A method for optimizing spatiotemporal scheduling and incentive price of V2G emergency power supply considering response rate uncertainty

Abstract

The present disclosure relates to a V2G emergency power supply space-time scheduling and incentive price optimization method considering response rate uncertainty. The method comprises: for the V2G emergency power supply, a deterministic response assumption is adopted, and it is difficult to reflect the randomness of user behavior. First, the vehicle spatial position is optimized based on GIS, and the optimal scheduling is determined by minimizing the user inconvenience. Second, an incentive price and response rate mapping model is established, and the Bayesian posterior probability is used to correct the random scene. Then, double clustering is used for hierarchical uncertainty decoupling and scene reduction, and the vehicle is equivalent to virtual energy storage. Finally, the expected income model is constructed by distinguishing between complete and incomplete power failure scenarios, and the optimal incentive price is determined based on marginal income analysis. The price and scheduling are optimized cooperatively, the uncertainty is reduced, the resource allocation efficiency and calculation speed are improved, and the power failure loss is reduced.

Description

Technical Field

[0001] This disclosure relates to the field of emergency power supply technology for distribution networks, and more specifically, to a method and apparatus for optimizing V2G emergency power supply scheduling and incentive prices considering response rate uncertainty. Background Technology

[0002] With the rapid development of electric vehicles (EVs), feeding EV power back to the grid through vehicle-to-grid (V2G) technology has gradually become an important means of enhancing the resilience of power systems. EVs possess exceptional flexibility, enabling emergency power supply during disasters or faults. However, user participation in V2G is highly uncertain. The V2G process relies on users' proactive responses, and their willingness to participate is closely related to the incentive price, as well as factors such as travel demand, battery status, and incentive price. This leads to significant fluctuations and uncertainties in response rates even at the same incentive price. Failure to reasonably quantify this uncertainty directly impacts the effectiveness and economy of V2G dispatch strategies in emergency scenarios. Therefore, how to construct a price and dispatch collaborative optimization mechanism, considering the uncertainty of user responses, to minimize system losses and improve power supply efficiency has become a critical issue that urgently needs to be addressed in current V2G emergency applications.

[0003] In existing technologies, scenario-based randomized optimization pricing methods obtain user response rates only through Monte Carlo sampling, lacking correction and update mechanisms and scenario reduction, resulting in numerous scenarios but insufficient typicality. Furthermore, when constructing the sampling relationship between electric vehicle response rates and incentive prices, the large vehicle scale and diverse incentive levels easily generate an exponentially growing scenario space, and existing methods fail to effectively decouple electric vehicles, leading to low computational efficiency. In addition, during electric vehicle scheduling optimization, the goal is usually only to minimize economic costs, ignoring the impact of Geographic Information System (GIS) information and scheduling strategies on user travel, making it difficult to balance user convenience with grid operation risks.

[0004] With the rapid growth of electric vehicle ownership, utilizing vehicle-to-grid (V2G) technology for emergency power supply in distribution networks has become an important means of enhancing network resilience. However, most existing V2G emergency power supply research is based on deterministic user response assumptions, making it difficult to accurately reflect the randomness and uncertainty of electric vehicle user participation behavior under incentive pricing. Furthermore, the uncertainty of user response rates further leads to uncertainty in the actual set of responding vehicles, available energy storage capacity, and spatial distribution, thus increasing the complexity of emergency power supply scheduling and incentive price optimization. Especially in scenarios involving large-scale electric vehicle participation, different prices correspond to multiple response rates, and different response rates correspond to multiple vehicle response scenarios, resulting in a high-dimensional optimization problem with a large computational scale, making it difficult to meet the needs of rapid decision-making in emergency power supply.

[0005] Therefore, one or more methods are needed to solve the above problems.

[0006] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0007] The purpose of this disclosure is to provide a method and apparatus for optimizing V2G emergency power supply scheduling and incentive prices considering response rate uncertainty, thereby overcoming at least to some extent one or more problems caused by the limitations and defects of related technologies.

[0008] According to one aspect of this disclosure, a method for optimizing V2G emergency power supply scheduling and incentive prices considering response rate uncertainty is provided, comprising: In emergency scenarios, GIS-based spatial location optimization for electric vehicles includes collecting multi-source data to construct a spatial layer and performing multi-layer filtering, establishing an electric vehicle location scheduling optimization model with the goal of minimizing user inconvenience, introducing availability constraints to screen feasible scheduling schemes and solving for the optimal scheduling result. The generation of random behavior scenarios induced by price and incorporating Bayesian posterior probability includes establishing a mapping relationship between incentive price and electric vehicle response rate, determining the incentive price range based on the system supply and demand gap, conducting random sampling within the range, and correcting the sampled scenarios based on multidimensional state coefficients using Bayesian posterior probability. Hierarchical uncertainty decoupling and scenario reduction of two-layer clustering includes extracting electric vehicle features and response rate samples for cluster analysis, determining the optimal number of categories and solving the clustering parameters, equating the electric vehicle clustering results to virtual energy storage resources, and obtaining the typical user response rate and probability of occurrence under different incentive prices to complete scenario reduction. The optimization of expected power loss risk and the calculation of optimal price include constructing differentiated expected return models for complete power loss and incomplete power loss scenarios, selecting candidate price ranges based on marginal return analysis, and determining the optimal incentive price.

[0009] In one exemplary embodiment of this disclosure, the objective function for optimizing electric vehicle location scheduling is:

[0010] Where D1 is the distance to the V2G site, S1 is the power consumption of the V2G site; D2 is the distance between the V2G site and the user's next destination, S2 is the power consumption between the V2G site and the user's next destination, and c1, c2, c3, and c4 are weighting coefficients.

[0011] In one exemplary embodiment of this disclosure, the available SOC constraint in the method is:

[0012] Among them, SOC after The remaining SOC after electric vehicles are dispatched, SOC min This is the minimum remaining SOC (State of Charge) of the electric vehicle after it arrives at the dispatch point. If the remaining SOC of the electric vehicle is lower than this minimum threshold after being dispatched, it is considered not dispatchable and the electric vehicle will not participate in the dispatch. The reachability constraint is:

[0013] Among them, SOC i The current SOC remains, SOC i,min The SOC is the SOC of the EV movement under the shortest path. If the remaining SOC of the electric vehicle is lower than the SOC of the EV movement under the shortest path, it is considered to be unreachable and the electric vehicle will not participate in the scheduling.

[0014] In one exemplary embodiment of this disclosure, the formula for calculating the power consumption along the route is:

[0015] in, dis To minimize the distance traveled, E dis The average power consumption per kilometer is expressed in kWh / km. The formula for calculating the remaining SOC of an electric vehicle is:

[0016] Where SOC is the remaining SOC before receiving scheduling; cap is the battery capacity, E travel Energy consumption along the shortest path, expressed in kWh; The formula for calculating the usable energy of an electric vehicle is: 。

[0017] In one exemplary embodiment of this disclosure, the mapping model between the incentive price and the electric vehicle response rate in the method is as follows:

[0018] Where x is the incentive price, f up f is the upper limit of response rate for EV users. down This represents the lower limit of response rate for EV users. The formula for calculating the response rate requirement of electric vehicles is:

[0019] Where, η resp Let ΔP be the target response rate for EV. load For the system power deficit, ΔE load For the system energy deficit, P EV,max E is the maximum power that a dispatchable electric vehicle can provide. EV,max The maximum energy that a dispatchable electric vehicle can provide; The formula for calculating the incentive price range is:

[0020]

[0021] in, X min It is the lowest incentive price. X max This is the highest incentive price.

[0022] In one exemplary embodiment of this disclosure, the method uses the following formula to calculate the response rate interval:

[0023] Monte Carlo sampling is performed, and random numbers are generated within the interval. N A sample set is obtained by taking response rate samples. ; After removing extreme data based on the 95% confidence interval, the calculation formula is as follows:

[0024]

[0025] Where, η avg z is the sample mean of the response rate. α / 2The value of quantile is 1.96 for the standard normal distribution, where N is the sample size and S is the number of samples. η CI represents the sample standard deviation of the response rate. low CI is the lower limit of the confidence interval. up This represents the upper limit of the confidence interval.

[0026] In one exemplary embodiment of this disclosure, the calculation formula for the Bayesian posterior probability correction strategy in the method is as follows:

[0027]

[0028]

[0029]

[0030] Where, p base The base probabilities are generated by Monte Carlo sampling, where α is the charging station availability coefficient, β is the road network accessibility coefficient, and γ is the battery availability coefficient; N online N represents the number of online, fault-free charging stations with V2G functionality. total N represents the total number of charging stations; v represents the current traffic flow on the road segment; c represents the maximum traffic capacity of the road segment; N sufficient,EV N represents the number of available EVs. total,EV This represents the total number of EVs.

[0031] In one exemplary embodiment of this disclosure, the electric vehicle is equivalent to virtual energy storage in the method according to the following formula:

[0032]

[0033]

[0034]

[0035] Among them, E g,total Let T be the total energy of the virtual energy storage g. g,avg P represents the travel time of the virtual energy storage g to each site. g,usable For virtual energy storage g, the available power, p g,usable Let N be the probability that the virtual energy storage g will travel to each site. g Let g be the number of EVs in the virtual energy storage.

[0036] In one exemplary embodiment of this disclosure, the expected return function under the complete power outage scenario is:

[0037] The expected payoff function for an incomplete power outage scenario is:

[0038] Where, π k For the response rate α k The probability of occurrence, C load P represents the economic loss per unit of power outage. load For power requirements, P V2G (t,α k ) indicates that the response rate is α k Let t represent the available power of the V2G system, E(L(x)) represent the expected economic loss when the scheduling incentive price is x, E(w(x)) represent the revenue when the scheduling incentive price is x, C represent the user scheduling cost, and x subsidy The price is subsidized by the superior platform; The formula for calculating marginal revenue is: in, x To incentivize prices, L ( x The expected return generated under the incentive price; The formula for selecting candidate intervals is:

[0039] in, w To smooth the width of the window; The formula for determining the optimal incentive price is: .

[0040] In one aspect of this disclosure, a device for optimizing V2G emergency power supply scheduling and incentive prices considering response rate uncertainty is provided, comprising: The spatial location optimization module is used for GIS-based spatial location optimization of electric vehicles in emergency scenarios. It includes collecting multi-source data to construct a spatial layer and performing multi-layer filtering, establishing an electric vehicle location scheduling optimization model with the goal of minimizing user inconvenience, introducing availability constraints to screen feasible scheduling schemes and solving for the optimal scheduling result. The scenario generation module is used to generate random behavior scenarios under price-induced induction by integrating Bayesian posterior probability. This includes establishing a mapping relationship between incentive price and electric vehicle response rate, determining the incentive price range based on the system supply and demand gap, randomly sampling within the range, and correcting the sampled scenarios based on multidimensional state coefficients using Bayesian posterior probability. The uncertainty decoupling module is used for hierarchical uncertainty decoupling and scenario reduction in two-layer clustering. It includes extracting electric vehicle features and response rate samples for cluster analysis, determining the optimal number of categories and solving the clustering parameters, converting the electric vehicle clustering results into virtual energy storage resources, and obtaining the typical user response rate and probability of occurrence under different incentive prices to complete scenario reduction. The price optimization module is used for optimizing the expected power loss risk and calculating the optimal price. This includes building differentiated expected revenue models for complete and incomplete power loss scenarios, selecting candidate price ranges based on marginal revenue analysis, and determining the optimal incentive price.

[0041] An exemplary embodiment of this disclosure provides a method for optimizing V2G emergency power supply scheduling and incentive prices while considering response rate uncertainty. The method includes: addressing the problem that V2G emergency power supply uses deterministic response assumptions and struggles to reflect the randomness of user behavior, firstly, optimizing vehicle spatial location based on GIS to determine the optimal scheduling with the goal of minimizing user inconvenience; secondly, establishing an incentive price-response rate mapping model, incorporating Bayesian posterior probability to correct for random scenarios; furthermore, employing two-layer clustering for hierarchical uncertainty decoupling and scenario reduction, treating vehicles as virtual energy storage; and finally, distinguishing between complete and incomplete power outage scenarios to construct an expected revenue model, determining the optimal incentive price based on marginal revenue analysis. This achieves synergistic optimization of price and scheduling, reduces uncertainty, improves resource allocation efficiency and computational speed, and minimizes power outage losses.

[0042] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0043] The above and other features and advantages of this disclosure will become more apparent from the detailed description of exemplary embodiments thereof with reference to the accompanying drawings.

[0044] Figure 1 A flowchart is shown below illustrating a method for optimizing V2G emergency power supply scheduling and incentive prices considering response rate uncertainty, according to an exemplary embodiment of this disclosure. Figure 2 The illustration shows a schematic diagram of a GIS-based EV spatial location optimization strategy in an emergency scenario, which is based on an exemplary embodiment of the present disclosure and considers the V2G emergency power supply time-conditioning and incentive price optimization method for response rate uncertainty. Figure 3 This illustration shows a schematic diagram of a random behavior scenario generation strategy that incorporates Bayesian posterior probabilities under price-induced optimization of V2G emergency power supply scheduling and incentive price, considering response rate uncertainty, according to an exemplary embodiment of this disclosure. Figure 4A schematic diagram of a hierarchical uncertainty decoupling and scenario reduction method of a two-layer GMM for optimizing V2G emergency power supply scheduling and incentive price considering response rate uncertainty is shown in an exemplary embodiment of the present disclosure. Figure 5 A schematic block diagram of a V2G emergency power supply scheduling and incentive price optimization device considering response rate uncertainty is shown according to an exemplary embodiment of the present disclosure. Detailed Implementation

[0045] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the embodiments set forth herein; rather, they are provided so that this disclosure will be thorough and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted.

[0046] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced without one or more of the specific details described, or other methods, components, materials, apparatuses, steps, etc., can be employed. In other instances, well-known structures, methods, apparatuses, implementations, materials, or operations are not shown or described in detail to avoid obscuring various aspects of this disclosure.

[0047] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, or in one or more software-hardened modules, or in different network and / or processor devices and / or microcontroller devices.

[0048] In this example embodiment, a method for optimizing V2G emergency power supply scheduling and incentive prices considering response rate uncertainty is first provided; (Refer to...) Figure 1 As shown, the method for optimizing V2G emergency power supply scheduling and incentive prices considering response rate uncertainty may include the following steps: Step S110, GIS-based spatial location optimization for electric vehicles in emergency scenarios, includes collecting multi-source data to construct a spatial layer and performing multi-layer filtering, establishing an electric vehicle location scheduling optimization model with the goal of minimizing user inconvenience, introducing availability constraints to screen feasible scheduling schemes and solving for the optimal scheduling result; Step S120: Generating random behavior scenarios under price-induced incorporation of Bayesian posterior probability, including establishing a mapping relationship between incentive price and electric vehicle response rate, determining the incentive price range based on the system supply and demand gap, randomly sampling within the range, and correcting the sampled scenarios based on multidimensional state coefficients using Bayesian posterior probability. Step S130, Hierarchical uncertainty decoupling and scenario reduction of two-layer clustering, includes extracting electric vehicle features and response rate samples for cluster analysis, determining the optimal number of categories and solving the clustering parameters, equating the electric vehicle clustering results to virtual energy storage resources, and obtaining the typical user response rate and probability of occurrence under different incentive prices to complete scenario reduction; Step S140, Optimization of expected power loss risk and calculation of optimal price, including constructing differentiated expected revenue models by distinguishing between complete power loss and incomplete power loss scenarios, selecting candidate price ranges based on marginal revenue analysis and determining the optimal incentive price.

[0049] An exemplary embodiment of this disclosure provides a method for optimizing V2G emergency power supply scheduling and incentive prices while considering response rate uncertainty. The method includes: addressing the problem that V2G emergency power supply, which employs deterministic response assumptions, struggles to reflect the randomness of user behavior; firstly, optimizing vehicle spatial location based on GIS to determine the optimal scheduling with the goal of minimizing user inconvenience; secondly, establishing an incentive price-response rate mapping model, incorporating Bayesian posterior probability to correct for random scenarios; furthermore, employing two-layer clustering for hierarchical uncertainty decoupling and scenario reduction, treating vehicles as virtual energy storage; and finally, distinguishing between complete and incomplete power outage scenarios to construct an expected revenue model, determining the optimal incentive price based on marginal revenue analysis. This achieves synergistic optimization of price and scheduling, reduces uncertainty, improves resource allocation efficiency and computational speed, and minimizes power outage losses.

[0050] The following will further explain a method for optimizing V2G emergency power supply scheduling and incentive prices that takes into account response rate uncertainty in this example embodiment.

[0051] Example 1: In step S110, GIS-based spatial location optimization for electric vehicles in emergency scenarios can be performed, including collecting multi-source data to construct a spatial layer and performing multi-layer filtering, establishing an electric vehicle location scheduling optimization model with the goal of minimizing user inconvenience, introducing availability constraints to screen feasible scheduling schemes and solving for the optimal scheduling result.

[0052] In this example embodiment, the objective function for optimizing electric vehicle location scheduling is:

[0053] Where D1 is the distance to the V2G site, S1 is the power consumption of the V2G site; D2 is the distance between the V2G site and the user's next destination, S2 is the power consumption between the V2G site and the user's next destination, and c1, c2, c3, and c4 are weighting coefficients.

[0054] In this example embodiment, the available SOC constraint in the method is:

[0055] Among them, SOC after The remaining SOC after electric vehicles are dispatched, SOC min This is the minimum remaining SOC (State of Charge) of the electric vehicle after arriving at the dispatch point. If the remaining SOC of the electric vehicle is lower than this minimum threshold after being dispatched, it is considered not dispatchable and the electric vehicle will not participate in the dispatch. The reachability constraint is:

[0056] Among them, SOC i The current SOC remains, SOC i,min The SOC is the SOC of the EV movement under the shortest path. If the remaining SOC of the electric vehicle is lower than the SOC of the EV movement under the shortest path, it is considered to be unreachable and the electric vehicle will not participate in the scheduling.

[0057] In this example embodiment, the formula for calculating the power consumption along the route is:

[0058] in, dis To minimize the distance traveled, E dis The average power consumption per kilometer is expressed in kWh / km. The formula for calculating the remaining SOC of an electric vehicle is:

[0059] Where SOC is the remaining SOC before being scheduled; cap is the battery capacity, E travel Energy consumption along the shortest path, expressed in kWh; The formula for calculating the usable energy of an electric vehicle is: 。

[0060] In step S120, a random behavior scenario can be generated under price-induced fusion of Bayesian posterior probability, including establishing a mapping relationship between incentive price and electric vehicle response rate, determining the incentive price range based on the system supply and demand gap, performing random sampling within the range, and correcting the sampled scenario based on multidimensional state coefficients using Bayesian posterior probability.

[0061] In this example embodiment, the mapping model between the incentive price and the electric vehicle response rate is as follows:

[0062] Where x is the incentive price, f up f is the upper limit of response rate for EV users. down This represents the lower limit of response rate for EV users. The formula for calculating the response rate requirement of electric vehicles is:

[0063] Where, η resp Let ΔP be the target response rate for EV. load For the system power deficit, ΔE load For the system energy deficit, P EV,max E is the maximum power that a dispatchable electric vehicle can provide. EV,max The maximum energy that a dispatchable electric vehicle can provide; The formula for calculating the incentive price range is:

[0064]

[0065] in, X min It is the lowest incentive price. X max This is the highest incentive price.

[0066] In this example embodiment, the formula for calculating the response rate interval in the method is:

[0067] Monte Carlo sampling is performed, and random numbers are generated within the interval. N A sample set is obtained by taking response rate samples. ; After removing extreme data based on the 95% confidence interval, the calculation formula is as follows:

[0068]

[0069] Where, η avgz is the sample mean of the response rate. α / 2 The value of quantile is 1.96 for the standard normal distribution, where N is the sample size and S is the number of samples. η CI represents the sample standard deviation of the response rate. low CI is the lower limit of the confidence interval. up This represents the upper limit of the confidence interval.

[0070] In this example embodiment, the calculation formula for the Bayesian posterior probability correction strategy in the method is:

[0071]

[0072]

[0073]

[0074] Where, p base The base probabilities are generated by Monte Carlo sampling, where α is the charging station availability coefficient, β is the road network accessibility coefficient, and γ is the battery availability coefficient; N online N represents the number of online, fault-free charging stations with V2G functionality. total N represents the total number of charging stations; v represents the current traffic flow on the road segment; c represents the maximum traffic capacity of the road segment; N sufficient,EV N represents the number of available EVs. total,EV This represents the total number of EVs.

[0075] In step S130, hierarchical uncertainty decoupling and scenario reduction of two-layer clustering can be achieved, including extracting electric vehicle features and response rate samples for cluster analysis, determining the optimal number of categories and solving the clustering parameters, equating the electric vehicle clustering results to virtual energy storage resources, and obtaining the typical user response rate and probability of occurrence under different incentive prices to complete scenario reduction.

[0076] In this example embodiment, the formula for the electric vehicle being equivalent to virtual energy storage is:

[0077]

[0078]

[0079]

[0080] Among them, E g,total Let T be the total energy of the virtual energy storage g. g,avg P represents the travel time of the virtual energy storage g to each site. g,usable For virtual energy storage g, the available power, p g,usableLet N be the probability that the virtual energy storage g will travel to each site. g Let g be the number of EVs in the virtual energy storage.

[0081] In step S140, the expected power loss risk can be optimized and the optimal price can be calculated, including distinguishing between complete power loss and incomplete power loss scenarios to construct a differentiated expected return model, selecting candidate price ranges based on marginal return analysis and determining the optimal incentive price.

[0082] In this example embodiment, the expected return function under the complete power outage scenario is:

[0083] The expected payoff function for an incomplete power outage scenario is:

[0084] Where, π k For the response rate α k The probability of occurrence, C load P represents the economic loss per unit of power outage. load For power requirements, P V2G (t,α k ) indicates that the response rate is α k Let t represent the available power of the V2G system, E(L(x)) represent the expected economic loss when the scheduling incentive price is x, E(w(x)) represent the revenue when the scheduling incentive price is x, C represent the user scheduling cost, and x subsidy The price is subsidized by the superior platform; The formula for calculating marginal revenue is: in, x To incentivize prices, L ( x The expected return generated under the incentive price; The formula for selecting candidate intervals is:

[0085] in, w To smooth the width of the window; The formula for determining the optimal incentive price is: .

[0086] Example 2: This invention presents a V2G emergency power supply scheduling and incentive price optimization method considering response rate uncertainty. It establishes a correlation mechanism between incentive prices and user response behavior, constructs a collaborative optimization framework for electric vehicle scheduling and prices under response rate uncertainty, and effectively represents and handles user behavior uncertainty. By reducing the impact of response rate uncertainty on scheduling decisions, it improves the efficiency of emergency resource allocation and the speed of optimization solutions, achieving efficient aggregation and rational scheduling of electric vehicle resources. This reduces power outage losses and power supply costs while ensuring the reliability of emergency power supply, providing a fast, economical, and engineering-approved solution for large-scale electric vehicle participation in distribution network emergency power supply.

[0087] In this example embodiment, the complete technical solution of the present invention is as follows: Figure 2 As shown, it mainly consists of three parts: GIS-based EV spatial location optimization in emergency scenarios, a random behavior scenario generation strategy that integrates Bayesian posterior probability under price-induced conditions, and a hierarchical uncertainty decoupling and scenario reduction method for two-layer GMM.

[0088] In this example embodiment, GIS-based EV spatial location optimization in emergency scenarios includes: Step 1.1 GIS Data Collection in Emergency Scenarios Obtain power grid fault information to determine whether the current state is a complete power outage or a partial power outage. After a power outage or shortage occurs, data collection is performed, including: electric vehicle ID, real-time location data, SOC, next destination, and battery capacity; charging pile number, location, rated discharge power, and current availability; road network segment ID, road alignment, segment capacity, and road traffic saturation; and data on the distribution network topology and load affected by the power outage or shortage.

[0089] Step 1.2 Construction of GIS Vector Layers Four vector layers were constructed based on the collected GIS data: the EV layer, charging pile layer, and load node layer were point features, and the road network layer was line features. A unified coordinate system was used to ensure the accuracy of distance calculation and spatial analysis.

[0090] Step 1.3 Spatial Range Filtering Based on the distribution network fault information or the affected range of load nodes, the power outage area is delineated by the convex hull of key load nodes and the area enclosed by adjacent road sections. It is then determined whether the EV layer, charging pile layer, load node layer and road network layer are within the power outage area. Only the data within the power outage area is retained to obtain a simplified dataset, which provides data support for subsequent analysis.

[0091] Step 1.4 Charging station availability filtering Data filtering is performed on the charging pile vector layer to remove offline, faulty, and V2G charging piles without discharge capability, thereby obtaining a set of available V2G charging piles.

[0092] Step 1.5 Road Network Accessibility Filtering Determine if the current road traffic saturation in the network is less than 0.95. If it is greater than 0.95, the road is considered congested and therefore unreachable.

[0093] Step 1.6 EV Battery Filtering If the current SOC of the EV is below 20%, it is considered unreachable; calculate the shortest distance from the EV to all charging stations, and if the remaining battery power is insufficient for the EV to reach all charging stations, it is considered unreachable.

[0094] Step 1.8 Design and optimize the objective function To optimize EV location scheduling, an objective function is established by acquiring the set of schedulable EVs, available charging stations, key load nodes, and feasible paths, with the goal of achieving the fastest EV scheduling while minimizing the impact on the user's subsequent journey. The degree of inconvenience to the user is defined using the distance offset between the scheduled destination and the user's next destination, as well as power consumption. The objective function formula is as follows:

[0095] Where D1 is the distance to the V2G site, S1 is the power consumption of the V2G site; D2 is the distance between the V2G site and the user's next destination, S2 is the power consumption between the V2G site and the user's next destination, and c1, c2, c3, and c4 are weighting coefficients. Step 1.9 Design constraints and filter out unreasonable scheduling schemes.

[0096] We introduce available SOC constraints to constrain the scheduling results and remove unreasonable schemes. The constraints are as follows:

[0097] SOC after The remaining SOC after electric vehicles are dispatched, SOC min This is the minimum remaining SOC value of an electric vehicle after it arrives at the dispatch point. If the remaining SOC of an electric vehicle is lower than this minimum threshold after it is dispatched, it is considered not dispatchable and the EV will not participate in the dispatch.

[0098] reachability constraints

[0099] SOC i The current SOC remains, SOC i,minThe SOC is the SOC of the EV's movement along the shortest path. If the remaining SOC of an electric vehicle is lower than the SOC of the EV's movement along the shortest path, it is considered unreachable and the EV will not participate in the scheduling.

[0100] Step 1.10 Solving the optimal EV scheduling scheme based on heuristic algorithms The optimization objective is used as the fitness function, and a heuristic algorithm is employed to solve it. EVs that do not meet the constraints are not scheduled. The optimal V2G site for each EV is obtained, and the solution with the minimum fitness function is taken as the optimal solution. Based on the optimal scheduling results, the available SOC, available battery power, and power consumption during the journey are calculated for each EV.

[0101] The formula for calculating power consumption during the journey is as follows:

[0102] in dis To minimize the distance traveled, E dis Average power consumption per kilometer, in kWh / km EV Remaining SOC Calculation Formula

[0103] Where SOC is the remaining SOC before receiving scheduling; cap is the battery capacity, E travel Energy consumption along the shortest path, expressed in kWh. The formula for calculating the usable energy of an EV is as follows:

[0104] In the embodiments of this example, as Figure 3 As shown, the strategies for generating random behavior scenarios based on price-induced Bayesian posterior probabilities include: Step 2.1 Modeling the Incentive Price-EV Response Rate Based on consumer psychology, an incentive price and discharge response model is established, with the following formula:

[0105] Where x is the incentive price, fup is the upper limit of EV user response rate, and fdown is the lower limit of EV user response rate.

[0106] Step 2.1 Solve for the incentive price plan range The electric vehicle response rate requirement is calculated based on the load power shortage and load energy demand, and the incentive price range is determined by combining the incentive price and discharge response model.

[0107] The formula for calculating EV response rate requirements is as follows:

[0108] Where ηresp is the target EV response rate, and ΔP load For the system power deficit, ΔE load For the system energy deficit, P EV,max The maximum power E that a dispatchable electric vehicle can provide EV,max This represents the maximum energy that a dispatchable electric vehicle can provide.

[0109] Grid interval calculation formula

[0110]

[0111] Where Xmin is the minimum incentive price and Xmax is the maximum incentive price.

[0112] Step 2.2 Generation of Price-Induced Random Behavior Scenarios Based on the incentive price and discharge response model, a response rate range is constructed, and the calculation formula is as follows:

[0113] Monte Carlo sampling is performed to randomly generate N response rate samples within the interval, resulting in a sample set.

[0114] After removing extreme data based on the 95% confidence interval, the calculation formula is as follows:

[0115]

[0116] Where ηavg is the sample mean of the response rate, z α / 2 denoted as the standard normal distribution quantile, with a value of 1.96, where N is the sample size and Sη is the sample standard deviation of the response rate. CI low CI is the lower limit of the confidence interval. up This represents the upper limit of the confidence interval.

[0117] Step 2.3 Random Behavior Scenario Correction Strategy Based on Bayesian Posterior Probability The random behavior sampling scenario is corrected based on Bayesian posterior probability, and the calculation formula is as follows:

[0118]

[0119]

[0120]

[0121] Where p base The base probabilities are generated by Monte Carlo sampling, where α is the charging station availability coefficient, β is the road network accessibility coefficient, and γ is the battery availability coefficient; N online N represents the number of online, fault-free charging stations with V2G functionality. total N represents the total number of charging stations; v represents the current traffic flow on the road segment; c represents the maximum traffic capacity of the road segment; N sufficient,EV N represents the number of available EVs. total,EV This represents the total number of EVs.

[0122] In the embodiments of this example, as Figure 4 As shown, the hierarchical uncertainty decoupling and scene reduction methods of the two-layer GMM include: Gaussian mixture clustering (GMM) was used to cluster the generated incentive price-EV user response rate and electric vehicle scheduling results, respectively.

[0123] Based on the EV scheduling optimization results, the distance to the V2G station, the distance to the next destination, the remaining dischargeable energy, the current SOC, and the probability of going to each station are extracted for each EV to construct an EV feature matrix. A response rate sample set is constructed by extracting the price-response rate set from the Monte Carlo generation scenario.

[0124] The EV feature matrix and the response rate sample set are input into the GMM model. Assuming the data follows K Gaussian distributions, AIC is used during clustering to determine the number of V2G response rate categories, and BIC is used to determine the number of EV categories. The calculation formula is as follows:

[0125]

[0126] The EM algorithm is used to solve for the GMM parameters, and the cluster centers and occurrence probabilities of response rates under different incentive prices are obtained. The EV cluster centers are also obtained.

[0127] Based on the clustering results, electric vehicles are equivalent to virtual energy storage, and the energy storage equivalence formula is as follows:

[0128]

[0129]

[0130]

[0131] Where E g,total Let T be the total energy of the virtual energy storage g. g,avg P represents the travel time of the virtual energy storage g to each site.g,usable For virtual energy storage g, the available power, p g,usable Let N be the probability that the virtual energy storage g will travel to each site. g Let g be the number of EVs in the virtual energy storage.

[0132] In this example embodiment, the expected optimization of power outage risk and the calculation of the optimal price include: Step 4.1 Construction of Expected Return Function in Multiple Scenarios In considering the emergency dispatch of electric vehicles participating in V2G, the system is divided into complete power outage scenarios and incomplete power outage scenarios for different power supply states, and differentiated expected revenue functions are constructed based on these scenarios.

[0133] In a complete power outage scenario, the optimization objective is to minimize the expected value of the economic loss due to power outage, as shown in the following formula:

[0134] In scenarios of incomplete power outages, the system optimization objective shifts from "restoring power supply" to "coordinated optimization of economics and safety." In this case, the participation of electric vehicles in V2G not only aims to reduce load losses but also requires comprehensive consideration of grid power purchase costs and user-side response costs. The formula is as follows:

[0135]

[0136] Where π k For the response rate α k The probability of occurrence, C load P represents the economic loss per unit of power outage. load For power requirements, P V2G (t,α k ) indicates that the response rate is α k Let t represent the available power of the V2G system, E(L(x)) represent the expected economic loss when the scheduling incentive price is x, E(w(x)) represent the revenue when the scheduling incentive price is x, C represent the user scheduling cost, and x subsidy The price is subsidized by the superior platform.

[0137] Step 4.2 Calculate the payoff function at different prices and solve for the optimal price. Based on two scenarios—complete power outage and incomplete power outage—the expected value of the system at different price levels is calculated according to the price-revenue function of the optimization objective framework. The marginal revenue curve, representing the revenue change per unit price change, is obtained by taking the first difference over continuous price points. The calculation formula is as follows:

[0138] Where x is the incentive price, and L(x) is the expected return generated at the incentive price.

[0139] Based on this, the price range with the smallest marginal revenue fluctuation is selected as the candidate range, representing that the impact of price adjustments on system revenue is relatively stable within this range, avoiding instability in scheduling decisions due to excessive fluctuations. The calculation formula is as follows:

[0140] Where w is the width of the smoothing window and the number of neighboring points considered during the smoothing process.

[0141] Ultimately, the lowest price value is selected from the candidate range, balancing system economy and user participation, to determine the optimal incentive price. The calculation formula is as follows:

[0142] In this example embodiment, the present invention proposes a method for optimizing V2G emergency power supply scheduling and incentive prices considering response rate uncertainty. Addressing the problem that existing emergency power supply research commonly employs deterministic response rate assumptions, which fail to accurately reflect actual user participation behavior, this invention establishes a mapping model between incentive prices and electric vehicle user response rates. Furthermore, it combines Monte Carlo random sampling techniques to generate user response behaviors across multiple scenarios, achieving a refined characterization of user participation uncertainty. Simultaneously, a Bayesian posterior probability correction mechanism is introduced, comprehensively considering multi-dimensional factors such as charging pile status, vehicle status, and road network accessibility to dynamically update the actual user response probability, improving the accuracy and reliability of response behavior modeling.

[0143] To address the problem of high dimensionality and computational complexity in scheduling models caused by the coupling between user response rate uncertainty and the uncertainty of the actual set of responding vehicles, this invention employs the Gaussian Mixture Model (GMM) clustering method to cluster and reduce a large number of random response scenarios, extracting representative typical response scenarios. This significantly reduces the computational scale while preserving the probability distribution characteristics of the scenarios. Simultaneously, based on the vehicle scheduling results, discrete-response electric vehicles are aggregated and equated to virtual energy storage resources, decoupling user response behavior from the emergency power supply scheduling process. This further reduces the computational difficulty caused by uncertainty propagation, improving the model's computational efficiency and engineering practicality.

[0144] Compared to existing technologies, this invention achieves coordinated optimization of incentive pricing and emergency power supply dispatch while fully considering the uncertainty of electric vehicle user responses. On the one hand, price guidance enhances user participation in V2G emergency power supply, increasing the scale of dispatchable resources. On the other hand, through response rate scenario modeling, probability correction, and virtual energy storage aggregation technologies, it effectively reduces the impact of prediction errors and random fluctuations on dispatch decisions, achieving rational allocation of power supply resources in power outage areas. Ultimately, while ensuring computational efficiency, it significantly reduces the expected power loss caused by power outages, improves the emergency power supply capacity, resource utilization efficiency, and system operation economy of the distribution network, providing an efficient, reliable, and engineering-approved solution for large-scale electric vehicle participation in distribution network emergency protection.

[0145] It should be noted that although the steps of the method in this disclosure are described in a specific order in the accompanying drawings, this does not require or imply that the steps must be performed in that specific order, or that all the steps shown must be performed to achieve the desired result. Additional or alternative steps may be omitted, multiple steps may be combined into one step, and / or a step may be broken down into multiple steps.

[0146] Furthermore, in this example embodiment, a device for optimizing V2G emergency power supply scheduling and incentive prices that considers response rate uncertainty is also provided. (Refer to...) Figure 5 As shown, the V2G emergency power supply scheduling and incentive price optimization device 200 considering response rate uncertainty may include: a spatial location optimization module 210, a scene generation module 220, an uncertainty decoupling module 230, and a price optimization module 240. Wherein: The spatial location optimization module 210 is used for GIS-based spatial location optimization of electric vehicles in emergency scenarios. It includes collecting multi-source data to construct a spatial layer and performing multi-layer filtering, establishing an electric vehicle location scheduling optimization model with the goal of minimizing user inconvenience, introducing availability constraints to screen feasible scheduling schemes and solving for the optimal scheduling result. The scene generation module 220 is used to generate random behavior scenes under price-induced induction and integrate Bayesian posterior probability. This includes establishing a mapping relationship between incentive price and electric vehicle response rate, determining the incentive price range based on the system supply and demand gap, randomly sampling within the range, and correcting the sampled scene based on multidimensional state coefficients and Bayesian posterior probability. Uncertainty decoupling module 230 is used for hierarchical uncertainty decoupling and scenario reduction in two-layer clustering. It includes extracting electric vehicle features and response rate samples for cluster analysis, determining the optimal number of categories and solving the clustering parameters, converting the electric vehicle clustering results into virtual energy storage resources, and obtaining the typical user response rate and probability of occurrence under different incentive prices to complete scenario reduction. The price optimization module 240 is used for optimizing the expected power loss risk and calculating the optimal price. This includes building a differentiated expected return model by distinguishing between complete power loss and incomplete power loss scenarios, selecting candidate price ranges based on marginal return analysis, and determining the optimal incentive price.

[0147] The specific details of each of the V2G emergency power supply scheduling and incentive price optimization device modules considering response rate uncertainty have been described in detail in the corresponding V2G emergency power supply scheduling and incentive price optimization method considering response rate uncertainty, so they will not be repeated here.

[0148] It should be noted that although several modules or units of a V2G emergency power supply scheduling and incentive price optimization device 200 considering response rate uncertainty have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0149] Furthermore, the above figures are merely illustrative of the processes included in the method according to exemplary embodiments of the present invention, and are not intended to be limiting. It is readily understood that the processes shown in the above figures do not indicate or limit the temporal order of these processes. Additionally, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0150] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and embodiments are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the claims.

[0151] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

Claims

1. A method for optimizing V2G emergency power supply scheduling and incentive prices considering response rate uncertainty, characterized in that, The method includes: In emergency scenarios, GIS-based spatial location optimization for electric vehicles includes collecting multi-source data to construct a spatial layer and performing multi-layer filtering, establishing an electric vehicle location scheduling optimization model with the goal of minimizing user inconvenience, introducing availability constraints to screen feasible scheduling schemes and solving for the optimal scheduling result. The generation of random behavior scenarios induced by price and incorporating Bayesian posterior probability includes establishing a mapping relationship between incentive price and electric vehicle response rate, determining the incentive price range based on the system supply and demand gap, conducting random sampling within the range, and correcting the sampled scenarios based on multidimensional state coefficients using Bayesian posterior probability. Hierarchical uncertainty decoupling and scenario reduction of two-layer clustering includes extracting electric vehicle features and response rate samples for cluster analysis, determining the optimal number of categories and solving the clustering parameters, equating the electric vehicle clustering results to virtual energy storage resources, and obtaining the typical user response rate and probability of occurrence under different incentive prices to complete scenario reduction. The optimization of expected power loss risk and the calculation of optimal price include constructing differentiated expected return models for complete power loss and incomplete power loss scenarios, selecting candidate price ranges based on marginal return analysis, and determining the optimal incentive price.

2. The method as described in claim 1, characterized in that, In the method, the objective function for optimizing electric vehicle location scheduling is: Where D1 is the distance to the V2G site, S1 is the power consumption of the V2G site; D2 is the distance between the V2G site and the user's next destination, S2 is the power consumption between the V2G site and the user's next destination, and c1, c2, c3, and c4 are weighting coefficients.

3. The method as described in claim 1, characterized in that, In the method described above, the available SOC constraint is: Among them, SOC after The remaining SOC after electric vehicles are dispatched, SOC min This is the minimum remaining SOC (State of Charge) of the electric vehicle after arriving at the dispatch point. If the remaining SOC of the electric vehicle is lower than this minimum threshold after being dispatched, it is considered not dispatchable and the electric vehicle will not participate in the dispatch. The reachability constraint is: Among them, SOC i The current SOC remains, SOC i,min The SOC is the SOC of the EV movement under the shortest path. If the remaining SOC of the electric vehicle is lower than the SOC of the EV movement under the shortest path, it is considered to be unreachable and the electric vehicle will not participate in the scheduling.

4. The method as described in claim 1, characterized in that, In the method, the formula for calculating the power consumption along the route is: in, dis To minimize the distance traveled, E dis The average power consumption per kilometer is expressed in kWh / km. The formula for calculating the remaining SOC of an electric vehicle is: Where SOC is the remaining SOC before being scheduled; cap is the battery capacity, E travel Energy consumption along the shortest path, expressed in kWh; The formula for calculating the usable energy of an electric vehicle is: 。 5. The method as described in claim 1, characterized in that, In the method, the mapping model between the incentive price and the electric vehicle response rate is as follows: Where x is the incentive price, f up f is the upper limit of response rate for EV users. down This represents the lower limit of response rate for EV users. The formula for calculating the response rate requirement of electric vehicles is: Where, η resp Let ΔP be the target response rate for EV. load For the system power deficit, ΔE load For the system energy deficit, P EV,max E is the maximum power that a dispatchable electric vehicle can provide. EV,max The maximum energy that a dispatchable electric vehicle can provide; The formula for calculating the incentive price range is: in, X min It is the lowest incentive price. X max This is the highest incentive price.

6. The method as described in claim 1, characterized in that, In the method, the formula for calculating the response rate interval is: Monte Carlo sampling is performed, and random numbers are generated within the interval. N A sample set is obtained by taking response rate samples. ; After removing extreme data based on the 95% confidence interval, the calculation formula is as follows: Where, η avg z is the sample mean of the response rate. α / 2 The value of quantile is 1.96 for the standard normal distribution, where N is the sample size and S is the number of samples. η CI represents the sample standard deviation of the response rate. low CI is the lower limit of the confidence interval. up This represents the upper limit of the confidence interval.

7. The method as described in claim 1, characterized in that, In the method, the calculation formula for the Bayesian posterior probability correction strategy is as follows: Where, p base The base probabilities are generated by Monte Carlo sampling, where α is the charging station availability coefficient, β is the road network accessibility coefficient, and γ is the battery availability coefficient; N online N represents the number of online, fault-free charging stations with V2G functionality. total N represents the total number of charging stations; v represents the current traffic flow on the road segment; c represents the maximum traffic capacity of the road segment; N sufficient,EV N represents the number of available EVs. total,EV This represents the total number of EVs.

8. The method as described in claim 1, characterized in that, In the method, the formula for equating the electric vehicle to virtual energy storage is: Among them, E g,total Let T be the total energy of the virtual energy storage g. g,avg P represents the travel time of the virtual energy storage g to each site. g,usable For virtual energy storage g, the available power, p g,usable Let N be the probability that the virtual energy storage g will travel to each site. g Let g be the number of EVs in the virtual energy storage.

9. The method as described in claim 1, characterized in that, In the method, the expected return function under the complete power outage scenario is: The expected payoff function for an incomplete power outage scenario is: Where, π k For the response rate α k The probability of occurrence, C load P represents the economic loss per unit of power outage. load For power requirements, P V2G (t,α k ) indicates that the response rate is α k Let t represent the available power of the V2G system, E(L(x)) represent the expected economic loss when the scheduling incentive price is x, E(w(x)) represent the revenue when the scheduling incentive price is x, C represent the user scheduling cost, and x subsidy The price is subsidized by the superior platform; The formula for calculating marginal revenue is: in, x To incentivize prices, L ( x The expected return generated under the incentive price; The formula for selecting candidate intervals is: in, w To smooth the width of the window; The formula for determining the optimal incentive price is: 。 10. A device for optimizing V2G emergency power supply scheduling and incentive price considering response rate uncertainty, characterized in that, The device includes: The spatial location optimization module is used for GIS-based spatial location optimization of electric vehicles in emergency scenarios. It includes collecting multi-source data to construct a spatial layer and performing multi-layer filtering, establishing an electric vehicle location scheduling optimization model with the goal of minimizing user inconvenience, introducing availability constraints to screen feasible scheduling schemes and solving for the optimal scheduling result. The scenario generation module is used to generate random behavior scenarios under price-induced induction by integrating Bayesian posterior probability. This includes establishing a mapping relationship between incentive price and electric vehicle response rate, determining the incentive price range based on the system supply and demand gap, randomly sampling within the range, and correcting the sampled scenarios based on multidimensional state coefficients using Bayesian posterior probability. The uncertainty decoupling module is used for hierarchical uncertainty decoupling and scenario reduction in two-layer clustering. It includes extracting electric vehicle features and response rate samples for cluster analysis, determining the optimal number of categories and solving the clustering parameters, converting the electric vehicle clustering results into virtual energy storage resources, and obtaining the typical user response rate and probability of occurrence under different incentive prices to complete scenario reduction. The price optimization module is used for optimizing the expected power loss risk and calculating the optimal price. This includes building differentiated expected revenue models for complete and incomplete power loss scenarios, selecting candidate price ranges based on marginal revenue analysis, and determining the optimal incentive price.

Images