Residential community charging pile planning method and system for operator revenue maximization

Abstract

The invention discloses a residential area charging pile planning method and system for operator revenue maximization in the technical field of charging pile planning. The method comprises the following steps: constructing a dynamic three-party master-slave game framework which takes an operator as a master party and takes property management and residents as slave parties; according to the dynamic three-party master-slave game framework, performing target priority division, and constructing a revenue maximization target function of each party; the improved NSGA-II algorithm is utilized to solve the benefit maximization objective function of each party in a balanced mode, and a leading edge solution set is obtained; and utilizing a Monte Carlo algorithm to carry out future income prediction and risk quantification on the frontier solution set to obtain a feasible charging pile construction scheme strategy set. According to the method, a scientific decision basis can be provided for an operator in a complex dynamic environment, construction and operation risks are reduced, and efficient utilization and sustainable development of charging facilities in residential areas are promoted.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of charging pile planning, in particular to a residential area charging pile planning method and system for maximizing operator revenue. BACKGROUND

[0002] With the rapid popularization of electric vehicles, the demand for residential area charging piles has grown significantly, and the planning and construction scale of supporting charging facilities continues to expand. However, the large-scale deployment of charging piles has also triggered a series of complex problems: imbalance between supply and demand of charging resources, intensification of multi-party interest conflicts, and frequent planning lag under dynamic subsidies and user demand fluctuations, resulting in high construction costs and low operational efficiency. Especially in the residential area scenario, the unreasonable layout of charging piles, the redundancy of capacity configuration, and the unquantified risk of later revenue fluctuations have become increasingly prominent, seriously hindering the efficient use and sustainable development of charging facilities.

[0003] The current charging pile planning method has significant technical bottlenecks: first, traditional single-objective optimization models are difficult to coordinate the interests of operators, property, and residents, resulting in low cooperation rate and difficulty in implementation of the pile construction scheme; second, existing methods lack adaptive modeling of key factors such as subsidy reduction and user charging behavior fluctuations, and the planning strategy is rigid, which cannot meet the long-term operation demand, resulting in high investment decision blindness and weak risk resistance. In the existing technology, although some planning methods based on game theory have been tried, few of them have considered the impact of property on charging facility planning, making it difficult to achieve balanced interests of multiple parties and optimal allocation of resources. Therefore, how to build a charging pile planning system that takes into account the interests of multiple parties, adapts to subsidy changes, and is risk-controllable in a complex dynamic environment has become a core problem that needs to be solved.

[0004] The existing charging pile planning technology has the following three problems:

[0005] Problem 1: Lack of property game subject; existing active game models (such as: a kind of electric vehicle optimization charging dispatching method and system based on master-slave game) only consider the game between operators and users, ignoring the key role of property in charging pile access and site coordination, resulting in practical problems such as property obstruction in construction in actual application.

[0006] Problem 2: Single target adaptation, solution efficiency and solution set distribution are difficult to balance; traditional NSGA-II algorithm has uneven solution set distribution when target dimension ≥3, and the convergence speed rapidly decreases when the solution set size exceeds 500. The strategy of existing research (such as: electric vehicle charging station site selection method based on improved NSGA-II algorithm) needs to calculate the ASF function and dynamic screening mechanism, and the algorithm running time is increased by about 20% compared with traditional NSGA-II.

[0007] Problem 3: Weak dynamic adaptability of the scheme; existing research (such as: charging service fee game pricing model research in residential area scenario) simplifies user behavior as static response, does not consider dynamic factors such as subsidy, demand, market, and is difficult to adapt to the real-time changing market environment. SUMMARY

[0008] The purpose of the present application is to overcome the problems existing in the prior art, provide a residential area charging pile planning method and system for maximizing the income of operators, effectively solve the core problems such as the difficulty in coordinating the conflict of interests of multiple parties in traditional planning, the insufficient adaptability of subsidies and demand fluctuations, and the lack of risk quantification, provide scientific decision-making basis for operators in complex dynamic environment, reduce construction and operation risks, and promote efficient use and sustainable development of residential area charging facilities.

[0009] To solve the above technical problems, the present application is realized by using the following technical solutions:

[0010] In a first aspect, the present application provides a residential area charging pile planning method for maximizing the income of operators, comprising:

[0011] A dynamic three-party master-slave game framework is constructed, with the operator as the leading party and the property and residents as the subordinate parties;

[0012] According to the dynamic three-party master-slave game framework, target priority is divided, and a maximum income objective function of each party is constructed;

[0013] The improved NSGA-II algorithm is used to balance the maximum income objective function of each party to obtain a set of frontier solutions;

[0014] The Monte Carlo algorithm is used to predict future income and quantify risks of the set of frontier solutions to obtain a set of feasible charging pile construction scheme strategies.

[0015] Optionally, the dynamic three-party master-slave game framework constructed with the operator as the leading party and the property and residents as the subordinate parties comprises:

[0016] The operator is the leading party, and the property and residents are the subordinate parties, a three-party game strategy is constructed, including the operator game strategy, the property game strategy and the resident game strategy;

[0017] According to the three-party game strategy, a three-party income matrix is constructed;

[0018] According to the three-party income matrix, a three-party income function is constructed, including the operator income function, the property income function and the resident income function.

[0019] Optionally, the expression of the operator game strategy is:

[0020] ,

[0021] wherein, represents an operator, represents an operator revenue, represents that the operator builds charging piles, obtains charging service fees, and pays equipment depreciation costs, operation costs, maintenance costs, and site rental costs; represents that the operator does not build charging piles, and has no costs and revenues;

[0022] The expression of the property game strategy is:

[0023] ,

[0024] wherein, represents a property, represents a property revenue, represents that the property cooperates with the operator, obtains charging service fee sharing and site rental income, and pays management costs; represents that the property does not cooperate with the operator, the operator cannot build piles, and has no income and no costs;

[0025] The expression of the resident game strategy is:

[0026] ,

[0027] wherein, represents a resident, represents a property revenue, represents that the resident uses charging piles, obtains fuel substitution revenue, and pays charging costs; represents that the resident does not use charging piles, and has no fuel substitution revenue;

[0028] The operator revenue function is:

[0029] ,

[0030] wherein, represents a charging service fee, represents a property sharing ratio, represents an average annual charging volume of a single vehicle of a resident, represents a community electric vehicle service volume, represents an electricity subsidy rate, represents an equipment annual depreciation rate, represents a single-pile construction cost, represents a single-pile annual operation cost, represents a single-pile site rental fee, represents a number of built piles, represents a single-pile annual maintenance cost, represents a constraint, represents the maximum power of a single charging pile, represents the grid capacity, represents the proportionality coefficient, represents the number of parking spaces, represents the upper limit of local service pricing, represents the upper limit of local electricity annual subsidies;

[0031] The property income function is:

[0032] ,

[0033] wherein, represents the upper limit of the property proportion, represents the management cost of a single charging pile;

[0034] The resident income function is:

[0035] ,

[0036] wherein, represents the annual average fuel replacement cost of a single car, represents the basic electricity price, represents the cost coefficient;

[0037] The community electric vehicle service quantity is obtained by the following formula:

[0038] ,

[0039] wherein, represents the average charging service coverage, represents the price sensitivity coefficient, represents the upper limit of local service pricing, represents the community electric vehicle ownership.

[0040] Optionally, the each-party income maximization objective function is:

[0041] ,

[0042] wherein, represents the total income of the three parties, represents the operator, represents the property, represents the resident, represents the Min-Max normalized operator income, represents the Min-Max normalized property income, represents the Min-Max normalized resident income, represents the operator income weight coefficient, represents a property income weight coefficient, represents a resident income weight coefficient, represents a minimum value function, represents a charging service fee, represents a local service fee pricing upper limit, represents a property sharing ratio, represents a property sharing ratio upper limit, represents a number of charging piles, represents a proportion coefficient, represents a number of parking spaces, represents a power grid capacity, represents a maximum power of a single charging pile.

[0043] Optionally, the improved NSGA-II algorithm is used to balance the solving of the maximum income objective function of each party to obtain a set of front solutions, including:

[0044] A population size K is set, and individuals satisfying all constraints are randomly generated according to the maximum income objective function of each party to obtain a set of feasible solutions satisfying the constraint conditions , each feasible solution containing decision variables ; wherein, , represents a feasible solution of the individual , represents a feasible solution of the individual , represents a number of charging piles, represents a charging service fee, represents a property sharing ratio;

[0045] According to the set of feasible solutions , the following steps are iteratively executed until a preset iteration stopping condition is met:

[0046] The set of feasible solutions is sorted by a non-dominated sorting algorithm, all non-dominated individuals in the population are classified into the first front layer , and their dominance levels are marked as 0, and the classified individuals are removed in turn, new non-dominated solutions are recursively identified in the remaining population, and are classified, until all individuals are classified, and the output classification results are output in descending order of dominance level ; wherein, , represents the th front layer, represents the th front layer;

[0047] According to the classification results Elite selection is performed from the front layer The number of individuals is accumulated until the first cumulative number exceeds the preset scale is located The front layer ;

[0048] The crowding distance of the individual in the front layer and its adjacent individual and is calculated respectively , the individuals in the front layer are sorted according to the crowding distance , the individuals with small crowding distance are deleted and the population is updated until the total number of individuals in the front layer is not greater than , recorded as the front layer , and the front layer result is obtained; wherein, , represents the number of individuals in the front layer ; ;

[0049] The front layer result is taken as the parent population to perform crossover and mutation to obtain the offspring population ; wherein, the mutation probability is , , represents the initial value of the mutation probability, represents the current iteration number, represents the total number of iterations, and the mutation range is , , represents the initial value of the mutation range;

[0050] The parent population and the offspring population are merged and non-dominated sorting is performed to obtain the new generation population by hierarchical truncation, which is taken as the initial population of the next round of iteration;

[0051] After the iteration is stopped, the front solution set is obtained, each front solution including the optimal income set and the optimal configuration solution set , wherein, represents the optimal operator income, represents the optimal property income, represents the optimal resident income, represents the optimal number of charging piles, represents the optimal charging service fee, This indicates the optimal property revenue sharing ratio.

[0052] Optionally, the step of using the Monte Carlo algorithm to predict future returns and quantify risks on the frontier solution set to obtain a set of feasible charging pile construction strategies includes:

[0053] Uncertain variables are selected based on the dynamic three-party master-slave game framework, and probability distribution assumptions are made for the uncertain variables to obtain a variable set;

[0054] Based on the aforementioned frontier solution set and variable set, the expected annual revenue and revenue risk of the operator in the next T years are calculated using M-time Monte Carlo simulations.

[0055] Based on the operator's expected annual revenue and revenue risk, the feasibility of the strategy is verified by constraint, and a set of feasible charging pile construction schemes is output.

[0056] Optionally, the uncertain variables include demand variables, subsidy variables, and market variables;

[0057] The demand variables include the volume of electric vehicle services in the community. Average annual charging volume per residential vehicle The electric vehicle service volume of the community Follow the mean Standard deviation is log-normal distribution ,in, Indicates the year sequence number. This indicates the average annual penetration growth rate. Indicates volatility; the average annual charging amount per residential vehicle. Follows uniform distribution , This indicates the minimum charging capacity for a single residential vehicle. This indicates the maximum charging capacity for a single residential vehicle.

[0058] The subsidy variables include the electricity subsidy rate. The electricity subsidy rate Linear decrease, slope reduction rate ;

[0059] The market variables include the local service fee pricing cap. Annual operating cost per pile Single pile site rental fee Annual maintenance cost per pile The local service fee pricing cap Follow the mean Standard deviation is normal distribution ,in, denotes the mean of local service fee, denotes the standard deviation of local service fee; the single-pile annual operation cost subject to a lognormal distribution with mean and standard deviation wherein, denotes the growth rate of single-pile annual operation cost, denotes the volatility rate of operation cost; the single-pile site rental fee subject to a lognormal distribution with mean and standard deviation wherein, denotes the growth rate of single-pile site rental fee, denotes the volatility rate of single-pile site rental fee; the single-pile annual maintenance cost subject to a gamma distribution with shape parameter and scale parameter wherein, denotes the single-pile life span, denotes the aging risk rate.

[0060] Optionally, the operator annual expected revenue is calculated by the following formula:

[0061] ,

[0062] wherein, denotes the optimal number of piles, denotes the optimal charging service fee, denotes the optimal property sharing ratio;

[0063] The revenue risk is calculated by the following formula:

[0064] ,

[0065] wherein, denotes the value at risk, denotes the conditional value at risk, denotes the average annual revenue within T years, denotes the average annual revenue variance within T years, denotes the confidence interval, denotes the corresponding normal quantile, denotes the standard normal density function.

[0066] Optionally, the strategy feasibility constraints include revenue benchmark constraints, multi-dimensional risk condition constraints, and multi-party cooperation sustainability condition constraints;

[0067] The revenue benchmark constraints are: ​​

[0068] ,

[0069] wherein, represents an industry benchmark return rate;

[0070] The multi-dimensional risk condition constraint is:

[0071] ,

[0072] wherein, represents an industry threshold, represents a risk threshold, represents a conditional risk threshold;

[0073] The multi-party cooperation sustainability condition constraint is:

[0074] ,

[0075] wherein, represents a property income, represents a property sharing ratio, represents a user income, represents an annual average fuel replacement cost of a single car, represents a resident income ratio.

[0076] In a second aspect, the present application provides a resident community charging pile planning system for maximizing the income of an operator, which is applicable to the resident community charging pile planning method for maximizing the income of an operator according to any one of the first aspect, and comprises:

[0077] A game framework modeling module is configured to: construct a dynamic three-party master-slave game framework with the operator as the leading party and the property and the residents as the subordinate parties.

[0078] A target function construction module is configured to: according to the dynamic three-party master-slave game framework, divide the target priorities, and construct a maximum income target function of each party.

[0079] A target function solving module is configured to: utilize an improved NSGA-II algorithm to balance the maximum income target function of each party, and obtain a front solution set.

[0080] A strategy planning module is configured to: utilize a Monte Carlo algorithm to predict future income and quantify risk of the front solution set, and obtain a feasible charging pile construction scheme strategy set.

[0081] Compared with the prior art, the present application has the following beneficial effects:

[0082] 1. By constructing a dynamic game framework dominated by the operator, subordinate to the property and residents, defining the three-party strategy space and revenue function, and introducing a hierarchical truncation screening and real-time congestion updating mechanism to optimize the convergence and uniformity of the multi-objective solution set, combined with the probability modeling and risk value analysis of multi-source uncertain variables, the multi-party benefit collaborative optimization, dynamic strategy adaptive adjustment and future income fluctuation accurate prediction are realized;

[0083] 2. Effectively solves the core problems of traditional planning such as multi-party interest conflict difficult to coordinate, subsidy and demand fluctuation adaptability insufficient, risk quantification missing, etc., provides scientific decision basis for the operator in complex dynamic environment, reduces the construction and operation risk, and promotes the efficient use and sustainable development of residential area charging facilities. BRIEF DESCRIPTION OF DRAWINGS

[0084] Figure 1 A flow chart of a residential area charging pile planning method for maximizing the operator's income according to an embodiment of the present application is provided.

[0085] Figure 2 A distribution diagram of the top 30 party incomes solved by the NSGA-II algorithm according to an embodiment of the present application is provided.

[0086] Figure 3 A first feasible strategy income result diagram according to an embodiment of the present application is provided.

[0087] Figure 4 A second feasible strategy income result diagram according to an embodiment of the present application is provided.

[0088] Figure 5 A third feasible strategy income result diagram according to an embodiment of the present application is provided.

[0089] Figure 6 A fourth feasible strategy income result diagram according to an embodiment of the present application is provided.

[0090] Figure 7 A fifth feasible strategy income result diagram according to an embodiment of the present application is provided. DETAILED DESCRIPTION

[0091] The technical scheme of the present application will be described in detail below with the aid of the accompanying drawings and specific embodiments. It should be understood that the specific features of the embodiments and the specific features in the embodiments are detailed descriptions of the technical scheme of the present application, and are not limitations of the technical scheme of the present application. In the case of no conflict, the technical features in the embodiments and the embodiments can be combined with each other.

[0092] It should be noted that the term "and / or" herein is merely an association relationship between the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the existence of A alone, the existence of A and B together, and the existence of B alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after it.

[0093] Embodiment one:

[0094] The embodiment of the application discloses a resident community charging pile planning method for maximizing the income of an operator, referring to Figure 1 as shown, comprising:

[0095] S1, a dynamic three-party master-slave game framework is constructed with the operator as the leading party and the property and residents as the subordinate parties;

[0096] S2, according to the dynamic three-party master-slave game framework, target priority division is carried out, and a maximum income target function of each party is constructed;

[0097] S3, the improved NSGA-II algorithm is used to balance the maximum income target function of each party to obtain a set of front solutions;

[0098] S4, the Monte Carlo algorithm is used to predict future income and quantify risk for the set of front solutions to obtain a set of feasible charging pile construction scheme strategies.

[0099] Specifically, in step S1, the dynamic three-party master-slave game framework is constructed with the operator as the leading party and the property and residents as the subordinate parties, comprising:

[0100] S1.1, a three-party game strategy is constructed with the operator as the leading party and the property and residents as the subordinate parties, including an operator game strategy, a property game strategy and a resident game strategy;

[0101] S1.2, a three-party income matrix is constructed according to the three-party game strategy;

[0102] S1.3, a three-party income function is constructed according to the three-party income matrix, including an operator income function, a property income function and a resident income function.

[0103] In step S1.1, the operator as the leading party, its income = charging service fee income - [equipment depreciation loss cost + operation cost + maintenance cost + site rental cost], and the strategy of the operator includes:

[0104] Building piles: the operator chooses to build charging piles, charging service fee income, and needs to pay equipment loss cost, operation cost, operation and maintenance cost, and site rental cost;

[0105] No pile building: the operator does not build charging piles, with no cost and income;

[0106] Thus, the expression of the operator game strategy is obtained:

[0107] ,

[0108] Wherein, represents the operator, represents the operator revenue, represents the operator to build charging piles, get charging service fees, pay equipment depreciation cost, operation cost, maintenance cost and site rental cost; represents that the operator does not build charging piles, no cost and income.

[0109] The property as the party, its revenue = [charging service fee sharing income + site rental income] - management input cost, the strategy of the property includes:

[0110] Co-construction: the property chooses to cooperate with the operator to obtain charging service fee sharing and site rental income, but needs to pay management cost;

[0111] Non-cooperation: the property does not cooperate with the operator, and the operator cannot build piles, with no income and cost;

[0112] Thus, the expression of the property game strategy is obtained:

[0113] ,

[0114] Wherein, represents the property, represents the property revenue, represents that the property cooperates with the operator to obtain charging service fee sharing and site rental income, and pays management cost; represents that the property does not cooperate with the operator, and the operator cannot build piles, with no income and cost.

[0115] The resident as the party, its revenue = annual savings of fuel cost - annual charging cost, the strategy of the resident includes:

[0116] Use charging piles: residents choose to use charging piles to obtain fuel substitution income, but need to pay charging cost.

[0117] Do not use charging piles: residents do not use charging piles to obtain fuel substitution income.

[0118] Thus, the expression of the resident game strategy is obtained:

[0119] ,

[0120] Wherein, represents the resident, represents the property revenue, represents that residents use charging piles, obtain fuel substitution benefits, and pay charging costs; represents that residents do not use charging piles and have no fuel substitution benefits.

[0121] In step S1.2, the three-party benefit matrix is shown in Table 1 below.

[0122] Table 1: Three-party benefit matrix

[0123]

[0124] In step S1.3, it is assumed that the community electric vehicle ownership is , is the local service fee pricing upper limit, is the average coverage rate of charging services, is the price sensitivity coefficient, and the community electric vehicle service volume is affected by the charging service price , the higher the price sensitivity, the lower , and the function formula is:

[0125] ,

[0126] The constraint condition is the upper limit of the electric vehicle service volume: wherein is the upper limit of the pile-to-vehicle ratio, represents the community electric vehicle service volume.

[0127] The operator as the leading party, its benefits mainly come from charging service fee income and electricity subsidy income, and the costs include equipment depreciation cost, operation cost, maintenance cost, and site rental cost, and its benefit function is:

[0128] ,

[0129] wherein represents the charging service fee, represents the property proportion, represents the annual average charging volume of a single vehicle of residents, represents the community electric vehicle service volume, represents the electricity subsidy rate, represents the annual depreciation rate of equipment, represents the construction cost of a single pile, represents the annual operation cost of a single pile, represents the site rental fee of a single pile, represents the number of piles, represents the annual maintenance cost of a single pile.

[0130] The constraint condition of the operator benefit function is:

[0131] 1. Grid capacity limit:

[0132] 2. Parking space limit:

[0133] 3. Local charging electricity price limit:

[0134] 4. Local charging fee annual subsidy upper limit:

[0135] wherein, represents the maximum power of a single charging pile, represents the grid capacity, represents a proportionality coefficient, represents the number of parking spaces, represents the local service fee pricing upper limit, represents the local electricity fee annual subsidy upper limit.

[0136] The property, as a follower, mainly obtains income from charging service fee sharing and site rental income, and the cost includes management investment, and the income function is:

[0137] ,

[0138] wherein, represents the property sharing proportion upper limit, represents the management cost of a single charging pile;

[0139] The constraint condition of the property income function is the sharing proportion upper limit: .

[0140] The residents, as a follower, mainly obtain income from saving fuel cost, and the cost is only charging cost, and the income function is:

[0141] ,

[0142] The constraint condition of the resident income function is the charging cost constraint: wherein, represents the annual average fuel replacement cost of a single vehicle, represents the basic electricity price, represents a cost coefficient.

[0143] In step S2, the target priority division mechanism is as shown in Table 2.

[0144] Table 2: Target priority division mechanism.

[0145]

[0146] The maximization target function of the parties is:

[0147] ,

[0148] Wherein, represents the total revenue of the three parties, represents the operator, represents the property, represents the residents, represents the Min-Max normalized operator revenue, represents the Min-Max normalized property revenue, represents the Min-Max normalized resident revenue, represents the operator revenue weight coefficient, represents the property revenue weight coefficient, represents the resident revenue weight coefficient, represents the minimum value function, represents the charging service fee, represents the local service fee pricing upper limit, represents the property sharing ratio, represents the upper limit of the property sharing ratio, represents the number of charging piles, represents the proportion coefficient, represents the number of parking spaces, represents the power grid capacity, represents the maximum power of a single charging pile.

[0149] In step S3, the traditional NSGA-II sorts the non-dominated solutions by calculating the crowding distance of the non-dominated solutions, removes individuals with small crowding distance, and improves the diversity of the solution set, but this method cannot reflect the change of the crowding distance in real time after the remaining individuals after the elimination of a single individual. The dynamic crowding sorting strategy is adopted in this embodiment to calculate the crowding distance in real time, to preferentially retain individuals with high non-dominated levels, and to calculate the crowding degree only in the necessary front layer to avoid repeated calculation of the whole population.

[0150] The improved NSGA-II algorithm is used to balance the maximization target function of the parties to obtain a front solution set, which includes:

[0151] The population size K and the iteration number G are set, and according to the maximization target function of the parties, based on the system parameter constraints (such as power grid capacity, parking space proportion, service fee upper limit, etc.), individuals satisfying all constraints are randomly generated to obtain a feasible solution set satisfying the constraint conditions , each feasible solution contains decision variables ; wherein, , represents an individual feasible solutions, representing individuals feasible solutions, representing the number of piles, representing the charging service fee, representing the property sharing ratio;

[0152] According to the feasible solution set , the following steps are iteratively performed until a preset iteration stopping condition is met:

[0153] The feasible solution set is sorted by a non-dominated sorting algorithm, all non-dominated individuals in the population are classified into the first front layer , their dominance levels are marked as 0, and the classified individuals are removed in turn, new non-dominated solutions are recursively identified in the remaining population, and are classified, until all individuals are classified, and the output classification results are output in descending order of dominance level ; wherein, , represents the th front layer, represents the th front layer;

[0154] According to the classification results , elitist selection is performed, and the number of individuals is accumulated from the front layer until the first front layer is located, which makes the cumulative number exceed a preset scale ;

[0155] The crowding distance of the individual and its adjacent individuals and in the front layer is calculated respectively, the individuals in the front layer are sorted according to the crowding distance , the individuals with small crowding distance are deleted and the population is updated, until the total number of individuals in the front layer is not greater than , which is recorded as the front layer , and the front layer result is obtained; wherein, , represents the number of individuals in the front layer ; ;

[0156] The front layer result is taken as the parent population Crossing and mutation are performed, simulated binary crossover is adopted, constraint-aware crossover is performed on parent individuals, and dynamic mutation coefficient is gradually attenuated with the increase of iteration number to obtain offspring population ; wherein the mutation probability is , , represents the initial value of the mutation probability, represents the current iteration number, represents the total number of iterations, and the mutation range is , , represents the initial value of the mutation range; this scheme realizes triggering the mutation operation with a higher probability at the initial iteration to speed up the population diversity exploration, and gradually converges in the later period to avoid destroying the found high-quality solution;

[0157] The parent population and the offspring population are merged and non-dominated sorting is performed, a new generation population is obtained by hierarchical truncation, and is used as the initial population for the next iteration;

[0158] The maximum iteration number or the change of the fitness function less than a threshold value are set as the algorithm termination conditions, and after iteration is stopped, a set of frontier solutions is obtained, each frontier solution includes a set of optimal revenue and a set of optimal configuration solutions , wherein represents the optimal operator revenue, represents the optimal property revenue, represents the optimal resident revenue, represents the optimal number of poles, represents the optimal charging service fee, represents the optimal property sharing ratio.

[0159] In step S4, the future revenue prediction and risk quantification of the set of frontier solutions are performed by using the Monte Carlo algorithm to obtain a set of feasible charging pile construction scheme strategies, including:

[0160] S4.1, according to the dynamic three-party principal-agent game framework, uncertain variables are screened, and probability distribution assumptions are made on the uncertain variables to obtain a set of variables;

[0161] S4.2, according to the set of frontier solutions and the set of variables, the operator's annual expected revenue and revenue risk in the future T years are calculated based on M times of Monte Carlo simulation;

[0162] S4.3, according to the operator's annual expected revenue and revenue risk, the strategy feasibility constraint verification is performed, and a set of feasible charging pile construction scheme strategies is output.

[0163] In step S4.1, the uncertain variables include:

[0164] 1) Demand variable: Electric vehicle service volume in the community Average annual charging volume per residential vehicle ;

[0165] 2) Subsidy variable: Electricity subsidy rate ;

[0166] 3) Market variables: Local service fee pricing ceiling Annual operating cost per pile Single pile site rental fee Annual maintenance cost per pile ;

[0167] The probability distribution assumptions for uncertain variables include:

[0168] 1) Electric vehicle service volume in the community Follow the mean Standard deviation is log-normal distribution ,in, Indicates the year sequence number. This indicates the average annual penetration growth rate. Indicates volatility;

[0169] 2) The average annual charging amount per residential vehicle Follows uniform distribution , This indicates the minimum charging capacity for a single residential vehicle. This indicates the maximum charging capacity for a single residential vehicle.

[0170] 3) The electricity subsidy rate Linear decrease, slope reduction rate ;

[0171] 4) The local service fee pricing cap Follow the mean Standard deviation is normal distribution ,in, This represents the average local service fee. This indicates the standard deviation of local service fees;

[0172] 5) Annual operating cost per pile Follow the mean Standard deviation is log-normal distribution ,in, This indicates the annual operating cost growth rate per pile. Indicates the volatility of operating costs;

[0173] 6) the single-pile site lease fee obeys mean , standard deviation lognormal distribution , wherein represents the single-pile site lease fee growth rate, represents the single-pile site lease fee volatility rate;

[0174] 7) the single-pile annual maintenance cost obeys gamma distribution with shape parameter , scale parameter , wherein represents the single-pile life span, represents the aging risk rate.

[0175] In step S4.2, the operator's annual expected revenue in T years is calculated based on Monte Carlo simulation of future revenue prediction and risk quantification , quantifying the revenue risk in the future T years;

[0176] For each Pareto frontier solution set, revenue prediction is performed, traversing the future T years, and dynamically adjusting the variables to bring them into the revenue function:

[0177] ,

[0178] wherein represents the optimal number of piles, represents the optimal charging service fee, represents the optimal property sharing ratio;

[0179] Cumulative simulation M times, output the average annual revenue of all Pareto frontier solution sets , wherein represents the average annual revenue of the frontier solution;

[0180] The expected revenue risk in the future T years is calculated by the following formula, assuming that the revenue risk is normally distributed:

[0181] ,

[0182] wherein represents the value at risk, represents the conditional value at risk, represents the average annual revenue in T years, represents the average annual revenue variance in T years, represents the confidence interval, generally 99%, represents The corresponding normal quantile (99% corresponds to 2.326), The standard normal density function is represented.

[0183] In step S4.3, the policy feasibility constraints include a yield benchmark constraint, a multi-dimensional risk condition constraint, and a multi-party cooperation sustainability condition constraint;

[0184] The yield benchmark constraint is:

[0185] ,

[0186] Wherein, The industry benchmark return rate is represented;

[0187] The multi-dimensional risk condition constraint is that the yield volatility does not exceed the industry tolerable threshold , screened by double risk thresholds And The scheme that meets the tail risk and extreme loss constraints at the same time; the multi-dimensional risk condition constraint is:

[0188] ,

[0189] Wherein, The industry tolerable threshold is represented, The risk threshold is represented, The conditional risk threshold is represented;

[0190] The multi-party cooperation sustainability condition constraint is that the property income meets a specific proportion not lower than the operator's income , and the resident income is not less than The proportion of fuel cost ensures the user's use enthusiasm and ensures the feasibility of the game equilibrium solution; the multi-party cooperation sustainability condition constraint is:

[0191] ,

[0192] Wherein, The property income is represented, The property sharing ratio is represented, The user income is represented, The annual average fuel replacement cost of a single car is represented, The resident income ratio is represented.

[0193] Through multiple constraint conditions screening, a feasible charging pile construction scheme strategy set is output, each strategy including:

[0194] 1) Charging pile initial construction scheme: ;

[0195] 2) Future T-year yield risk indicators: ;

[0196] If there is a combination strategy that meets the requirements, it is determined that it is feasible to build charging piles in the cell, and the combination of strategies that meet all the constraint conditions and the yield risk quantification results are output, providing more reliable auxiliary decision-making for the initial charging pile planning of the operator in the residential area.

[0197] There are potential alternative solutions for each technical solution proposed in the embodiment, as shown in Table 3.

[0198] Table 3: Potential alternative solutions

[0199]

[0200] In order to verify the effectiveness of the charging pile planning method proposed in the embodiment, the following experiments are performed in the embodiment:

[0201] A residential area plans to build electric vehicle charging piles, with a total of 100 public parking spaces and an initial number of 300 charging vehicles. First, the NSGA-II algorithm is used to solve the maximum revenue objective function of each party. The initial population size is set to 200, the number of iterations is 200, the initial mutation probability is 0.5, the initial mutation range is 0.1, and the top 200 are cut off each time. The Pareto frontier solution set of the top 30 is output, as shown in Table 4 below, which shows the top 5 charging pile planning schemes and the revenue of each party, Figure 2 The distribution of the revenue of each party in the top 30 solutions obtained by the NSGA-II algorithm is shown.

[0202] Table 4: Top 5 charging pile planning schemes and revenue of each party

[0203]

[0204] The Monte Carlo simulation period T is set to 5 years, and the simulation is performed 1000 times. After risk value calculation and constraint convergence, the following feasible solutions are obtained:

[0205] The first feasible strategy is to build 9 charging piles, set the service fee to 0.80 yuan, and the distribution ratio to 10.0%. The future T-year revenue risk index is shown in Table 5. Figure 3

[0206] Table 5: Future T-year revenue risk index of the first feasible strategy

[0207]

[0208] The second feasible strategy is to build 9 charging piles, set the service fee to 0.76 yuan, and the distribution ratio to 10.0%. The future T-year revenue risk index is shown in Table 6. Figure 4

[0209] ​​Table 6: Future T-year income risk indicators of the second feasible strategy

[0210]

[0211] The third feasible strategy is to build 9 piles, service fee 0.79 yuan, and the proportion of sharing is 10.0%. The future T-year income risk indicators are shown in Table 7. Figure 5

[0212] Table 7: Future T-year income risk indicators of the third feasible strategy

[0213]

[0214] The fourth feasible strategy is to build 9 piles, service fee 0.78 yuan, and the proportion of sharing is 12.5%. The future T-year income risk indicators are shown in Table 8. Figure 6

[0215] Table 8: Future T-year income risk indicators of the fourth feasible strategy

[0216]

[0217] The fifth feasible strategy is to build 9 piles, service fee 0.77 yuan, and the proportion of sharing is 12.1%. The future T-year income risk indicators are shown in Table 9. Figure 7

[0218] Table 9: Future T-year income risk indicators of the fifth feasible strategy

[0219]

[0220] In summary, the residential area charging pile planning method for maximizing the income of the operator proposed in this embodiment combines dynamic master-slave game modeling, improved NSGA-II algorithm balanced solution, and Monte Carlo income risk quantification methods. The property is included in the subordinate game party, a three-party dynamic income matrix dominated by the operator is constructed, and a strategy space is defined. Through hierarchical truncation screening and real-time congestion degree updating mechanism, dynamic adjustment of mutation amplitude is adopted, and the amplitude gradually decreases with the increase of iteration number, which accelerates the convergence speed and improves the solution set quality. The Monte Carlo simulation is introduced to jointly model the probability of multiple uncertain variables such as subsidies and demand, and to quantify the income fluctuation interval under extreme risk. This method effectively solves the core problems such as the difficulty of coordinating the conflict of interests of multiple parties in traditional planning, the lack of adaptability of subsidy and demand fluctuations, and the lack of risk quantification, etc. It provides a scientific decision-making basis for operators in complex dynamic environment, reduces the construction and operation risk, and promotes the efficient use and sustainable development of residential area charging facilities.

[0221] Embodiment two:

[0222] ​​​Based on the same inventive concept as embodiment one, the embodiment of the present application discloses a resident community charging pile planning system facing the maximization of operator revenue, which is applicable to any resident community charging pile planning method facing the maximization of operator revenue in embodiment one, and comprises:

[0223] A game framework modeling module is configured to model a dynamic three-party master-slave game framework with an operator as a leading party and a property and residents as subordinate parties.

[0224] A target function modeling module is configured to perform target priority division according to the dynamic three-party master-slave game framework, and model a maximum revenue target function of each party.

[0225] A target function solving module is configured to solve the maximum revenue target function of each party by using an improved NSGA-II algorithm to obtain a front solution set.

[0226] A strategy planning module is configured to perform future revenue prediction and risk quantification on the front solution set by using a Monte Carlo algorithm to obtain a feasible charging pile construction scheme strategy set.

[0227] The specific function implementation of each module is referred to the related content in the method of embodiment one, and will not be described here.

[0228] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

[0229] The present application is described with reference to flowcharts and / or block diagrams according to the method, device (system), and computer program product of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The device that implements the functions specified in one or more flows and / or blocks. Figure 1 The device that implements the functions specified in one or more flows and / or blocks.

[0230] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flow Figure 1 The functions of the flow or flows and / or blocks Figure 1 The functions of the flow or flows and / or blocks

[0231] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flow Figure 1 The functions of the flow or flows and / or blocks Figure 1 The functions of the flow or flows and / or blocks

[0232] The embodiments of the present application described above are merely intended to illustrate the present application, but are not intended to limit the present application. The above-described embodiments are merely illustrative, and are not intended to limit the present application, and any person skilled in the art can make many modifications without departing from the spirit and scope of the present application, and these modifications are also intended to fall within the scope of the present application.

Claims

1. A method for planning charging piles in residential communities to maximize operator revenue, characterized in that, include: Construct a dynamic three-party master-slave game framework with operators as the leading party and property management and residents as subordinate parties; Based on the dynamic three-party master-slave game framework, the objectives are prioritized and a target function for maximizing the benefits of each party is constructed. The improved NSGA-II algorithm is used to solve the objective function of maximizing the benefits of the parties in equilibrium, and the frontier solution set is obtained. The Monte Carlo algorithm is used to predict future returns and quantify risks in the frontier solution set, resulting in a set of feasible charging pile construction strategies.

2. The residential community charging pile planning method for maximizing operator revenue as described in claim 1, characterized in that, The aforementioned construction of a dynamic three-party master-slave game framework with the operator as the leading party and property management and residents as subordinate parties includes: With the telecom operator as the leading party and property management companies and residents as the supporting parties, a three-way game strategy is constructed, including the operator's game strategy, the property management company's game strategy, and the residents' game strategy. Based on the aforementioned three-party game strategy, construct a three-party payoff matrix; Based on the aforementioned tripartite revenue matrix, a tripartite revenue function is constructed, including the operator's revenue function, the property's revenue function, and the resident's revenue function.

3. The residential community charging pile planning method for maximizing operator revenue as described in claim 2, characterized in that, The expression for the operator's game strategy is: ， in, Indicates the operator, Indicates operator revenue. This means that operators build charging stations, receive charging service fees, and pay for equipment depreciation costs, operating costs, maintenance costs, and site rental costs. This indicates that operators do not build charging stations, thus incurring no costs or revenue. The expression for the property game strategy is: ， in, Indicates property management, Indicates property income, This indicates that the property management company collaborates with the operator, receiving a share of charging service fees and site rental income to cover management costs. This indicates that the property management company does not cooperate with the operator, so the operator cannot install charging piles and therefore has no revenue and no costs. The expression for the resident game strategy is: ， in, Indicates residents, Indicates property income, This means that residents use charging stations to obtain benefits from replacing fuel and pay for charging costs. This indicates that residents who do not use charging stations will not receive any revenue from fuel substitution. The operator's revenue function is: ， in, This indicates the charging service fee. Indicates the profit-sharing ratio of the property. This represents the average annual charging amount per vehicle in a residential community. This indicates the number of electric vehicles served in the community. Indicates the electricity subsidy rate. This indicates the annual depreciation rate of the equipment. This indicates the construction cost of a single pile. This indicates the annual operating cost per pile. This indicates the cost of renting a single pile site. Indicates the number of piles constructed. This indicates the annual maintenance cost per pile. Indicates constraints. This indicates the maximum power of a single charging station. Indicates grid capacity. This represents the proportionality coefficient. Indicates the number of parking spaces. This indicates the upper limit for local service fee pricing. This indicates the maximum annual subsidy for local electricity bills; The property revenue function is: ， in, This indicates the upper limit of the property revenue sharing ratio. This indicates the management cost of a single charging station; The resident income function is: ， in, This represents the average annual fuel substitution cost per vehicle. Indicates the base electricity price. Indicates the cost coefficient; Electric vehicle service volume in the community It can be obtained through the following formula: ， in, This represents the average coverage of charging services. Indicates the price sensitivity coefficient. This indicates the upper limit for local service fee pricing. This indicates the number of electric vehicles in the community.

4. The residential community charging pile planning method for maximizing operator revenue as described in claim 1, characterized in that, The objective function for maximizing the benefits of each party is: ， in, This represents the total benefit to all three parties. Indicates the operator, Indicates property management, Indicates residents, This represents the operator's revenue after Min-Max normalization. This represents the property yield after Min-Max normalization. This represents the residents' income after Min-Max normalization. This represents the operator's revenue weighting coefficient. This represents the property income weighting coefficient. This represents the weighting coefficient of residents' income. This represents the function that takes the minimum value. This indicates the charging service fee. This indicates the upper limit for local service fee pricing. Indicates the profit-sharing ratio of the property. This indicates the upper limit of the property revenue sharing ratio. Indicates the number of piles constructed. This represents the proportionality coefficient. Indicates the number of parking spaces. Indicates grid capacity. This indicates the maximum power of a single charging station.

5. The residential community charging pile planning method for maximizing operator revenue as described in claim 1, characterized in that, The improved NSGA-II algorithm is used to solve the objective function of maximizing the benefits of all parties, resulting in a frontier solution set, including: Set the population size to K, and based on the objective function of maximizing the benefits of all parties, randomly generate individuals that satisfy all constraints to obtain a set of feasible solutions that meet the constraints. Each feasible solution contains decision variables. ;in, , Represents an individual A feasible solution. Represents an individual A feasible solution. Indicates the number of piles constructed. This indicates the charging service fee. Indicates the percentage of property ownership shared; According to the feasible solution set The following steps are executed iteratively until the preset iteration stopping condition is met: The feasible solution set is sorted using a non-dominated sorting algorithm. Sort the population and assign all non-dominant individuals to the first front layer. Mark the dominance level as 0, and remove the stratified individuals one by one. Recursively identify new non-dominated solutions in the remaining population and stratify them until all individuals have been divided. Output the results in descending order of dominance level. Output layer results ;in, , Indicates the first A frontier layer, Indicates the first One frontier layer; According to the stratification results Selecting elites, starting from the forefront Start accumulating the number of individuals until the first individual is found that causes the cumulative number to exceed the preset size. Frontier layer ; Calculate the front layer separately medium-sized individuals Its neighboring individuals and Crowding distance Distance based on congestion For the frontier layer Individuals are sorted by crowding, and individuals with low crowding are removed and the population is updated until the frontier layer is reached. The total number of individuals is not greater than , denoted as the frontier layer To obtain the frontier layer results ;in, , Represents the frontier layer The number of individuals, ; The front layer results As the parental population Crossover and mutation are performed to obtain offspring populations. The mutation probability is: , , This represents the initial value of the mutation probability. Indicates the current iteration number. This represents the total number of iterations, with a variation magnitude of . , , Indicates the initial value of the variation amplitude; Merging parent populations and offspring population It then performs non-dominated sorting and obtains a new generation of population through stratified truncation. This will serve as the initial population for the next iteration. After the iteration stops, the frontier solution set is obtained. Each frontier solution includes the optimal payoff set. and optimal configuration solution set ,in, This represents the optimal operator revenue. Indicates the optimal property return. This represents the optimal resident income. This indicates the optimal number of piles to be constructed. This indicates the optimal charging service fee. This indicates the optimal property revenue sharing ratio.

6. The residential community charging pile planning method for maximizing operator revenue as described in claim 1, characterized in that, The Monte Carlo algorithm is used to predict future returns and quantify risks in the frontier solution set, resulting in a set of feasible charging pile construction strategies, including: Uncertain variables are selected based on the dynamic three-party master-slave game framework, and probability distribution assumptions are made for the uncertain variables to obtain a variable set; Based on the aforementioned frontier solution set and variable set, the expected annual revenue and revenue risk of the operator in the next T years are calculated using M-time Monte Carlo simulations. Based on the operator's expected annual revenue and revenue risk, the feasibility of the strategy is verified by constraint, and a set of feasible charging pile construction schemes is output.

7. The residential community charging pile planning method for maximizing operator revenue as described in claim 6, characterized in that, The uncertain variables include demand variables, subsidy variables, and market variables; The demand variables include the volume of electric vehicle services in the community. Average annual charging volume per residential vehicle The electric vehicle service volume of the community Follow the mean Standard deviation is log-normal distribution ,in, Indicates the year sequence number. This indicates the average annual penetration growth rate. Indicates volatility; the average annual charging amount per residential vehicle. Follows uniform distribution , This indicates the minimum charging capacity for a single residential vehicle. This indicates the maximum charging capacity for a single residential vehicle. The subsidy variables include the electricity subsidy rate. The electricity subsidy rate Linear decrease, slope reduction rate ; The market variables include the local service fee pricing cap. Annual operating cost per pile Single pile site rental fee Annual maintenance cost per pile The local service fee pricing cap Follow the mean Standard deviation is normal distribution ,in, This represents the average local service fee. This represents the standard deviation of local service fees; the annual operating cost per pile. Follow the mean Standard deviation is log-normal distribution ,in, This indicates the annual operating cost growth rate per pile. This represents the volatility of operating costs; the single-pile site rental fee. Follow the mean Standard deviation is log-normal distribution ,in, This indicates the growth rate of single-pile site rental costs. This represents the volatility of site rental costs per pile; the annual maintenance cost per pile. Obeying shape parameters Scale parameters are The gamma distribution, in which, Indicates the lifespan of a single pile. This indicates the aging risk rate.

8. The residential community charging pile planning method for maximizing operator revenue as described in claim 7, characterized in that, The operator's expected annual revenue is calculated using the following formula: ， in, This indicates the optimal number of piles to be constructed. This indicates the optimal charging service fee. Indicates the optimal property revenue sharing ratio; The risk-reward ratio is calculated using the following formula: ， in, Value at risk (VaR) Represents conditional risk value. This represents the average annual return over year T. This represents the variance of the average annual return over year T. Indicates the confidence interval. express The corresponding normal quantile, This represents the standard normal density function.

9. The residential community charging pile planning method for maximizing operator revenue as described in claim 8, characterized in that, The feasibility constraints of the strategy include return benchmark constraints, multi-dimensional risk condition constraints, and multi-party cooperation sustainability condition constraints. The return benchmark constraint is: ， in, This represents the industry benchmark rate of return; The multidimensional risk condition constraint is as follows: ， in, This indicates the industry's acceptable threshold. Indicates the risk threshold. Indicates the conditional risk threshold; The constraints on the sustainability of the multi-party cooperation are as follows: ， in, Indicates property income, Indicates the profit-sharing ratio of the property. Indicates user benefits, This represents the average annual fuel substitution cost per vehicle. This indicates the proportion of residents' income.

10. A residential community charging pile planning system aimed at maximizing operator revenue, characterized in that, include: The game framework construction module is used to: construct a dynamic three-party master-slave game framework with the operator as the dominant party and property management and residents as subordinate parties; The objective function construction module is used to: prioritize objectives and construct an objective function that maximizes the gains of each party based on the dynamic three-party master-slave game framework; The objective function solving module is used to: solve the objective function for maximizing the benefits of the parties using the improved NSGA-II algorithm to obtain the frontier solution set; The strategy planning module is used to: use the Monte Carlo algorithm to predict future returns and quantify risks on the frontier solution set to obtain a set of feasible charging pile construction strategies.

Images