Intelligent agent-based virtual power plant adjustable resource aggregation scheduling method

Abstract

The invention discloses a virtual power plant adjustable resource aggregation scheduling method based on an intelligent agent, and the method comprises the steps: carrying out the electric energy supply and demand evaluation analysis of each subarea managed by a power distribution network, and forming electric energy regulation and control subtasks of different time sequences of each subarea; a central scheduling agent issues a scheduling task package after considering the electric energy regulation and control sub-task; aggregating distributed power supplies, industrial and commercial solid-state energy storage equipment, controllable load elastic resources and mobile charging and discharging resources in respective partitions through a plurality of partition adjustable resource aggregation agents, and performing bidding quotation on a scheduling task packet based on a game mechanism of each partition, self adjustable resource characteristic parameters and benefit parameters; the central scheduling agent synthesizes the bidding quotation and the adjustable resource characteristic parameters of the adjustable resource aggregation agent of each partition, and preferably selects an adjustable resource scheduling scheme with the minimum global cost and the highest task completion rate; and according to the task completion rate of each partition adjustable resource aggregation agent, allocating the income of each partition.

Description

Technical Field

[0001] This invention belongs to the field of virtual power plant scheduling technology, specifically relating to a method for aggregated scheduling of adjustable resources in a virtual power plant based on intelligent agents. Background Technology

[0002] The new power system is currently facing the dual pressures of new energy consumption and power supply security. As a flexible hub for adjustable resources, virtual power plants can coordinate and manage various types of flexible adjustable resources to alleviate the supply and demand balance problem, and aggregate scattered adjustable resources such as charging piles, photovoltaics, and air conditioners to participate in the balance regulation of electricity.

[0003] However, various types of adjustable resources are distributed across different regions, and the resource adjustment capacity and adjustable time periods also vary. They are affected by uncertain factors such as weather and user behavior, making it difficult to aggregate and coordinate the scheduling and management of resources. In addition, traditional scheduling strategies do not consider the differences in supply and demand between regions and time periods, as well as the adjustable resource parameters of each region. Furthermore, they lack effective game theory and cooperation mechanisms. When various resources are aggregated, there are conflicts of interest, making it difficult to achieve global optimization. When coordinating across regions, the allocation of rights, responsibilities, and benefits is ambiguous, resulting in low willingness to cooperate.

[0004] Based on the above technical problems, a new agent-based method for the aggregation and scheduling of adjustable resources in a virtual power plant is needed. Summary of the Invention

[0005] The technical problem to be solved by this invention is to overcome the shortcomings of the prior art and provide a virtual power plant adjustable resource aggregation and scheduling method based on intelligent agents. This method can perform power supply and demand assessment by region and time sequence, aggregate resources within a region through multi-region adjustable resource aggregation intelligent agents, and participate in competitive and cooperative game-playing of scheduling tasks through decision-making. Furthermore, it uses a central scheduling intelligent agent to coordinate and manage the resources of each region, thereby tapping the adjustable potential of resources, avoiding resource waste, improving the overall control capacity of the virtual power plant, and simultaneously taking into account both global cost optimization and individual interests of each region.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: This invention provides a method for aggregated scheduling of adjustable resources in a virtual power plant based on intelligent agents, comprising: S1. Conduct power supply and demand assessment and analysis for each zone under the jurisdiction of the distribution network, obtain the power supply and demand parameters of each zone, form power control sub-tasks for each zone at different times, and transmit them to the virtual power plant dispatching platform. S2. Set up a central dispatching agent and multiple partition adjustable resource aggregation agents in the virtual power plant dispatching platform, and release dispatching task packages after considering the power regulation sub-tasks of different time sequences of each partition through the central dispatching agent. S3. Multiple partitioned adjustable resource aggregation intelligent agents aggregate distributed power sources, industrial and commercial solid-state energy storage devices, controllable load elastic resources, and mobile charging and discharging resources within their respective partitions, and bid on scheduling task packages based on the game mechanism of each partition, its own adjustable resource characteristic parameters, and benefit parameters; wherein, the game mechanism of each partition includes: multiple partitioned adjustable resource aggregation intelligent agents forming a competitive game for the same task package, or cross-partition collaboration forming a cooperative game; S4. By integrating the bidding quotations and adjustable resource characteristic parameters of the adjustable resource aggregation agents in each partition through the central scheduling agent, the optimal adjustable resource scheduling scheme with the lowest global cost and the highest task completion rate is selected. S5. After executing the adjustable resource scheduling scheme, the central scheduling agent allocates the revenue to each region based on the task completion rate of the adjustable resource aggregation agents in each region.

[0007] Furthermore, S1 includes: Data from the supply side, demand side, and network side of each zone under the jurisdiction of the distribution network are collected to obtain data on the power supply, power load demand, voltage, line loss rate, and operating status of the distribution network in each zone. The supply and demand status of each zone is divided according to the power supply and load demand of each zone, including supply exceeding demand, supply falling short of demand, and supply and demand in balance. For zones with supply-demand imbalances, risk levels are assigned based on preset voltage deviation and line loss rate indicators: low risk is assigned when both voltage deviation and line loss rate are within safe limits; medium risk is assigned when voltage deviation is within the first voltage range and line loss rate is within the first line loss range; and high risk is assigned when voltage deviation is within the second voltage range and line loss rate is within the second line loss range. Based on the suddenness, duration, and risk level of the supply-demand imbalance, the control zones, control duration intervals, control targets, control supply-demand imbalance measures, and control task types are determined, forming power control sub-tasks with different time sequences for each zone, which are then encapsulated as structured data and transmitted to the virtual power plant dispatch platform.

[0008] Furthermore, the control timing sequence includes ultra-short-term timing, short-term timing, and medium-to-long-term timing; the triggering scenario for the ultra-short-term timing is a sudden change in supply and demand status and a high-risk zone, with the control objective being to quickly smooth out supply and demand imbalances and avoid grid failures, and the control duration being within 15 minutes; the triggering scenario for the short-term timing is a zone with a persistent supply and demand imbalance and a medium-risk zone, with the control objective being to smoothly regulate supply and demand and maintain grid stability, and the control duration being 1 to 4 hours; the triggering scenario for the medium-to-long-term timing is a zone with a predictable supply and demand imbalance and a low-risk zone, with the control objective being to prevent the expansion of supply and demand imbalances, and the control duration being 4 to 24 hours; The types of control tasks include emergency energy absorption tasks, routine energy absorption tasks, pre-stored energy tasks, emergency energy replenishment tasks, routine energy replenishment tasks, and load transfer tasks.

[0009] Furthermore, S2 includes: A central dispatching agent is set up in the virtual power plant dispatching platform, which integrates a task parsing module, a resource matching module, a game coordination module, and a global optimization module. It is responsible for coordinating the overall control tasks. Multiple regional adjustable resource aggregation agents are deployed according to the distribution network zones, which integrate a resource status acquisition module and a local decision-making module. They are responsible for managing the adjustable resources within their respective zones. After receiving power regulation sub-tasks of different time sequences from the distribution network side, the central dispatching intelligent agent parses the sub-tasks through the task parsing module, merges multiple sub-tasks of the same task type and level under the same regulation time interval, and classifies multiple sub-tasks of different task types or different task levels, generates the overall dispatching task package under different regulation time intervals, and publishes it to the adjustable resource aggregation intelligent agent of each zone.

[0010] Furthermore, S3 includes: The system aggregates distributed power sources, commercial and industrial solid-state energy storage, controllable load elastic resources, and mobile charging and discharging resources within each zone through a multi-zone adjustable resource aggregation intelligent agent, and labels core characteristic parameters for each type of resource. Distributed power sources include photovoltaic generators and wind turbine generators; commercial and industrial solid-state energy storage utilizes the adjustability of solid-state energy storage devices and electricity consumption periods to adjust charging and discharging periods and capacity; controllable load elastic resources include the transfer and reduction of flexible loads, and the regulation of energy demand through energy conversion equipment; mobile charging and discharging resources include mobile charging and discharging vehicles that achieve flexible adjustment and interaction of power resources with the distribution network and electricity users through charging piles. The core characteristic parameters include the output of distributed power sources at different times, the maximum charging and discharging power and remaining capacity of commercial and industrial solid-state energy storage, the adjustable power range and response delay of controllable load elastic resources, the available charging and discharging periods, charging and discharging power, the number of vehicles that can be connected, and the vehicle's response willingness for mobile charging and discharging resources, as well as the scheduling cost for each type of resource. Each partition's adjustable resource aggregation agent learns whether there are tasks in its own partition based on the scheduling task package. If there are, it first determines whether the adjustable resources in this partition can meet the partition's task requirements. If they can, the adjustable resources in the partition complete the task independently. If they cannot, it sends a request to the central scheduling agent to inform it of the partition's power shortage. If there are no tasks, it waits for the central scheduling agent to issue task competition requests or task cooperation requests from other partitions and then bids for them. In this system, when each partition's adjustable resource aggregation agent can independently complete the tasks within its own partition using its adjustable resources, it does not need to bid for a price; instead, the local adjustable resources within the partition are scheduled by that partition's adjustable resource aggregation agent. When the central scheduling agent receives a request for a partition's power shortage from a partition's adjustable resource aggregation agent, if the partition's power shortage is in a low range, it uses a competitive game mechanism to form a new competitive game task package with a reward mechanism, and then publishes it. Other partition's adjustable resource aggregation agents then compete for this competitive game task package and its reward. The incentive mechanism considers the characteristics of its own adjustable resources and the benefit parameters after resource scheduling to conduct competitive game and bidding. If the power shortage of a region is in a high range, it means that the power shortage of that region requires cross-regional cooperation of multiple adjustable resource aggregation agents. The power shortage request of that region is formed into a new cooperative game task package with a reward mechanism attached and released. Other adjustable resource aggregation agents of the region conduct cooperative game and bidding based on the cooperative game task package and reward mechanism, considering their own adjustable resource characteristics and the benefit parameters after resource scheduling.

[0011] Furthermore, S4 includes: The central dispatching agent classifies the bidding proposals submitted by the adjustable resource aggregation agents of each region into competitive game task packages and collaborative game task packages. It verifies whether the bidding proposals meet the control response time limit, control objectives, control supply and demand imbalance and control task type of the task package. It filters out bidding proposals that do not meet the relevant indicators and retains valid proposals that meet the relevant indicators, forming an initial selection set of adjustable resource scheduling schemes for each region corresponding to competitive game task packages and collaborative game task packages. Calculate the global cost and task completion rate of each scheduling scheme in the preliminary selection set of adjustable resource scheduling schemes for each partition corresponding to the competitive game task package and the cooperative game task package; the global cost includes the sum of partition scheduling costs calculated by the adjustable resource aggregation agent of each partition based on the bidding resource scheduling parameters and bidding information, and the communication cost of cross-partition cooperation; the task completion rate is predicted based on the resource characteristic parameters of each scheduling scheme and the bidding resource scheduling parameters. The scheduling scheme with the lowest global cost and the highest task completion rate is selected as the preferred adjustable resource scheduling scheme, and then distributed to the corresponding adjustable resource aggregation agents in each partition for execution.

[0012] Furthermore, when the partitioned adjustable resource aggregation agent engages in competitive games and bidding, it comprehensively considers the resource schedulable parameters, resource competition response speed, resource scheduling cost, and task type indicators of each partition, and introduces dynamic weight coefficients for each indicator in the game, automatically adjusting the resource scheduling parameters and bidding strategies during the competitive game. When the partitioned adjustable resource aggregation agent engages in collaborative game and bidding, it considers the complementarity index of each partition's adjustable resources, resource collaboration scheduling cost, task type, resource collaboration response speed, and historical collaboration trust index, and introduces dynamic weight coefficients for each index to automatically adjust the resource collaboration scheduling parameters and bidding strategy during collaborative game.

[0013] Furthermore, when selecting the scheduling scheme with the minimum global cost and the highest task completion rate as the preferred adjustable resource scheduling scheme, a combination of multi-objective optimization algorithm and game equilibrium algorithm is used to solve the dual objectives of minimizing global cost and maximizing task completion rate. Among them, the multi-objective optimization algorithm is the non-dominated sorting genetic algorithm NSGA-III, combined with the TOPSIS approximation ideal solution sorting method. After encoding each scheduling scheme, non-dominated sorting, generating the next generation population, and iteratively generating the Pareto optimal solution set using NSGA-III, TOPSIS is used to construct a decision matrix with the Pareto optimal solution set, determine the positive and negative ideal solutions and calculate the closeness, and then select the scheme with the highest closeness as the final preferred adjustable resource scheduling scheme. The application of the game equilibrium algorithm includes: in competitive games, adding a reasonableness constraint on the objective of NSGA-III, where the bid price of each partition is less than or equal to the reward cap of the task package, and the bid price is greater than or equal to the partition resource scheduling cost; adjustable resource scheduling schemes that do not meet the bid constraints are directly marked as infeasible solutions; in cooperative games, adding cooperative parameters to the encoding of NSGA-III, including the number of cooperative partitions and the resource contribution ratio of each partition, and embedding Nash bargaining equilibrium constraints in the non-dominated sorting and iterative calculation of NSGA-III, so that the payoff of each cooperative partition is greater than or equal to the payoff of that partition performing the task alone, and the total cooperative payoff is greater than or equal to the sum of the payoffs of each partition performing the task alone, indicating that the adjustable resource scheduling scheme must satisfy the Nash bargaining equilibrium, so that the cooperative schemes in the Pareto solution set meet the game fairness.

[0014] Furthermore, in the TOPSIS screening process, a new dimension of game strategy effectiveness is added to the decision matrix, including the bidding competitiveness of the solution in competitive games and the satisfaction of the benefit distribution in cooperative games. After assigning weights to global cost, task completion rate, and game strategy effectiveness, the closeness of TOPSIS is calculated, so that the preferred adjustable resource scheduling solution is stably executed in the game.

[0015] Furthermore, S5 includes: After executing the adjustable resource scheduling scheme, each partition's adjustable resource aggregation intelligent agent reports the actual completed data. The central scheduling intelligent agent then compares the actual completed data with the scheduling task package to calculate the task completion rate. For partitions participating in competitive games, the first revenue rule is applied: Partition revenue = Task package reward × Task completion rate; For partitions participating in collaborative games, the second revenue rule is applied: Partition revenue = (Task package reward × Total task completion rate) × (Resource usage of this partition / Total resource usage of the collaborative group). Based on the benefit parameters of each partition, the partition revenue is adjusted twice: if the partition revenue is less than the resource scheduling cost, the minimum compensation mechanism is triggered, and the virtual power plant scheduling platform makes up the difference; if the partition's task completion rate is ≥100% and there are no constraints or violations, the excess reward is triggered, and the virtual power plant scheduling platform adds an extra preset incentive revenue. The central dispatching agent generates a settlement statement for each partition's adjustable resource aggregation agent, which includes a task package identifier, task completion rate, revenue amount, and correction instructions. The settlement statement is automatically executed and the revenue is distributed using a blockchain smart contract mechanism. At the same time, the settlement statement is stored on the blockchain and the settlement results are synchronized to each partition's adjustable resource aggregation agent.

[0016] The beneficial effects of this invention are: (1) This invention accurately identifies the power surplus and shortage status of each region by dividing the power supply and demand into regions and time sequences, and generates targeted control sub-tasks to avoid local supply and demand imbalance caused by general control. (2) This invention achieves global overall scheduling and local decision-making and execution in each partition through a central scheduling agent and multiple partition adjustable resource aggregation agents. It can accurately analyze and aggregate the adjustable resources of each partition, which facilitates the central scheduling agent to perform global unified scheduling management. (3) This invention aggregates various types of resources within a partition through a partition-adjustable resource aggregation intelligent agent, thereby achieving the coordination of multiple flexible and adjustable resources. It also sets up competitive and cooperative game mechanisms to incentivize each partition-adjustable resource aggregation intelligent agent to actively explore the adjustable potential of local resources. Furthermore, through the cooperation of each partition, resource waste can be avoided and the overall control capacity of the virtual power plant can be improved. (4) This invention optimizes the scheme with the lowest global cost and the highest task completion rate by the central scheduling agent, which reduces the overall control cost of the virtual power plant and ensures the task execution effect. At the same time, the game bidding mechanism allows the partitioned adjustable resource aggregation agent to bid based on its own resource scheduling cost and scheduling parameters, so as to take into account both the global optimal and the individual interests of the partition. (5) The present invention has a revenue distribution mechanism based on task completion rate, in which partitions with high completion rates receive higher revenue, and partitions that violate regulations have their revenue reduced, thus forming a positive incentive. In the collaborative game scenario, revenue is distributed according to the proportion of resource contribution, ensuring the fairness of cross-partition collaboration and enhancing the willingness of partition adjustable resource aggregation agents to participate in long-term regulation. (6) The intelligent agent architecture of the present invention has good scalability. When adding a new partition or resource type, it is only necessary to connect the corresponding partition adjustable resource aggregation intelligent agent, without reconstructing the overall scheduling system. The game bidding mechanism is compatible with the power market trading rules and can be directly connected to the spot market and the ancillary service market, enabling virtual power plants to participate in market transactions to obtain additional benefits and improve the feasibility of the business model.

[0017] Other features and advantages will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention are realized and obtained through the structures particularly pointed out in the description and the drawings.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a flowchart of a virtual power plant adjustable resource aggregation and scheduling method based on intelligent agents according to the present invention; Figure 2 This is a schematic block diagram illustrating the adjustable resource aggregation and scheduling of the virtual power plant according to the present invention. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] like Figure 1 , Figure 2 As shown, this embodiment provides a method for aggregated scheduling of adjustable resources in a virtual power plant based on intelligent agents, which includes: S1. Conduct power supply and demand assessment and analysis for each zone under the jurisdiction of the distribution network, obtain the power supply and demand parameters of each zone, form power control sub-tasks for each zone at different times, and transmit them to the virtual power plant dispatching platform. S2. Set up a central dispatching agent and multiple partition adjustable resource aggregation agents in the virtual power plant dispatching platform, and release dispatching task packages after considering the power regulation sub-tasks of different time sequences of each partition through the central dispatching agent. S3. Multiple partitioned adjustable resource aggregation intelligent agents aggregate distributed power sources, industrial and commercial solid-state energy storage devices, controllable load elastic resources, and mobile charging and discharging resources within their respective partitions, and bid on scheduling task packages based on the game mechanism of each partition, its own adjustable resource characteristic parameters, and benefit parameters; wherein, the game mechanism of each partition includes: multiple partitioned adjustable resource aggregation intelligent agents forming a competitive game for the same task package, or cross-partition collaboration forming a cooperative game; S4. By integrating the bidding quotations and adjustable resource characteristic parameters of the adjustable resource aggregation agents in each partition through the central scheduling agent, the optimal adjustable resource scheduling scheme with the lowest global cost and the highest task completion rate is selected. S5. After executing the adjustable resource scheduling scheme, the central scheduling agent allocates the revenue to each region based on the task completion rate of the adjustable resource aggregation agents in each region.

[0023] In this embodiment, S1 includes: Data from the supply side, demand side, and network side of each zone under the jurisdiction of the distribution network are collected to obtain data on the power supply, power load demand, voltage, line loss rate, and operating status of the distribution network in each zone. The supply and demand status of each zone is divided according to the power supply and load demand of each zone, including supply exceeding demand, supply falling short of demand, and supply and demand in balance. For zones with supply-demand imbalances, risk levels are assigned based on preset voltage deviation and line loss rate indicators: low risk is assigned when both voltage deviation and line loss rate are within safe limits; medium risk is assigned when voltage deviation is within the first voltage range and line loss rate is within the first line loss range; and high risk is assigned when voltage deviation is within the second voltage range and line loss rate is within the second line loss range. Based on the suddenness, duration, and risk level of the supply-demand imbalance, the control zones, control duration intervals, control targets, control supply-demand imbalance measures, and control task types are determined, forming power control sub-tasks with different time sequences for each zone, which are then encapsulated as structured data and transmitted to the virtual power plant dispatch platform.

[0024] In this embodiment, the control timing includes ultra-short-term timing, short-term timing, and medium-to-long-term timing. The ultra-short-term timing is triggered by sudden changes in supply and demand and a high-risk partition, with the control objective being to quickly smooth out supply and demand imbalances and avoid grid failures, and the control duration is within 15 minutes. The short-term timing is triggered by a persistent supply and demand imbalance and a medium-risk partition, with the control objective being to smoothly regulate supply and demand and maintain grid stability, and the control duration is between 1 and 4 hours. The medium-to-long-term timing is triggered by a predictable supply and demand imbalance and a low-risk partition, with the control objective being to prevent the supply and demand imbalance from expanding, and the control duration is between 4 and 24 hours. The types of control tasks include emergency energy absorption tasks, routine energy absorption tasks, pre-stored energy tasks, emergency energy replenishment tasks, routine energy replenishment tasks, and load transfer tasks.

[0025] It should be noted that the following tasks are categorized: Emergency power consumption: Utilizing controllable load elastic resources and energy storage charging to consume excess power; Regular power consumption: Slightly increasing the power of controllable load elastic resources to slowly consume excess power; Pre-storage: Prioritizing the storage of excess power in energy storage for later use; Emergency power replenishment: Utilizing industrial and commercial solid-state energy storage and mobile charging / discharging resources for discharge, and reducing unnecessary controllable load elastic resources to supplement the power gap; Regular power replenishment: Utilizing distributed power sources to increase output and slightly reducing controllable load elastic resources; Load transfer: Transferring some load to off-peak hours to reduce the current gap. These are just examples of flexible resources that can be scheduled for each task type. During the scheduling of adjustable resources in the virtual power plant, these can be dynamically changed based on the actual available adjustable resources.

[0026] In this embodiment, S2 includes: A central dispatching agent is set up in the virtual power plant dispatching platform, which integrates a task parsing module, a resource matching module, a game coordination module, and a global optimization module. It is responsible for coordinating the overall control tasks. Multiple regional adjustable resource aggregation agents are deployed according to the distribution network zones, which integrate a resource status acquisition module and a local decision-making module. They are responsible for managing the adjustable resources within their respective zones. After receiving power regulation sub-tasks of different time sequences from the distribution network side, the central dispatching intelligent agent parses the sub-tasks through the task parsing module, merges multiple sub-tasks of the same task type and level under the same regulation time interval, and classifies multiple sub-tasks of different task types or different task levels, generates the overall dispatching task package under different regulation time intervals, and publishes it to the adjustable resource aggregation intelligent agent of each zone.

[0027] In this embodiment, S3 includes: The system aggregates distributed power sources, commercial and industrial solid-state energy storage, controllable load elastic resources, and mobile charging and discharging resources within each zone through a multi-zone adjustable resource aggregation intelligent agent, and labels core characteristic parameters for each type of resource. Distributed power sources include photovoltaic generators and wind turbine generators; commercial and industrial solid-state energy storage utilizes the adjustability of solid-state energy storage devices and electricity consumption periods to adjust charging and discharging periods and capacity; controllable load elastic resources include the transfer and reduction of flexible loads, and the regulation of energy demand through energy conversion equipment; mobile charging and discharging resources include mobile charging and discharging vehicles that achieve flexible adjustment and interaction of power resources with the distribution network and electricity users through charging piles. The core characteristic parameters include the output of distributed power sources at different times, the maximum charging and discharging power and remaining capacity of commercial and industrial solid-state energy storage, the adjustable power range and response delay of controllable load elastic resources, the available charging and discharging periods, charging and discharging power, the number of vehicles that can be connected, and the vehicle's response willingness for mobile charging and discharging resources, as well as the scheduling cost for each type of resource. Each partition's adjustable resource aggregation agent learns whether there are tasks in its own partition based on the scheduling task package. If there are, it first determines whether the adjustable resources in this partition can meet the partition's task requirements. If they can, the adjustable resources in the partition complete the task independently. If they cannot, it sends a request to the central scheduling agent to inform it of the partition's power shortage. If there are no tasks, it waits for the central scheduling agent to issue task competition requests or task cooperation requests from other partitions and then bids for them. In this system, when each partition's adjustable resource aggregation agent can independently complete the tasks within its own partition using its adjustable resources, it does not need to bid for a price; instead, the local adjustable resources within the partition are scheduled by that partition's adjustable resource aggregation agent. When the central scheduling agent receives a request for a partition's power shortage from a partition's adjustable resource aggregation agent, if the partition's power shortage is in a low range, it uses a competitive game mechanism to form a new competitive game task package with a reward mechanism, and then publishes it. Other partition's adjustable resource aggregation agents then compete for this competitive game task package and its reward. The incentive mechanism considers the characteristics of its own adjustable resources and the benefit parameters after resource scheduling to conduct competitive game and bidding. If the power shortage of a region is in a high range, it means that the power shortage of that region requires cross-regional cooperation of multiple adjustable resource aggregation agents. The power shortage request of that region is formed into a new cooperative game task package with a reward mechanism attached and released. Other adjustable resource aggregation agents of the region conduct cooperative game and bidding based on the cooperative game task package and reward mechanism, considering their own adjustable resource characteristics and the benefit parameters after resource scheduling.

[0028] It should be noted that the regulation logic of distributed power sources is to participate in the power supply regulation of the power grid by adjusting the output at different times; industrial and commercial solid-state energy storage is based on solid-state energy storage technology and industrial and commercial energy storage equipment, such as energy storage systems in factories and commercial buildings. The regulation logic is to utilize the adjustability of energy storage equipment and electrical equipment to adjust the charging and discharging time and capacity; controllable load elastic resources are mainly flexible loads, such as industrial production loads and commercial air conditioning loads. Flexible regulation is achieved through load transfer (such as shifting production load from peak to off-peak), load reduction (such as temporarily reducing unnecessary loads), and adjustment of energy conversion equipment (such as adjusting the power of electric heat pumps, reducing the electrical energy input of electric heat pumps, and reducing heat output); mobile charging and discharging resources include mobile charging and discharging vehicles and charging piles. The regulation logic is that mobile charging and discharging vehicles interact with the distribution network and users through charging piles. For example, when there is excess power, the vehicle charges to consume the power, and when there is a power shortage, the vehicle discharges to replenish the power.

[0029] In this embodiment, S4 includes: The central dispatching agent classifies the bidding proposals submitted by the adjustable resource aggregation agents of each region into competitive game task packages and collaborative game task packages. It verifies whether the bidding proposals meet the control response time limit, control objectives, control supply and demand imbalance and control task type of the task package. It filters out bidding proposals that do not meet the relevant indicators and retains valid proposals that meet the relevant indicators, forming an initial selection set of adjustable resource scheduling schemes for each region corresponding to competitive game task packages and collaborative game task packages. Calculate the global cost and task completion rate of each scheduling scheme in the preliminary selection set of adjustable resource scheduling schemes for each partition corresponding to the competitive game task package and the cooperative game task package; the global cost includes the sum of partition scheduling costs calculated by the adjustable resource aggregation agent of each partition based on the bidding resource scheduling parameters and bidding information, and the communication cost of cross-partition cooperation; the task completion rate is predicted based on the resource characteristic parameters of each scheduling scheme and the bidding resource scheduling parameters. The scheduling scheme with the lowest global cost and the highest task completion rate is selected as the preferred adjustable resource scheduling scheme, and then distributed to the corresponding adjustable resource aggregation agents in each partition for execution.

[0030] In this embodiment, when the partitioned adjustable resource aggregation agent conducts competitive games and bids, it comprehensively considers the resource schedulable parameters, resource competition response speed, resource scheduling cost, and task type indicators of each partition, and introduces dynamic weight coefficients for each indicator to automatically adjust the resource scheduling parameters and bidding strategies during the competitive games. When the partitioned adjustable resource aggregation agent engages in collaborative game and bidding, it considers the complementarity index of each partition's adjustable resources, resource collaboration scheduling cost, task type, resource collaboration response speed, and historical collaboration trust index, and introduces dynamic weight coefficients for each index to automatically adjust the resource collaboration scheduling parameters and bidding strategy during collaborative game.

[0031] In this embodiment, when selecting the scheduling scheme with the minimum global cost and the highest task completion rate as the preferred adjustable resource scheduling scheme, a combination of multi-objective optimization algorithm and game equilibrium algorithm is used to solve the dual objectives of minimizing global cost and maximizing task completion rate. Among them, the multi-objective optimization algorithm is the non-dominated sorting genetic algorithm NSGA-III, combined with the TOPSIS approximation ideal solution sorting method. After encoding each scheduling scheme, non-dominated sorting, generating the next generation population, and iteratively generating the Pareto optimal solution set using NSGA-III, TOPSIS is used to construct a decision matrix with the Pareto optimal solution set, determine the positive and negative ideal solutions and calculate the closeness, and then select the scheme with the highest closeness as the final preferred adjustable resource scheduling scheme. The application of the game equilibrium algorithm includes: in competitive games, adding a reasonableness constraint on the objective of NSGA-III, where the bid price of each partition is less than or equal to the reward cap of the task package, and the bid price is greater than or equal to the partition resource scheduling cost; adjustable resource scheduling schemes that do not meet the bid constraints are directly marked as infeasible solutions; in cooperative games, adding cooperative parameters to the encoding of NSGA-III, including the number of cooperative partitions and the resource contribution ratio of each partition, and embedding Nash bargaining equilibrium constraints in the non-dominated sorting and iterative calculation of NSGA-III, so that the payoff of each cooperative partition is greater than or equal to the payoff of that partition performing the task alone, and the total cooperative payoff is greater than or equal to the sum of the payoffs of each partition performing the task alone, indicating that the adjustable resource scheduling scheme must satisfy the Nash bargaining equilibrium, so that the cooperative schemes in the Pareto solution set meet the game fairness.

[0032] In practical applications, using NSGA-III to solve Pareto optimal solution sets includes: Encoding and initializing the population: Each scheduling scheme is encoded as a chromosome of adjustable resource scheduling combinations and partition allocation ratios, and an initial scheme population is randomly generated; Non-dominated sorting and crowding calculation: The schemes in the population are non-dominatedly sorted and divided into different levels (the lower the level, the better the scheme); the crowding of schemes at the same level is calculated (measures the degree of dispersion of schemes in the solution set) to ensure the diversity of the solution set; Genetic operations and iterative optimization: The next generation population is generated through selection, crossover, and mutation operations; the process is iterated to a preset number of times to finally obtain the Pareto optimal solution set.

[0033] The TOPSIS-based selection criteria include: Using the Pareto solution set as the decision matrix, rows represent alternatives, and columns represent global cost and task completion rate; Determine the positive ideal solution: the adjustable resource scheduling scheme with the minimum global cost and the highest task completion rate; determine the negative ideal solution: the adjustable resource scheduling scheme with the maximum global cost and the lowest task completion rate. Calculate the Euclidean distance between each scheme and the positive and negative ideal solutions to obtain the proximity score (the closer to 1, the better the scheme); select the adjustable resource scheduling scheme with the highest proximity score as the final preferred scheduling scheme.

[0034] In this embodiment, during the TOPSIS screening process, a new dimension of game strategy effectiveness is added to the decision matrix, including the bidding competitiveness of the scheme in competitive games and the satisfaction of the benefit distribution in cooperative games. After allocating the weights of global cost, task completion rate, and game strategy effectiveness, the closeness of TOPSIS is calculated, so that the preferred adjustable resource scheduling scheme is stably executed in the game.

[0035] It should be noted that the bidding competitiveness = (task package reward cap - solution bid) / task package reward cap; the higher the value, the stronger the competitiveness. Revenue distribution satisfaction = the average matching degree between the revenue and expected revenue of each collaborative partition; the higher the value, the stronger the collaboration stability. For competitive game scenarios: NSGA-III (embedded with bidding rationality constraints) – generates Pareto solution set of competitive solutions – TOPSIS (adds bidding competitiveness dimension) – selects the optimal partition competitive solution. For collaborative game scenarios: NSGA-III (embedded with collaboration parameters and Nash bargaining equilibrium constraints) – generates Pareto solution set of competitive solutions – TOPSIS (adds revenue distribution satisfaction dimension) – selects the optimal partition collaborative solution. This integrated solution method retains the dual-objective optimization capabilities of NSGA-III and TOPSIS, and ensures that the adjustable resource scheduling solution conforms to market and collaboration rules through game constraints, matching the needs of competitive and collaborative games.

[0036] In this embodiment, S5 includes: After executing the adjustable resource scheduling scheme, each partition's adjustable resource aggregation intelligent agent reports the actual completed data. The central scheduling intelligent agent then compares the actual completed data with the scheduling task package to calculate the task completion rate. For partitions participating in competitive games, the first revenue rule is applied: Partition revenue = Task package reward × Task completion rate; For partitions participating in collaborative games, the second revenue rule is applied: Partition revenue = (Task package reward × Total task completion rate) × (Resource usage of this partition / Total resource usage of the collaborative group). Based on the benefit parameters of each partition, the partition revenue is adjusted twice: if the partition revenue is less than the resource scheduling cost, the minimum compensation mechanism is triggered, and the virtual power plant scheduling platform makes up the difference; if the partition's task completion rate is ≥100% and there are no constraints or violations, the excess reward is triggered, and the virtual power plant scheduling platform adds an extra preset incentive revenue. The central dispatching agent generates a settlement statement for each partition's adjustable resource aggregation agent, which includes a task package identifier, task completion rate, revenue amount, and correction instructions. The settlement statement is automatically executed and the revenue is distributed using a blockchain smart contract mechanism. At the same time, the settlement statement is stored on the blockchain and the settlement results are synchronized to each partition's adjustable resource aggregation agent.

[0037] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can also be implemented in other ways. The system embodiments described above are merely illustrative; for example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.

[0038] Furthermore, the functional modules in the various embodiments of this invention can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part. If the function is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0039] Based on the above-described preferred embodiments of the present invention, and through the foregoing description, those skilled in the art can make various changes and modifications without departing from the inventive concept. The technical scope of this invention is not limited to the contents of the specification, but must be determined according to the scope of the claims.

Claims

1. A method for aggregated scheduling of adjustable resources in a virtual power plant based on intelligent agents, characterized in that, It includes: S1. Conduct power supply and demand assessment and analysis for each zone under the jurisdiction of the distribution network, obtain the power supply and demand parameters of each zone, form power control sub-tasks for each zone at different times, and transmit them to the virtual power plant dispatching platform. S2. Set up a central dispatching agent and multiple partition adjustable resource aggregation agents in the virtual power plant dispatching platform, and release dispatching task packages after considering the power regulation sub-tasks of different time sequences of each partition through the central dispatching agent. S3. Multiple partitioned adjustable resource aggregation intelligent agents aggregate distributed power sources, industrial and commercial solid-state energy storage devices, controllable load elastic resources, and mobile charging and discharging resources within their respective partitions, and bid on scheduling task packages based on the game mechanism of each partition, its own adjustable resource characteristic parameters, and benefit parameters; wherein, the game mechanism of each partition includes: multiple partitioned adjustable resource aggregation intelligent agents forming a competitive game for the same task package, or cross-partition collaboration forming a cooperative game; S4. By integrating the bidding quotations and adjustable resource characteristic parameters of the adjustable resource aggregation agents in each partition through the central scheduling agent, the optimal adjustable resource scheduling scheme with the lowest global cost and the highest task completion rate is selected. S5. After executing the adjustable resource scheduling scheme, the central scheduling agent allocates the revenue to each region based on the task completion rate of the adjustable resource aggregation agents in each region.

2. The virtual power plant adjustable resource aggregation and scheduling method according to claim 1, characterized in that, S1 includes: Data from the supply side, demand side, and network side of each zone under the jurisdiction of the distribution network are collected to obtain data on the power supply, power load demand, voltage, line loss rate, and operating status of the distribution network in each zone. The supply and demand status of each zone is divided according to the power supply and load demand of each zone, including supply exceeding demand, supply falling short of demand, and supply and demand in balance. For zones with supply-demand imbalances, risk levels are assigned based on preset voltage deviation and line loss rate indicators: low risk is assigned when both voltage deviation and line loss rate are within safe limits; medium risk is assigned when voltage deviation is within the first voltage range and line loss rate is within the first line loss range; and high risk is assigned when voltage deviation is within the second voltage range and line loss rate is within the second line loss range. Based on the suddenness, duration, and risk level of the supply-demand imbalance, the control zones, control duration intervals, control targets, control supply-demand imbalance measures, and control task types are determined, forming power control sub-tasks with different time sequences for each zone, which are then encapsulated as structured data and transmitted to the virtual power plant dispatch platform.

3. The virtual power plant adjustable resource aggregation and scheduling method according to claim 2, characterized in that, The control timing sequence includes ultra-short-term timing, short-term timing, and medium-to-long-term timing; the triggering scenario for the ultra-short-term timing is a sudden change in supply and demand status and a high-risk zone, with the control objective being to quickly restore supply and demand imbalance and avoid grid failure, and the control duration being within 15 minutes; the triggering scenario for the short-term timing is a zone with a continuous supply and demand imbalance and a medium-risk zone, with the control objective being to smoothly regulate supply and demand and maintain grid stability, and the control duration being 1 to 4 hours; The triggering scenario for the medium- and long-term time series is a zone with predictive supply and demand imbalance and a low risk level. The control objective is to prevent the supply and demand imbalance from expanding, and the control duration is 4 to 24 hours. The types of control tasks include emergency energy absorption tasks, routine energy absorption tasks, pre-stored energy tasks, emergency energy replenishment tasks, routine energy replenishment tasks, and load transfer tasks.

4. The virtual power plant adjustable resource aggregation and scheduling method according to claim 1, characterized in that, S2 includes: A central dispatching agent is set up in the virtual power plant dispatching platform, which integrates a task parsing module, a resource matching module, a game coordination module, and a global optimization module. It is responsible for coordinating the overall control tasks. Multiple regional adjustable resource aggregation agents are deployed according to the distribution network zones, which integrate a resource status acquisition module and a local decision-making module. They are responsible for managing the adjustable resources within their respective zones. After receiving power regulation sub-tasks of different time sequences from the distribution network side, the central dispatching intelligent agent parses the sub-tasks through the task parsing module, merges multiple sub-tasks of the same task type and level under the same regulation time interval, and classifies multiple sub-tasks of different task types or different task levels, generates the overall dispatching task package under different regulation time intervals, and publishes it to the adjustable resource aggregation intelligent agent of each zone.

5. The virtual power plant adjustable resource aggregation and scheduling method according to claim 1, characterized in that, S3 includes: The system aggregates distributed power sources, commercial and industrial solid-state energy storage, controllable load elastic resources, and mobile charging and discharging resources within each zone through a multi-zone adjustable resource aggregation intelligent agent, and labels core characteristic parameters for each type of resource. Distributed power sources include photovoltaic generators and wind turbine generators; commercial and industrial solid-state energy storage utilizes the adjustability of solid-state energy storage devices and electricity consumption periods to adjust charging and discharging periods and capacity; controllable load elastic resources include the transfer and reduction of flexible loads, and the regulation of energy demand through energy conversion equipment; mobile charging and discharging resources include mobile charging and discharging vehicles that achieve flexible adjustment and interaction of power resources with the distribution network and electricity users through charging piles. The core characteristic parameters include the output of distributed power sources at different times, the maximum charging and discharging power and remaining capacity of commercial and industrial solid-state energy storage, the adjustable power range and response delay of controllable load elastic resources, the available charging and discharging periods, charging and discharging power, the number of vehicles that can be connected, and the vehicle's response willingness for mobile charging and discharging resources, as well as the scheduling cost for each type of resource. Each partition's adjustable resource aggregation agent learns whether there are tasks in its own partition based on the scheduling task package. If there are, it first determines whether the adjustable resources in this partition can meet the partition's task requirements. If they can, the adjustable resources in the partition complete the task independently. If they cannot, it sends a request to the central scheduling agent to inform it of the partition's power shortage. If there are no tasks, it waits for the central scheduling agent to issue task competition requests or task cooperation requests from other partitions and then bids for them. In this system, when each partition's adjustable resource aggregation agent can independently complete the tasks within its own partition using its adjustable resources, it does not need to bid for a price; instead, the local adjustable resources within the partition are scheduled by that partition's adjustable resource aggregation agent. When the central scheduling agent receives a request for a partition's power shortage from a partition's adjustable resource aggregation agent, if the partition's power shortage is in a low range, it uses a competitive game mechanism to form a new competitive game task package with a reward mechanism, and then publishes it. Other partition's adjustable resource aggregation agents then compete for this competitive game task package and its reward. The incentive mechanism considers the characteristics of its own adjustable resources and the benefit parameters after resource scheduling to conduct competitive game and bidding. If the power shortage of a region is in a high range, it means that the power shortage of that region requires cross-regional cooperation of multiple adjustable resource aggregation agents. The power shortage request of that region is formed into a new cooperative game task package with a reward mechanism attached and released. Other adjustable resource aggregation agents of the region conduct cooperative game and bidding based on the cooperative game task package and reward mechanism, considering their own adjustable resource characteristics and the benefit parameters after resource scheduling.

6. The virtual power plant adjustable resource aggregation and scheduling method according to claim 5, characterized in that, S4 includes: The central dispatching agent classifies the bidding proposals submitted by the adjustable resource aggregation agents of each region into competitive game task packages and collaborative game task packages. It verifies whether the bidding proposals meet the control response time limit, control objectives, control supply and demand imbalance and control task type of the task package. It filters out bidding proposals that do not meet the relevant indicators and retains valid proposals that meet the relevant indicators, forming an initial selection set of adjustable resource scheduling schemes for each region corresponding to competitive game task packages and collaborative game task packages. Calculate the global cost and task completion rate of each scheduling scheme in the preliminary selection set of adjustable resource scheduling schemes for each partition corresponding to the competitive game task package and the cooperative game task package; the global cost includes the sum of partition scheduling costs calculated by the adjustable resource aggregation agent of each partition based on the bidding resource scheduling parameters and bidding information, and the communication cost of cross-partition cooperation; the task completion rate is predicted based on the resource characteristic parameters of each scheduling scheme and the bidding resource scheduling parameters. The scheduling scheme with the lowest global cost and the highest task completion rate is selected as the preferred adjustable resource scheduling scheme, and then distributed to the corresponding adjustable resource aggregation agents in each partition for execution.

7. The virtual power plant adjustable resource aggregation and scheduling method according to claim 5, characterized in that, When the partitioned adjustable resource aggregation agent engages in competitive games and bidding, it comprehensively considers the resource schedulable parameters, resource competition response speed, resource scheduling cost, and task type indicators of each partition, and introduces dynamic weight coefficients for each indicator in the game to automatically adjust the resource scheduling parameters and bidding strategies during the competitive game. When the partitioned adjustable resource aggregation agent engages in collaborative game and bidding, it considers the complementarity index of each partition's adjustable resources, resource collaboration scheduling cost, task type, resource collaboration response speed, and historical collaboration trust index, and introduces dynamic weight coefficients for each index to automatically adjust the resource collaboration scheduling parameters and bidding strategy during collaborative game.

8. The virtual power plant adjustable resource aggregation and scheduling method according to claim 5, characterized in that, When selecting the scheduling scheme with the minimum global cost and the highest task completion rate as the preferred adjustable resource scheduling scheme, a combination of multi-objective optimization algorithm and game equilibrium algorithm is used to solve the dual objectives of minimizing global cost and maximizing task completion rate. Among them, the multi-objective optimization algorithm is the non-dominated sorting genetic algorithm NSGA-III, combined with the TOPSIS approximation ideal solution sorting method. After encoding each scheduling scheme, non-dominated sorting, generating the next generation population, and iteratively generating the Pareto optimal solution set using NSGA-III, TOPSIS is used to construct a decision matrix with the Pareto optimal solution set, determine the positive and negative ideal solutions and calculate the closeness, and then select the scheme with the highest closeness as the final preferred adjustable resource scheduling scheme. The application of the game equilibrium algorithm includes: in competitive games, adding a reasonableness constraint on the objective of NSGA-III, where the bid price of each partition is less than or equal to the reward cap of the task package, and the bid price is greater than or equal to the partition resource scheduling cost; adjustable resource scheduling schemes that do not meet the bid constraints are directly marked as infeasible solutions; in cooperative games, adding cooperative parameters to the encoding of NSGA-III, including the number of cooperative partitions and the resource contribution ratio of each partition, and embedding Nash bargaining equilibrium constraints in the non-dominated sorting and iterative calculation of NSGA-III, so that the payoff of each cooperative partition is greater than or equal to the payoff of that partition performing the task alone, and the total cooperative payoff is greater than or equal to the sum of the payoffs of each partition performing the task alone, indicating that the adjustable resource scheduling scheme must satisfy the Nash bargaining equilibrium, so that the cooperative schemes in the Pareto solution set meet the game fairness.

9. The virtual power plant adjustable resource aggregation and scheduling method according to claim 8, characterized in that, In the TOPSIS screening process, a new dimension of game strategy effectiveness is added to the decision matrix, including the bidding competitiveness of the solution in competitive games and the satisfaction of the benefit distribution in cooperative games. After assigning weights to global cost, task completion rate and game strategy effectiveness, the closeness of TOPSIS is calculated to ensure that the preferred adjustable resource scheduling solution is stably executed in the game.

10. The virtual power plant adjustable resource aggregation and scheduling method according to claim 1, characterized in that, S5 includes: After executing the adjustable resource scheduling scheme, each partition's adjustable resource aggregation intelligent agent reports the actual completed data. The central scheduling intelligent agent then compares the actual completed data with the scheduling task package to calculate the task completion rate. For partitions participating in competitive games, the first revenue rule is applied: Partition revenue = Task package reward × Task completion rate; For partitions participating in collaborative games, the second revenue rule is applied: Partition revenue = (Task package reward × Total task completion rate) × (Resource usage of this partition / Total resource usage of the collaborative group). Based on the benefit parameters of each partition, the partition revenue is adjusted twice: if the partition revenue is less than the resource scheduling cost, the minimum compensation mechanism is triggered, and the virtual power plant scheduling platform makes up the difference; if the partition's task completion rate is ≥100% and there are no constraints or violations, the excess reward is triggered, and the virtual power plant scheduling platform adds an extra preset incentive revenue. The central dispatching agent generates a settlement statement for each partition's adjustable resource aggregation agent, which includes a task package identifier, task completion rate, revenue amount, and correction instructions. The settlement statement is automatically executed and the revenue is distributed using a blockchain smart contract mechanism. At the same time, the settlement statement is stored on the blockchain and the settlement results are synchronized to each partition's adjustable resource aggregation agent.

Images