CN-121981385-A - Multi-agent-based AC/DC hybrid power distribution network dynamic bearing capacity assessment method

CN121981385ACN 121981385 ACN121981385 ACN 121981385ACN-121981385-A

Abstract

The invention relates to a multi-agent-based alternating-current/direct-current hybrid power distribution network dynamic bearing capacity assessment method which comprises the following steps of S1, obtaining and preprocessing multi-source heterogeneous data of an alternating-current/direct-current hybrid power distribution network to obtain a standardized data packet, S2, building a unified model of the alternating-current/direct-current hybrid power distribution network after the standardized data packet is obtained, carrying out topology analysis and structural feature extraction to obtain partitions and boundaries, S3, building a layered multi-agent architecture based on the partitions and the boundaries obtained in S2, obtaining mapping relations between agent catalogues and responsibility tables and area-boundary-equipment, S4, obtaining a global state based on the layered multi-agent architecture, S5, entering dynamic bearing capacity assessment calculation after obtaining the global state, and combining time sequence prediction and operation constraint to obtain a new load capacity capable of being safely borne in a future time window. The invention effectively improves the operation efficiency of the power distribution network and promotes the distributed energy consumption.

Inventors

YANG BO
LIU SHENGFU
LIANG XIAOLIN
ZHANG XIAOLEI
LU ZHIPENG
YANG YANG
YANG SHENQUAN
LIU SHULI
LI ZHAO
DENG SHAOZHI
LIU ZHAO
YU QIUYANG
CUI CAN
WANG KEXIN
ZHAO REN
WANG YANSHUO
LI HAO
WANG YAOLEI
ZHANG WEN
LI KAI
QI LUJIE
AN PENG
ZHAO LONG
WANG CHUNYI
ZHENG ZHIJIE
LIANG RONG
WANG CHEN

Assignees

国网山东省电力公司经济技术研究院

Dates

Publication Date: 20260505
Application Date: 20260120

Claims (10)

1. The method for evaluating the dynamic bearing capacity of the alternating current-direct current series-parallel power distribution network based on the multiple intelligent agents is characterized by comprising the following steps of: S1, acquiring and preprocessing multisource heterogeneous data of an alternating current-direct current hybrid power distribution network to obtain a standardized data packet; s2, after a standardized data packet is obtained, constructing a unified model of the alternating-current/direct-current hybrid distribution network, and carrying out topology analysis and structural feature extraction to obtain partitions and boundaries; S3, constructing a layered multi-agent architecture based on the subareas and boundaries obtained in the S2, and obtaining mapping relations between agent catalogs and responsibility tables and between areas and boundaries and devices; S4, based on a layered multi-agent architecture, each management layer agent estimates by using a measurement set and a unified model in the area and sends a boundary state to a coordination layer, and the coordination layer agent fuses and coordinately coordinates the boundary information of multiple areas so that the whole network state meets physical coupling and power balance at the boundary, thereby obtaining a global state; s5, after the global state is obtained, the dynamic bearing capacity evaluation calculation is carried out, and the new load capacity which can be safely borne in a future time window is obtained by combining time sequence prediction and operation constraint.
2. The method for evaluating the dynamic bearing capacity of the multi-agent-based alternating-current/direct-current hybrid power distribution network is characterized by comprising the steps of establishing a multi-source data access port for the alternating-current/direct-current hybrid power distribution network, synchronously acquiring operation related data from different system and equipment sides to form an original data set, wherein the original data set comprises static standing account and model data, real-time measurement and state data and external and planning data of the power distribution network, and completing data identification unification, unit and dimension unification, sampling frequency marking and integrity verification in an access layer.
3. The multi-agent-based alternating current-direct current hybrid power distribution network dynamic bearing capacity assessment method according to claim 2 is characterized by comprising the steps of firstly performing time alignment and synchronization, aligning multi-sampling rate data according to uniform time stamps, processing time scale drift caused by communication delay through resampling, performing missing value processing and dead point identification, identifying abnormal points based on threshold value out-of-limit, mutation detection and physical consistency constraint, and performing replenishment based on trusted values, and then performing consistency check and fusion, wherein the consistency check and fusion comprises whether a switching state is matched with a topology, whether measured power is consistent with measured data, whether alternating current-direct current interface power direction is consistent with symbol convention, so as to obtain a preprocessed standardized data packet, the content of which is organized in a structured form, and the standardized data packet comprises a current time topology snapshot measurement set and time stamps thereof, equipment operation state vectors, short/medium time sequence prediction data and data quality reports.
4. The method for evaluating the dynamic bearing capacity of the multi-agent-based alternating-current/direct-current hybrid distribution network according to claim 3, wherein after the standardized data packet is obtained, a unified model of the alternating-current/direct-current hybrid distribution network is constructed, specifically as follows: Firstly, constructing a unified resolvable model of an alternating current-direct current hybrid power distribution network based on a topology snapshot and a device parameter library: the alternating current side adopts a Newton-Laportson jacobian matrix of a power flow model of the power distribution network, and a relation between node power balance and branch power is established; the direct current side adopts a node voltage-branch current balance model to describe direct current bus voltage, branch current and power distribution; the AC-DC coupling is realized by a converter interface model, the power relation, control mode and operation boundary of the AC side and the DC side of the converter are explicitly written into a unified equation set, and the real-time set values in a standardized data packet, including the converter reference power, reactive power setting, the tapping position of the transformer and the energy storage available power boundary, are mapped into the initial values of the model to ensure that the model is consistent with the real-time operation, so that the unified model can be used for the subsequent state estimation and the bearing capacity boundary calculation.
5. The method for evaluating the dynamic bearing capacity of the multi-agent-based alternating-current/direct-current hybrid distribution network is characterized by comprising the steps of carrying out topology analysis and structural feature extraction, obtaining partitions and boundaries, specifically, generating a current operation topology according to switch switching positions, contact switch states and equipment outage information in a standardized data packet, carrying out topology analysis and structural feature extraction on the current operation topology, specifically, constructing a node-branch association relation and a connectivity matrix, identifying islands, radial/ring network structural changes and alternating-current/direct-current coupling point distribution caused by switching operation, identifying key nodes and key branches through graph theory indexes, extracting potential bottleneck candidate sets by combining equipment rated capacity and historical trend statistics, and simultaneously analyzing standby power supply paths and redundancy.
6. The method for evaluating the dynamic bearing capacity of the multi-agent-based alternating-current/direct-current hybrid distribution network is characterized in that the layered multi-agent architecture adopts a distributed computing and collaborative decision-making framework comprising an equipment layer, a management layer and a coordination layer, wherein the equipment layer agents are deployed according to the principle of one equipment agent, cover an inverter, an important switch, voltage regulating equipment, an energy storage system and an important feeder line, each equipment layer agent is responsible for data acquisition, health status monitoring and control instruction execution and equipment protection of the equipment, the management layer agents are deployed one to one according to the electrical partition determined by the S2, the management layer agents comprise a platform area agent, a feeder line agent, a station agent, a looped network unit agent and a direct-current sub-network agent, all equipment layer agents in the management area are managed by each management layer agent, the state estimation, local power flow calculation, fault diagnosis and bearing capacity estimation and control strategy are generated in the bearing area, and the coordination layer agents comprise a global coordination agent, an alternating-current/direct-current interface coordination agent and an emergency linkage agent, and the coordination layer realizes the goal of balancing and the balance of the power and the convergence of all-phase/frequency and the power coordination network are achieved through the information exchange with each management layer.
7. The method for evaluating the dynamic bearing capacity of the alternating-current/direct-current hybrid distribution network based on the multiple intelligent agents according to claim 6 is characterized in that the equipment layer intelligent agents are used for data sensing and control execution, collecting equipment operation parameters in real time, evaluating equipment health, executing control instructions issued by upper levels, feeding back execution results and abnormal alarms in time, and meanwhile having basic self-protection capacity, the management layer intelligent agents are used for regional analysis and local optimization, carrying out regional state estimation, regional load and new energy output prediction, carrying out local fault diagnosis and out-of-limit analysis, calculating regional bearing capacity and bottleneck recognition, generating regional control candidate strategies, carrying out information exchange with other management layers and coordination layers through boundary information, and the coordination layer intelligent agents are used for global coordination and system optimization, collecting boundary state information reported by each management layer, carrying out boundary consistency coordination, processing cross-regional conflicts, executing system-level emergency linkage, and carrying out whole network bearing capacity optimization and resource coordination allocation.
8. The method for evaluating the dynamic bearing capacity of the multi-agent-based alternating-current/direct-current hybrid distribution network according to claim 7, wherein the agent catalogue and the responsibility table record basic information, function responsibility lists and performance index requirements of each agent in a structured form to form an agent specification, the area-boundary-device mapping relation establishes a multi-layer attribution and coupling relation, the area mapping table defines the boundary of an electric area managed by each management layer agent, the contained node and branch collection, the equipment list and equipment layer agent allocation in the area, the boundary mapping table defines cross-area coupling interfaces, boundary variable definition and direction convention, boundary information exchange frequency and priority, the equipment mapping table establishes the dependency relation between an equipment entity and an agent, the management layer and the equipment layer, and allocation of equipment control authority, and the flexible and reliable information exchange mechanism of communication topology and routing rule design is to define the communication connection relation between the agents, message type and data format standardization (communication frequency and priority setting, and fault tolerance and degradation mechanism, and stable cooperative operation of the multi-agent system under complex network environments is ensured.
9. The method for evaluating the dynamic bearing capacity of the multi-agent-based alternating-current/direct-current hybrid distribution network according to claim 1, wherein after the global state is obtained, the method enters dynamic bearing capacity evaluation calculation, and the new load capacity capable of being safely borne in a future time window is obtained by combining time sequence prediction and operation constraint, and is specifically as follows: After the current running state with the global consistency is obtained in S4, the state is used as an initial condition of rolling evaluation, dynamic bearing capacity calculation is carried out, an evaluation time window T and a rolling step length delta T are set, a time sequence prediction result and a confidence interval thereof are read from a standardized data packet and are organized into a future time sequence for input, the target of the dynamic bearing capacity is defined as a newly increased load capacity delta P add which can be additionally accepted by a system on the premise that all moments in the time window meet running constraint and control feasibility, the newly increased load capacity delta P add can be further thinned into the maximum accessible power, accessible time interval, accessible position distribution and safety margin of the newly increased load, a reference running track is used as a reference in the calculation process, the reference load sequence is obtained by rolling deduction according to the predicted load and power output under the condition of not newly increased load, and the candidate newly increased load injection scheme is optimized and solved on the basis, and the maximum newly increased load capacity which is not triggered to exceed the limit in the whole time window is obtained.
10. The system for evaluating the dynamic bearing capacity of the multi-agent-based alternating-current/direct-current hybrid distribution network is characterized by comprising a processor, a memory and a computer program stored on the memory, wherein the steps in the method for evaluating the dynamic bearing capacity of the multi-agent-based alternating-current/direct-current hybrid distribution network are specifically executed when the processor executes the computer program.

Description

Multi-agent-based AC/DC hybrid power distribution network dynamic bearing capacity assessment method Technical Field The invention relates to the field of power distribution network evaluation, in particular to a multi-agent-based AC/DC series-parallel power distribution network dynamic bearing capacity evaluation method. Background With the deep construction of a novel power system, a power distribution network is undergoing unprecedented revolution. The traditional single alternating current power distribution network architecture gradually evolves towards an alternating current-direct current hybrid form, and a large number of novel loads such as distributed photovoltaic, wind power, energy storage and electric automobile charging piles are connected with power supply equipment, so that the power distribution network has the novel characteristics of deep coupling of 'source network charge storage', 'alternating current-direct current coexistence', 'bidirectional tide', and the like. The traditional power distribution network planning and design is mainly based on a static checking method of maximum load, so that the real bearing capacity of a system under a dynamic operation condition is difficult to accurately evaluate, the problems of 'equipment light load but partial overload' or 'resource waste caused by conservative design' and the like are easy to occur, and a more accurate and dynamic bearing capacity evaluation technology needs to be developed. The alternating current-direct current series-parallel power distribution network realizes flexible interconnection of alternating current and direct current systems through the power electronic converter, can better adapt to the direct current characteristics of a distributed power supply, improves the power supply reliability and reduces the line loss. Currently, the mainstream AC/DC interface technology includes a Voltage Source Converter (VSC), a bidirectional DC/DC converter, an AC/DC hybrid bus, and the like, and the control strategy covers various modes such as constant Power (PQ) control, constant direct current Voltage (VDC) control, droop (drop) control, and the like. However, the coupling characteristic of the ac-dc system makes the traditional ac-power flow-based analysis method no longer applicable, and a unified mathematical model considering the characteristics of the converter, control constraint and loss characteristic needs to be established. Meanwhile, technical challenges still exist in the aspects of fault propagation mechanism, protection coordination, stability analysis and the like of an alternating current-direct current series-parallel system, and particularly under the condition of high-proportion renewable energy source access, the time variability and uncertainty of the running state of the system are obviously increased, and higher requirements are put forward on the instantaneity and the accuracy of the bearing capacity evaluation. Disclosure of Invention In order to solve the problems, the invention aims to provide a multi-agent-based dynamic bearing capacity evaluation method for an alternating-current/direct-current hybrid power distribution network, which can effectively improve the operation efficiency of the power distribution network and promote the consumption of distributed energy sources. In order to achieve the above purpose, the present invention adopts the following technical scheme: a multi-agent-based AC/DC hybrid power distribution network dynamic bearing capacity assessment method comprises the following steps: S1, acquiring and preprocessing multisource heterogeneous data of an alternating current-direct current hybrid power distribution network to obtain a standardized data packet; s2, after a standardized data packet is obtained, constructing a unified model of the alternating-current/direct-current hybrid distribution network, and carrying out topology analysis and structural feature extraction to obtain partitions and boundaries; S3, constructing a layered multi-agent architecture based on the subareas and boundaries obtained in the S2, and obtaining mapping relations between agent catalogs and responsibility tables and between areas and boundaries and devices; S4, based on a layered multi-agent architecture, each management layer agent estimates by using a measurement set and a unified model in the area and sends a boundary state to a coordination layer, and the coordination layer agent fuses and coordinately coordinates the boundary information of multiple areas so that the whole network state meets physical coupling and power balance at the boundary, thereby obtaining a global state; s5, after the global state is obtained, the dynamic bearing capacity evaluation calculation is carried out, and the new load capacity which can be safely borne in a future time window is obtained by combining time sequence prediction and operation constraint. The method comprises the steps of obtaining multi-source hetero