the sub-microgrids.

[0077] The dual residual of the globally shared variables refers to the change in the globally consistent objective value between two adjacent iterations. In each iteration, the microgrid group coordination center compares the globally consistent objective value calculated in the current iteration with the objective value in the previous iteration, calculates the deviation between the two, and takes the L2 norm as the dual residual. The dual residual reflects whether the globally consistent objective value has stabilized; the larger the value, the less stable the current solution is.

[0078] Furthermore, the corresponding preset thresholds for the original residuals and dual residuals of the globally shared variables can be set according to actual needs.

[0079] After each sub-micronet completes the solution of its local subproblem, it compares the local variables obtained in this round with those obtained in the previous round, calculating the change in each local variable between two adjacent iterations. If the change in all local variables is less than their respective preset change thresholds, the sub-micronet's lower-level optimization objective is considered converged. If the change in any type of local variable exceeds the corresponding threshold, the sub-micronet's lower-level optimization objective is considered not yet converged, and it needs to proceed to the next iteration. Furthermore, each sub-micronet independently determines its own lower-level convergence status, and its convergence determination is not affected by the convergence status of other sub-micronets.

[0080] Furthermore, the preset change thresholds for each local variable can be set according to actual needs. However, considering that the physical properties of the demand response variable differ from those of other local variables, and that the adjustment of power generation output and energy storage charging and discharging power is limited by the physical characteristics of the equipment, its change thresholds are set based on the rated parameters and operating requirements of the equipment, and are usually fixed values. In contrast, the demand response variable reflects the proactive adjustment of users' electricity consumption behavior, and its convergence judgment criteria need to consider the differences in users' willingness to participate.

[0081] Therefore, based on the user interaction willingness factor of each sub-micronet, a corresponding preset change threshold is set for the change in the corresponding demand response call. The larger the user interaction willingness factor, the larger the corresponding preset change threshold.

[0082] For users with high willingness to participate, their demand response cost coefficient is low, and the objective function is not sensitive to changes in this variable. Even if the demand response variable changes significantly between adjacent iterations, the impact on the total cost is minimal. Therefore, it can be determined to be stable even with large fluctuations, i.e., a larger threshold is set to fix it early and reduce computational load. For users with low willingness to participate, their demand response cost coefficient is high, and the objective function is sensitive to changes in this variable. Fixing its value before it is fully optimized could lead to a significant increase in cost. Therefore, a smaller threshold is set to ensure that it is fixed only after a high level of accuracy has been achieved through refined calculations.

[0083] Based on this, after each sub-micronet completes a local solution, it also performs: Based on the demand response call volume of the previous local solution and the current local solution, determine the change in demand response call volume of each sub-micronet; When the change in demand response call is less than the corresponding preset change threshold, it is determined that the demand response scheduling quantity meets the corresponding convergence condition, and the value of the current demand response scheduling quantity is fixed in the local variable.

[0084] Once the demand response scheduling amount is fixed, if the changes in other local variables have not yet met the corresponding preset change thresholds, the corresponding sub-microgrid will enter the next iteration, updating only other local variables such as power generation output, energy storage charging and discharging power, and power purchased and sold with the distribution network.

[0085] The entire iterative process terminates only when the lower-level optimization objectives of all sub-micronets are determined to have converged, and the upper-level optimization objectives are determined to have converged simultaneously. The scheduling scheme and power interaction strategy of each sub-micronet are then output.

[0086] If the lower-level optimization objective of any sub-micronet has not yet converged or the upper-level optimization objective has not yet converged, then continue to the next round of iteration to solve the problem.

[0087] The dispatching scheme includes a power generation output scheme for the corresponding microgrid, an energy storage charging and discharging scheme, a power purchase and sale scheme with the distribution network, and a demand response dispatch scheme. The power generation output scheme is the output plan for each generator unit in each time period within the dispatching cycle, including the unit's start-up and shutdown status and output magnitude. The energy storage charging and discharging scheme is the charging and discharging power plan and state-of-charge change trajectory of each energy storage device in each time period within the dispatching cycle. The power purchase and sale scheme with the distribution network is the power plan for purchasing electricity from the superior distribution network and the power plan for selling electricity to the distribution network in each time period within the dispatching cycle. The demand response dispatch scheme is the plan for the amount of load that can be shifted in each time period within the dispatching cycle and the plan for the amount of load that can be reduced at each level.

[0088] The power interaction strategy includes the power transmitted by each sub-microgrid to other sub-microgrids and the power received from other sub-microgrids during each time period within the scheduling cycle.

[0089] Another aspect of this embodiment provides a microgrid group coordinated control system that takes into account demand response, including: The data acquisition module is used to acquire equipment operating parameters, new energy output forecast data, load data, electricity price information, and user interaction willingness factors for each sub-microgrid. The model building module is used to build a demand response model based on user interaction willingness factors and electricity price elasticity coefficients, obtain the demand response cost of each sub-microgrid, and set equipment operation constraints based on the equipment operation parameters of each sub-microgrid, so as to build a two-layer game scheduling model with minimizing the total operating cost of the microgrid group as the upper-layer optimization objective and minimizing the local operating cost of each sub-microgrid as the lower-layer optimization objective. The optimization solution module is used to solve the two-layer game scheduling model through distributed iteration to obtain the scheduling scheme and power interaction strategy of each sub-micronet.

[0090] The data acquisition module consists of intelligent monitoring and control terminals installed at various sub-microgrid photovoltaic grid connection points, wind turbine grid connection points, energy storage grid connection points, charging pile access points, and public contact points. It is connected to the model building module and can collect equipment operating parameters, new energy output data, load data, and electricity price information and upload them to the model building module.

[0091] The model building module is deployed on the server cluster of the microgrid group coordination center. It communicates with the optimization solution module and stores various data uploaded by the data acquisition module and user interaction willingness factors. It can build demand response models and two-level game scheduling models and transmit them to the optimization solution module via Ethernet.

[0092] The optimization solution module consists of a computing server deployed at the microgrid group coordination center and local controllers deployed in each sub-microgrid. The computing server and each local controller are connected via fiber optic Ethernet or 5G communication network. The computing server runs the optimization solver to solve the upper-level problem and update the global variables. Each local controller receives the inter-group interactive price coefficients issued by the computing server, calls the local optimization solver to complete the solution of the lower-level problem, and sends the updated data back to the computing server.

[0093] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any way. Other variations and modifications are possible without departing from the technical solutions described in the claims.

Claims

1. A microgrid group coordinated control method considering demand response, characterized in that, include: A demand response model is constructed based on the user interaction willingness factor and the electricity price elasticity coefficient, and the demand response cost of each sub-microgrid is obtained based on the demand response model. Based on the equipment operation parameters of each sub-microgrid in the microgrid group, the equipment operation constraints are set. The upper-level optimization objective is to minimize the total operating cost of the microgrid group, including the penalty for waste of new energy and the cost of carbon emissions. The lower-level optimization objective is to minimize the local operating cost of each sub-microgrid, including the corresponding demand response cost. A two-level game scheduling model is constructed. The two-level game scheduling model is solved based on distributed iteration to obtain the scheduling scheme and power interaction strategy of each sub-microgrid in the microgrid group.

2. The microgrid group coordinated control method considering demand response according to claim 1, characterized in that, The demand response model constructed based on user interaction willingness factors and electricity price elasticity coefficient includes: Based on the basic electricity consumption data of the transferable load in each time period within each sub-microgrid and the electricity price elasticity coefficient in each time period, a transferable load response sub-model is established, with the constraints being that the total electricity consumption of the transferable load is conserved within the preset scheduling cycle and that the transfer amount does not exceed the preset upper and lower limits in each time period. Based on the original load data of each level of load that can be reduced within each sub-microgrid and the preset maximum reduction ratio of each level, a load reduction response sub-model is established. A demand response model is formed by combining the shiftable load response sub-model and the reduceable load response sub-model.

3. The microgrid group coordinated control method considering demand response according to claim 2, characterized in that, The process of obtaining the demand response cost of each sub-micronet based on the demand response model includes: The response of the shiftable load in each time period under the influence of electricity price is obtained based on the shiftable load response sub-model; Obtain the maximum load call limit for each level of load reduction, and use the corresponding maximum load call limit as the response amount for each level of load reduction; Based on the user interaction willingness factor of each user in each sub-micronet, set the unit demand response cost parameter for each sub-micronet; Based on the response volume of the shiftable load and the response volume of the load that can be reduced at each level in each time period, and combined with the unit demand response cost parameter, the demand response cost of each sub-microgrid in each time period is calculated. Among them, the response volume of the shiftable load and the response volume of the load that can be reduced at each level in each time period are the demand response call volumes to be optimized. Accumulate the demand response costs for each time period within the preset scheduling period to obtain the total demand response cost for each sub-micronet within the preset scheduling period.

4. The microgrid group coordinated control method considering demand response according to claim 1, characterized in that, The above describes a two-layer game-theoretic scheduling model, which takes minimizing the total operating cost of the microgrid group (including penalties for renewable energy waste and carbon emission costs) as the upper-level optimization objective and minimizing the local operating cost of each sub-microgrid (including corresponding demand response costs) as the lower-level optimization objective. The model includes: Using the power generation output, energy storage charging and discharging power, power purchase and sale with the distribution network, inter-group interaction power, and demand response call volume of each sub-microgrid as optimization variables, and combining equipment operating parameters, fuel cost coefficient, power purchase and sale price and inter-group interaction price coefficient, respectively establish the power generation fuel cost item, energy storage aging cost item, distribution network transaction cost item, inter-group interaction cost item and demand response cost item for each sub-microgrid, and obtain the local operating cost of each sub-microgrid after summing them; To minimize the corresponding local operating cost, the lower-level optimization objectives for each sub-micronet are determined. Based on the local operating cost and inter-group interaction power of each sub-microgrid corresponding to the lower-level optimization objective, the total operating cost item of the sub-microgrid and the inter-group carbon emission cost item are established respectively. Using the curtailment of wind and solar power in each sub-microgrid at each time period as the optimization variable, and combining the predicted power output of new energy sources with the preset curtailment penalty coefficient, a new energy waste penalty item is established. Based on the total operating cost of the sub-microgrid, the cost of renewable energy waste, and the cost of carbon emissions between groups, the total operating cost of the microgrid group is obtained, and minimizing the total operating cost of the microgrid group is the upper-level optimization objective. Based on the upper-level optimization objective and the lower-level optimization objective of each sub-microgrid, a two-level game scheduling model is constructed, with the microgrid group as the upper-level decision-maker and each sub-microgrid as the lower-level decision-maker.

5. The microgrid group coordinated control method considering demand response according to claim 1, characterized in that, The method of solving the two-layer game scheduling model based on distributed iteration to obtain the scheduling scheme and power interaction strategy of each sub-microgrid in the microgrid group includes: The power generation output, energy storage charging and discharging power, power purchase and sale with the distribution network, and demand response call volume of each sub-microgrid are used as local variables, and the corresponding inter-group interaction power is used as a globally shared variable. Each sub-microgrid initializes its corresponding local variables and global shared variables, and uploads the initialized global shared variables to the microgrid group; The microgrid group solves the upper-level optimization objective based on all received global shared variables, and obtains the updated inter-group interaction price coefficient based on the solution results; The microgrid cluster distributes the updated inter-group interaction price coefficients to each sub-microgrid. Each sub-microgrid solves the corresponding lower-level optimization objective based on the updated inter-group interaction price coefficients, and obtains the updated local variables and globally shared variables based on the solution results. Each sub-microgrid uploads the updated global shared variables to the microgrid cluster, and repeatedly executes the upper-level solution of the microgrid cluster and the local solution of each sub-microgrid until the solution results of the upper-level optimization objective and the solution results of the lower-level optimization objective of each sub-microgrid satisfy the corresponding convergence conditions. Based on the local variables and globally shared variables of each sub-microgrid, the scheduling scheme and power interaction strategy of each sub-microgrid in the microgrid group are obtained.

6. The microgrid group coordinated control method considering demand response according to claim 5, characterized in that, The convergence condition for the solution of the upper-level optimization objective includes that the original residuals and dual residuals of the globally shared variables are both less than the corresponding preset thresholds. The convergence condition for the solution of the lower-level optimization objective includes that the change in each local variable is less than the corresponding preset change threshold.

7. The microgrid group coordinated control method considering demand response according to claim 6, characterized in that, After each sub-micronet completes a local solution, it also performs the following: Based on the demand response call volume of the previous local solution and the current local solution, determine the change in demand response call volume of each sub-micronet; When the change in demand response call is less than the corresponding preset change threshold, it is determined that the demand response scheduling quantity meets the corresponding convergence condition, and the value of the current demand response scheduling quantity is fixed in the local variable.

8. The microgrid group coordinated control method considering demand response according to claim 7, characterized in that, Based on the user interaction willingness factor of each sub-micronet, a corresponding preset change threshold is set for the change in the corresponding demand response call. The larger the user interaction willingness factor, the larger the corresponding preset change threshold.

9. The microgrid group coordinated control method considering demand response according to claim 1, characterized in that, The equipment operation constraints include at least the upper limit constraints on the renewable energy output of each sub-microgrid in the microgrid group at each time period, as well as the charging and discharging power constraints and state of charge constraints of the energy storage devices of each sub-microgrid in the microgrid group at each time period.

10. A microgrid group coordinated control system considering demand response, used to execute the microgrid group coordinated control method according to any one of claims 1 to 9, characterized in that, include: The data acquisition module is used to acquire equipment operating parameters, new energy output forecast data, load data, electricity price information, and user interaction willingness factors for each sub-microgrid. The model building module is used to build a demand response model based on user interaction willingness factors and electricity price elasticity coefficients, obtain the demand response cost of each sub-microgrid, and set equipment operation constraints based on the equipment operation parameters of each sub-microgrid, so as to build a two-layer game scheduling model with minimizing the total operating cost of the microgrid group as the upper-layer optimization objective and minimizing the local operating cost of each sub-microgrid as the lower-layer optimization objective. The optimization solution module is used to solve the two-layer game scheduling model through distributed iteration to obtain the scheduling scheme and power interaction strategy of each sub-micronet.

Images