Grid-connected ac-dc hybrid microgrid optimal operation control method and system

Abstract

The application provides a grid-connected AC-DC hybrid microgrid optimal operation control method and system, and belongs to the technical field of power system automation and energy management. The method comprises the following steps: constructing an AC microgrid agent, a DC microgrid agent and an AC-DC microgrid general agent; the AC and DC agents respectively perform internal autonomous optimization on the respective subnets, and report the subnet power balance state and internal marginal cost information to the general agent; the general agent generates a decision instruction for determining the exchange power with the external grid and the mutual aid power among the subnets according to the received AC and DC subnet power balance state and internal marginal cost information and the real-time acquired large grid price information, and the decision instruction is sent to each subnet agent to adjust the internal power generation unit output plan. The application can maximize the economic benefits of the microgrid by constructing three intelligent agents that work cooperatively and optimizing the interaction with the large grid.

Description

Technical Field

[0001] This application belongs to the field of power system automation and energy management technology, specifically relating to a method and system for optimizing the operation and control of a grid-connected AC / DC hybrid microgrid. Background Technology

[0002] With the continuous increase in the proportion of renewable energy power generation and the rapid development of power electronics technology, AC / DC hybrid microgrids have become a research hotspot due to their ability to efficiently accommodate DC sources such as photovoltaic power generation and energy storage systems, and to flexibly serve diverse AC / DC loads. Currently, energy management systems for AC / DC hybrid microgrids are mainly divided into two categories: centralized and distributed. Early research often used a centralized controller to perform unified optimization scheduling of the entire microgrid. This method requires collecting network-wide operational data, with the central processing unit performing global optimization calculations. However, the centralized architecture has the following inherent drawbacks: first, it puts heavy pressure on the communication network, requiring extremely high real-time performance in data acquisition and transmission; second, it has high computational complexity, making it difficult to adapt to the dynamic expansion of the microgrid scale; and third, it has the risk of single-point failure, as failure of the central controller will lead to the loss of control of the entire system, resulting in poor robustness and reliability.

[0003] In recent years, distributed control methods based on multi-agent systems have gradually attracted attention. These methods divide the microgrid into several autonomous regions, each optimized locally by an independent agent, and achieve global coordination through information exchange between agents. However, existing distributed methods are mostly designed for islanded operation modes, focusing on maintaining system survivability and minimizing costs. They cannot effectively utilize grid price signals and optimize interaction with the main grid in grid-connected modes, thus failing to maximize the overall economic benefits of the microgrid.

[0004] Therefore, designing an optimized operation and control method suitable for grid-connected AC / DC hybrid microgrids to fully utilize grid price signals to optimize interaction with the main grid and maximize the overall economic benefits of the microgrid has become an urgent technical problem to be solved. Summary of the Invention

[0005] The purpose of this application is to solve the problems existing in the prior art and to provide a method and system for optimizing the operation and control of grid-connected AC / DC hybrid microgrids.

[0006] This application is achieved through the following technical solution:

[0007] The first aspect of this application provides a method for optimized operation control of a grid-connected AC / DC hybrid microgrid based on multi-agent cooperation, comprising: An AC microgrid agent, a DC microgrid agent, and a general agent for both AC and DC microgrids are constructed. The AC microgrid agent is used to perform internal autonomous optimization of the AC subgrid and reports the optimized AC subgrid power balance status and internal marginal cost information to the general agent. The DC microgrid agent is used to perform internal autonomous optimization of the DC subgrid and reports the optimized DC subgrid power balance status and internal marginal cost information to the general agent. The general agent, based on the received AC subgrid power balance status and internal marginal cost information, DC subgrid power balance status and internal marginal cost information, and real-time large grid electricity price information, generates decision instructions to determine the exchange power with the external grid and the mutual assistance power between each subgrid, with the goal of maximizing the economic profit of the microgrid as a whole in the transaction with the large grid. The general agent will issue the decision instructions to the AC microgrid agent and the DC microgrid agent; The AC microgrid agent and the DC microgrid agent adjust the output plan of their respective internal power generation units according to the decision command.

[0008] Optionally, the method by which the general agent generates the decision instruction includes: When any subgrid has a power deficit and its internal marginal cost is higher than the purchase price of electricity from the main grid, the decision is made to purchase electricity from the main grid to make up for the power deficit. When any subgrid has a power surplus and its internal marginal cost is lower than the electricity sales price of the main grid, the decision is made to sell the surplus power to the main grid. When the AC subgrid and the DC subgrid have power deficit and power surplus respectively, the internal mutual assistance cost is compared with the power price of the main grid to decide whether to carry out power mutual assistance between subgrids or to trade with the main grid respectively.

[0009] Optionally, the internal autonomous optimization of the AC microgrid agent aims to minimize the operating cost of the AC subgrid and satisfies AC power flow constraints; the internal autonomous optimization of the DC microgrid agent aims to minimize the operating cost of the DC subgrid and satisfies DC power flow constraints; the AC power flow constraints and the DC power flow constraints include node voltage constraints, line capacity constraints, power output constraints, energy storage operation constraints, and power balance constraints.

[0010] Optionally, the AC subgrid operating cost includes at least one of the following: diesel generator power generation cost, energy storage charging and discharging loss cost, and AC grid loss cost; the DC subgrid operating cost includes at least one of the following: renewable energy power generation conversion cost, energy storage cost, and DC / DC conversion loss cost.

[0011] Optionally, the AC microgrid agent and the DC microgrid agent use at least one of the improved particle swarm optimization algorithm, linear programming and mixed integer programming methods to solve the internal optimization problem; the general agent uses linear programming or decision tree methods to generate the decision instructions.

[0012] A second aspect of this application provides an optimized operation control system for a grid-connected AC / DC hybrid microgrid, used to implement the method described in any one of the first aspects above, comprising: One or more processors, individually or collectively configured to: An AC microgrid proxy module is implemented, which is used to perform internal autonomous optimization of the AC subgrid and report the power balance status and internal marginal cost information of the AC subgrid. A DC microgrid proxy module is implemented, which is used to perform internal autonomous optimization of the DC subgrid and report the power balance status and internal marginal cost information of the DC subgrid. An AC / DC microgrid general agent module is implemented. The AC / DC microgrid general agent module is used to generate decision instructions to determine the exchange power with the external grid and the mutual assistance power between each subgrid based on the received AC subgrid power balance status and internal marginal cost information, DC subgrid power balance status and internal marginal cost information, and real-time acquired grid electricity price information, with the goal of maximizing the economic profit of the microgrid as a whole in the transaction with the grid. The decision instructions are then sent to the AC microgrid agent module and the DC microgrid agent module. The AC microgrid agent and the DC microgrid agent are also used to adjust the output plan of their respective internal power generation units according to the decision instructions; And a communication interface, used to enable information exchange between various agent modules.

[0013] Optionally, it also includes an electricity price information collection module that is communicatively connected to the AC / DC microgrid general agent module; The electricity price information acquisition module is used to acquire electricity price information of the large power grid; the electricity price information includes time-of-use electricity price or real-time electricity price data.

[0014] Optionally, the communication interface supports at least one of fieldbus communication, industrial Ethernet communication, or wireless communication to enable data interaction between the one or more processors.

[0015] Optionally, the AC microgrid agent module and the DC microgrid agent module are deployed in the local controllers of the AC subgrid and the DC subgrid, respectively, and the AC / DC microgrid master agent module is deployed in the microgrid central controller; the local controller and the microgrid central controller are connected through the communication interface.

[0016] Compared with the prior art, the beneficial effects of this application are: 1. Innovative architecture for layered collaboration: By constructing a two-layer architecture of AC agent, DC agent and general agent, the complex global optimization problem is decomposed into two levels: autonomous optimization within the subnet and global transaction decision-making. The responsibilities of each agent are clearly defined, which not only ensures the independence of subnet operation, but also achieves global economic coordination, significantly improving the system's scalability and robustness.

[0017] 2. Optimization of dual economic objectives to improve operational profitability: The subgrid agent aims to minimize internal operating costs to ensure technical and economic efficiency; the general agent aims to maximize profits from transactions with the main grid, using internal marginal costs and external electricity price signals for market-based decision-making, achieving dual optimization of minimizing internal costs and maximizing external profits, effectively improving the overall economic benefits of the microgrid.

[0018] 3. Efficient information exchange and sufficient decision-making basis: The subnet agent only reports key economic signals such as power balance status and marginal cost to the general agent, rather than all operating data, which greatly reduces communication pressure; the general agent makes the optimal trading decision based on these economic signals and real-time electricity prices, realizing the deep integration of technology and market, and the decision-making process is fast and accurate.

[0019] 4. Optimized Problem Decomposition for Enhanced Computational Efficiency and System Scalability: The complex global optimization problem of AC / DC hybrid microgrids is decomposed into an internal security and economic scheduling problem solved by subgrid agents and an external market transaction optimization problem solved by a master agent. This decomposition significantly reduces the dimensionality of a single optimization problem, increasing computational speed. Simultaneously, each subgrid agent can solve the problem independently in parallel, further improving overall optimization efficiency. Furthermore, the modular solution architecture allows for the addition of new subgrids or devices by simply connecting the corresponding agent, without requiring reconstruction of the global optimization model, thus greatly enhancing system scalability. Attached Figure Description

[0020] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The accompanying drawings are used to provide a further understanding of the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof.

[0021] Figure 1 A flowchart illustrating an optimized operation control method for a grid-connected AC / DC hybrid microgrid based on multi-agent collaboration, provided for an embodiment of this application; Figure 2 This is a schematic diagram of the logical relationship between intelligent agents provided in the embodiments of this application; Figure 3A schematic diagram illustrating the workflow of an AC / DC microgrid agent provided in an embodiment of this application; Figure 4 This is a schematic diagram of the workflow of the AC / DC microgrid general agent provided in the embodiments of this application. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this application more apparent, exemplary embodiments according to this application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein. Based on the embodiments of this application described herein, all other embodiments obtained by those skilled in the art without inventive effort should fall within the protection scope of this application.

[0023] To address the problem that existing methods cannot effectively utilize grid price signals and optimize interaction with the main grid, thus failing to maximize the overall economic benefits of microgrids, this application proposes an optimized operation control method and system for grid-connected AC / DC hybrid microgrids. This method can fully utilize grid price signals to optimize interaction with the main grid and maximize the overall economic benefits of the microgrid.

[0024] The present application will now be described in further detail with reference to the accompanying drawings.

[0025] Figure 1 A flowchart illustrating an optimized operation control method for a grid-connected AC / DC hybrid microgrid based on multi-agent cooperation, provided in this application embodiment, is shown below. The following first refers to... Figure 1 This application describes an optimized operation control method for a grid-connected AC / DC hybrid microgrid based on multi-agent collaboration, provided by an embodiment of this application.

[0026] like Figure 1 As shown, the method for optimized operation control of grid-connected AC / DC hybrid microgrid based on multi-agent cooperation in this application includes at least the following steps S100 to S400.

[0027] Step S100: Construct an AC microgrid agent, a DC microgrid agent, and a general agent for AC and DC microgrids. The AC microgrid agent is used to perform internal autonomous optimization of the AC subgrid and report the optimized AC subgrid power balance status and internal marginal cost information to the general agent. The DC microgrid agent is used to perform internal autonomous optimization of the DC subgrid and report the optimized DC subgrid power balance status and internal marginal cost information to the general agent.

[0028] In this application, the AC microgrid agent (AC-Agent) refers to the control entity deployed on the AC subgrid side. It can be a software module or a hardware controller, responsible for the internal operation optimization of the AC subgrid, including distributed power dispatch, energy storage charging and discharging management, load forecasting and balancing, etc. It possesses independent decision-making capabilities and reports the AC subgrid power balance status and internal marginal cost information to the central agent. The DC microgrid agent (DC-Agent) refers to the control entity deployed on the DC subgrid side, with similar functions to the AC agent. It is responsible for the internal optimization operation of the DC subgrid, satisfying DC power flow constraints, and reporting the DC subgrid power balance status and internal marginal cost. The AC / DC microgrid central agent (Hybrid-Agent) refers to the coordinating control entity located at the center of the microgrid. It is responsible for receiving the status information reported by each subgrid agent, and, in conjunction with the main grid electricity price, making power purchase and sale decisions with the external grid and power mutual assistance decisions between subgrids with the goal of maximizing overall economic benefits, and issuing instructions to the subgrid agents.

[0029] In this application, internal autonomous optimization refers to each subnet agent independently solving the optimal power flow or economic dispatch problem within its subnet without relying on external information, minimizing operating costs while satisfying physical constraints. Internal marginal cost refers to the additional operating cost required to increase unit load demand when the subnet's internal optimization reaches its optimal operating point. It reflects the scarcity of power generation resources within the subnet and is a key economic signal for the general agent's market transaction decisions. Power balance status refers to the difference between the total power generation and total load power within the subnet after internal optimization. If power generation exceeds load, there is a power surplus; if power generation is less than load, there is a power deficit. This status information includes whether the subnet is balanced and the quantitative values ​​for imbalance.

[0030] Step S200: Based on the received AC subgrid power balance status and internal marginal cost information, DC subgrid power balance status and internal marginal cost information, and real-time large grid electricity price information, the general agent generates decision instructions to determine the exchange power with the external grid and the mutual assistance power between each subgrid, with the goal of maximizing the economic profit of the microgrid as a whole in the transaction with the large grid.

[0031] Here, the power grid price information refers to the time-of-use or real-time electricity price data released by the external power grid, which the general agent uses as a reference for power purchase and sale decisions. Decision instructions refer to the control commands derived by the general agent based on optimized calculations, including the exchange power setpoint (power purchased or sold) with the external power grid, the transmission power instructions of the bidirectional converters between sub-grids, etc. The sub-grid agent adjusts the output plans of each internal power generation unit accordingly to achieve globally optimal operation.

[0032] Step S300: The general agent sends the decision instruction to the AC microgrid agent and the DC microgrid agent.

[0033] In this step, after the AC / DC microgrid general agent completes the global transaction optimization decision, it generates specific control commands and sends the commands to the AC microgrid agents and DC microgrid agents through the communication network.

[0034] The decision-making instructions issued by the general agent shall include at least the following two types of control information: Power exchange instructions with the external power grid: including power purchase setting value or power sales setting value, in kilowatts (kW) or megawatts (MW), with positive and negative signs indicating the power direction (e.g., positive value means purchasing power from the grid, negative value means selling power to the grid).

[0035] Inter-subnet power transfer instruction: When the general agent decides to perform power transfer between the AC subnet and the DC subnet, the instruction also includes the transmission power setting value of the bidirectional converter, specifying the power flow direction (such as from the DC subnet to the AC subnet) and the amount of transmission power.

[0036] In step S400, the AC microgrid agent and the DC microgrid agent adjust the output plan of their respective internal power generation units according to the decision instructions.

[0037] In this step, the AC microgrid agent and the DC microgrid agent respectively re-optimize and schedule their internal operations based on the received decision instructions, adjusting the output plans of each power generation unit so that the overall operating state of the microgrid approaches the global optimal target determined by the general agent.

[0038] Each subnet agent will break down the received master agent instructions into output adjustment amounts for each controllable unit within its network: AC microgrid agent: Based on the "power exchange command with the external grid" and possible "inter-subgrid mutual assistance power command", combined with the current load forecast, renewable energy generation forecast, and energy storage state of charge (SOC) information within the AC subgrid, re-solve the optimal scheduling model of the AC subgrid to determine the output plan of at least one of the following units: Active power output of diesel generators or micro gas turbines; The charging and discharging power and charging and discharging status of the energy storage system; Output limits for renewable energy generation units (such as wind power) (when wind curtailment is required); The amount of load shedding or adjustment of interruptible or adjustable loads.

[0039] DC microgrid agent: Similarly, based on the received instructions, the optimal scheduling model of the DC subgrid is resolved to determine the output plan of at least one of the following units: Maximum power point tracking control or power limiting control for photovoltaic power generation units; The charging and discharging power of the energy storage system; The transmission power of the DC / DC converter; The adjustment amount of DC load.

[0040] When adjusting the output plan, each subnet agent must re-satisfy all internal physical constraints, including but not limited to: Power balance constraints: The adjusted power generation, load power, power exchanged with the external power grid, and power exchanged between subgrids must satisfy the nodal power balance equations.

[0041] Equipment output constraints: The output of each power generation unit and energy storage unit must be within the minimum / maximum limits allowed by technology.

[0042] Voltage constraints: After adjustment, the voltage of each node must be within the allowable fluctuation range (the AC subnet must meet the voltage amplitude constraint, and the DC subnet must meet the voltage deviation constraint).

[0043] Line capacity constraints: The transmission power of each feeder or tie line must not exceed the thermal stability limit.

[0044] Energy storage constraints: The charging and discharging power, SOC upper and lower limits, and charge / discharge conversion times of the energy storage system must meet the operational requirements.

[0045] After the optimization calculation is completed, each subnet agent converts the new output plan into specific control commands and sends them to terminal execution devices such as power generation units, energy storage converters, and load switches via the local communication bus. For generators, issue active power setpoints or speed regulation commands; For energy storage converters, charge and discharge power commands and operating modes (charging / discharging / standby) are issued. For the converter, a transmission power command is issued; For adjustable loads, issue power adjustment or on / off commands.

[0046] After executing the instructions, the terminal device feeds back the actual operating status to the subnet agent, forming a closed-loop control.

[0047] This application employs a rolling optimization mechanism: within each control cycle (e.g., 15 minutes), steps S300 and S400 are executed once; at the start of the next cycle, the subgrid agent re-reports the latest status, and the main agent optimizes the decision again based on the new electricity price and status information, forming a dynamic, real-time updated optimization control closed loop. This mechanism can effectively cope with the volatility of renewable energy and the uncertainty of load, ensuring that the microgrid always operates in a globally economically optimal state.

[0048] It should be noted that the AC microgrid agent, DC microgrid agent, and AC / DC microgrid master agent described in this application are all specific implementations of intelligent agents. Each agent possesses the core characteristics of an intelligent agent, such as autonomy, responsiveness, initiative, and sociality, and achieves optimized operation of the microgrid through mutual cooperation. Those skilled in the art will understand that the agents described in this application are substantially equivalent to intelligent agents.

[0049] This application provides an optimized operation control method for grid-connected AC / DC hybrid microgrids based on multi-agent collaboration. By constructing three collaborative agents, it achieves an organic combination of distributed autonomous optimization and global economic coordination, optimizes the interaction behavior with the large power grid, and enables the grid-connected microgrid to actively participate in the power grid market while meeting internal safety operation constraints, thereby maximizing overall operating benefits.

[0050] The method of this application will be specifically described below with reference to specific embodiments.

[0051] Example 1 This embodiment provides an optimized operation control method for grid-connected AC / DC hybrid microgrids based on multi-agent cooperation. The method includes the following steps: Step S110: Construct three collaborative intelligent agents: AC microgrid agent, DC microgrid agent, and AC / DC microgrid master agent.

[0052] like Figure 2 As shown, the AC microgrid agent, DC microgrid agent, and AC / DC microgrid master agent form a two-layer architecture at the control level. The AC and DC microgrid agents belong to the lower execution layer and are at the same level, responsible for the internal autonomous optimization of the AC and DC microgrids respectively. The AC and DC microgrids can be connected via an AC / DC bidirectional converter. The AC / DC microgrid master agent is located at the upper coordination layer and is responsible for global economic coordination of the two peer-level microgrid agents. The three agents achieve collaborative operation through information exchange.

[0053] Step S120: The AC microgrid agent performs internal autonomous optimization of the AC subgrid, and the DC microgrid agent performs internal autonomous optimization of the DC subgrid, and reports the corresponding status information respectively.

[0054] The AC microgrid agent is responsible for the autonomous optimization of the AC subgrid's internal operation, aiming to minimize the subgrid's internal operating costs. These costs include, but are not limited to, the generation costs of distributed power sources (such as diesel generators), energy storage charging and discharging losses, and network losses. The solution must strictly satisfy the physical constraints of AC power flow, including node voltage upper and lower limits, line transmission capacity, generator output upper and lower limits, energy storage operation constraints, and power balance (KCL / KVL).

[0055] The DC microgrid agent is responsible for the autonomous optimization of the DC subgrid's internal operation, aiming to minimize the subgrid's internal operating costs. These costs include, but are not limited to, the converted costs of renewable energy sources such as photovoltaics and wind power, energy storage costs, and DC / DC conversion losses. The solution must satisfy the physical constraints of DC power flow, including voltage stability range, line capacity, power output limits, energy storage constraints, and power balance.

[0056] After completing internal optimization, the AC / DC microgrid agent does not directly report all optimization results to the main agent. Instead, it only reports two types of key status information to reduce communication burden and protect subgrid privacy: 1. Power balance state The AC / DC microgrid agent first determines whether power balance has been achieved after internal optimization. The criterion is: under the premise of satisfying all constraints, does there exist a feasible solution that makes the generated power equal to the load power?

[0057] If a feasible solution exists and the generating power equals the load power, then report the "power balance" flag.

[0058] If no feasible solution exists, or if a feasible solution exists but requires external power support to satisfy the constraints, then report the "power imbalance" flag and calculate the imbalance amount.

[0059] 2. Power imbalance and its marginal cost When power balance cannot be achieved within the AC / DC subgrid, the AC / DC microgrid performs proxy calculations: Power deficit: This indicates the amount of power that the AC / DC subgrid needs to inject from external sources to meet internal load demands and operational constraints. The deficit is equal to the minimum active power that needs to be injected externally for the system to be feasible.

[0060] Power surplus: This indicates that the internal generation capacity of the AC / DC subgrid exceeds the load demand, resulting in surplus power that can be exported. The surplus is equal to the maximum active power that can be exported while satisfying all constraints.

[0061] Meanwhile, the internal marginal cost of the AC / DC microgrid is calculated by proxy and defined as the additional operating cost required to increase the unit load demand at the optimal operating point. Mathematically, the marginal cost is equal to the Lagrange multiplier (dual variable) corresponding to the power balance constraint in the optimization problem.

[0062] The following explanation uses an AC microgrid as an example.

[0063] like Figure 3 As shown, the specific steps for dynamic balance management of load and renewable energy generation in an AC microgrid within the next 15 minutes are as follows: S1, Basic Data Prediction First, two core forecasting operations are performed: one is to complete the load forecast for the next 15 minutes, obtaining the power load demand data that will be generated in the microgrid; the other is to complete the forecast of renewable energy generation for the next 15 minutes, clarifying the output power data of renewable energy generation during this period.

[0064] S2, Balanced Optimization Calculation Based on the two forecast data mentioned above, the load balance calculation of the AC microgrid in the next 15 minutes is carried out, and the optimal matching relationship between load and power generation is solved by using an optimization calculation method.

[0065] AC microgrid agents can employ various optimization algorithms, including but not limited to: Linear programming: Linearizes the nonlinear AC power flow equations (e.g., by using DC power flow approximation), and is suitable for radial distribution networks or scenarios where high accuracy is not required.

[0066] Mixed-integer linear programming: Introducing integer variables to handle discrete decisions such as the charging and discharging state of energy storage and the switching of interruptible loads.

[0067] Improved Particle Swarm Optimization Algorithm: Applicable to non-convex and nonlinear problems, it searches for the global optimum through swarm intelligence.

[0068] In practical engineering, the above algorithms can be flexibly selected according to the size of the AC subnet and the requirements for solution speed.

[0069] S3, Load Balance Determination Determining whether the load is unbalanced based on optimized calculations: If the determination result is yes, then the balance status information will be directly reported to the general agent; If the result is negative, the process proceeds to the next level of power generation adequacy assessment.

[0070] S4. Determination of Power Generation Adequacy In the case of load imbalance, determine whether the power generation is insufficient: If the determination result is yes, then the additional power generation value that needs to be supplemented shall be reported to the general agent; If the determination result is negative, the current power generation data that the microgrid can provide will be reported to the general agent.

[0071] The operation process in a DC microgrid is the same as that in an AC microgrid, and will not be explained further here.

[0072] Step S130: The AC / DC microgrid general agent makes a globally optimal transaction decision based on the grid price and the received status information, and issues an instruction.

[0073] The AC / DC microgrid general agent (hereinafter referred to as the general agent) is the upper-level coordination layer intelligent agent in the two-layer intelligent agent architecture of this application. It can be deployed in the microgrid central controller or cloud coordination platform, serving as the unified coordinator and decision-maker for the microgrid's external interactions. The core responsibility of the general agent is to optimize the interaction between the microgrid as a whole and the external power grid, and to maximize the overall economic benefits of the microgrid through market-based decision-making while ensuring internal safe operation.

[0074] like Figure 4 As shown, firstly, the general agent receives the status information of the DC microgrid, the status information of the AC microgrid, and the power purchase and sale information of the power grid. Then, based on all the collected information, and according to the preset objective function, it calculates and determines the power dispatch commands for the "tether line" and the "AC / DC bidirectional converter" to achieve economic operation and power balance of the microgrid.

[0075] Here, the general agent does not directly participate in the detailed internal scheduling of each subgrid. Instead, it receives power surplus / deficit information and their respective marginal costs reported by AC and DC agents, while also obtaining electricity price information from the main grid, such as time-of-use pricing and real-time pricing. It then makes globally optimal trading decisions and issues these decisions to the subgrid agents for execution. This design maintains the autonomy of each subgrid while achieving overall economic coordination.

[0076] The core optimization objective of the general agent is to maximize the overall economic profit of the microgrid in its transactions with the main grid within an optimization cycle. Profit is calculated as: Profit = Electricity Sales Revenue - Electricity Purchase Cost ± Internal Mutual Aid Cost (Internal Transfer Pricing).

[0077] The general agent's decision-making logic is based on the principle of marginal cost comparison: comparing the internal marginal cost of each sub-network with the external grid electricity price, and selecting the power source with the lowest cost or the power destination with the highest revenue. The specific decision-making rules are as follows: When a subgrid experiences a power shortage and its internal marginal cost exceeds the grid price, the general agent decides to purchase electricity from the main grid to meet the shortage, which is more economical.

[0078] When a subgrid has a surplus and its internal marginal cost is lower than the grid electricity price, the general agent can make a profit by selling the surplus electricity to the main grid.

[0079] When one AC subgrid is short of power and the other DC subgrid is in surplus, the general agent will compare the internal power transfer costs (with grid electricity prices) to decide whether to transfer power through internal bidirectional converters or to trade with the grid independently.

[0080] After the general agent completes the optimization decision, it generates two types of control commands and issues them to each sub-network agent: power exchange command with the main power grid and power mutual assistance command between sub-networks.

[0081] Step S140: The AC / DC microgrid agent adjusts its internal power generation plan according to the new instructions, ultimately achieving the economic optimization of the overall operation of the microgrid.

[0082] In actual operation, within an operating cycle (e.g., 15 minutes), the three agents execute in parallel or iteratively. The AC / DC-Agent first performs internal optimization, reporting the supply and demand status and marginal cost to the Hybrid-Agent. The Hybrid-Agent makes a globally optimal trading decision based on grid prices and internal information, and issues instructions. The AC / DC-Agent adjusts its internal generation plan according to the new instructions, ultimately achieving the economic optimization of the overall microgrid operation.

[0083] This embodiment proposes a two-layer agent architecture (AC-Agent, DC-Agent, Hybrid-Agent) for grid-connected operation, clearly defining the new division of responsibilities among agents in the grid-connected mode. The objective function of the subgrid agents (AC-Agent, DC-Agent) focuses on minimizing internal operating costs to achieve technical optimization. The objective function of the general agent (Hybrid-Agent) focuses on maximizing electricity purchase and sale profits to achieve economic optimization. This two-layer objective design of "minimizing internal costs + maximizing external profits" perfectly aligns with the commercial positioning of grid-connected microgrids as "prosumers," representing a core innovation that distinguishes it from the islanded mode (which aims at survival and cost minimization). Furthermore, the communication between the subgrid agents and the general agent is no longer the "load transfer request" or "load reduction instruction" of the islanded mode, but rather standardized power imbalance (surplus / deficit) and its marginal cost. Based on these economic signals and external electricity prices, the general agent makes market-based trading decisions, achieving efficient integration of technical operation and market transactions.

[0084] Furthermore, the complex global optimization problem of AC / DC hybrid microgrids is decomposed into an internal safe and economical dispatch problem (solved by an AC / DC-Agent) and an external market transaction optimization problem (solved by a Hybrid-Agent). This decomposition reduces the dimensionality of the problem solution, improves computational speed and system modularity, and facilitates expansion.

[0085] This application also provides a multi-agent cooperative optimized operation control system for grid-connected AC / DC hybrid microgrids. The following detailed description, in conjunction with specific embodiments, illustrates this multi-agent cooperative optimized operation control system for grid-connected AC / DC hybrid microgrids, which is used to implement the multi-agent cooperative optimized operation control method for grid-connected AC / DC hybrid microgrids of this application.

[0086] Example 2 The optimized operation control system for grid-connected AC / DC hybrid microgrids based on multi-agent collaboration provided in this embodiment is fundamentally based on a multi-agent collaborative architecture to achieve hierarchical autonomy and global optimization of the AC / DC hybrid microgrid. Its specific structure includes one or more processors and communication interfaces, and the functions and collaborative relationships of each component are as follows: One or more processors can be configured individually or collaboratively. Their core function is to implement the various intelligent agent agent modules of the system. The specific configuration is as follows: (1) AC Microgrid Agent Module: Implemented by the processor, this module is specifically designed for internal autonomous optimization of the AC subgrid in a hybrid AC / DC microgrid. Specifically, this module collects real-time operating data of the power generation units (such as AC wind turbines, gas turbines, etc.) and load units within the AC subgrid. Based on a preset autonomous optimization strategy, it autonomously adjusts the power generation output and load distribution within the AC subgrid to achieve power balance within the AC subgrid. At the same time, this module will collect and report the power balance status of the AC subgrid (including three states: balance, surplus, and shortage) and internal marginal cost information (i.e., the cost required to generate an additional unit of electricity within the AC subgrid) in real time, providing basic data support for global optimization decisions.

[0087] (2) DC Microgrid Agent Module: Also implemented by a processor, its function corresponds to that of the AC microgrid agent module. It is specifically used for internal autonomous optimization of the DC subgrid in the AC / DC hybrid microgrid. This module collects the operating parameters of the power generation units (such as photovoltaic modules, energy storage batteries, etc.) and load units in the DC subgrid in real time. Through the internal autonomous optimization algorithm, it adjusts the output of each power generation unit in the DC subgrid to ensure the power balance within the DC subgrid. At the same time, it reports the power balance status and internal marginal cost information of the DC subgrid and, in collaboration with the AC microgrid agent module, provides data to the global decision-making module.

[0088] (3) AC / DC Microgrid General Agent Module: Implemented by the processor, this module serves as the global decision-making core of the entire system. It is responsible for receiving the AC subgrid power balance status and internal marginal cost information reported by the AC microgrid agent module, the DC subgrid power balance status and internal marginal cost information reported by the DC microgrid agent module, and the real-time electricity price information of the main grid. This module aims to maximize the economic profit of the microgrid's overall transactions with the main grid. Through a preset optimization decision algorithm, it calculates the exchange power between the microgrid and the external main grid (i.e., the power sold by the microgrid to or purchased from the main grid) and the mutual assistance power between the AC and DC subgrids (i.e., the power allocation between the two when one subgrid has a power surplus and the other has a power shortage), generates corresponding decision instructions, and sends these instructions to the AC and DC microgrid agent modules.

[0089] Furthermore, after receiving decision instructions from the AC / DC microgrid general agent module, the AC microgrid agent module and the DC microgrid agent module will adjust the output plan of their respective internal power generation units according to the instructions. For example, when the decision instruction requires the AC subgrid to increase its output to meet the power shortage of the DC subgrid, the AC microgrid agent module will adjust the output of its internal AC power generation units to ensure that the instruction is implemented. When the decision instruction requires the microgrid to sell electricity to the main grid, the two agent modules will coordinate to adjust the output of their respective internal power generation units to ensure that the electricity sold meets the target.

[0090] The communication interface is the core component for enabling information exchange between various agent modules. It establishes communication links between the AC microgrid agent modules, the DC microgrid agent modules, and the AC / DC microgrid master agent module, ensuring real-time and accurate information transmission between modules. Specifically, the communication interface transmits the power balance status and internal marginal cost information of the AC and DC subgrids, as well as decision-making instructions issued by the AC / DC microgrid master agent module, providing communication support for multi-agent collaborative optimization.

[0091] Example 3 This embodiment further optimizes the system structure based on embodiment two by adding an electricity price information collection module. The specific improvements are as follows: The system also includes a price information acquisition module that is communicatively connected to the AC / DC microgrid general agent module. The core function of this module is to acquire the electricity price information of the main power grid in real time, providing price data support for the overall optimization decision-making of the AC / DC microgrid general agent module. The electricity price information includes, but is not limited to, time-of-use price data (i.e., different electricity prices set by the main power grid at different times, such as peak-hour price, normal-hour price, and off-peak price) or real-time price data (i.e., the electricity price dynamically adjusted by the main power grid according to the real-time power supply load).

[0092] The electricity price information acquisition module transmits the acquired electricity price information to the AC / DC microgrid general agent module in real time. The general agent module combines this electricity price information with the power balance status and internal marginal cost information of each subgrid to optimize the calculation of the exchange power with the main grid. For example, when the main grid electricity price is in a valley (lower price), if the microgrid has a power shortage, it can prioritize purchasing electricity from the main grid; when the main grid electricity price is in a peak (higher price), if the microgrid has a power surplus, it can prioritize selling electricity to the main grid, thereby maximizing the economic profit of the microgrid's transaction with the main grid.

[0093] Example 4 This embodiment, based on Embodiment 2, further defines the communication method of the communication interface, as follows: The communication interface supports at least one of fieldbus communication, industrial Ethernet communication, or wireless communication to enable data interaction between the one or more processors. Fieldbus communication can use bus protocols such as RS485 or CAN, suitable for short-distance communication between controllers within a microgrid, offering advantages such as strong anti-interference capabilities and stable transmission. Industrial Ethernet communication can use the Ethernet protocol, suitable for long-distance, high-volume data transmission, ensuring efficient information interaction between agent modules. Wireless communication can use wireless communication methods such as WiFi, LoRa, or 5G, suitable for scenarios where laying wired communication lines is inconvenient, improving the flexibility of system deployment.

[0094] By flexibly selecting from the above-mentioned multiple communication methods, it can be ensured that the communication interface can adapt to different microgrid deployment scenarios, realize real-time and reliable data transmission between processors (i.e., between agent modules), and ensure the smooth realization of multi-agent collaborative optimization.

[0095] Example 5 This embodiment, based on embodiment two, further defines the deployment location of each agent module and clarifies the hardware deployment architecture of the system, as follows: The AC microgrid agent module is deployed in the local controller of the AC subgrid, the DC microgrid agent module is deployed in the local controller of the DC subgrid, and the AC / DC microgrid general agent module is deployed in the microgrid central controller; wherein, the local controllers of the AC subgrid, the local controllers of the DC subgrid, and the microgrid central controller establish a communication connection through the communication interface, forming a hierarchical control architecture of "central-local".

[0096] The advantage of this deployment method is that the AC microgrid agent module and the DC microgrid agent module are deployed near the local controller of their respective subgrids, which can quickly collect the operating data of the local subgrids and realize rapid response of internal autonomous optimization; the AC and DC microgrid general agent module is deployed on the central controller, which can globally coordinate the operating status of the two subgrids, generate global decision commands in combination with the power grid price information, and send them to each local controller through the communication interface, realizing the synergy of hierarchical autonomy and global optimization, and improving the system's operating efficiency and control accuracy.

[0097] In summary, the multi-agent collaborative grid-connected AC / DC hybrid microgrid optimized operation control system of this invention realizes the functions of each agent module through a processor and realizes information interaction between modules through a communication interface. It can add a price information collection module, optimize the communication method and the deployment location of the modules according to actual needs, effectively realize the control method proposed in this application, ensure the optimized operation of the AC / DC hybrid microgrid, and improve the economy and stability of the system.

[0098] Finally, it should be noted that the above technical solution is only one implementation method of this application. For those skilled in the art, based on the application methods and principles disclosed in this application, it is easy to make various types of improvements or modifications, and not limited to the methods described in the specific implementation methods above. Therefore, the methods described above are only preferred and have no limiting significance.

Claims

1. A method for optimized operation control of grid-connected AC / DC hybrid microgrids based on multi-agent cooperation, characterized in that, include: Establish AC microgrid agents, DC microgrid agents, and general agents for AC and DC microgrids; The AC microgrid agent is used to perform internal autonomous optimization of the AC subgrid and report the optimized AC subgrid power balance status and internal marginal cost information to the main agent; the DC microgrid agent is used to perform internal autonomous optimization of the DC subgrid and report the optimized DC subgrid power balance status and internal marginal cost information to the main agent. The general agent, based on the received AC subgrid power balance status and internal marginal cost information, DC subgrid power balance status and internal marginal cost information, and real-time large grid electricity price information, generates decision instructions to determine the exchange power with the external grid and the mutual assistance power between each subgrid, with the goal of maximizing the economic profit of the microgrid as a whole in the transaction with the large grid. The general agent will issue the decision instructions to the AC microgrid agent and the DC microgrid agent; The AC microgrid agent and the DC microgrid agent adjust the output plan of their respective internal power generation units according to the decision command.

2. The method according to claim 1, characterized in that, The method by which the general agent generates the decision instruction includes: When any subgrid has a power deficit and its internal marginal cost is higher than the purchase price of electricity from the main grid, the decision is made to purchase electricity from the main grid to make up for the power deficit. When any subgrid has a power surplus and its internal marginal cost is lower than the electricity sales price of the main grid, the decision is made to sell the surplus power to the main grid. When the AC subgrid and the DC subgrid have power deficit and power surplus respectively, the internal mutual assistance cost is compared with the power price of the main grid to decide whether to carry out power mutual assistance between subgrids or to trade with the main grid respectively.

3. The method according to claim 1 or 2, characterized in that, The internal autonomous optimization of the AC microgrid agent aims to minimize the operating cost of the AC subgrid and satisfies AC power flow constraints; the internal autonomous optimization of the DC microgrid agent aims to minimize the operating cost of the DC subgrid and satisfies DC power flow constraints; the AC power flow constraints and the DC power flow constraints include node voltage constraints, line capacity constraints, power output constraints, energy storage operation constraints, and power balance constraints.

4. The method according to claim 3, characterized in that, The operating cost of the AC subgrid includes at least one of the following: diesel generator power generation cost, energy storage charging and discharging loss cost, and AC grid loss cost; the operating cost of the DC subgrid includes at least one of the following: renewable energy power generation conversion cost, energy storage cost, and DC / DC conversion loss cost.

5. The method according to claim 1, characterized in that, The AC microgrid agent and the DC microgrid agent solve the internal optimization problem using at least one of the improved particle swarm optimization algorithm, linear programming, and mixed integer programming methods; the general agent generates the decision instructions using linear programming or decision tree methods.

6. A multi-agent collaborative optimized operation control system for grid-connected AC / DC hybrid microgrids, characterized in that, To implement the method according to any one of claims 1 to 5, comprising: One or more processors, individually or collectively configured to: An AC microgrid proxy module is implemented, which is used to perform internal autonomous optimization of the AC subgrid and report the power balance status and internal marginal cost information of the AC subgrid. A DC microgrid proxy module is implemented, which is used to perform internal autonomous optimization of the DC subgrid and report the power balance status and internal marginal cost information of the DC subgrid. An AC / DC microgrid general agent module is implemented. The AC / DC microgrid general agent module is used to generate decision instructions to determine the exchange power with the external grid and the mutual assistance power between each subgrid based on the received AC subgrid power balance status and internal marginal cost information, DC subgrid power balance status and internal marginal cost information, and real-time acquired grid electricity price information, with the goal of maximizing the economic profit of the microgrid as a whole in the transaction with the grid. The decision instructions are then sent to the AC microgrid agent module and the DC microgrid agent module. The AC microgrid agent and the DC microgrid agent are also used to adjust the output plan of their respective internal power generation units according to the decision instructions; And a communication interface, used to enable information exchange between various agent modules.

7. The system according to claim 6, characterized in that, It also includes an electricity price information collection module that is communicatively connected to the AC / DC microgrid general agent module; The electricity price information acquisition module is used to acquire electricity price information of the large power grid; the electricity price information includes time-of-use electricity price or real-time electricity price data.

8. The system according to claim 6, characterized in that, The communication interface supports at least one of fieldbus communication, industrial Ethernet communication, or wireless communication, and is used to realize data interaction between the one or more processors.

9. The system according to claim 6, characterized in that, The AC microgrid agent module and the DC microgrid agent module are respectively deployed in the local controllers of the AC subgrid and the DC subgrid, and the AC / DC microgrid master agent module is deployed in the microgrid central controller; the local controller and the microgrid central controller are connected through the communication interface.

Images