Optimal configuration method and device for light-storage-vehicle-diesel multi-source coordination for terminal power supply

Abstract

The application discloses a light-storage-vehicle-diesel multi-source collaborative optimization configuration method and device for terminal power supply, and belongs to the technical field of novel power distribution network planning. The application constructs a distributed photovoltaic, energy storage, electric vehicle, diesel generator and hierarchical load coupling mathematical model and an island operation constraint system; a three-dimensional weighted comprehensive optimization function is constructed with economy, system adjustability and hierarchical load guarantee rate as sub-indicators; an improved genetic algorithm with an embedded gradient descent local search is adopted for solution, and multi-source optimal capacity configuration results are output; performance checking is carried out in combination with island working conditions, power supply guarantee indexes of zero power loss of primary load and controllable power supply of secondary load are verified, and new energy utilization efficiency and power supply quality are comprehensively evaluated. The application realizes a multi-element power supply guarantee pattern of "normal efficient consumption and reliable bottom in fault", and significantly improves the power supply resilience and emergency power supply capacity of mountain terminal areas.

Description

Technical Field

[0001] This invention relates to the field of power system distribution network planning technology, specifically a method and device for multi-source collaborative optimization configuration of photovoltaic-storage-vehicle-diesel for end-point power supply. Background Technology

[0002] With the implementation of dual-carbon targets and the advancement of new power system construction, the penetration rate of distributed renewable energy in end-point distribution networks continues to increase. However, ensuring continuous power supply to end-point distribution networks, especially in remote county-level areas, has become a severe challenge for current distribution network construction and operation. Due to the complex terrain and weak grid structure in mountainous areas of Hubei, end-point distribution areas generally suffer from long power supply radii, high line maintenance difficulty, and weak disaster resistance. During frequent extreme disasters such as rain, snow, ice, and landslides, pole collapses and line breaks are highly likely, leading to prolonged power outages in the main power grid. Existing end-point power supply models lack targeted emergency power supply mechanisms, making it difficult to maintain the operation of critical livelihood and industrial loads after disconnection from the main grid, resulting in a persistently high risk of "power outage leading to paralysis." With the large-scale development of distributed renewable energy and the mature application of microgrid technology, the photovoltaic-storage synergy and multi-source complementary microgrid power supply model provides a new technical path for improving the power supply capacity of end-point distribution areas and achieving emergency power supply in fault conditions. Currently, the installed capacity of distributed photovoltaic (PV) power in Hubei Province has exceeded 22 million kilowatts. The high penetration rate of new energy grid connection provides a sufficient clean power foundation for end-point microgrids, but it also brings new operational challenges such as reverse power transmission from PV, random load fluctuations, and voltage exceeding limits. In particular, the power grid architecture in remote mountainous areas of Hubei is weak, and the emergency repair cycle for extreme weather faults is long. How to optimize the allocation of multi-source resources in microgrids and balance normal operating efficiency with emergency power supply capabilities has become an urgent engineering requirement for the upgrading of county-level distribution networks.

[0003] To improve the operational performance of end-point microgrids, extensive research on resource optimization and allocation has been conducted both domestically and internationally. Domestic mainstream solutions primarily focus on optimizing resource combinations such as photovoltaic (PV) + energy storage and PV + electric vehicles. Related research and applications also largely revolve around the capacity configuration and operational control of single or two types of distributed resources. However, as the penetration rate of PV within provinces continues to rise, the limitations of the traditional dual-resource allocation model are becoming increasingly apparent. It struggles to adapt to complex grid conditions and differentiated load supply demands, failing to meet the high-quality, high-supply-guarantee operational requirements of end-point microgrids.

[0004] Currently, mainstream microgrid multi-source configuration technologies generally suffer from shortcomings in three aspects during engineering applications: resource coordination, load management, and optimization models. These shortcomings limit the accuracy of resource configuration and the power supply guarantee capability of the end-point microgrids. Specific problems include:

[0005] (1) The degree of multi-resource synergy is low, and the potential for resource complementarity has not been fully explored. Most configuration technologies only use diesel generators as emergency equipment after a failure and do not incorporate them into the early planning and coordination system. At the same time, very few systems utilize the flexible energy storage regulation capabilities of electric vehicle clusters. Only local optimization is carried out for photovoltaics and energy storage, and a global coordinated control system for the four types of distributed resources has not been formed. The temporal complementarity and spatial distribution advantages of various resources cannot be effectively integrated. In the face of extreme conditions such as large power grid failures, sudden drops in photovoltaic output, and sudden increases in load, the system's regulation margin and backup support capabilities are insufficient, making it difficult to resist the power grid operation risks caused by extreme weather in mountainous areas, and power outages are likely to occur.

[0006] (2) The load management model is crude and lacks differentiated supply guarantee logic. Existing optimization allocation methods generally use a uniform load weight for calculation, without distinguishing the livelihood guarantee attributes and economic value of different loads, and there is no hierarchical power supply constraint mechanism. When the power grid supply and demand are tight and power curtailment and load shedding operations need to be carried out, there is a lack of scientific basis for hierarchical load shedding, which easily leads to the problem of simultaneous power outage of critical guarantee loads and ordinary loads, which can easily cause adverse consequences such as interruption of important livelihood guarantee and industrial economic losses. It cannot adapt to the actual operation and maintenance needs of the mountainous areas of Hubei Province where critical loads are dispersed and supply guarantee priorities are differentiated.

[0007] (3) The optimization target dimension is too singular, and the dynamic adaptability to operating conditions is insufficient. Current configuration models mostly use the total life cycle cost and basic power supply reliability as core optimization indicators, without incorporating the dynamic adjustability of the system into an independent optimization dimension. The optimization design focuses on static economic and steady-state power supply indicators, neglecting the dynamic adjustment margin of the system in response to new energy fluctuations and load disturbances, resulting in configuration schemes generally having the problem of "emphasizing economy and neglecting adjustment" and insufficient supply margin. In high-penetration photovoltaic access scenarios, problems such as high curtailment rate and weak system anti-disturbance capability are likely to occur, which is not compatible with the development plan of large-scale development of regional new energy and improvement of grid quality and efficiency.

[0008] In summary, current microgrid configuration technologies generally suffer from problems such as insufficient adaptability to operating conditions, inefficient resource utilization, lack of precision in power supply reliability management, and one-sided optimization systems. They are unable to simultaneously meet the needs of efficient consumption of new energy and emergency power supply in case of faults. The industry urgently needs a configuration scheme that can achieve global coordination of multiple resources, precise supply of graded loads, and balanced optimization of multi-dimensional performance in order to improve the operational resilience and resource utilization efficiency of end-point microgrids. Summary of the Invention

[0009] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and device for multi-source coordinated optimization configuration of photovoltaic-storage-vehicle-diesel power supply for end-point power supply. By constructing a multi-source coordinated end-point power supply system, it focuses on solving the problem of continuous power supply in mountainous power grids under extreme disasters. While improving the absorption capacity of new energy sources, it significantly enhances the absolute power supply capacity for primary critical loads (such as hospitals and emergency stations) and the emergency support capacity for secondary loads, achieving a multi-dimensional optimal balance between economic efficiency and the effectiveness of power supply resources.

[0010] To achieve the above technical objectives, the present invention adopts the following technical solution:

[0011] A method for coordinated optimization of multi-source power supply (optical, energy storage, vehicle, and diesel) for end-point power supply includes the following steps:

[0012] S1: Construct a mathematical model for coupling multiple power sources and hierarchical loads in the terminal microgrid, as well as a constraint system for islanded operation conditions. Specifically, this includes: establishing a mathematical model for the stochastic output of distributed photovoltaic power, an energy storage operation constraint model, a time-series charging and discharging model for electric vehicle clusters, a diesel generator output-fuel consumption characteristic model, and a three-level load quantification constraint model based on hierarchical load weights; at the same time, construct a set of islanded operation constraint conditions, including power outage duration constraints and an emergency power supply load screening mechanism based on load importance, and output multi-source equipment characteristic parameters and islanded operation condition boundary conditions.

[0013] S2: Based on the multi-source equipment characteristic parameters and islanding boundary conditions output in step S1, a three-dimensional weighted comprehensive optimization function is constructed with economic efficiency, system adjustability, and graded load guarantee rate as optimization sub-indicators. It is also equipped with system power balance constraints, equipment operation constraints, graded power supply control constraints, and node voltage constraints to form a multi-objective optimization model.

[0014] S3: An improved genetic algorithm with embedded gradient descent local search is used to iteratively solve the multi-objective optimization model constructed in step S2, and output the optimal configuration results of photovoltaic installed capacity, energy storage rated power and capacity, electric vehicle aggregate adjustable capacity, and diesel generator installed capacity.

[0015] S4: Based on the optimal configuration results output in step S3, performance verification is performed in conjunction with the islanding condition during abnormal power outages. Emergency supply capability is then verified in reverse based on the ratio between the guaranteed power supply load and the maximum load of the distribution area. This verifies the supply guarantee indicators of zero power loss for primary loads, controllable power supply for secondary loads, and on-demand disconnection of tertiary residential loads. This also assists in calculating the photovoltaic absorption rate. and power supply reliability of loads at all levels The system comprehensively evaluates the new energy utilization efficiency and graded power supply quality of the configuration scheme, and outputs the final verified configuration scheme.

[0016] Furthermore, the three-level load quantification constraint model described in step S1 divides the load into primary critical loads, secondary industrial and commercial loads, and tertiary residential loads according to load importance, and assigns importance weights to each. ,satisfy ;

[0017] The three-level load quantification constraint model sets the following: Level 1 critical loads are subject to a hard zero-power-out constraint; Level 2 industrial and commercial loads are allowed a very small proportion of short-term power outages to adapt to grid fluctuations; and Level 3 residential loads are treated as switchable loads under islanding conditions to release power supply capacity for core power supply protection. The core equations are as follows:

[0018] In the formula: The permissible maximum power loss for Level 3 loads is listed in descending order; the permissible maximum power loss ratio for Level 2 loads is listed in descending order. Maximum permissible power loss ratio for Level 3 loads ;

[0019] The islanded operation constraint set includes power outage duration constraints and emergency power supply load screening formulas. The formula for the power outage duration constraint is as follows:

[0020] ; In the formula: This refers to the actual duration of the abnormal power outage; To preset the maximum fault repair time, For the output of distributed photovoltaic power during power outages, These are energy storage during power outages and the discharge output of electric vehicles, respectively. Provide power for the diesel engine during power outages; The total power supply load required to be guaranteed under island conditions; The total rated power for the first-level critical load is guaranteed to be fully supplied under abnormal operating conditions; The rated power is set for the total load of secondary industrial and commercial vehicles; This is the secondary load guarantee factor.

[0021] Furthermore, in step S1,

[0022] The mathematical model for distributed photovoltaic stochastic output adopts the Beta probability distribution model to quantify the temporal fluctuation characteristics of photovoltaic stochastic output. The core calculation formula is as follows: ;

[0023] ;

[0024] In the formula: Provide real-time power for photovoltaic systems; This refers to the rated installed capacity of photovoltaic power. The shape coefficient of the Beta distribution is obtained by fitting local measured irradiation data; For beta functions; It is a gamma function;

[0025] The energy storage operation constraint model includes a time-series energy iterative equation, upper and lower limits of state of charge constraints, charging and discharging power limit constraints, and charging and discharging mutual exclusion constraints. The core formulas are as follows: ;

[0026] ;

[0027] ;

[0028] In the formula: for Constant energy storage state of charge; These are the energy storage charging and discharging power, respectively. These refer to the energy storage charging and discharging efficiencies, respectively. These are the rated energy storage capacity and power, respectively.

[0029] The electric vehicle cluster time-series charging and discharging model fits the individual vehicle's power consumption and remaining power based on the log-normal distribution of driving mileage, and aggregates them to obtain the cluster's time-sharing charging and discharging power. The core formula is as follows: ;

[0030] ;

[0031] ;

[0032] ; ; ;

[0033] In the formula: This refers to the daily mileage traveled by a single vehicle. These are the logarithmic mean and logarithmic standard deviation of mileage, respectively. Electricity consumption per unit distance of vehicle; Rated capacity of the vehicle's power battery; Let m be the rated discharge power of the m-th vehicle; This represents the lowest state of charge for electric vehicles, ensuring a minimum amount of power available for users' travel needs. They are respectively A fleet of vehicles that are constantly charging and discharging; For the first The time-sharing charging and discharging power of the vehicles; The total charging and discharging power of the electric vehicle cluster during time-sharing;

[0034] The diesel generator output-fuel consumption characteristic model adopts a quadratic function. Describe fuel consumption and set minimum stable output constraints. In the formula: This refers to the fuel consumption per unit time of the generator unit. These are the characteristic coefficients for fitting diesel engine fuel consumption; To provide real-time active power output for diesel generators; This refers to the rated maximum output of the diesel generator. This is the minimum stable operating output of the diesel generator, used to constrain the unit's low-load shutdown conditions.

[0035] Furthermore, the three-dimensional weighted synthesis optimization function mentioned in step S2 is: ;

[0036] ;

[0037] ;

[0038] ;

[0039] In the formula: F is the total value of the system's overall optimization objective, derived from economic efficiency. Adjustability Supply success rate The three sub-objectives are weighted together, and the optimization objective is to minimize the value. These are weighted coefficients for economic efficiency, adjustability, and supply success rate, respectively. These are the average annual investment and maintenance costs for photovoltaic, energy storage, electric vehicle supporting facilities, and diesel generators, respectively. To reduce the cost of purchasing electricity from the higher-level power grid, Revenue from selling surplus electricity online; This is the overall adjustable margin coefficient of the system.

[0040] Furthermore, the solution process of the improved genetic algorithm described in step S3 includes:

[0041] S31: Constructing individual chromosomes using real-number encoding. The initial population is randomly generated, corresponding to photovoltaic installed capacity, rated energy storage power, rated energy storage capacity, adjustable aggregate capacity of electric vehicles, and installed capacity of diesel generators, respectively.

[0042] S32: Using the fitness function Evaluate the merits and demerits of individuals, where F is the comprehensive optimization target value described in step S2, and ε is a minimal constant to avoid a denominator of zero;

[0043] S33: Adaptive crossover and mutation operations are performed using crossover and mutation probabilities that dynamically decrease with the number of iterations. The formula is as follows: In the formula: These represent the initial crossover probability and the initial mutation probability of the algorithm, respectively; t is the current iteration number. This represents the maximum number of iterations.

[0044] S34: Select elite individuals, perform gradient descent local search on them, and update the formula as follows: In the formula: , These are the individual configuration solutions before and after the update, respectively; This is the gradient learning step size; The gradient vector of a multi-objective function;

[0045] S35: Repeatedly select, crossover, mutate, and gradient optimization until the number of iterations reaches the maximum iteration threshold or the objective function tends to converge, and finally output the optimal configuration result of the light-storage-vehicle-diesel capacity.

[0046] Furthermore, in step S4, an islanded power supply capacity verification model is set up. Based on the abnormal power outage duration and the scale of the power supply load, the power supply pressure is quantified through the ratio of the power supply load to the maximum load of the distribution area, and the emergency adaptability of the multi-source configuration capacity is verified in reverse. The calculation formula is as follows:

[0047] ;

[0048] ;

[0049] In the formula: To ensure the proportion of power supply load, this is used to quantify the emergency power supply pressure; This represents the maximum total load of the transformer substation connected to the grid. To maximize the combined output of photovoltaic, energy storage, vehicle, and diesel power sources; These are the effective outputs of photovoltaics, energy storage, electric vehicle clusters, and diesel generators, respectively. To ensure the total power supply load required under isolated operating conditions, the capacity verification for abnormal operating conditions is completed by matching the total output of multiple sources with the power supply load.

[0050] The performance of the configuration scheme is evaluated using multiple quantitative indicators. The core evaluation formula is as follows:

[0051] In the formula: Photovoltaic grid integration rate, used to characterize the utilization rate of new energy sources. To absorb the photovoltaic power output, Contribute to the overall photovoltaic power generation; The load guarantee rates are categorized into three levels: Level 1, Level 2, and Level 3. For the power shortage of loads at all levels, Power for each load level; For load level indexing.

[0052] A multi-source collaborative optimization configuration device for end-point power supply, comprising: light-storage-vehicle-diesel, for executing the method described above.

[0053] The model building module is used to build distributed photovoltaic stochastic output models, energy storage operation constraint models, electric vehicle cluster time-series charging and discharging models, diesel generator consumption-output models, and three-level load quantification constraint models based on graded load weights. It also builds a set of islanded operation constraint conditions and outputs multi-source equipment characteristic parameters, graded load weight coefficients, and islanded operating condition boundary conditions.

[0054] The optimization module is used to receive the characteristic parameters of the multi-source equipment under the islanding condition boundary conditions, construct a three-dimensional weighted comprehensive optimization function with economic efficiency, system adjustability and graded load guarantee rate as sub-objectives, and provide a set of equations for power balance, equipment operation, graded power supply control and node voltage constraints to form a multi-objective optimization model.

[0055] The solution module is used to iteratively solve the multi-objective optimization model using an improved genetic algorithm with embedded gradient descent local search, and output the optimal configuration results of photovoltaic installed capacity, rated power and capacity of energy storage, aggregated adjustable capacity of electric vehicles, and installed capacity of diesel generators.

[0056] The verification module is used to perform performance verification based on the optimal configuration results and the abnormal power outage islanding conditions. It reverse-verifies the emergency supply capability by comparing the ratio of the guaranteed power supply load to the maximum load of the distribution area, and verifies the supply guarantee indicators of zero power loss for primary loads, controllable power supply for secondary loads, and on-demand disconnection for tertiary residential loads. It also assists in calculating the photovoltaic absorption rate. and power supply reliability of loads at all levels The system comprehensively evaluates the new energy utilization efficiency and graded power supply quality of the configuration scheme, and outputs the final verified configuration scheme.

[0057] Furthermore, in the model construction module: the three-level load quantification constraint model divides the load into primary critical load, secondary industrial and commercial load, and tertiary residential load according to load importance, and assigns importance weights. ,satisfy ;

[0058] The three-level load quantification constraint model sets the following: Level 1 critical loads are subject to a hard zero-power-out constraint; Level 2 industrial and commercial loads are allowed a very small proportion of short-term power outages to adapt to grid fluctuations; and Level 3 residential loads are treated as switchable loads under islanding conditions to release power supply capacity for core power supply protection. The core equations are as follows:

[0059] In the formula: The permissible maximum power loss for Level 3 loads is listed in descending order; the permissible maximum power loss ratio for Level 2 loads is listed in descending order. Maximum permissible power loss ratio for Level 3 loads ;

[0060] The islanded operation constraint set includes power outage duration constraints and emergency power supply load screening formulas. The formula for the power outage duration constraint is as follows:

[0061] ; In the formula: This refers to the actual duration of the abnormal power outage; To preset the maximum fault repair time, For the output of distributed photovoltaic power during power outages, These are energy storage during power outages and the discharge output of electric vehicles, respectively. Provide power for the diesel engine during power outages; The total power supply load required to be guaranteed under island conditions; The total rated power for the first-level critical load is guaranteed to be fully supplied under abnormal operating conditions; The rated power is set for the total load of secondary industrial and commercial vehicles; This is the secondary load guarantee factor.

[0062] Furthermore, in the optimization construction module, the three-dimensional weighted comprehensive optimization function is min F= · + · +α · ,and + + =1, where The three-dimensional weighted comprehensive optimization function for the comprehensive economic analysis of annual average investment and operation and maintenance costs, electricity purchase costs, and electricity sales revenue is as follows: ;

[0063] ;

[0064] ;

[0065] ;

[0066] In the formula: F is the total value of the system's overall optimization objective, derived from economic efficiency. Adjustability Supply success rate The three sub-objectives are weighted together, and the optimization objective is to minimize the value. These are weighted coefficients for economic efficiency, adjustability, and supply success rate, respectively. These are the average annual investment and maintenance costs for photovoltaic, energy storage, electric vehicle supporting facilities, and diesel generators, respectively. To reduce the cost of purchasing electricity from the higher-level power grid, Revenue from selling surplus electricity online; This is the overall adjustable margin coefficient of the system.

[0067] Furthermore, in the solution module, the improved genetic algorithm includes: a population initialization unit, used to construct individual chromosomes using a real-number encoding method. The initial population is randomly generated, corresponding to photovoltaic installed capacity, rated energy storage power, rated energy storage capacity, adjustable aggregate capacity of electric vehicles, and installed capacity of diesel generators, respectively.

[0068] Fitness evaluation unit, used to evaluate fitness function The evaluation of individual performance is based on F, where F is the overall optimization target value and ε is a minimal constant to avoid a denominator of zero.

[0069] The adaptive crossover and mutation unit is used to perform adaptive crossover and mutation operations using crossover and mutation probabilities that dynamically decrease with the number of iterations. The formula is as follows: In the formula: These represent the initial crossover probability and the initial mutation probability of the algorithm, respectively; t is the current iteration number. This represents the maximum number of iterations.

[0070] Elite gradient optimization unit, used to select elite individuals, performs gradient descent local search on them, and updates them using the following formula: In the formula: , These are the individual configuration solutions before and after the update, respectively; This is the gradient learning step size; The gradient vector of a multi-objective function;

[0071] The convergence determination unit is used for repeated selection, crossover, mutation, and gradient optimization until the number of iterations reaches the maximum iteration threshold or the objective function tends to converge, and finally outputs the optimal configuration result of the light-storage-vehicle-diesel capacity.

[0072] Compared with the prior art, the present invention has the following significant advantages:

[0073] 1. Constructing a multi-source collaborative architecture of photovoltaic-storage-vehicle-diesel, achieving high resource complementarity and utilization. This invention integrates four types of distributed resources—photovoltaics, energy storage, electric vehicles, and diesel generators—into a unified planning and optimization system, breaking the limitations of the traditional dual-resource allocation model and fully integrating the temporal complementarity and spatial distribution advantages of various resources. Distributed photovoltaics provide clean basic power, energy storage enables rapid power regulation, electric vehicle clusters play a flexible support role in mobile energy storage, and diesel generators undertake the emergency backup function under extreme conditions, forming a diversified supply guarantee pattern of "efficient consumption of new energy under normal conditions and reliable backup under fault conditions." This effectively smooths out the output fluctuations of high-penetration photovoltaics, significantly improves the system's regulation margin and anti-disturbance capability, and can effectively cope with complex conditions such as extreme weather in mountainous areas and large power grid failures, ensuring continuous survivability after disconnection from the main grid.

[0074] 2. Establish a load importance-based hierarchical management mechanism with high accuracy and success rate in hierarchical power supply guarantee. This invention abandons the traditional microgrid's "one-size-fits-all" equal power supply and load shedding management model. It divides load priorities into three levels based on the public safety attributes and industrial economic value of the end-user loads, constructing a quantifiable and controllable differentiated power outage constraint and hierarchical power supply guarantee system. Under normal grid-connected operation, refined and differentiated stable power supply can be achieved based on load level. Under the fault island operation condition of mountainous power grids, through emergency load screening strategies, priority is given to fully guaranteeing the power supply of primary critical loads, selectively guaranteeing the orderly operation of secondary industrial and commercial loads, and controllably shedding tertiary ordinary residential loads. This achieves the best power supply guarantee effect for the entire network with minimal power outage losses, effectively avoiding the risk of power outages of critical loads, and adapting to the differentiated power supply guarantee tasks under the complex conditions of mountainous areas in Hubei.

[0075] 3. Employing a three-dimensional collaborative optimization model, the system boasts excellent overall adaptability. This invention breaks through the one-sidedness of traditional microgrid optimization based solely on economics and dual objectives. It innovatively incorporates economic efficiency, adjustability, and supply success rate into a unified equilibrium optimization system. While strictly controlling investment and operation and maintenance costs throughout the project's lifecycle, it maximizes the reserved dynamic adjustment margin of the system. This effectively solves the problems of traditional configuration schemes that prioritize economics over adjustment, have high renewable energy curtailment rates, and insufficient end-point supply margins. It is adaptable to large-scale grid connection scenarios with high penetration rates of distributed photovoltaic power, aligning with regional development plans for improving the quality and efficiency of renewable energy.

[0076] 4. Adaptable to dual-condition operation scenarios in mountainous areas, with strong engineering feasibility and operational robustness. This invention adds an islanding condition constraint and verification system adapted to the characteristics of mountainous power grids. Through power outage duration constraints, emergency power supply load screening, and multi-source capacity reverse verification mechanisms, it achieves full coverage verification of both grid-connected normal and fault-isolated scenarios, making up for the technical shortcomings of traditional technologies that only adapt to grid-connected conditions and lack extreme condition power supply capability verification. At the same time, it adopts an improved genetic algorithm with embedded gradient descent local search, effectively overcoming the defects of premature convergence and local optima in traditional algorithms. The solution accuracy and convergence efficiency for complex scenarios with multiple constraints are higher, and the configuration scheme fits the actual operating conditions of terminal transformer areas in mountainous areas, significantly improving engineering adaptability and practical value. Attached Figure Description

[0077] Figure 1 This is a flowchart of a multi-source collaborative optimization configuration method for end-point power supply based on the present invention, which involves optical, energy storage, vehicle, and diesel power sources.

[0078] Figure 2 This is a diagram of the multi-source-hierarchical load coupling topology of the terminal microgrid of this invention;

[0079] Figure 3 This is a hierarchical architecture diagram of the three-dimensional multi-objective optimization model of the present invention;

[0080] Figure 4 This is a flowchart of the improved genetic algorithm for solving the local search problem using embedded gradient descent in this invention.

[0081] Figure 5 This is a schematic diagram of the load classification and supply guarantee rules for transformer areas in this invention;

[0082] Figure 6 This is a schematic diagram of the dual-mode operation of the mountainous area grid-connected / islanded transformer area according to the present invention. Detailed Implementation

[0083] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0084] This invention provides a multi-source collaborative optimization configuration method for photovoltaic-storage-vehicle-diesel power supply for end-point power supply. This method follows a closed-loop execution logic of "modeling-optimization-solving-verification," with the core design concept of "refined characterization of source and load characteristics, multi-dimensional collaborative optimization objectives, intelligent solution to the optimal scheme, and engineering verification of configuration effects." It can be directly applied to the planning and design of end-point microgrids in Hubei counties / industrial parks, adapting to the needs of high-penetration distributed photovoltaic access and tiered power supply guarantee. The overall method flow of this invention is as follows: Figure 1 As shown, the steps include the following.

[0085] Step 1: Construct a mathematical model for coupling multiple power sources and tiered loads in the end-point microgrid.

[0086] Basic data for the target distribution area was collected, including grid parameters (transformer capacity, line impedance), load data (load data for the past year, distribution of load nodes at all levels), photovoltaic data (existing installed capacity, irradiance data for the past three years), and equipment parameters (technical parameters and investment and maintenance costs of various power sources). Combining typical meteorological data and equipment market prices in Hubei Province, core parameters for five types of models—distributed photovoltaic, energy storage, electric vehicle clusters, diesel generators, and three-level loads—were assigned values, completing the construction of a multi-source and load coupling mathematical model. The multi-source-hierarchical load coupling topology of the terminal microgrid in this invention is as follows: Figure 2 As shown.

[0087] Based on the technical solution of this invention, the core formulas of each model are derived in a targeted manner to ensure that the formulas are accurately adapted to the optimization requirements, providing core parameter support for the implementation of the configuration process:

[0088] 1. Formula for the stochastic output model of distributed photovoltaic power:

[0089]

[0090]

[0091] In the formula, Provide real-time power for photovoltaic systems; This refers to the rated installed capacity of photovoltaic power. The shape coefficient of the Beta distribution was obtained by fitting local irradiance data from the past three years. For beta functions, This is a gamma function.

[0092] 2. Iterative formula for energy storage time series capacity:

[0093]

[0094]

[0095] In the formula, The energy storage state of charge at time t; These are the energy storage charging and discharging power, respectively. For charging and discharging efficiency; This refers to the rated capacity of the energy storage.

[0096] 3. Three-level load quantization constraint model:

[0097] In the formula: The permissible maximum power loss for Level 3 loads is listed in descending order; the permissible maximum power loss ratio for Level 2 loads is listed in descending order. Maximum permissible power loss ratio for Level 3 loads ;

[0098] Based on the above quantitative constraint formula, a standardized and implementable full-condition graded power supply guarantee mechanism is constructed, clarifying the power supply control rules for different load levels: Level 1 critical loads cover public safety loads such as township medical facilities and emergency stations, with a zero-power-loss hard constraint to ensure uninterrupted power supply under all operating conditions; Level 2 industrial and commercial loads are local micro-industries and commercial operations, allowing for a very small proportion of short-term power outages, adapting to grid power fluctuations while ensuring basic industrial operation; Level 3 ordinary residential loads have high redundancy and a small impact range, and can be orderly disconnected under extreme conditions such as grid supply and demand tensions and fault islanding, prioritizing the power supply capacity of core loads, effectively improving the technical shortcomings of indiscriminate load shedding in traditional microgrids. See Figure 5 for details of the load classification and power supply guarantee rules for each distribution area.

[0099] In response to the frequent power grid failures and uncontrollable islanding operation durations in mountainous areas, this invention establishes a dedicated power supply constraint system for islanding conditions, specifically adapted to special scenarios such as large power grid outages and extreme disasters that disconnect from the main grid. This is achieved through power outage duration constraints. Define the maximum emergency power supply duration to adapt to both short-term and extreme disaster scenarios; rely on load screening formulas. Accurately select isolated emergency power loads, solidify the emergency power supply logic of "Level 1 must be protected, Level 2 is selected preferentially, and Level 3 is cut off", maximize the reliability of core load power supply under the condition of limited distributed power supply capacity, and achieve precise emergency backup in fault conditions.

[0100] Step 2: Construct the three-dimensional weighted comprehensive optimization function and multi-dimensional constraint equations

[0101] Based on the actual needs of the transformer substation, the weighting coefficients for the three sub-objectives—economic efficiency, adjustability, and tiered load guarantee rate—were determined, and their values ​​were set to [values ​​to be filled in]. A comprehensive optimization objective function is constructed. Supporting system constraints include real-time active power balance, node voltage safety constraints (0.95~1.05 pu), operational boundary constraints for various equipment, and three-level load supply guarantee constraints, forming a complete multi-objective optimization mathematical system. The hierarchical architecture of the multi-objective optimization model of this invention is as follows: Figure 3 As shown.

[0102] The derivation of the core formula is as follows:

[0103] 1. Comprehensive optimization objective formula:

[0104]

[0105] In the formula, For economic sub-objectives; As a sub-objective of adjustability; To ensure the success rate of supply delivery; These are the corresponding weighting coefficients.

[0106] 2. Real-time active power balance formula for the system:

[0107]

[0108] In the formula, For power exchange; This represents the total load power of the system.

[0109] Step 3: Solve using an improved genetic algorithm with embedded gradient descent and local search.

[0110] Based on the MATLAB simulation platform, an improved genetic algorithm solver is used to iteratively solve a multi-objective optimization model. The core parameters of the algorithm are set as follows: population size 100, maximum number of iterations 200, initial crossover probability 0.8, initial mutation probability 0.05, gradient learning step size 0.01, and elite individual retention ratio 10%. During algorithm execution, the change in the objective function value is monitored in real time. When the number of iterations reaches the maximum threshold or the change in the objective function value is less than 1e-6 for 30 consecutive generations, the algorithm is considered to have converged, and the optimal capacity configuration results for photovoltaic, energy storage, electric vehicles, and diesel generators are output. The improved genetic algorithm solution logic flow of this invention is as follows: Figure 4 As shown.

[0111] The derivation of the core formula is as follows:

[0112] 1. Individual chromosome coding formula:

[0113]

[0114] In the formula, the array elements correspond to the photovoltaic installed capacity, the rated power of energy storage, the rated capacity of energy storage, the aggregated adjustable capacity of electric vehicles, and the installed capacity of diesel generators, respectively.

[0115] 2. Elite gradient local optimization formula:

[0116]

[0117] In the formula, These are the individual configuration solutions before and after the update, respectively; This is the gradient learning step size; This is the gradient vector of a multi-objective function.

[0118] Step 4: Verification of End-of-Line Power Supply Capability: Based on the optimal configuration results output in Step S3, performance verification is performed in conjunction with the islanding scenario of abnormal power outages. The emergency power supply capability is verified in reverse based on the ratio of the guaranteed power supply load to the maximum load of the distribution area. This verifies the power supply indicators of zero power loss for primary loads and controllable power supply for secondary loads, and assists in calculating the photovoltaic absorption rate. and power supply reliability of loads at all levels The system comprehensively evaluates the new energy utilization efficiency and graded power supply quality of the configuration scheme, and outputs the final verified configuration scheme.

[0119] Specifically, it includes: (1) Verification of isolated power supply capacity. Step S4 sets up an isolated power supply capacity verification model. Based on the abnormal power outage duration and the scale of the power supply load, the power supply pressure is quantified by the ratio of the power supply load to the maximum load of the distribution area. The emergency adaptability of the multi-source configuration capacity is verified in reverse. The calculation formula is as follows:

[0120]

[0121]

[0122] In the formula: To ensure the proportion of power supply load, this is used to quantify the emergency power supply pressure; This represents the maximum total load of the transformer substation connected to the grid. To maximize the combined output of photovoltaic, energy storage, vehicle, and diesel power sources; These are the effective outputs of photovoltaics, energy storage, electric vehicle clusters, and diesel generators, respectively. To ensure the total power supply load required under islanded operating conditions, the capacity verification for abnormal operating conditions is completed by matching the total output of multiple sources with the power supply load. The verification passes when the maximum combined output of multiple sources is not less than the total power supply load required under islanded operating conditions.

[0123] (2) Verification of Tiered Supply Guarantee Indicators. Simulating two types of islanding conditions: routine short-term faults and extreme disaster faults, the tiered supply guarantee rules based on load importance were strictly enforced to verify the supply guarantee indicators: full supply guarantee with zero power loss for Tier 1 critical loads, controllable power supply for Tier 2 industrial and commercial loads (power loss ratio ≤ 5%), and on-demand power cut-off for Tier 3 residential loads (power loss ratio ≤ 20%). Simultaneously, the photovoltaic absorption rate was calculated as an auxiliary calculation. and power supply reliability of loads at all levels A comprehensive evaluation of the renewable energy utilization efficiency and tiered power supply quality of the configuration scheme is conducted.

[0124] The core formula is as follows:

[0125] 1. Formula for photovoltaic power absorption rate:

[0126]

[0127] In the formula, This enables local consumption of photovoltaic power.

[0128] 2. Power supply reliability formulas for each load level:

[0129]

[0130] In the formula, This represents the power shortage of the k-th level load. The total rated power for the k-th level load.

[0131] (3) Output of verification results. If the optimal configuration scheme meets the above supply guarantee indicators under both types of islanded operating conditions, the scheme is deemed to have passed the performance verification, and the final configuration result (including the capacity parameters of photovoltaic, energy storage, electric vehicle, and diesel generator) is output. If either operating condition fails to meet the requirements, the optimization weights or constraints are adjusted, and the solution is iterated again until the verification is passed. At the same time, the scheme of this invention can be compared with traditional configuration schemes (such as photovoltaic-storage, photovoltaic-storage-vehicle, etc.) to quantitatively verify the comprehensive advantages of this invention in terms of multi-source synergy, hierarchical supply guarantee, and economy.

[0132] Implementation Cases

[0133] To verify the engineering practicality and technological advancement of the proposed multi-source coordinated optimization configuration method for photovoltaic-storage-vehicle-diesel systems that takes into account load importance, this embodiment selects a central distribution area in a mountainous region of Hubei Province as the simulation verification object. This distribution area has typical characteristics such as high-penetration distributed photovoltaic access, high line impedance in mountainous areas, weak support capacity of the large power grid, low spatiotemporal matching degree of source and load, and coexistence of residential and industrial / commercial loads, which perfectly match the applicable scenario of this invention.

[0134] 4.1 Basic Operating Conditions and Parameter Settings of the Case Study Substation

[0135] The specific operating parameters of the transformer substation in this case are as follows: The substation is equipped with an 800kVA distribution transformer, and the overall structure is a single busbar radial 0.4kV distribution network, divided into 1 common grid-connected busbar node and 12 low-voltage load branch nodes.

[0136] Tiered load parameters: Based on the annual 8760h measured load data of the transformer area, and according to the load importance grading rules of this invention, the load is divided into three levels: Level 1 emergency load (township health centers, emergency material stations) with a total rated load of 70kW and a load importance coefficient. To ensure zero power outages across the entire area; the total rated load of secondary industrial and commercial loads (township agricultural product processing plants, convenience stores) is 180kW, with a load importance coefficient of [missing information]. Maximum allowable power loss percentage The total rated load for ordinary residential buildings at level three is 240kW, with a load importance coefficient of [missing information]. Maximum allowable power loss percentage The load classification and supply guarantee rules for transformer areas in this invention are as follows: Figure 5 As shown.

[0137] Multi-source equipment engineering parameters: The existing distributed photovoltaic capacity in the area is 720kW, which can be used to regulate 35 private cars via V2G. The energy storage and diesel generator sets adopt the general equipment parameters of Hubei county distribution network. The energy storage's normal state of charge range is as follows. The safe operating output range of the diesel generator set is defined. Based on the local 5-year line maintenance log, the maximum power outage time for regular faults is defined as 4 hours, and the maximum power outage time for extreme disasters is defined as 12 hours, matching the islanded operation constraints of this invention.

[0138] 4.2 Simulation Conditions and Control Group Settings

[0139] (1) Division of global simulation working conditions

[0140] Based on the actual operation and maintenance mode of power distribution networks in mountainous areas, two simulation scenarios are divided:

[0141] Operating Condition 1: Normal Grid-Connected Operating Condition: The distribution substation is connected to the upstream main grid for power supply, all loads are put into operation, the voltage constraint range of the substation nodes is 0.95~1.05pu, the diesel generator set is in hot standby, the power loss scale of each level of load is controlled according to the importance of the load, and the local consumption of photovoltaic power and the economic operation of the substation are taken into account.

[0142] Operating Condition 2: Fault Island Operating Condition: Affected by landslides, rain and snow, and line faults, the transformer area is separated from the main grid and operates independently. It covers two types of fault durations: regular short-term faults (maximum 4 hours) and extreme disaster faults (maximum 12 hours). The load priority supply guarantee rules are uniformly implemented, giving priority to ensuring uninterrupted power supply to the first-level emergency loads, controlling the power supply scale of the second-level loads as needed, and orderly cutting off the third-level loads to reduce power supply pressure.

[0143] (2) Optimize the solution algorithm parameters

[0144] Based on the previously established source-load simulation model of the power distribution area, dual-condition operation constraints, and three-level load supply constraints, and by substituting all measured boundary parameters, a multi-objective capacity optimization iterative solution was carried out using the proposed improved genetic algorithm. This simulation involved a total of 172 iterations, achieving an iterative convergence accuracy of 10^-6, and ultimately outputting the optimal coordinated configuration scheme of photovoltaic-storage-vehicle-diesel power distribution area.

[0145] The configuration capacity and simulation adaptation results of each unit are as follows: 1) Photovoltaic capacity expansion configuration: Based on the calculation of the maximum daytime surplus load of 116kW and the normal curtailment load in the distribution area, the original 720kW photovoltaic capacity will be expanded to 800kW. The local photovoltaic consumption capacity under grid-connected conditions can match the power demand of the entire area; 2) Energy storage configuration: The simulation calculation shows that the average short-term island power gap of 4h is 78kW. A 120kW / 240kWh energy storage system will be configured to fully cover the short-term island load gap and meet the peak shaving demand of the grid at night; 3) Electric vehicle aggregated output: The simulation extracts the optimal output range of flexible energy replenishment under off-grid conditions. The aggregated adjustable discharge power of 35 private cars in the area is locked at 90kW to help make up for the dynamic power fluctuation of the island; 4) Emergency diesel generator set: For the 124kW supply load of the first level of island and the second level of rigid demand in 12h extreme disasters, a 60kW diesel generator set will be configured to combine energy storage, photovoltaic and electric vehicle output to ensure long-term power supply. Simulation verification shows that this configuration can simultaneously meet the requirements of normal grid connection voltage compliance, load power failure rate compliance, and the mandatory requirements of 4h / 12h fault islanding graded power supply guarantee, providing the optimal simulation basis for subsequent operational verification and scheme benchmarking.

[0146] 4.3 Dual-condition macroscopic adaptation verification

[0147] The results were verified against the two simulation conditions defined above, and the specific operation diagram is shown below. Figure 6 As shown:

[0148] 1. Operating Condition 1: Normal Grid Connection Adaptability Verification: The voltage of the entire distribution area is maintained within the qualified range, the power loss scale of each level of load meets the constraints of this invention, and the multi-source coordination can smoothly complete the local photovoltaic consumption and load peak-valley smoothing, meeting the long-term grid connection operation and management constraints throughout the process.

[0149] 2. Operating Condition 2 Fault Islanding Adaptability Verification: During the off-grid islanding operation of the transformer area, relying on the adjustable output of the photovoltaic-storage-vehicle-diesel system, it can adapt to two types of power outage durations: 4 hours of normal faults and 12 hours of extreme disaster faults. It strictly follows the graded supply guarantee according to the importance of the load, with zero power loss for the first-level emergency load throughout the entire process, ensuring the power supply of the second-level essential load as needed, and flexibly cutting off the third-level secondary load, thus meeting the local fault emergency supply guarantee requirements throughout the entire area.

[0150] 4.4 Analysis of the results of benchmarking multiple schemes

[0151] To highlight the core innovative advantages of this invention—tiered load importance guarantee and multi-source complementary synergy—three commonly used county-level power distribution network configuration schemes were set up as control groups. Control group 1 had the output of electric vehicles and diesel generators set to zero, retaining only photovoltaic-storage operation; control group 2 had the output of diesel generators set to zero, retaining photovoltaic-storage-vehicle synergy operation; control group 3 had the V2G discharge output of electric vehicles set to zero, retaining photovoltaic-storage-diesel synergy operation; the experimental group had all four power parameters assigned normal values ​​without any zeroing process, representing the complete synergy scheme of this invention. The simulation calculation results and benchmark indicators for each group are detailed in the table below.

[0152] Table 1 Comparison of simulation calculation results for each group

[0153]

[0154] The benchmarking results show that each single combination scheme has its shortcomings. The photovoltaic-storage-vehicle scheme has outstanding flexibility and photovoltaic absorption capacity, but its long-term emergency supply capacity is insufficient. The photovoltaic-storage-diesel scheme has stronger islanded power supply capacity, but its new energy absorption and power flow optimization effects are relatively weak.

[0155] This invention also provides a multi-source collaborative optimization configuration device for end-point power supply, comprising: a light-storage-vehicle-diesel multi-source configuration, for executing the method described above, characterized in that it includes:

[0156] The model building module is used to build distributed photovoltaic stochastic output models, energy storage operation constraint models, electric vehicle cluster time-series charging and discharging models, diesel generator consumption-output models, and three-level load quantification constraint models based on graded load weights. It also builds a set of islanded operation constraint conditions and outputs multi-source equipment characteristic parameters, graded load weight coefficients, and islanded operating condition boundary conditions.

[0157] The optimization module is used to receive the characteristic parameters of the multi-source equipment under the islanding condition boundary conditions, construct a three-dimensional weighted comprehensive optimization function with economic efficiency, system adjustability and graded load guarantee rate as sub-objectives, and provide a set of equations for power balance, equipment operation, graded power supply control and node voltage constraints to form a multi-objective optimization model.

[0158] The solution module is used to iteratively solve the multi-objective optimization model using an improved genetic algorithm with embedded gradient descent local search, and output the optimal configuration results of photovoltaic installed capacity, rated power and capacity of energy storage, aggregated adjustable capacity of electric vehicles, and installed capacity of diesel generators.

[0159] The verification module is used to perform performance verification based on the optimal configuration results and the abnormal power outage islanding conditions. It reverse-verifies the emergency supply capability by comparing the ratio of the guaranteed power supply load to the maximum load of the distribution area, and verifies the supply guarantee indicators of zero power loss for primary loads, controllable power supply for secondary loads, and on-demand disconnection for tertiary residential loads. It also assists in calculating the photovoltaic absorption rate. and power supply reliability of loads at all levels The system comprehensively evaluates the new energy utilization efficiency and graded power supply quality of the configuration scheme, and outputs the final verified configuration scheme.

[0160] This invention integrates the advantages of four types of power sources, and through collaborative optimization, rationally controls the configuration capacity of individual devices to avoid redundant expansion. At the same time, it improves the photovoltaic absorption rate, reducing the economic losses from curtailment, and diesel generator sets are only used under extreme fault conditions, with limited increases in fuel and maintenance costs under normal circumstances. The combination of multiple factors results in this invention achieving the optimal overall cost over its entire life cycle, despite integrating multiple types of equipment, while also considering tiered load guarantee rates, renewable energy utilization, and emergency supply capabilities.

[0161] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for multi-source coordinated optimization configuration of photovoltaic-storage-vehicle-diesel power supply for end-point power supply, characterized in that, Includes the following steps: S1: Construct a mathematical model for coupling multiple power sources and hierarchical loads in the terminal microgrid, as well as a constraint system for islanded operation conditions. Specifically, this includes: establishing a mathematical model for the stochastic output of distributed photovoltaic power, an energy storage operation constraint model, a time-series charging and discharging model for electric vehicle clusters, a diesel generator output-fuel consumption characteristic model, and a three-level load quantification constraint model based on hierarchical load weights; at the same time, construct a set of islanded operation constraint conditions, including power outage duration constraints and an emergency power supply load screening mechanism based on load importance, and output multi-source equipment characteristic parameters and islanded operation condition boundary conditions. S2: Based on the multi-source equipment characteristic parameters and islanding boundary conditions output in step S1, a three-dimensional weighted comprehensive optimization function is constructed with economic efficiency, system adjustability, and graded load guarantee rate as optimization sub-indicators. It is also equipped with system power balance constraints, equipment operation constraints, graded power supply control constraints, and node voltage constraints to form a multi-objective optimization model. S3: An improved genetic algorithm with embedded gradient descent local search is used to iteratively solve the multi-objective optimization model constructed in step S2, and output the optimal configuration results of photovoltaic installed capacity, energy storage rated power and capacity, electric vehicle aggregate adjustable capacity, and diesel generator installed capacity. S4: Based on the optimal configuration results output in step S3, performance verification is performed in conjunction with the islanding condition during abnormal power outages. Emergency supply capability is then verified in reverse based on the ratio between the guaranteed power supply load and the maximum load of the distribution area. This verifies the supply guarantee indicators of zero power loss for primary loads, controllable power supply for secondary loads, and on-demand disconnection of tertiary residential loads. This also assists in calculating the photovoltaic absorption rate. and power supply reliability of loads at all levels The system comprehensively evaluates the new energy utilization efficiency and graded power supply quality of the configuration scheme, and outputs the final verified configuration scheme.

2. The method according to claim 1, characterized in that, The three-level load quantification constraint model described in step S1 divides the load into three levels: Level 1 critical load, Level 2 industrial and commercial load, and Level 3 residential load, according to load importance, and assigns importance weights to each level. ,satisfy The three-level load quantification constraint model sets the following: Level 1 critical loads are subject to a hard zero-power-loss guarantee; Level 2 industrial and commercial loads are allowed a very small proportion of short-term power outages to adapt to grid fluctuations; and Level 3 residential loads are treated as switchable loads under islanding conditions to release power capacity for core power supply. The core equations are as follows: In the formula: The permissible maximum power loss for Level 3 loads is listed in descending order; the permissible maximum power loss ratio for Level 2 loads is listed in descending order. Maximum permissible power loss ratio for Level 3 loads ; The islanded operation constraint set includes power outage duration constraints and emergency power supply load screening formulas. The formula for the power outage duration constraint is as follows: ; In the formula: This refers to the actual duration of the abnormal power outage; To preset the maximum fault repair time, For the output of distributed photovoltaic power during power outages, These are energy storage during power outages and the discharge output of electric vehicles, respectively. Provide power for the diesel engine during power outages; The total power supply load required to be guaranteed under island conditions; The total rated power for the first-level critical load is guaranteed to be fully supplied under abnormal operating conditions; The rated power is set for the total load of secondary industrial and commercial vehicles; This is the secondary load guarantee factor.

3. The method according to claim 1, characterized in that, In step S1, The mathematical model for distributed photovoltaic stochastic output adopts the Beta probability distribution model to quantify the temporal fluctuation characteristics of photovoltaic stochastic output. The core calculation formula is as follows: ; ； In the formula: Provide real-time power for photovoltaic systems; This refers to the rated installed capacity of photovoltaic power. The shape coefficient of the Beta distribution is obtained by fitting local measured irradiation data; For beta functions; It is a gamma function; The energy storage operation constraint model includes a time-series energy iterative equation, upper and lower limits of state of charge constraints, charging and discharging power limit constraints, and charging and discharging mutual exclusion constraints. The core formulas are as follows: ; ; ; In the formula: for Constant energy storage state of charge; These are the energy storage charging and discharging power, respectively. These refer to the energy storage charging and discharging efficiencies, respectively. These are the rated energy storage capacity and power, respectively. The electric vehicle cluster time-series charging and discharging model fits the individual vehicle's power consumption and remaining power based on the log-normal distribution of driving mileage, and aggregates them to obtain the cluster's time-sharing charging and discharging power. The core formula is as follows: ; ； ； ； ； ； In the formula: This refers to the daily mileage traveled by a single vehicle. These are the logarithmic mean and logarithmic standard deviation of mileage, respectively. Electricity consumption per unit distance of vehicle; Rated capacity of the vehicle's power battery; Let m be the rated discharge power of the m-th vehicle; This represents the lowest state of charge for electric vehicles, ensuring a minimum amount of power available for users' travel needs. They are respectively A fleet of vehicles that are constantly charging and discharging; For the first The time-sharing charging and discharging power of the vehicles; The total charging and discharging power of the electric vehicle cluster during time-sharing; The diesel generator output-fuel consumption characteristic model adopts a quadratic function. Describe fuel consumption and set minimum stable output constraints. In the formula: This refers to the fuel consumption per unit time of the generator unit. These are the characteristic coefficients for fitting diesel engine fuel consumption; To provide real-time active power output for diesel generators; This refers to the rated maximum output of the diesel generator. This is the minimum stable operating output of the diesel generator, used to constrain the unit's low-load shutdown conditions.

4. The method according to claim 2, characterized in that, The three-dimensional weighted synthesis optimization function mentioned in step S2 is: ; ； ； ； In the formula: F is the total value of the system's overall optimization objective, derived from economic efficiency. Adjustability Supply success rate The three sub-objectives are weighted together, and the optimization objective is to minimize the value. These are weighted coefficients for economic efficiency, adjustability, and supply success rate, respectively. These are the average annual investment and maintenance costs for photovoltaic, energy storage, electric vehicle supporting facilities, and diesel generators, respectively. To reduce the cost of purchasing electricity from the higher-level power grid, Revenue from selling surplus electricity online; This is the overall adjustable margin coefficient of the system.

5. The method according to claim 1, characterized in that, The solution process of the improved genetic algorithm described in step S3 includes: S31: Constructing individual chromosomes using real-number encoding. The initial population is randomly generated, corresponding to photovoltaic installed capacity, rated energy storage power, rated energy storage capacity, adjustable aggregate capacity of electric vehicles, and installed capacity of diesel generators, respectively. S32: Using the fitness function Evaluate the merits and demerits of individuals, where F is the comprehensive optimization target value described in step S2, and ε is a minimal constant to avoid a denominator of zero; S33: Adaptive crossover and mutation operations are performed using crossover and mutation probabilities that dynamically decrease with the number of iterations. The formula is as follows: In the formula: These represent the initial crossover probability and the initial mutation probability of the algorithm, respectively; t is the current iteration number. This represents the maximum number of iterations. S34: Select elite individuals, perform gradient descent local search on them, and update the formula as follows: In the formula: , These are the individual configuration solutions before and after the update, respectively; This is the gradient learning step size; The gradient vector of a multi-objective function; S35: Repeatedly select, crossover, mutate, and gradient optimization until the number of iterations reaches the maximum iteration threshold or the objective function tends to converge, and finally output the optimal configuration result of the light-storage-vehicle-diesel capacity.

6. The method according to claim 1, characterized in that, Step S4 sets up an islanded power supply capacity verification model. Based on the abnormal power outage duration and the scale of the power supply load, the power supply pressure is quantified by the ratio of the power supply load to the maximum load of the distribution area. The emergency adaptability of the multi-source configuration capacity is verified in reverse. The calculation formula is as follows: ； ； In the formula: To ensure the proportion of power supply load, this is used to quantify the emergency power supply pressure; This represents the maximum total load of the transformer substation connected to the grid. To maximize the combined output of photovoltaic, energy storage, vehicle, and diesel power sources; These are the effective outputs of photovoltaics, energy storage, electric vehicle clusters, and diesel generators, respectively. To ensure the total power supply load required under isolated operating conditions, the capacity verification for abnormal operating conditions is completed by matching the total output of multiple sources with the power supply load. The performance of the configuration scheme is evaluated using multiple quantitative indicators. The core evaluation formula is as follows: In the formula: Photovoltaic grid integration rate, used to characterize the utilization rate of new energy sources. To absorb the photovoltaic power output, Contribute to the overall photovoltaic power generation; The load guarantee rates are categorized into three levels: Level 1, Level 2, and Level 3. For the power shortage of loads at all levels, Power for each load level; For load level indexing.

7. A multi-source collaborative optimization configuration device for end-point power supply, comprising optical-storage-vehicle-diesel, for executing the method according to any one of claims 1-6, characterized in that, include: The model building module is used to build distributed photovoltaic stochastic output models, energy storage operation constraint models, electric vehicle cluster time-series charging and discharging models, diesel generator consumption-output models, and three-level load quantification constraint models based on graded load weights. It also builds a set of islanded operation constraint conditions and outputs multi-source equipment characteristic parameters, graded load weight coefficients, and islanded operating condition boundary conditions. The optimization module is used to receive the characteristic parameters of the multi-source equipment under the islanding condition boundary conditions, construct a three-dimensional weighted comprehensive optimization function with economic efficiency, system adjustability and graded load guarantee rate as sub-objectives, and provide a set of equations for power balance, equipment operation, graded power supply control and node voltage constraints to form a multi-objective optimization model. The solution module is used to iteratively solve the multi-objective optimization model using an improved genetic algorithm with embedded gradient descent local search, and output the optimal configuration results of photovoltaic installed capacity, rated power and capacity of energy storage, aggregated adjustable capacity of electric vehicles, and installed capacity of diesel generators. The verification module is used to perform performance verification based on the optimal configuration results and the abnormal power outage islanding conditions. It reverse-verifies the emergency supply capability by comparing the ratio of the guaranteed power supply load to the maximum load of the distribution area, and verifies the supply guarantee indicators of zero power loss for primary loads, controllable power supply for secondary loads, and on-demand disconnection for tertiary residential loads. It also assists in calculating the photovoltaic absorption rate. and power supply reliability of loads at all levels The system comprehensively evaluates the new energy utilization efficiency and graded power supply quality of the configuration scheme, and outputs the final verified configuration scheme.

8. The apparatus according to claim 7, characterized in that, In the model construction module: the three-level load quantification constraint model divides the load into primary critical load, secondary industrial and commercial load, and tertiary residential load according to load importance, and assigns importance weights accordingly. ,satisfy ; The three-level load quantification constraint model sets the following: Level 1 critical loads are subject to a hard zero-power-out constraint; Level 2 industrial and commercial loads are allowed a very small proportion of short-term power outages to adapt to grid fluctuations; and Level 3 residential loads are treated as switchable loads under islanding conditions to release power supply capacity for core power supply protection. The core equations are as follows: In the formula: The permissible maximum power loss for Level 3 loads is listed in descending order; the permissible maximum power loss ratio for Level 2 loads is listed in descending order. Maximum permissible power loss ratio for Level 3 loads ; The islanded operation constraint set includes power outage duration constraints and emergency power supply load screening formulas. The formula for the power outage duration constraint is as follows: ; In the formula: This refers to the actual duration of the abnormal power outage; To preset the maximum fault repair time, For the output of distributed photovoltaic power during power outages, These are energy storage during power outages and the discharge output of electric vehicles, respectively. Provide power for the diesel engine during power outages; The total power supply load required to be guaranteed under island conditions; The total rated power for the first-level critical load is guaranteed to be fully supplied under abnormal operating conditions; The rated power is set for the total load of secondary industrial and commercial vehicles; This is the secondary load guarantee factor.

9. The apparatus according to claim 7, characterized in that, In the optimization construction module, the three-dimensional weighted comprehensive optimization function is min F= · + · +α · ,and + + =1, where The three-dimensional weighted comprehensive optimization function for the comprehensive economic analysis of annual average investment and operation and maintenance costs, electricity purchase costs, and electricity sales revenue is as follows: ; ； ； ； In the formula: F is the total value of the system's overall optimization objective, derived from economic efficiency. Adjustability Supply success rate The three sub-objectives are weighted together, and the optimization objective is to minimize the value. These are weighted coefficients for economic efficiency, adjustability, and supply success rate, respectively. These are the average annual investment and maintenance costs for photovoltaic, energy storage, electric vehicle supporting facilities, and diesel generators, respectively. To reduce the cost of purchasing electricity from the higher-level power grid, Revenue from selling surplus electricity online; This is the overall adjustable margin coefficient of the system.

10. The apparatus according to claim 7, characterized in that, The improved genetic algorithm in the solution module includes: a population initialization unit, used to construct individual chromosomes using a real-number encoding method. The initial population is randomly generated, corresponding to photovoltaic installed capacity, rated energy storage power, rated energy storage capacity, adjustable aggregate capacity of electric vehicles, and installed capacity of diesel generators, respectively. Fitness evaluation unit, used to evaluate fitness function The evaluation of individual performance is based on F, where F is the overall optimization target value and ε is a minimal constant to avoid a denominator of zero. The adaptive crossover and mutation unit is used to perform adaptive crossover and mutation operations using crossover and mutation probabilities that dynamically decrease with the number of iterations. The formula is as follows: In the formula: These are the initial crossover probability and the initial mutation probability of the algorithm, respectively. t is the current iteration number. This represents the maximum number of iterations. Elite gradient optimization unit, used to select elite individuals, performs gradient descent local search on them, and updates them using the following formula: In the formula: , These are the individual configuration solutions before and after the update, respectively; This is the gradient learning step size; The gradient vector of a multi-objective function; The convergence determination unit is used for repeated selection, crossover, mutation, and gradient optimization until the number of iterations reaches the maximum iteration threshold or the objective function tends to converge, and finally outputs the optimal configuration result of the light-storage-vehicle-diesel capacity.

Images