An agent cooperation-based regional energy system planning monitoring method, system, device and medium

Abstract

The application discloses a kind of based on agent cooperation's regional energy system planning monitoring method, system, equipment and medium, belong to information technology field, including: obtaining the original geographic boundary data of target area and energy load data, according to data determine initial spatial division range;Energy supply and demand data are constructed;Through the information interaction between each agent node, dynamically adjust the spatial division boundary, and based on the boundary of adjustment collaborative generation interregional energy collaborative allocation strategy;Efficiency feedback data are obtained, and the spatial division range and energy collaborative allocation strategy are iteratively optimized according to efficiency feedback data, to determine the final resource allocation adjustment scheme.The application reduces the prediction bias, realizes the balanced allocation and efficient use of interregional energy resources.

Description

Technical Field

[0001] This invention relates to the field of information technology, and specifically to a method, system, equipment, and medium for regional energy system planning and monitoring based on intelligent agent collaboration. Background Technology

[0002] In the field of energy planning and resource allocation, research on energy measurement and coordination mechanisms at multiple spatial scales is of paramount strategic importance. This field is directly related to the optimization and sustainable development of regional energy structures, and its research value is increasingly prominent, especially against the backdrop of rapid integration of new energy sources and constantly changing electricity demand. However, existing methods often fall short in addressing complex environmental changes. Many schemes neglect the spatial interplay between regional energy distribution and demand characteristics, lacking in-depth exploration of the dynamic relationships between different regions. This is particularly problematic when facing shifts in regional development models, making it difficult to adapt to new energy patterns and demand characteristics. This limitation leads to a disconnect between planning results and actual needs, resulting in inefficient resource allocation. A deeper challenge lies in the failure to effectively address two key factors: the flexibility of spatial division and the adaptability of predictive models. The granularity of spatial division directly determines the accuracy of measurement. If the division method cannot be dynamically adjusted according to regional characteristics, data for some regions will be too coarse or too fragmented, affecting the rationality of the overall plan. This spatial division problem further exacerbates the difficulty of adjusting the parameters of the prediction model, because the model needs to capture the evolution of the characteristics of new energy output and electricity consumption behavior at different spatial scales. If the parameters cannot adapt to changes in a timely manner, it will lead to prediction bias, thereby affecting the scientific nature of resource allocation. Therefore, how to dynamically adjust the spatial division granularity and prediction model parameters in the context of regional development pattern transformation, so as to accurately capture the spatial coupling relationship between new energy capacity and load demand in different regions, has become a key problem that this study urgently needs to solve. Summary of the Invention

[0003] In view of the above-mentioned problems, the present invention is proposed.

[0004] Therefore, the technical problem solved by this invention is: how to achieve dynamic and coordinated optimization of spatial division boundaries and energy coordination strategies in regional energy system planning, so as to solve the problems of low resource allocation efficiency and poor model adaptability caused by rigid spatial division and insufficient inter-regional correlation analysis in existing methods.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a regional energy system planning and monitoring method based on intelligent agent collaboration, comprising, Obtain the original geographic boundary data and energy load data of the target area, and determine the initial spatial division range based on the data. The initial spatial division range includes several sub-regions. Construct intelligent agent nodes corresponding to each sub-region, and each intelligent agent node stores the energy supply and demand data of the corresponding sub-region; through information interaction between intelligent agent nodes, dynamically adjust the spatial division boundary between several sub-regions, and collaboratively generate an energy collaborative allocation strategy between regions based on the adjusted boundary. Obtain performance feedback data after implementing the energy collaborative allocation strategy, and iteratively optimize the spatial division range and energy collaborative allocation strategy based on the performance feedback data to determine the final resource allocation adjustment plan.

[0006] As a preferred embodiment of the regional energy system planning and monitoring method based on intelligent agent collaboration described in this invention, the determination of the initial spatial division range includes, Acquire time-series electricity consumption data, industry distribution data, and traffic flow data for each sub-region; Based on time-series electricity consumption data, industry distribution data, and traffic flow data, the geographical boundaries of the initial spatial division range are refined and adjusted to obtain a refined spatial division granularity scheme.

[0007] As a preferred embodiment of the regional energy system planning and monitoring method based on intelligent agent collaboration described in this invention, the information interaction between each intelligent agent node includes: Each agent node exchanges the information required for decision-making with at least one neighboring agent node based on the stored power generation capacity data and load demand data of the corresponding sub-region; Each intelligent agent node generates a local scheduling plan based on locally stored data and received information, and iteratively adjusts the scheduling plan through multiple rounds of information interaction until the coordination conditions are met.

[0008] As a preferred embodiment of the regional energy system planning and monitoring method based on intelligent agent collaboration described in this invention, the preset collaboration conditions include: There are no conflicts in power flow or capacity between the scheduling plans generated by each intelligent agent node; The local scheduling plans of each intelligent agent node all satisfy the corresponding local environment constraints and facility capacity constraints.

[0009] As a preferred embodiment of the regional energy system planning and monitoring method based on intelligent agent collaboration described in this invention, the dynamic adjustment of the spatial division boundaries between several sub-regions includes: Based on the energy load density distribution data of each sub-region obtained during the information exchange process, identify the boundary locations where the load density difference between adjacent sub-regions exceeds a preset threshold. Adjust the coordinates of the boundary positions and calculate the load density variance of each sub-region after adjustment; The boundary coordinates that minimize the load density variance are selected as the adjusted spatial partition boundaries.

[0010] This invention solves the technical problem in the prior art where the difference in load density between adjacent sub-regions exceeds a preset threshold by identifying the boundary position of the load density difference between adjacent sub-regions based on the energy load density distribution data of each sub-region obtained during the information interaction process, and adjusting the boundary coordinates to minimize the variance of the load density of each sub-region after adjustment. This solves the technical problem that the spatial division is based on fixed administrative boundaries and cannot reflect the actual distribution characteristics of energy load within the region, resulting in the mixing of high-load and low-load areas in the same region and the difficulty in accurately matching the planning scheme with actual needs. It achieves the goal of making the energy characteristics within each sub-region more consistent and improving the accuracy of spatial division.

[0011] As a preferred embodiment of the regional energy system planning and monitoring method based on intelligent agent collaboration described in this invention, the step of acquiring performance feedback data after executing the energy collaborative allocation strategy, and iteratively optimizing the spatial division range and energy collaborative allocation strategy based on the performance feedback data, includes: Based on the adjusted spatial delineation boundaries, an updated energy collaborative allocation strategy is generated and executed. Obtain performance feedback data after implementing the updated energy collaborative allocation strategy; Based on performance feedback data, the adjusted spatial boundary and updated energy collaborative allocation strategy are iteratively adjusted until the preset optimization target is met, resulting in the final resource allocation adjustment scheme.

[0012] As a preferred embodiment of the regional energy system planning and monitoring method based on intelligent agent collaboration described in this invention, the iterative adjustment includes: Based on performance feedback data, the adjusted spatial division boundary and the updated energy collaborative allocation strategy are evaluated; If the evaluation results do not meet the preset optimization objectives, the spatial division boundary is adjusted again, and the energy collaborative allocation strategy is updated according to the adjusted boundary. Then, the process returns to the step of obtaining performance feedback data.

[0013] This invention addresses the technical problems of existing planning methods, which are mostly one-time static decisions, lack feedback mechanisms for execution effects, and cannot be adjusted in a timely manner when actual operation deviates from planning assumptions, resulting in low resource allocation efficiency. This approach achieves continuous optimization, gradually bringing boundary division and strategy formulation closer to the optimal, and ensuring that the system always remains consistent with actual operating conditions.

[0014] This invention provides a regional energy system planning and monitoring system based on intelligent agent collaboration.

[0015] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a regional energy system planning and monitoring system based on intelligent agent collaboration, comprising: a spatial division module, an energy collaborative allocation strategy generation module, and an output module; The spatial division module acquires the original geographical boundary data and energy load data of the target area, and determines the initial spatial division range based on the data. The initial spatial division range includes several sub-regions. The module for generating energy collaborative allocation strategy constructs intelligent agent nodes corresponding to each sub-region, and each intelligent agent node stores the energy supply and demand data of the corresponding sub-region; through information interaction between the intelligent agent nodes, the spatial division boundary between several sub-regions is dynamically adjusted, and an energy collaborative allocation strategy between regions is generated based on the adjusted boundary. The output module acquires performance feedback data after executing the energy collaborative allocation strategy, iteratively optimizes the spatial division range and energy collaborative allocation strategy based on the performance feedback data, and determines the final resource allocation adjustment scheme.

[0016] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the aforementioned regional energy system planning and monitoring method based on intelligent agent collaboration.

[0017] The present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the aforementioned regional energy system planning and monitoring method based on intelligent agent collaboration.

[0018] The beneficial effects of this invention are as follows: By acquiring geographical boundary and energy load data, and combining it with factors such as land use and population density for spatial division, and refining the granularity of the division based on historical data such as electricity consumption patterns and industrial structure, this invention effectively solves the problem of measurement accuracy caused by insufficient flexibility and unreasonable granularity in spatial division. By establishing inter-regional energy supply and demand relationships, analyzing spatial coupling strength, and dynamically updating energy facility demand and environmental constraints, this invention compensates for the shortcomings of existing methods that ignore the dynamic correlation of energy between regions. Through a multi-agent distributed decision-making algorithm, this invention maps the long-term trend of the proportion of new energy sources, integrates the demand for distribution network expansion and energy storage layout, obtains an energy collaborative allocation strategy, and finally combines energy utilization efficiency assessment data and real-time changes to determine resource allocation adjustment schemes. This improves the adaptability of the prediction model to changes in regional development patterns, reduces prediction bias, and achieves balanced allocation and efficient utilization of energy resources between regions. Attached Figure Description

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

[0020] Figure 1 The above is a flowchart of a regional energy system planning and monitoring method based on intelligent agent collaboration, provided as an embodiment of the present invention. Detailed Implementation

[0021] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0022] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides a regional energy system planning and monitoring method based on intelligent agent collaboration, comprising: S1. Obtain the original geographical boundary data and energy load data of the target area, and determine the initial spatial division range based on the data. The initial spatial division range includes several sub-regions. S2. Construct intelligent agent nodes corresponding to each sub-region. Each intelligent agent node stores the energy supply and demand data of the corresponding sub-region. Through information interaction between intelligent agent nodes, dynamically adjust the spatial division boundary between several sub-regions, and collaboratively generate an energy collaborative allocation strategy between regions based on the adjusted boundary. S3. Obtain performance feedback data after implementing the energy collaborative allocation strategy, and iteratively optimize the spatial division range and energy collaborative allocation strategy based on the performance feedback data to determine the final resource allocation adjustment plan.

[0023] Example 2, an embodiment of the present invention, provides a regional energy system planning and monitoring method based on intelligent agent collaboration, based on the previous embodiment, including: Step S1 involves acquiring the original geographic boundary data and energy load data of the target area, and determining the initial spatial division range based on the data. This includes the following steps: S11. Obtain time-series electricity consumption data, industry distribution data, and traffic flow data for each sub-region.

[0024] Specifically, the first step is to acquire the geographic boundary coordinates and land use classification information of the target area. The geographic boundary coordinates are obtained by integrating administrative division data released by government departments, vector boundary data from geographic information systems, and satellite remote sensing imagery. The land use classification information comes from land use planning maps and land use status survey data from urban planning departments, and is divided into residential areas, commercial areas, industrial areas, and mixed-use areas based on land use type.

[0025] In the overlay analysis of population density data, spatial interpolation methods are used to transform discrete population statistics into a continuous density distribution map. Kriging interpolation or inverse distance weighted interpolation are employed to estimate population density variations within each land use category, identifying population distribution differences within the same land use category. Adaptive grid division is applied to the target area based on population density distribution characteristics. Smaller grid units (e.g., 500m × 500m) are used for high-population-density areas, while larger grid units (e.g., 2km × 2km) are used for low-population-density areas. This variable-scale grid division method improves data processing efficiency while maintaining computational accuracy. Historical power load data and real-time load monitoring data for each grid are used to calculate the load density value for each geographic grid unit. The load density value is obtained by dividing the total load by the grid area. Based on the load density value, the region is divided into high-load-density, medium-load-density, and low-load-density areas.

[0026] Spatially overlaying analysis is performed on the load density level distribution with the location information of existing renewable energy power generation facilities in each region. Renewable energy power generation facilities include rooftop photovoltaics, small wind turbines, and distributed energy storage systems. By statistically analyzing the installed capacity and annual power generation of renewable energy in each load density level region, and comparing it with the annual electricity consumption of that region, the proportion of renewable energy is obtained. If the proportion of renewable energy in a certain region is lower than a preset threshold, the region is marked as a key region for renewable energy development, forming key region distribution data that includes regional location and renewable energy proportion values. Combining the land use category and load density level information of each key region, the boundary coordinates of the initial spatial division range are determined, resulting in a basic distribution feature dataset of regional renewable energy proportion that includes geographical boundaries, load density levels, and renewable energy proportion values.

[0027] S12. Based on time-series electricity consumption data, industry distribution data, and traffic flow data, the geographical boundaries of the initial spatial division range are refined and adjusted to obtain a refined spatial division granularity scheme.

[0028] Specifically, the hourly electricity load curves of each region on weekdays, rest days, and holidays are extracted from historical data. The time-domain signal is converted into a frequency-domain signal using Fourier transform. The electricity consumption time sequence patterns of daily, weekly, and seasonal cycles are determined by identifying the main frequency components in the spectrum, resulting in a dataset of electricity consumption time sequence patterns containing the frequency and amplitude of each cycle.

[0029] Based on the periodic differences in the data of electricity consumption time series patterns, and combined with the initial spatial division range, information on the distribution of industrial types and traffic flow data of traffic monitoring stations in each region within the range are obtained. By comparing the changes in the proportion of industrial electricity consumption and the time distribution of traffic flow in different periods, indicators for industrial structure adjustment and trends in traffic flow are determined.

[0030] Using industrial structure adjustment indicators and traffic flow change trends, the geographical boundaries within the initial spatial division are adjusted a second time. If the change in the proportion of industrial electricity consumption in a certain area exceeds a preset threshold or the growth rate of traffic flow exceeds a threshold, the area is further subdivided according to industrial agglomeration and traffic density to obtain a refined set of geographical boundary coordinates.

[0031] The original energy load data is remapped and aggregated by refining the set of geographic boundary coordinates. The correlation coefficient of the electricity consumption time sequence between adjacent refined areas determines whether to merge or further subdivide, resulting in a spatial division granularity scheme that includes the final boundary coordinates and corresponding load distribution characteristics.

[0032] For example, in practice, the extraction of hourly electricity load curves involves the systematic collection and processing of historical electricity data.

[0033] By recording the electricity consumption data hourly by smart meters, a load curve is formed for 24 consecutive hours. The load curve on weekdays usually shows a double-peak characteristic, with the morning peak occurring between 8-10 am and the evening peak occurring between 18-21 pm. On rest days, it shows a single-peak characteristic, with the peak concentrated between 19-22 pm. The electricity consumption pattern on holidays is more dispersed, and the overall load level is lower. This differentiated electricity consumption pattern reflects the regular changes in residents' life and production activities at different times.

[0034] The specific application of Fourier transform in this process is to convert the time-domain load curve to the frequency domain for analysis. By performing a Fast Fourier Transform on 8760 hours of load data over a year, the energy distribution of different frequency components can be identified. The frequency corresponding to the daily cycle is 1 / 24 hour, the weekly cycle is 1 / 168 hour, and the seasonal cycle is 1 / 2190 hour. In the spectral analysis results, the frequency components with larger amplitudes represent the significance of this periodicity.

[0035] For example, the daily periodic component amplitude in commercial areas is usually much larger than the weekly periodic component, indicating that their electricity consumption behavior is mainly affected by daily business hours.

[0036] The determination of industrial structure adjustment indicators is based on the differences in electricity consumption characteristics of different industry types. Manufacturing exhibits a sustained high load characteristic with a relatively stable electricity consumption curve; the service industry shows a clear concentration of electricity consumption during business hours; and high-tech industries are characterized by all-weather but large load fluctuations. By calculating the proportion of electricity consumption of different industry types in the total electricity consumption at different times and comparing the changes in the proportions over different years, the evolution trend of industrial structure can be quantified. When the proportion of electricity consumption of the service industry in a certain region increases from 30% to 50%, it indicates that the region is undergoing a transformation from industry-led to service-led. The analysis of traffic flow trends relies on vehicle flow data collected by road monitoring equipment. Traffic flow during morning and evening peak hours directly affects the electricity demand of surrounding commercial and office areas. Statistical analysis shows that for every 1,000 vehicles / hour increase in traffic flow, the average electricity load within a 500-meter radius increases by 2-3 megawatts. This correlation makes traffic data an important reference factor for electricity load forecasting.

[0037] The geographical boundaries within the initial spatial division area are adjusted using industrial structure adjustment indicators and traffic flow trends.

[0038] For example, the daily cycle component amplitude in commercial areas is usually much larger than the weekly cycle component, indicating that their electricity consumption behavior is mainly affected by daily business hours. The determination of industrial structure adjustment indicators is based on the differences in electricity consumption characteristics of different industry types. Manufacturing shows a continuous high load characteristic, and the electricity consumption curve is relatively stable. The service industry shows obvious concentrated electricity consumption during business hours. High-tech industries have the characteristics of all-weather but large load fluctuations. By calculating the proportion of electricity consumption of different industry types in the total electricity consumption at each time period and comparing the changes in the proportion in different years, the evolution trend of industrial structure can be quantified. When the proportion of electricity consumption of the service industry in a certain area increases from 30% to 50%, it indicates that the area is undergoing a transformation from industry-led to service-led. The analysis of traffic flow change trends relies on the traffic flow data collected by road monitoring equipment. The traffic flow during morning and evening peak hours directly affects the electricity demand of surrounding commercial and office areas. Statistical analysis shows that for every 1,000 vehicles / hour increase in traffic flow, the electricity load within a 500-meter radius increases by an average of 2-3 megawatts. This correlation makes traffic data an important reference factor for electricity load forecasting.

[0039] If the change in the proportion of industrial electricity consumption in a certain region exceeds 20% or the annual growth rate of traffic flow exceeds 15%, the region is further subdivided according to industrial agglomeration and traffic density. The original grid unit is subdivided into four sub-grids, resulting in a refined set of geographic boundary coordinates. The original energy load data is then remapped and aggregated using the refined set of geographic boundary coordinates. Based on the correlation coefficient of electricity consumption time series patterns between adjacent refined regions, it is determined whether merging or further subdivision is necessary, generating a spatial partitioning granularity scheme that includes the final boundary coordinates and corresponding load distribution characteristics.

[0040] In step S2, intelligent agent nodes are constructed corresponding to each sub-region, and each intelligent agent node stores the energy supply and demand data of its corresponding sub-region. Through information interaction between the intelligent agent nodes, the spatial division boundaries between several sub-regions are dynamically adjusted, and an energy collaborative allocation strategy between regions is collaboratively generated based on the adjusted boundaries, including the following steps: S21. Construct intelligent agent nodes corresponding to each sub-region, and each intelligent agent node stores the energy supply and demand data of the corresponding sub-region.

[0041] Specifically, after completing the spatial partitioning granularity scheme, agent nodes representing each sub-region are constructed. Each agent node stores a complete energy profile of that sub-region, including: Inherit the geographic boundary coordinate data and land use classification information (residential area, commercial area, industrial area or mixed-use area) from S12; population density distribution characteristics, load density level and load density value calculated in S12; new energy proportion value and key areas for new energy development marked in S12.

[0042] Further storage includes data on the installed power generation capacity of the sub-region, including the rated power values ​​of thermal power, hydropower, wind power, and photovoltaic power generation facilities; real-time load demand data and historical electricity load curves, including hourly electricity load data for weekdays, rest days, and holidays; information on transmission lines connecting the sub-region, including voltage level, line length, and conductor cross-sectional area; information on the distribution of industry types (manufacturing, service industry, high-tech industry, etc.) and traffic flow data from traffic monitoring stations; the area distribution and historical electricity consumption records of residential, commercial, and industrial buildings; initial carbon emission quota data; and land use status maps and planning maps to identify the area and location of land available for energy facility construction.

[0043] These data are continuously updated through a real-time monitoring system, serving as the basis for information interaction and collaborative decision-making among intelligent agent nodes. The spatial coupling strength, updated constraints, energy facility requirements, load density distribution, supply and demand balance, and other data calculated in subsequent steps will be dynamically stored in the corresponding intelligent agent nodes after the calculations are completed, to support the next round of iterative optimization.

[0044] S22. Through information interaction between various intelligent agent nodes, dynamically adjust the spatial division boundaries between several sub-regions.

[0045] S221. Each agent node exchanges the information required for decision-making with at least one neighboring agent node based on the stored power generation capacity data and load demand data of the corresponding sub-region.

[0046] Specifically, each intelligent agent node exchanges energy flow information and load forecast data with neighboring intelligent agent nodes through a message passing mechanism.

[0047] First, each intelligent agent node calculates the maximum active power transmission capacity of each line based on the stored power generation capacity data (including the rated power values ​​of thermal power, hydropower, wind power, and photovoltaic power generation facilities) and transmission line information (voltage level, line length, conductor cross-sectional area), and obtains the inter-regional power transmission capacity matrix.

[0048] Secondly, each intelligent agent node calculates the difference between the power generation capacity and the power demand of each sub-region based on the stored real-time load demand data and power generation capacity. When the difference is positive, it is identified as a power surplus region, and when the difference is negative, it is identified as a power deficit region. An inter-regional energy supply and demand relationship network is established, which includes supply regions, demand regions, and transmission channel capacity. This network information is exchanged with neighboring intelligent agent nodes through a message passing mechanism.

[0049] The information exchanged by each intelligent agent node also includes: the proportion of new energy power generation in each sub-region, load demand forecast data, power surplus or deficit data, transmission line transmission capacity data, carbon emission quota usage, distribution network capacity margin, the rate of change of the proportion of new energy (calculated by dividing the difference in proportion between adjacent time periods by the time interval), and the standard deviation of energy consumption per unit area for different building types as the energy consumption fluctuation range, etc.

[0050] For example, in practical applications, the acquisition of power generation capacity data covers detailed information on various power generation types. The rated power of thermal power units is usually between 300 and 1000 megawatts. Hydropower stations determine their installed capacity based on reservoir capacity and head height. The total installed capacity of wind farms is obtained by multiplying the single unit capacity by the number of wind turbines. Photovoltaic power stations calculate their total capacity based on module area and conversion efficiency. These basic data form the basis for regional power supply capacity assessment.

[0051] The calculation of the maximum transmission capacity of a transmission line involves the comprehensive consideration of multiple electrical parameters. The rated voltage determines the insulation level and voltage drop limit of the line, the conductor cross-sectional area affects the current carrying capacity, and the line length is related to the resistance loss and reactive power compensation requirements. For example, a 500 kV transmission line using 4-split conductors with a single conductor cross-section of 400 square millimeters can carry a current of 2400 amperes at an ambient temperature of 25 degrees Celsius, corresponding to an active power transmission capacity of approximately 2000 megawatts. This calculation method ensures the accuracy of transmission capacity assessment. In the process of establishing an inter-regional energy supply and demand network, there are clear quantitative standards for determining power surplus and deficit. When a region's power generation capacity is 800 MW and its real-time load demand is 600 MW, a difference of 200 MW indicates that the region is a power surplus region. Conversely, if the power generation capacity is 500 MW and the load demand reaches 700 MW, then a deficit of 200 MW makes it a power demand region. This supply and demand matching relationship directly determines the direction of power flow between regions.

[0052] The calculation of the rate of change in the proportion of new energy sources adopts the time series analysis method. Taking photovoltaic power generation as an example, the photovoltaic output gradually rises from zero in the morning to a peak at noon during the daytime, and then drops to zero in the evening. This intraday variation pattern makes the proportion of new energy sources significantly different at different times. By calculating the difference in the proportion between adjacent hours, such as if the proportion of new energy sources is 30% at 10:00 am and rises to 35% at 11:00 am, the change rate for that hour is 5%. This dynamic change characteristic puts forward higher requirements for grid dispatch. The differences in energy consumption characteristics of building types are an important factor affecting regional electricity consumption patterns. The electricity consumption of residential buildings is mainly concentrated in the morning and evening, with relatively low energy consumption per unit area but significant fluctuations. Commercial buildings maintain a high level of energy consumption during business hours, with air conditioning and lighting loads accounting for a large proportion. Industrial buildings exhibit continuous high energy consumption characteristics, and the start and stop of production equipment directly affect energy consumption fluctuations. By statistically analyzing the electricity consumption data of different types of buildings at different times and calculating their standard deviation, the degree of energy consumption fluctuation can be quantified. The calculation of the spatial coupling strength index comprehensively considers the two dimensions of frequency and scale of power exchange.

[0053] For example, there are 48 power exchanges between region A and region B every day, with an average exchange of 50 megawatt-hours each time, and a total daily exchange of 2,400 megawatt-hours. If the total daily load of region B is 10,000 megawatt-hours, then the exchange volume accounts for 24%. Multiplying this proportion by the exchange frequency of 48, the resulting value reflects the degree of energy dependence between the two regions. The larger the value, the tighter the coupling.

[0054] S222. Each intelligent agent node generates a local scheduling plan based on the data stored locally and the information received, and iteratively adjusts the scheduling plan through multiple rounds of information interaction until the coordination conditions are met.

[0055] Specifically, each intelligent agent node generates a preliminary local scheduling plan based on locally stored data and information received from neighboring nodes. The scheduling plan includes the time periods during which power will be supplied to or received from neighboring areas, as well as the expected power values.

[0056] Subsequently, each intelligent agent node counts the number of power transmissions per unit time between adjacent regions as the exchange frequency based on the transmission channel data received from the inter-regional energy supply and demand relationship network, uses the cumulative transmitted power as the exchange quantity, calculates the proportion of the exchange quantity to the total load of the power receiving area, multiplies it by the exchange frequency, and obtains the spatial coupling strength.

[0057] The spatial coupling strength is compared with a preset threshold. If the spatial coupling strength is higher than the preset threshold, the region is marked as a high coupling region.

[0058] Identify the existing transformer capacity and line current carrying capacity data of the high-coupling area, calculate the current load rate, and combine the historical load growth rate to estimate the future load value, thereby obtaining the value of the distribution network capacity expansion demand.

[0059] Based on the expansion area determined by the distribution network capacity expansion demand, time-series data of new energy power generation and load power within this area are obtained. The difference between the maximum and minimum load values ​​is calculated as the peak-valley difference, and the standard deviation of new energy power generation is calculated as the fluctuation amplitude. The capacity demand of energy storage facilities is determined by combining the peak-valley difference and the fluctuation amplitude. The location of energy storage facilities is determined based on the load density distribution.

[0060] By using data on energy storage facility capacity demand and location, the carbon emission coefficient per unit of electricity during the charging and discharging process of thermal power and energy storage is obtained. The new carbon emissions are calculated based on the charging and discharging efficiency of energy storage facilities and the expected number of operating hours. Combined with the original carbon emissions and total emission control indicators of the region, the updated carbon emission limit is obtained by subtracting the existing emissions and new emissions from the total emissions.

[0061] By obtaining the updated carbon emission limits and energy storage facility location requirements, we can obtain regional land use status maps and planning maps, identify the area and location of land available for energy facility construction, and use the information on buildable land as land use restrictions, which together with carbon emission limits constitute the updated environmental constraints.

[0062] Subsequently, the agents iteratively adjust the scheduling plan through multiple rounds of information exchange: In the first round, each node sends its preliminary local plan to its neighboring nodes. After receiving the plan, the neighboring nodes check whether it conflicts with their own plans (such as whether the transmission capacity is exceeded or whether the direction is contradictory).

[0063] If the requirements are not met, modification suggestions are proposed; each node updates its local plan based on the received suggestions and proceeds to the next round of interaction.

[0064] This process is repeated until the preset coordination conditions are met.

[0065] The coordination conditions include: (1) there is no conflict between the power flow or capacity of the scheduling plans generated by each intelligent agent node; (2) each intelligent agent node’s local scheduling plan meets the corresponding local environmental constraints (such as updated carbon emission limits, land use restrictions) and facility capacity constraints (such as distribution network transformer capacity, line current carrying capacity, and energy storage facility capacity).

[0066] To give another example: the threshold determination of spatial coupling strength is a key trigger point for the adjustment of the entire energy facility planning.

[0067] When a region exchanges electricity frequently with neighboring regions and the exchange volume accounts for a large proportion, it indicates that the region's energy supply and demand are heavily dependent on external inputs or outputs.

[0068] An industrial park purchases large amounts of electricity from surrounding residential areas during the day and supplies excess electricity back to the residential areas at night. This frequent two-way energy flow creates a close energy dependency between the two areas. When the intensity of this dependency exceeds a preset threshold, it means that the existing energy infrastructure can no longer meet the needs of regional development.

[0069] The calculation of distribution network capacity expansion demand is based on multi-dimensional data analysis. Transformer capacity utilization reflects the load level of equipment; when the utilization rate consistently exceeds 80%, it indicates that the equipment is operating close to full load. Line current carrying capacity reflects the transmission margin of transmission lines. By monitoring the ratio of line current to rated current carrying capacity, it is possible to determine whether there is an overload risk on the line. The analysis of historical load growth rate uses electricity consumption data from the past three years, and predicts future load growth trends through linear regression or exponential smoothing methods. This prediction method based on historical data can reflect the inherent laws of regional economic development and electricity demand.

[0070] Determining the capacity of energy storage facilities requires comprehensive consideration of the volatility of new energy sources and load characteristics. The standard deviation of new energy power generation reflects the degree of instability of its output. Photovoltaic power generation drops sharply when blocked by clouds, while wind power generation is significantly affected by wind speed changes. The peak-valley load difference reflects the time-varying differences in electricity demand. The peak-valley difference is usually larger in commercial areas and relatively stable in industrial areas. Energy storage facilities can mitigate this supply-demand mismatch by charging during off-peak hours and discharging during peak hours. The calculation of energy storage capacity not only considers the peak-valley difference but also needs to take into account the volatility of new energy sources to ensure that power supply stability can be maintained even under extreme conditions. The process of updating carbon emission limits involves complex emission accounting. The carbon emission coefficient of thermal power units depends on the fuel type and power generation efficiency. Coal-fired units produce about 0.8 kg of carbon dioxide per kilowatt-hour of electricity generated, while natural gas units produce about 0.4 kg. Although energy storage facilities themselves do not directly generate carbon emissions, energy losses occur during their charging and discharging processes, indirectly increasing the system's carbon emissions.

[0071] By calculating the carbon footprint of energy storage facilities throughout their entire lifecycle, including emissions during manufacturing, operation, and recycling, their environmental impact can be assessed more accurately. Regional carbon emission control targets are usually set by higher-level environmental protection departments, and emission quotas for various facilities need to be allocated reasonably under these constraints. Land use restrictions, as an important component of environmental constraints, directly affect the spatial layout feasibility of energy facilities. By overlaying and analyzing land use maps and urban planning maps, land parcels suitable for energy facility construction can be identified. Agricultural land, ecological protection zones, and areas surrounding residential areas have strict construction restrictions, while industrial land and some commercial land are relatively more lenient. These land constraints, together with carbon emission limits, constitute the boundary conditions for the layout of energy facilities, ensuring the coordination and unity between energy development and environmental protection.

[0072] S223. Based on the energy load density distribution data of each sub-region obtained during the information exchange process, identify the boundary locations where the load density difference between adjacent sub-regions exceeds a preset threshold.

[0073] Specifically, during the information exchange process, each intelligent agent node calculates the energy load density distribution data of each sub-region based on the exchanged data (including load demand, power generation capacity, proportion of new energy sources, etc.) and the updated energy facility requirements and environmental constraints.

[0074] First, we acquire peak electricity consumption data for industrial areas during working hours, trough electricity consumption data for residential areas during nighttime rest periods, and peak electricity consumption data for commercial areas during the cooling and heating seasons. Based on the electricity consumption characteristics of different types of areas, we establish a time-series electricity consumption pattern matrix that includes time series and power values, thereby obtaining a typical electricity consumption pattern dataset for various types of users.

[0075] Secondly, based on the power demand values ​​for each time period in the time-series electricity consumption pattern matrix, combined with the updated carbon emission limits and land use restrictions, the maximum electricity load that each sub-region can bear under environmental constraints is calculated. The upper limit of power supply is determined by the distribution capacity, and the peak-shaving capacity is determined by the energy storage layout. The upper limit of the actual power supply capacity of each sub-region for each time period is obtained by combining these factors.

[0076] Then, by comparing the upper limit of the actual power supply capacity with the demand value in the time-series electricity consumption pattern matrix, the supply and demand gap for each period is calculated. If the supply and demand gap for a certain period exceeds the preset threshold, the output of the new energy power generation facilities is adjusted according to the size of the gap. The change in the proportion of new energy power generation to total power generation before and after the adjustment is recorded to obtain the energy demand difference data after weighing environmental constraints.

[0077] Finally, using energy demand difference data, peak-valley difference in time dimension, regional type identifier in spatial dimension, and temperature correlation coefficient in seasonal dimension are extracted as feature vectors. Principal component analysis is used to extract the first three principal components as comprehensive features. The comprehensive feature value is divided by the corresponding regional area to obtain the energy load density distribution of each sub-region that reflects multi-dimensional features.

[0078] Based on this, the boundary locations where the load density difference between adjacent sub-regions exceeds a preset threshold are identified, and these locations are key candidate regions for boundary optimization.

[0079] Further example: Constructing a time-series electricity consumption pattern matrix is ​​fundamental to understanding the characteristics of regional energy demand.

[0080] Peak electricity consumption in industrial areas typically occurs during periods of concentrated production shifts, exhibiting a relatively stable rectangular wave pattern with power levels remaining at a high level. Residential areas, on the other hand, show a distinct bimodal characteristic, with two peak periods: morning wake-up and evening return home, while consumption drops to its lowest point late at night. The seasonal differences are particularly pronounced in commercial areas, with cooling loads accounting for over 40% of the total load in summer and heating demand relatively low in winter. This time-series characteristic matrix not only records power values ​​but also includes time stamps, forming a two-dimensional data structure.

[0081] Environmental constraints limit power supply capacity in multiple ways. Carbon emission limits directly restrict the operating hours and output of thermal power units. When regional carbon emissions approach the limit, thermal power output must be limited even if there is sufficient power generation capacity. Land use restrictions affect the location and scale of new energy facilities. For example, large-scale energy storage power stations cannot be built around ecological protection zones. Distribution capacity determines the physical transmission limit of the power grid, while energy storage layout provides time-dimensional adjustment capabilities. Together, they determine the upper limit of the actual available power supply.

[0082] Calculating the supply-demand gap requires comparative analysis over time periods. When an industrial zone demands 800 MW during peak production hours, but the maximum power supply capacity for that period is only 650 MW, a 150 MW gap occurs. At this point, the renewable energy output adjustment mechanism comes into play. By improving the power factor of photovoltaic inverters and adjusting wind turbine blade angles, renewable energy output is increased without exceeding the installed capacity. This adjustment changes the proportion of renewable energy in total power generation, increasing it from 25% to 35%. This change in proportion is recorded in detail as a crucial input for subsequent analysis.

[0083] Principal component analysis (PCA) plays a role in dimensionality reduction and information condensation in multi-dimensional feature extraction. The original feature vector contains dozens of dimensions, including power values ​​at 24 time points, four regional type identifiers, and temperature data for 12 months. By calculating the covariance matrix of these features, the directions with the largest variance are identified as principal components. The first principal component typically reflects the overall electricity consumption level, the second principal component reflects peak-valley differences, and the third principal component is related to seasonal variations. After extracting the first three principal components, the originally complex multi-dimensional data is compressed into three comprehensive indicators, which retains the main information and facilitates subsequent processing. The final calculation of energy load density distribution integrates both spatial and energy dimensions. The comprehensive feature value obtained from PCA is used as the numerator, representing the energy demand intensity of the region; the region area is used as the denominator, achieving comparability between regions of different sizes. For example, although a commercial center has a large total electricity consumption, its load density may be lower than that of a data center with a smaller footprint but concentrated electricity consumption due to its large land area.

[0084] S224. Adjust the coordinates of the boundary positions and calculate the load density variance of each sub-region after adjustment.

[0085] Specifically, for the candidate boundary locations identified by S223, adjustments are made by moving the boundary coordinates point by point. The boundary optimization process for energy load density distribution is essentially about finding the optimal way to divide regions, ensuring that the energy characteristics within each region are as consistent as possible, while the differences between regions are as pronounced as possible. When the load densities of two adjacent regions differ significantly—for example, one side is a high-energy-consuming industrial area and the other a low-energy-consuming residential area—the original administrative boundary may not be suitable as a dividing line for energy management, requiring optimization through point-by-point boundary movement.

[0086] Each time the boundary is moved, the load density of the sub-regions on both sides is recalculated, and the variance of the load density within each sub-region (i.e., the mean of the sum of squares of the differences between the load density at each point within the region and the average load density of the region) is calculated. The variance value after each adjustment is recorded as an indicator for evaluating the rationality of the boundary. The point-by-point boundary moving method can find the optimal boundary position that minimizes the differences within the region by calculating the variance of the load density after each small adjustment. This goal of minimizing variance ensures the homogeneity within each region after division.

[0087] During boundary adjustments, constraints related to regional area and total load must also be considered. Decisions regarding regional merging and re-division are based on clear quantitative criteria. Based on a preset minimum area threshold, areas that are too small are merged with adjacent areas. For example, when an area is less than 0.5 square kilometers and has similar energy characteristics to adjacent areas, merging them can simplify the management structure and avoid increasing management complexity due to excessively small areas without bringing corresponding benefits.

[0088] Based on a preset maximum load capacity threshold, areas with excessive load are further subdivided. For example, when the total load of an industrial park approaches the substation's capacity limit, it needs to be divided into multiple sub-regions, each supplied by a different power line, ensuring that the energy demand of a single region does not exceed the carrying capacity of the power distribution facilities. Through the above merging and subdivision operations, it is ensured that the adjusted regions meet management accuracy requirements and facility carrying capacity.

[0089] An optimized geographical boundary delineation scheme was adopted to calculate the difference between energy supply and demand power in each sub-region at different time periods. The calculation of the supply-demand balance index fully considers the dynamic characteristics of the time dimension: during the day when photovoltaic power generation is sufficient, some regions may experience a situation of supply exceeding demand; while at night when photovoltaic power generation stops, the same region may experience a situation of supply falling short of demand. By calculating the absolute value of the supply-demand difference for each time period within the day and then taking the time average, the degree of supply-demand matching in the region can be comprehensively reflected. This average difference is normalized by dividing it by the average demand power, making the balance index of regions of different sizes comparable, thus obtaining the supply-demand balance index of each sub-region.

[0090] The assessment of resource allocation balance employs a comparison method between adjacent regions. If one adjacent region consistently experiences energy surplus while the other experiences energy shortage, it indicates an unreasonable regional division or insufficient energy transmission channels between regions. By calculating the sum of squared differences in balance indices between all pairs of adjacent regions, the degree of imbalance in the entire system can be quantified. If the sum of squared differences is less than a preset balance threshold, the resource allocation balance requirement is considered met. When this sum of squared differences is small, it indicates a relatively balanced supply and demand situation in each region, with energy resources rationally allocated. This balance not only improves energy utilization efficiency but also enhances the stability and reliability of the system.

[0091] S225. Select the boundary coordinates that minimize the load density variance as the adjusted spatial division boundary.

[0092] Specifically, for all the boundary coordinate sets generated by the trial adjustments, the corresponding load density variance values ​​are compared, and the boundary coordinates that minimize the variance are selected as the final adjustment result for that location. This optimization objective ensures that the energy characteristics within each sub-region are as consistent as possible after adjustment, while maximizing the differences between regions, thereby improving the accuracy of subsequent energy planning.

[0093] The real-time nature of boundary adjustment is reflected in its rapid response to dynamic changes. With urban development and industrial transformation, the energy characteristics of certain areas will change significantly. For example, a former commercial area may be transformed into a high-energy-consuming area due to the construction of a large data center. When changes in energy characteristics are identified, this step can respond in a timely manner and reselect the optimal boundary.

[0094] After adjusting all candidate boundaries, the updated spatial partition boundary is obtained.

[0095] S23. Based on the adjusted boundary coordination, generate an energy coordination allocation strategy between regions.

[0096] Specifically, after the spatial boundary adjustment is completed, each intelligent agent node reorganizes its own energy supply and demand data based on the new boundary, and generates an inter-regional energy collaborative allocation strategy through a multi-agent distributed decision-making algorithm.

[0097] The construction of intelligent agent nodes is based on the core concept of distributed system architecture, where each sub-region is abstracted as an intelligent entity with autonomous decision-making capabilities. The intelligent agent internally stores a complete energy profile of that sub-region, including multi-dimensional data such as historical electricity load curves, renewable energy power generation output records, and equipment capacity parameters.

[0098] This data is not stored statically, but is continuously updated through a real-time monitoring system to ensure that the decisions of the agent nodes are based on the latest information. The message passing mechanism enables agents in neighboring areas to exchange key information. For example, an agent in an industrial area can send information about surplus electricity at night to an agent in a neighboring residential area, while the residential agent can report the electricity demand gap during morning and evening peak hours.

[0099] Each intelligent agent node uses an autoregressive moving average model to predict the proportion of new energy in each sub-region for a predetermined period, based on the exchanged energy flow information and historical proportion data. The application of the autoregressive moving average model in predicting the proportion of new energy fully considers the temporal characteristics of historical data. By analyzing the changing patterns of the proportion of new energy over the past few years, the model identifies the trend components, seasonal components, and random fluctuation components.

[0100] For example, the proportion of photovoltaic power generation is significantly higher in summer than in winter. This seasonal pattern is captured by the model and used for future predictions. The model can also identify policy-driven trend changes, such as the accelerated growth of the proportion after the implementation of new energy subsidy policies. By comprehensively considering these complex factors and arranging the predicted proportion values ​​by time, a trend mapping data of the proportion of new energy reflecting long-term development trends is formed, and the resulting predictions are closer to the actual development trajectory.

[0101] By using the trend mapping data of the proportion of new energy sources to identify sub-regions whose growth rate exceeds a preset threshold, the existing capacity and load growth forecast of the distribution network in these sub-regions are obtained, and the required expansion capacity of the distribution network is calculated.

[0102] Meanwhile, the determination of energy storage capacity requirements employed statistical methods: the standard deviation of renewable energy power generation reflects the degree of instability in its output; a larger standard deviation indicates more severe fluctuations. Dividing the standard deviation by the average power yields the coefficient of variation, a dimensionless indicator that objectively measures the relative degree of fluctuation. When the coefficient of variation of photovoltaic power generation in a certain region reaches 0.6, it means that its output fluctuations are significant, requiring the configuration of energy storage facilities with corresponding capacity to mitigate these fluctuations. The selection of energy storage locations comprehensively considers the grid structure, load center location, and land availability, prioritizing locations near substations that meet construction conditions.

[0103] Using data on power distribution capacity expansion needs and energy storage deployment locations, each agent generates an initial allocation plan based on local constraints. A multi-round negotiation mechanism demonstrates the advantages of distributed decision-making: In the initial stage, each agent generates a preliminary plan based on local information. For example, if a region plans to purchase 200 MW of electricity from a neighboring region during off-peak hours, this plan is sent to relevant agents via message passing. The receiving agents evaluate the plan based on their own constraints. If the power supply region does have surplus capacity during that period, the request is accepted; otherwise, modification suggestions are proposed, such as adjusting the purchase period or reducing the purchase volume. Through multiple rounds of such information exchange and plan adjustments, the needs and constraints of all parties gradually reach a balance, ultimately forming a collaborative plan acceptable to all parties.

[0104] The formation of an energy collaborative allocation strategy is a dynamic optimization process. Through multiple rounds of information exchange and scheme adjustments, a convergent inter-regional energy collaborative allocation strategy is obtained, incorporating cross-regional energy dispatch timing, transmission capacity constraints, and energy storage charging and discharging plans. The cross-regional energy dispatching sequence clarifies the direction of power flow at different times. For example, during the day when photovoltaic power generation is sufficient, power is transmitted from photovoltaic-rich areas to industrial areas, while at night it may flow in the opposite direction. The transmission capacity limit ensures that the dispatching scheme does not exceed the physical limit of the power grid. The energy storage charging and discharging plan provides a buffer in the time dimension, charging and storing when supply exceeds demand, and discharging to supplement when supply falls short of demand.

[0105] This multi-dimensional collaborative mechanism enhances the flexibility and reliability of the entire regional energy system, enabling efficient collaboration of energy resources across regions while meeting all constraints.

[0106] The process of obtaining performance feedback data after implementing the energy collaborative allocation strategy in step S3, and iteratively optimizing the spatial division range and energy collaborative allocation strategy based on the performance feedback data to determine the final resource allocation adjustment scheme, includes the following steps: S31. Based on the adjusted spatial division boundary, generate an updated energy collaborative allocation strategy and execute it.

[0107] Specifically, after adjusting the spatial boundary, each agent node reorganizes its energy supply and demand data based on the new boundary. Agent nodes representing each sub-region are constructed, with each agent storing energy supply and demand data, renewable energy installed capacity, and historical percentage changes for that sub-region. Through a message passing mechanism, agents exchange energy flow information and load forecast data between regions.

[0108] Based on the energy flow information and historical proportion data exchanged by each intelligent agent, an autoregressive moving average model is used to predict the proportion of new energy in each sub-region for a future preset period. The predicted proportion values ​​are arranged by time to form a new energy proportion trend mapping data that reflects the long-term development trend.

[0109] By mapping the trend of new energy proportion, sub-regions whose proportion growth rate exceeds a preset threshold are identified. The existing capacity and load growth forecast of the distribution network in these sub-regions are obtained, and the required expansion capacity of the distribution network is calculated. At the same time, the energy storage capacity demand and the optimal layout location are determined based on the ratio of the standard deviation to the average value of new energy power generation.

[0110] Using data on power distribution capacity expansion needs and energy storage location data, each agent generates an initial allocation scheme based on local constraints. After multiple rounds of information exchange and scheme adjustment, a regional energy collaborative allocation strategy is obtained, which includes cross-regional energy dispatch timing, transmission capacity limitations, and energy storage charging and discharging plans. This strategy is then executed.

[0111] S32. Obtain performance feedback data after implementing the updated energy collaborative allocation strategy.

[0112] Specifically, this involves obtaining energy efficiency assessment data for each sub-region after implementing the energy collaborative allocation strategy. Obtaining this energy efficiency assessment data involves the comprehensive calculation of several key indicators: Energy transmission loss rate reflects the energy loss during electricity transmission. When current passes through a transmission line, some electrical energy is converted into heat and dissipated due to the resistance of the conductor. For example, a 10-kilometer-long transmission line may experience a 20-megawatt line loss when transmitting 1000 megawatts of power, with a loss rate of 2%. This indicator directly affects the economics of the energy system; a higher loss rate means more energy waste.

[0113] Calculating the renewable energy absorption rate is more complex. It not only considers the amount of renewable energy generated but also analyzes whether this electricity is effectively utilized. During certain periods, wind and solar power may generate a large amount of electricity, but if local load is insufficient and it cannot be transmitted to other regions in a timely manner, the phenomenon of "wind and solar curtailment" occurs. By calculating the portion of renewable energy generated that is actually absorbed and dividing it by the total generated electricity, the absorption rate is obtained. When the solar photovoltaic absorption rate in a certain region is only 70%, it indicates that 30% of clean energy is wasted, which requires adjustments to resource allocation to improve the situation.

[0114] Dynamic monitoring of spatial coupling strength can promptly detect changes in energy dependence between regions. For example, an industrial zone that was originally energy-sufficient may experience a significant increase in coupling strength due to the establishment of several energy-intensive enterprises that frequently purchase electricity from surrounding areas. By comparing the coupling strength values ​​before and after implementing a collaborative allocation strategy, the effectiveness of the strategy can be assessed. If the coupling strength actually increases, it indicates that the current resource allocation may have exacerbated the imbalance between regions, requiring further adjustments.

[0115] At the same time, it is necessary to identify the critical state of environmental constraints: carbon emissions approaching the limit means that there is very little room for fossil energy use in the region, and the process of replacing energy sources must be accelerated; insufficient available land for construction limits the possibility of adding new energy facilities, and it is necessary to improve efficiency by optimizing the layout of existing facilities.

[0116] S33. Based on the performance feedback data, iteratively adjust the adjusted spatial division boundary and the updated energy collaborative allocation strategy until the preset optimization target is met, and obtain the final resource allocation adjustment scheme.

[0117] Based on the performance feedback data obtained from S32 (including energy transmission loss rate, renewable energy absorption rate, updated spatial coupling strength, energy facility relocation needs, and load shifting needs), the adjusted spatial boundary delineation and the updated energy collaborative allocation strategy are evaluated. The evaluation process proceeds in the following hierarchical manner: First, determine whether the supply and demand balance of each sub-region is less than a preset threshold, and whether the sum of the squares of the differences in the supply and demand balance indices between all adjacent regions is less than a preset equilibrium threshold. If both conditions are met, the optimization objective is considered achieved, the iteration terminates, and the current solution becomes the final resource allocation adjustment solution.

[0118] If the core equilibrium objective is not achieved, further evaluation is conducted to determine if there is room for optimization in the current performance indicators. Specifically, this includes: assessing whether the energy transmission loss rate is higher than the preset standard (if higher, there is room to reduce transmission loss by adjusting the boundaries); assessing whether the renewable energy absorption rate is lower than the preset standard (if lower, there is room to increase the absorption rate by adjusting the strategy); and assessing whether the change in spatial coupling strength exceeds the preset threshold (if it does, the inter-regional dependency has changed, and the boundaries need to be re-optimized). If there is no room for optimization in any of the above three performance indicators (i.e., the loss rate is not higher than the standard, the absorption rate is not lower than the standard, and the change in coupling strength does not exceed the threshold), then even if the core equilibrium objective is not achieved, further adjustments will not improve system performance. In this case, the iteration is terminated, and the current solution is taken as the final resource allocation adjustment solution.

[0119] If there is room for efficiency optimization, the next step is to determine whether further adjustments are supported under the current constraints. Specifically, this includes determining whether the carbon emission balance is positive (i.e., the current cumulative carbon emissions have not exceeded the limit, allowing new facilities to operate) and whether the available land area meets the layout requirements for new energy facilities (such as energy storage power stations and power distribution facilities). If the carbon emission balance is positive and there is sufficient available land, it is determined that iterative adjustments can continue; if the carbon emission balance is negative or there is insufficient available land, the physical constraints prevent further adjustments, at which point the iteration is terminated, and the current solution is taken as the final resource allocation adjustment solution.

[0120] If it is determined that iterative adjustments can continue, the following operations will be performed: First, by adjusting the geographical boundaries of relevant areas to centralize the management of high-energy-consuming facilities based on the demand for energy facility relocation and load transfer, the locations of energy storage and power distribution facilities will be reconfigured to reduce transmission losses.

[0121] Secondly, based on the revised boundary update energy coordination and allocation strategy, the energy dispatching sequence between regions is modified to improve the renewable energy absorption rate, and the energy storage charging and discharging plan is adjusted to adapt to the new load distribution.

[0122] Then, return to step S32 to obtain performance feedback data on the adjusted strategy, and re-enter the evaluation process of this step.

[0123] This process is iterated repeatedly until the core equilibrium objective of the first level is met, or if the core equilibrium objective is not met but the second-level judgment shows no room for optimization, or the third-level judgment shows that the constraints are not met, then the iteration terminates, resulting in the final resource allocation adjustment scheme. This scheme includes the final determined spatial division boundaries, energy facility layout (expansion of power distribution capacity, location of energy storage facilities), and inter-regional energy collaborative allocation strategies (cross-regional energy dispatch timing, transmission capacity limits, energy storage charging and discharging plans).

[0124] Example 3 is an embodiment of the present invention, which provides a regional energy system planning and monitoring system based on intelligent agent collaboration, including: a spatial division module, an energy collaborative allocation strategy generation module, and an output module; The spatial division module acquires the original geographical boundary data and energy load data of the target area, and determines the initial spatial division range based on the data. The initial spatial division range includes several sub-regions. The module for generating energy collaborative allocation strategy constructs intelligent agent nodes corresponding to each sub-region, and each intelligent agent node stores the energy supply and demand data of the corresponding sub-region; through information interaction between the intelligent agent nodes, the spatial division boundary between several sub-regions is dynamically adjusted, and an energy collaborative allocation strategy between regions is generated based on the adjusted boundary. The output module acquires performance feedback data after executing the energy collaborative allocation strategy, iteratively optimizes the spatial division range and energy collaborative allocation strategy based on the performance feedback data, and determines the final resource allocation adjustment scheme.

[0125] This embodiment also provides an electronic device applicable to a regional energy system planning and monitoring method based on intelligent agent collaboration, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to realize the regional energy system planning and monitoring method based on intelligent agent collaboration as proposed in the above embodiment.

[0126] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements a regional energy system planning and monitoring method based on intelligent agent collaboration as proposed in the above embodiments.

[0127] The storage medium proposed in this embodiment belongs to the same inventive concept as the method for planning and monitoring regional energy systems based on intelligent agent collaboration proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0128] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0129] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A regional energy system planning and monitoring method based on agent-based collaboration, characterized in that: include, Obtain the original geographic boundary data and energy load data of the target area, and determine the initial spatial division range based on the data. The initial spatial division range includes several sub-regions. Construct intelligent agent nodes corresponding to each sub-region, and each intelligent agent node stores the energy supply and demand data of the corresponding sub-region; through information interaction between intelligent agent nodes, dynamically adjust the spatial division boundary between several sub-regions, and collaboratively generate an energy collaborative allocation strategy between regions based on the adjusted boundary. Obtain performance feedback data after implementing the energy collaborative allocation strategy, and iteratively optimize the spatial division range and energy collaborative allocation strategy based on the performance feedback data to determine the final resource allocation adjustment plan.

2. The regional energy system planning and monitoring method based on agent collaboration as described in claim 1, characterized in that: Determining the initial spatial division range includes, Acquire time-series electricity consumption data, industry distribution data, and traffic flow data for each sub-region; Based on time-series electricity consumption data, industry distribution data, and traffic flow data, the geographical boundaries of the initial spatial division range are refined and adjusted to obtain a refined spatial division granularity scheme.

3. The regional energy system planning and monitoring method based on agent collaboration as described in claim 2, characterized in that: The information interaction between the various intelligent agent nodes includes Each agent node exchanges the information required for decision-making with at least one neighboring agent node based on the stored power generation capacity data and load demand data of the corresponding sub-region; Each intelligent agent node generates a local scheduling plan based on locally stored data and received information, and iteratively adjusts the scheduling plan through multiple rounds of information interaction until the coordination conditions are met.

4. The regional energy system planning and monitoring method based on agent collaboration as described in claim 3, characterized in that: The preset coordination conditions include, There are no conflicts in power flow or capacity between the scheduling plans generated by each intelligent agent node; The local scheduling plans of each intelligent agent node all satisfy the corresponding local environment constraints and facility capacity constraints.

5. The regional energy system planning and monitoring method based on agent collaboration as described in claim 4, characterized in that: The dynamic adjustment of the spatial boundaries between several sub-regions includes... Based on the energy load density distribution data of each sub-region obtained during the information exchange process, identify the boundary locations where the load density difference between adjacent sub-regions exceeds a preset threshold. Adjust the coordinates of the boundary positions and calculate the load density variance of each sub-region after adjustment; The boundary coordinates that minimize the load density variance are selected as the adjusted spatial partition boundaries.

6. The regional energy system planning and monitoring method based on agent collaboration as described in claim 5, characterized in that: The step of obtaining performance feedback data after implementing the energy collaborative allocation strategy, and iteratively optimizing the spatial division range and energy collaborative allocation strategy based on the performance feedback data, includes... Based on the adjusted spatial delineation boundaries, an updated energy collaborative allocation strategy is generated and executed. Obtain performance feedback data after implementing the updated energy collaborative allocation strategy; Based on performance feedback data, the adjusted spatial boundary and updated energy collaborative allocation strategy are iteratively adjusted until the preset optimization target is met, resulting in the final resource allocation adjustment scheme.

7. The regional energy system planning and monitoring method based on agent collaboration as described in claim 6, characterized in that: The iterative adjustments include, Based on performance feedback data, the adjusted spatial division boundary and the updated energy collaborative allocation strategy are evaluated; If the evaluation results do not meet the preset optimization objectives, the spatial division boundary is adjusted again, and the energy collaborative allocation strategy is updated according to the adjusted boundary. Then, the process returns to the step of obtaining performance feedback data.

8. A regional energy system planning and monitoring system based on intelligent agent collaboration, employing the regional energy system planning and monitoring method based on intelligent agent collaboration as described in any one of claims 1 to 7, characterized in that, include: The module includes a spatial partitioning module, a module for generating energy collaborative allocation strategies, and an output module. The spatial division module acquires the original geographical boundary data and energy load data of the target area, and determines the initial spatial division range based on the data. The initial spatial division range includes several sub-regions. The module for generating energy collaborative allocation strategy constructs intelligent agent nodes corresponding to each sub-region, and each intelligent agent node stores the energy supply and demand data of the corresponding sub-region; through information interaction between the intelligent agent nodes, the spatial division boundary between several sub-regions is dynamically adjusted, and an energy collaborative allocation strategy between regions is generated based on the adjusted boundary. The output module acquires performance feedback data after executing the energy collaborative allocation strategy, iteratively optimizes the spatial division range and energy collaborative allocation strategy based on the performance feedback data, and determines the final resource allocation adjustment scheme.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the regional energy system planning and monitoring method based on intelligent agent collaboration as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the regional energy system planning and monitoring method based on intelligent agent collaboration as described in any one of claims 1 to 7.

Images