A transformer area electric vehicle resource and photovoltaic consumption coordination decision method and device

Abstract

The application discloses a kind of table area electric automobile resources and photovoltaic consumption coordination decision method and device, belong to electric power system and its automation field, the method includes obtaining electric automobile driving law, charge-discharge characteristics and state of charge data, input fine load model, obtain the charging power and state of charge of each single body each time;In the mode of charging as soon as possible, constraint is applied, and charging load is aggregated to obtain.The load and table area photovoltaic predicted output are based, and the matching value is calculated by cosine similarity, a dynamic electricity price discount model is established, and a dynamic time-of-use electricity price is output.Combining consumer psychology model and user difference, an interactive response model is constructed, and a response decision is output.Finally, the maximum charging station revenue and the maximum photovoltaic consumption are taken as the target, a coordination decision model is constructed and solved, and the optimal charging strategy and photovoltaic consumption scheme are output.The application can significantly improve the photovoltaic consumption rate, greatly reduce the electricity purchase cost and light abandonment penalty, and balance the user economy and grid safety.

Description

Technical Field

[0001] This invention relates to a method and apparatus for collaborative decision-making on electric vehicle resources and photovoltaic power consumption in a power distribution area, belonging to the field of power systems and their automation. Background Technology

[0002] Traditional electric vehicle charging methods often employ a disorderly, on-demand charging approach, lacking scientific guidance and overall control over user charging behavior. This results in a relatively uneven distribution of charging load. During midday, photovoltaic power output is at its peak, while electric vehicle charging demand is at its lowest. Consequently, a large amount of photovoltaic power cannot be consumed locally, leading to curtailment and energy waste. This also easily causes problems such as overload of distribution transformers and voltage exceeding limits, significantly reducing the economic efficiency and safety of distribution transformer operation. This contradicts the requirements of ensuring safe power supply and promoting the local consumption of new energy in the high-quality development of the distribution network.

[0003] Existing power distribution area source-load regulation technologies have significant limitations and are difficult to adapt to the dynamic changes in both the source and load sides: At the price incentive level, existing regulation methods mostly employ fixed peak-valley pricing mechanisms. This mechanism lacks flexibility and dynamic adaptability, and cannot dynamically adjust based on the real-time matching status of photovoltaic output and charging load within the distribution area. Its guiding effect on user charging behavior is limited, making it difficult to achieve source-load coordinated optimization. At the user interaction level, existing user response models do not fully consider individual user attribute differences, including user income level, age structure, awareness of electricity pricing policies, and the urgency of charging needs. This results in insufficient model prediction accuracy and an inability to achieve precise interaction and personalized guidance for different user groups. At the optimization decision-making level, existing distribution area optimization models are mostly centered on a single objective, either focusing on improving photovoltaic absorption rate or on grid operation safety, without comprehensively considering the needs of local photovoltaic absorption, charging station operation revenue, and grid operation constraints. Furthermore, existing solution methods often suffer from the problem of difficulty in balancing efficiency and multi-objective optimization performance, making it impossible to quickly obtain the globally optimal regulation decision scheme.

[0004] In summary, the current technical deficiencies in source-load coordinated regulation and optimization decision-making in low-voltage distribution substations can no longer meet the needs of high-quality development of distribution networks in the context of large-scale distributed photovoltaic access and the popularization of electric vehicles. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and device for coordinated decision-making on electric vehicle resources and photovoltaic consumption in distribution areas, so as to realize the improvement of local photovoltaic consumption rate, maximize the synergistic benefits of multiple parties, and intelligent control of electric vehicle resources, and provide technical support for the safe, efficient and low-carbon operation of low-voltage distribution network areas under the new power system.

[0006] To achieve the above objectives, the present invention is implemented using the following technical solution:

[0007] In a first aspect, the present invention provides a method for coordinated decision-making regarding electric vehicle resources and photovoltaic power consumption in a power distribution area, comprising:

[0008] Acquire data on the driving patterns of electric vehicles, battery charging and discharging characteristics, and state of charge.

[0009] The driving pattern data, battery charging and discharging characteristic data, and state of charge data are input into a pre-constructed fine load model of electric vehicles to obtain the charging power and state of charge of each individual electric vehicle at each time.

[0010] The charging power and state of charge of each individual electric vehicle at each time are constrained in the instant charging mode, and the charging load of the electric vehicle in the instant charging mode is obtained by aggregating them.

[0011] Based on the electric vehicle charging load under the instant charging mode and the pre-acquired photovoltaic power output data of the transformer area, the matching value between photovoltaic power output and the total charging load is calculated by cosine similarity. Based on the matching value, a dynamic electricity price discount model is established, and a dynamic time-of-use electricity price is output.

[0012] Based on the dynamic time-of-use electricity price, combined with consumer psychology models and user multi-attribute difference data, an electric vehicle interactive response model is constructed to output the electric vehicle's response decision.

[0013] Based on the response decision of the electric vehicle and the pre-constructed district collaborative decision-making model with the objectives of maximizing charging station revenue and maximizing photovoltaic absorption, the optimal electric vehicle charging strategy and photovoltaic absorption scheme are output under preset constraints.

[0014] Furthermore, the driving pattern data, battery charge / discharge characteristic data, and state of charge data are input into a pre-constructed fine-grained load model of electric vehicles to obtain the charging power and state of charge of each individual electric vehicle at each time point, including:

[0015] Using the electric vehicle's grid connection time, grid disconnection time, and mileage as driving pattern data, user travel behavior is fitted based on a log-normal distribution;

[0016] Using the battery capacity, maximum charge / discharge power, and charging efficiency of electric vehicles as battery charge / discharge characteristic data, and combining this with the current state of charge, a charge / discharge factor model is constructed, as shown in the following formula:

[0017] ;

[0018] In the formula, and The first electric vehicles The charging power and discharging power at any given time; and The first The maximum charging power and maximum discharging power of a vehicle; and The first The charging factor and discharging factor of an electric vehicle;

[0019] The charge / discharge factor includes a charge factor and a discharge factor, and its relationship with the state of charge is as follows:

[0020] ;

[0021] In the formula, For the first The current state of charge of the electric vehicle; and These are the state-of-charge reference values ​​corresponding to the charging factor and the discharging factor, respectively; and These are the coefficients of change for the charging factor and the discharging factor, respectively.

[0022] Obtain the charging power of each individual electric vehicle at each time point. and state of charge .

[0023] Furthermore, the charging power and state of charge of each individual electric vehicle at each time moment are constrained under the instant charging mode, and aggregated to obtain the electric vehicle charging load under the instant charging mode, including:

[0024] Using the charging power and state of charge as inputs, the instant charging mode is set as follows: each electric vehicle is charged immediately at the current maximum available power after entering the network until the preset off-grid conditions are reached. The preset off-grid conditions include reaching the user-set expected state of charge or reaching the latest off-grid time.

[0025] The following constraints are applied during the calculation process:

[0026] State of charge equilibrium constraints:

[0027] ;

[0028] In the formula, For the first electric vehicles State of charge at any given moment; Improve electric vehicle charging efficiency; For the first Battery capacity of an electric vehicle; This refers to the charging time.

[0029] Upper and lower limits of state of charge constraints:

[0030] ;

[0031] In the formula, , The first The minimum and maximum states of charge of an electric vehicle;

[0032] Network entry and exit time constraints:

[0033] ;

[0034] In the formula, and The first The registration and deregistration times of electric vehicles; and The first The earliest registration time and the latest deregistration time of each electric vehicle;

[0035] Based on the above constraints, the charging power of each electric vehicle at each time point is adjusted, and the charging load of all electric vehicles in the on-demand charging mode is obtained by summing up the data.

[0036] ;

[0037] in For electric vehicles in the instant charging mode Electric vehicle charging load at any given time.

[0038] Furthermore, based on the electric vehicle charging load under the instant charging mode and the pre-acquired photovoltaic power output forecast data of the substation, a matching value between the photovoltaic power output and the total charging load is calculated using cosine similarity. A dynamic electricity price discount model is then established based on this matching value, outputting a dynamic time-of-use electricity price, including:

[0039] The electric vehicle charging load under the instant charging mode With the obtained photovoltaic power generation forecast data of the substation area As input, cosine similarity is used to calculate the matching value between the two curves:

[0040] ;

[0041] In the formula, For the matching value, The cosine similarity between the two curves;

[0042] Obtain peak-valley time-of-use electricity pricing from the power grid By adding electricity price discount percentages The dynamic time-of-use electricity price is obtained using the following formula:

[0043] ;

[0044] ;

[0045] In the formula, The original peak-valley electricity price; , , These are their peak electricity price, normal electricity price, and off-peak electricity price, respectively. , , These are their corresponding time periods; Dynamic time-of-use pricing;

[0046] The hyperbolic tangent function is used to describe the nonlinear mapping between the matching value deviation and the electricity price discount ratio, thus obtaining the electricity price discount ratio. The calculation formula is as follows:

[0047] ;

[0048] In the formula: This refers to the electricity price discount percentage; The baseline threshold for the matching value; Discount factor; This represents the rate of change in electricity price discounts.

[0049] Furthermore, based on the aforementioned dynamic time-of-use electricity pricing, combined with consumer psychology models and user multi-attribute difference data, an electric vehicle interactive response model is constructed to output the electric vehicle's response decision, including:

[0050] Obtain the electricity price discount ratio in the dynamic time-of-use electricity price. ;

[0051] Based on a consumer psychology model, a function for the average user response rate is constructed, taking the electricity price discount percentage as input, and calculating the average user response rate through a piecewise function.

[0052] ;

[0053] In the formula, Average user response rate; and These are the dead zone response rate and the saturation zone response rate, respectively. and These are the dead zone inflection point and the saturation zone inflection point, respectively.

[0054] We obtained multi-attribute difference data on users, including economic income, age, and awareness of electricity pricing policies, and constructed economic income influencing factors for each. Age influencing factor Factors influencing awareness of electricity pricing policies The values ​​of each influencing factor are constrained to the interval (0,1);

[0055] The product of the three attribute influence factors is normalized to obtain the comprehensive influence factor. :

[0056] ;

[0057] User calculation based on dynamic electricity price exist Base response rate at any given time:

[0058] ;

[0059] In the formula, For users exist Baseline response rate at any given time; For users exist Expected electricity price at any given time; The highest electricity price;

[0060] By combining the baseline response rate with comprehensive influencing factors, an individual user response rate model is constructed:

[0061] ;

[0062] In the formula, For users exist Individual response rate at any given time;

[0063] Based on the preset average user response rate and individual response rate Output the first result according to the following judgment rules. Response decision of an electric vehicle:

[0064] ;

[0065] In the formula, For the first Response decisions of electric vehicles; This indicates that the user is not participating; This indicates user participation.

[0066] Furthermore, the objective function of the transformer area collaborative decision-making model includes:

[0067] With maximizing the total daily operating revenue of the charging stations in the area as the optimization objective, an objective function is constructed. Total revenue Including charging revenue from users Electricity purchase costs paid to the upstream main grid And the penalty losses due to insufficient solar power consumption :

[0068] ;

[0069] ;

[0070] In the formula, Dynamic time-of-use pricing; for The total charging power of all electric vehicles at any given time; Peak-valley electricity pricing on the grid side; for The amount of electricity purchased from the power grid by the charging stations in the designated area at any given time; This refers to the amount of curtailed solar power generated by distributed photovoltaic (PV) systems, of which:

[0071] ;

[0072] In the formula, and They are respectively The charging power of electric vehicles that participate in and do not participate in the scheduling at all times; for Distributed photovoltaic power generation capacity at all times; Discarding loss factor:

[0073] ;

[0074] ;

[0075] In the formula, The percentage of solar power curtailment; , , These represent low, moderate, and high wastelight loss factors, respectively. and These are low and high light curtailment ratios, respectively.

[0076] Furthermore, the constraints of the aforementioned district collaborative decision-making model include:

[0077] Average charging time constraint:

[0078] ;

[0079] In the formula, The average charging time for electric vehicles; and These are the upper and lower limits of the average charging time.

[0080] Expected state of charge constraints:

[0081] ;

[0082] In the formula, and The first The expected state of charge and maximum state of charge of an electric vehicle; For the first The expected charge coefficient of an electric vehicle;

[0083] Extended offline time constraints:

[0084] ;

[0085] In the formula, For the first The off-grid time for electric vehicles participating in the dispatch was extended. For the first The extended off-grid time is allowed for electric vehicles participating in the dispatch;

[0086] Charging time window constraint:

[0087] ;

[0088] In the formula, For the first vehicle The charging power of electric vehicles at all times; and The first The time when an electric vehicle enters or leaves the network;

[0089] Electric power balance constraints:

[0090] At any given time, the total charging demand within the distribution area must be met jointly by photovoltaic power supply and electricity purchased from the main power grid.

[0091] ;

[0092] In the formula, for The total charging power of all electric vehicles at any given time; for The charging power constantly comes from photovoltaics; for The amount of electricity purchased from the power grid by the charging stations in the designated area at any given time;

[0093] Photovoltaic output constraints:

[0094] The actual photovoltaic power used for charging cannot exceed its maximum output limit:

[0095] ;

[0096] In the formula, for Photovoltaic output power at any given time; This represents the maximum output power of the photovoltaic system.

[0097] Grid-side interactive power constraints:

[0098] The concentrated charging of a large number of electric vehicles may lead to a large peak-valley difference in the power grid, necessitating the limitation of the power exchange between charging stations and the upstream distribution network.

[0099] ;

[0100] In the formula, for The charging power constantly supplied by the power grid; This represents the maximum peak-to-valley difference allowed by the power grid.

[0101] Furthermore, the solution to the transformer area collaborative decision-making model employs an improved particle swarm optimization algorithm, including:

[0102] Initialize the population and set the population size to [value]. ,generate Each particle represents a charging power allocation scheme. The upper bound of the variable is dynamically determined according to the allowed charging time window. Power is allocated only during the allowed charging period, and the upper bound of the power is 0 during other periods.

[0103] Perform a three-step constraint repair for each particle:

[0104] Time window reset fix: Sets the power exceeding the allowed charging time window to 0;

[0105] State of charge upper limit constraint correction: If the charging power causes the state of charge to exceed the limit, the power during that period will be reduced;

[0106] Energy demand replenishment and repair: If the expected state of charge is not met after repair, the power will be increased to make up the energy gap within the feasible period.

[0107] The fitness value of each repaired particle is calculated based on the objective function.

[0108] Introducing a linearly decreasing inertial weight strategy to update particle velocity and position:

[0109] The inertia weight decreases linearly:

[0110] ;

[0111] In the formula, This represents the current iteration number. The maximum number of iterations, As the initial inertia weight, This refers to the inertia weight at the end of the iteration;

[0112] Speed ​​updates:

[0113] ;

[0114] In the formula, , As a learning factor, , It is a random number. This represents the optimal position in the history of an individual particle. The optimal position for the entire group;

[0115] Location update:

[0116] ;

[0117] After the update, the particle positions will be clipped to within the preset feasible region:

[0118] ;

[0119] When the maximum number of iterations is reached Or the change in the continuously preset algebraic global optimum is less than a threshold. When the iteration terminates, the global optimal solution is output. The global optimal solution is used as the optimal electric vehicle charging power allocation scheme and the corresponding photovoltaic absorption capacity.

[0120] Secondly, the present invention provides a device for coordinated decision-making on electric vehicle resources and photovoltaic power consumption in a transformer substation, used to implement the coordinated decision-making method for electric vehicle resources and photovoltaic power consumption in a transformer substation as described in any of the preceding claims, comprising:

[0121] The acquisition module is used to acquire data on the driving patterns of electric vehicles, battery charging and discharging characteristics, and state of charge.

[0122] The first processing module is used to input the driving pattern data, battery charging and discharging characteristic data and state of charge data into a pre-constructed fine load model of electric vehicles to obtain the charging power and state of charge of each individual electric vehicle at each time.

[0123] The aggregation module is used to apply constraints to the charging power and state of charge of each individual electric vehicle at each time in the instant charging mode, and aggregate them to obtain the electric vehicle charging load in the instant charging mode.

[0124] The second processing module is used to calculate the matching value between photovoltaic output and total charging load based on the electric vehicle charging load under the instant charging mode and the pre-acquired photovoltaic power output data of the transformer area through cosine similarity, establish a dynamic electricity price discount model based on the matching value, and output dynamic time-of-use electricity price.

[0125] The third processing module is used to construct an electric vehicle interactive response model based on the dynamic time-of-use electricity price, combined with the consumer psychology model and the user's multi-attribute difference data, and output the electric vehicle's response decision.

[0126] The decision-making module is used to solve, under preset constraints, the optimal electric vehicle charging strategy and photovoltaic absorption scheme based on the response decision of the electric vehicle and a pre-built district collaborative decision-making model with the objectives of maximizing charging station revenue and maximizing photovoltaic absorption.

[0127] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0128] Fourthly, the present invention provides an electronic device, comprising:

[0129] Memory, used to store computer programs / instructions;

[0130] A processor for executing the computer program / instructions to implement the steps of any of the methods described above.

[0131] Fifthly, the present invention provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of any of the methods described above.

[0132] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0133] 1. This invention provides a method and device for coordinated decision-making between electric vehicle resources and photovoltaic (PV) consumption in a distribution area. It employs a cosine similarity metric to quantify the matching degree between PV output and PV charging load, and constructs a dynamic electricity price discount model using a hyperbolic tangent function. The electricity price is adjusted in real time based on the source-load matching status, effectively guiding PV charging load to shift to peak PV generation periods. Simultaneously, a tiered curtailment penalty mechanism is incorporated into the optimization objective to differentiate and constrain curtailment behavior. Simulation results show that this scheme can significantly reduce curtailed PV power and curtailment rate, greatly improve the local consumption level of distributed PV, and fully tap the potential of new energy utilization.

[0134] 2. This invention combines consumer psychology to construct a basic response model and incorporates three individual attribute factors: user income, age, and electricity price awareness. This enables differentiated modeling and precise guidance of response behavior for different user groups. While ensuring users' normal charging needs, dynamic electricity price incentives enhance users' enthusiasm for participating in interactive scheduling, while simultaneously reducing users' average charging costs, achieving a win-win situation for both grid scheduling and user interests.

[0135] 3. This invention aims to maximize both charging station revenue and photovoltaic (PV) consumption, comprehensively considering multiple economic indicators such as charging revenue, grid power purchase costs, and curtailment penalties. Through load timing optimization, the cost of electricity purchase for the distribution area and curtailment penalty expenditures are significantly reduced; the total amount of electric vehicle charging increases after dispatching, further increasing charging revenue, ultimately resulting in a substantial increase in the overall revenue of the charging station, demonstrating good economic viability for engineering applications.

[0136] 4. This invention strictly considers multiple operational constraints such as battery state of charge, charging time window, power balance, and grid interaction power. By optimizing and dispersing centralized charging loads through orderly charging, it effectively reduces the peak-valley difference of the power grid and avoids problems such as distribution transformer overload and voltage exceeding limits. It fully leverages the adjustment capability of electric vehicle flexible loads to achieve coordinated operation of distributed photovoltaics, electric vehicles, charging stations, and the distribution network, providing a feasible technical solution for the interaction of "source, grid, load, and storage" in the new power system. Attached Figure Description

[0137] Figure 1 This is a flowchart of a collaborative decision-making method for electric vehicle resources and photovoltaic power consumption in a transformer substation, provided by an embodiment of the present invention;

[0138] Figure 2 This is a framework diagram for collaborative decision-making regarding electric vehicle resources and photovoltaic power consumption in the power grid area.

[0139] Figure 3 This is a diagram of the electric vehicle user response decision-making architecture;

[0140] Figure 4 It is a curve of average user response rate based on consumer psychology;

[0141] Figure 5 This is a comparison chart of photovoltaic power output and electric vehicle aggregated charging power before and after optimization;

[0142] Figure 6 This is a heat map of electric vehicle charging scheduling;

[0143] Figure 7 This is a schematic diagram of the single-vehicle charging curve and the aggregated charging curve of an electric vehicle. Detailed Implementation

[0144] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0145] Example 1: This example introduces a collaborative decision-making method for electric vehicle resources and photovoltaic power consumption in a power distribution area, including:

[0146] Acquire data on the driving patterns of electric vehicles, battery charging and discharging characteristics, and state of charge.

[0147] The driving pattern data, battery charging and discharging characteristic data, and state of charge data are input into a pre-constructed fine load model of electric vehicles to obtain the charging power and state of charge of each individual electric vehicle at each time.

[0148] The charging power and state of charge of each individual electric vehicle at each time are constrained in the instant charging mode, and the charging load of the electric vehicle in the instant charging mode is obtained by aggregating them.

[0149] Based on the electric vehicle charging load under the instant charging mode and the pre-acquired photovoltaic power output data of the transformer area, the matching value between photovoltaic power output and the total charging load is calculated by cosine similarity. Based on the matching value, a dynamic electricity price discount model is established, and a dynamic time-of-use electricity price is output.

[0150] Based on the dynamic time-of-use electricity price, combined with consumer psychology models and user multi-attribute difference data, an electric vehicle interactive response model is constructed to output the electric vehicle's response decision.

[0151] Based on the response decision of the electric vehicle and the pre-constructed district collaborative decision-making model with the objectives of maximizing charging station revenue and maximizing photovoltaic absorption, the optimal electric vehicle charging strategy and photovoltaic absorption scheme are output under preset constraints.

[0152] like Figure 1-3 As shown in the figure, the collaborative decision-making method for electric vehicle resources and photovoltaic power consumption in the transformer substation provided in this embodiment involves the following steps in its application process:

[0153] Step 1: Construct a refined load model for electric vehicles that considers charge / discharge factors based on the log-normal distribution and the Sigmoid function; details are as follows:

[0154] The charge / discharge factor of an electric vehicle directly affects its charge / discharge power, rate of change of state of charge, and charging time, thereby altering the vehicle's time window for participating in substation interactions and its response capability. An electric vehicle model considering the charge / discharge factor is constructed as follows:

[0155]

[0156] In the formula, For the first electric vehicles The charging power at any given moment; and The first Maximum charging and discharging power of a single electric vehicle; and For the first The constraint factors for the maximum charging and discharging power of an electric vehicle, and their relationship with the electric vehicle's State of Charge (SOC), are as follows:

[0157]

[0158] In the formula, For the first The current state of charge of the electric vehicle; and These are the state-of-charge reference values ​​corresponding to the charging and discharging factors, respectively; and These are the variation coefficients of the charging and discharging factors, respectively, used to characterize the sensitivity of the charging and discharging factors to changes in SOC.

[0159] Step 2: Based on the refined electric vehicle load model from Step 1, obtain the electric vehicle charging load under the on-demand charging mode; introduce a cosine similarity metric to quantify the matching degree between photovoltaic power output and the electric vehicle charging load under the on-demand charging mode, and establish a dynamic electricity price discount model based on the matching degree; details are as follows:

[0160] Based on the refined load model of electric vehicles in step 1, and considering constraints such as electric vehicle SOC balance, electric vehicle battery state of charge, and grid connection and disconnection time, the charging load of electric vehicles under the on-demand charging mode is obtained.

[0161] 1) Constraints on SOC balancing:

[0162]

[0163] In the formula, For the first electric vehicles State of charge at any given moment; Improve electric vehicle charging efficiency; For the first Battery capacity of an electric vehicle; This refers to the charging time.

[0164] 2) Electric vehicle battery state of charge constraints:

[0165] The state of charge (SOC) value cannot be lower than the minimum permissible SOC, nor can it exceed the maximum permissible SOC.

[0166]

[0167] In the formula, , The first The minimum and maximum states of charge of an electric vehicle.

[0168] 3) Network entry and exit time constraints:

[0169]

[0170] In the formula, and The first The registration and deregistration times of electric vehicles; and The first The earliest time an electric vehicle is registered with the grid and the latest time it is deregistered.

[0171] Obtain the charging load of electric vehicles in the on-demand charging mode .

[0172] Select typical daily photovoltaic power output forecast data Using the initial aggregate charging power data of electric vehicles under the "on-demand charging" mode, a cosine similarity index is used to evaluate and calculate the matching value between photovoltaic output and the initial charging curve of electric vehicles, and this value is used as an important parameter for setting charging price discounts in the future.

[0173] Cosine similarity reflects the directional similarity between two vectors, and its value ranges from 1 to 2. The formula for calculating the interval is as follows:

[0174]

[0175] In the formula, For matching values; Let be the cosine similarity between the two curves.

[0176] To further explore the interactive potential of flexible resources in the distribution area, a price discount model based on photovoltaic-load matching value is constructed. This is based on the grid's peak-valley time-of-use pricing. Based on this, by adding electricity price discounts Guide users' charging behavior. The calculation method for dynamic time-of-use electricity pricing is as follows: As shown:

[0177]

[0178]

[0179] In the formula, The original peak-valley electricity price; , , These are their peak electricity price, normal electricity price, and off-peak electricity price, respectively. , , These are their corresponding time periods; This refers to dynamic time-of-use pricing. Dynamic time-of-use pricing enables real-time adjustment of electricity market prices, exhibiting significant immediacy. A hyperbolic tangent function is used to describe the nonlinear mapping between the matching value deviation and the electricity price discount ratio. The calculation formula is as follows:

[0180]

[0181] In the formula: This refers to the electricity price discount percentage; The baseline threshold for the matching value; Discount factor; This represents the rate of change in electricity price discounts.

[0182] Step 3: As Figure 4 As shown, based on consumer psychology and differences in multiple user attributes, the dynamic electricity price discount model from step 2 is introduced to construct an electric vehicle interaction response model that takes into account income, age, and electricity price awareness, thereby obtaining the judgment result on whether electric vehicle users participate in the interaction; specifically as follows:

[0183] (1) User average response rate model based on consumer psychology

[0184] Based on a consumer psychology model of demand response, this study quantitatively analyzes the participation intentions of electric vehicle users. The relationship between user demand response rate and incentive intensity is modeled as a piecewise function as follows:

[0185]

[0186] In the formula, Average user response rate; and These are the dead zone response rate and the saturation zone response rate, respectively. and These represent the dead zone inflection point and the saturation zone inflection point, respectively. The growth zone is characterized by a quadratic function depicting the increase in response rate with electricity price discounts.

[0187] (2) User response rate model considering user attributes

[0188] To accurately characterize the response rates of different electric vehicle users under varying electricity prices, a response rate model for electric vehicle users is constructed, with electricity price differences as the core factor and considering individual differences. This model introduces three influencing factors: economic income, age, and awareness of electricity price policies. A comprehensive correction factor is designed to quantify the differentiated response rates. The model first characterizes the basic response rate of individual users using dynamic electricity prices, and then corrects for these rates using the three attribute factors, ultimately yielding the user's response rate. exist The response rate at any given time is determined by the following steps:

[0189] 1) User base response rate based on dynamic time-of-use pricing

[0190] The formula for calculating the baseline response rate is:

[0191]

[0192] In the formula, For users exist Baseline response rate at any given time; For users exist Expected electricity price at any given time; This is the highest electricity price.

[0193] The specific design of the response influence factor considering user-differential attributes in step 3 is as follows:

[0194] To quantify the differentiated response rates of different electric vehicle users, influencing factors were designed for three attributes: economic income, age, and awareness of electricity pricing policies. Each factor's value was constrained to the (0,1) range; a higher factor value indicates a stronger positive impact of that attribute on the user's response rate. The product of the three attribute factors was normalized to ensure that the value of the comprehensive influencing factor was strictly constrained to the (0,1) range, guaranteeing a reasonable correction to the electricity price-driven baseline response rate. The calculation formula is as follows:

[0195]

[0196] In the formula, For users The comprehensive influencing factors; , , users respectively The factors influencing the price of electricity include economic income, age, and awareness of electricity pricing policies. The calculation methods for each influencing factor are as follows:

[0197] ① Factors affecting economic income

[0198] Considering that the impact of income changes on the electricity price sensitivity of ultra-high-income users tends to saturate, a piecewise exponentially decreasing function is designed to characterize the impact of economic income. This reflects both the negative nonlinear correlation between income and regulatory intentions and avoids the extreme case where ultra-high income causes the factor value to approach 0. The formula for calculating the economic income impact factor is as follows:

[0199]

[0200] In the formula, For users Economic income; The coefficient of change of the economic income influencing factor; , and These are categorized as low-income, middle-income, and high-income users, respectively.

[0201] ② Age-related factors

[0202] Average age of all electric vehicle users A benchmark is established, and a Gaussian kernel function is introduced to characterize the non-linear effect of age. The closer a user's age is to the benchmark value, the higher the age factor value and the higher the response rate; the further away from the benchmark value, the more exponentially the factor value decays. The formula for calculating the age influence factor is:

[0203]

[0204] In the formula, The baseline coefficient for the age factor; For users Age; and These represent the mean age and standard deviation, respectively.

[0205] ③ Factors influencing public awareness of electricity pricing policies

[0206] Electricity price policy awareness is a discrete categorical variable, unsuitable for continuous function characterization. Therefore, a step-wise piecewise function is designed to divide awareness into three levels based on a threshold, each corresponding to a different factor value. The factor value increases sequentially with increasing awareness. The formula for calculating the factor influencing electricity price policy awareness is as follows:

[0207]

[0208] In the formula, , , Awareness factor of electricity pricing policy; ; , , The threshold for public awareness of electricity pricing policies; For users Awareness of electricity pricing policies.

[0209] The specific model for the overall individual electric vehicle user response rate in step 3 is as follows:

[0210] By combining the baseline response rate with a comprehensive influencing factor of individual attributes, a user response rate model that considers individual differences is constructed as follows:

[0211]

[0212] In the formula, For users exist Response rate at any given moment.

[0213] The electric vehicle user response decision is set as a binary choice. Based on the aforementioned average user response rate model based on consumer psychology and the user response rate model considering individual differences, the average user response rate and individual user response rate are derived. These are then used to determine whether an individual user participates in adjusting the charging time. When the actual user response rate meets the criteria... When the electricity price discount exceeds the dead zone inflection point value, the user is determined to participate in optimized dispatch; when When this happens, it is determined that the user will not participate in optimized scheduling, and the expression is:

[0214]

[0215] In the formula, For the first Response decisions of electric vehicles; This indicates that the user is not participating; This indicates user participation; Average user engagement rate; For users exist Response rate at any given moment; Electricity price discount; This refers to the electricity price discount corresponding to the dead zone inflection point.

[0216] Step 4: Based on the results of electric vehicle participation in Step 3, and with the dual objectives of maximizing charging station revenue and maximizing photovoltaic absorption, establish a district collaborative decision-making model with constraints such as power balance, curtailment penalty, and charging window.

[0217] The objective function and constraints of the collaborative decision-making model for the transformer substations are as follows:

[0218] (1) Objective function

[0219] With the goal of maximizing the total daily operating revenue of the charging stations in the area, a decision objective function is constructed, and the total revenue is... This includes charging users for their charging revenue. Electricity purchase costs paid to the upstream main grid And the penalty losses due to insufficient solar power consumption :

[0220]

[0221]

[0222] In the formula, Dynamic time-of-use pricing; for The total charging power of all electric vehicles at any given time; Peak-valley electricity pricing on the grid side; for The amount of electricity purchased from the power grid by the charging stations in the designated area at any given time; This refers to the amount of curtailed solar power generated by distributed photovoltaic (PV) systems, of which:

[0223]

[0224] In the formula, and They are respectively The charging power of electric vehicles that participate in and do not participate in the scheduling at all times; for Distributed photovoltaic power generation capacity at all times; Discarding loss factor:

[0225]

[0226]

[0227] In the formula, The percentage of solar power curtailment; , , These represent low, moderate, and high wastelight loss factors, respectively. and These represent low and high curtailment rates, respectively. Curtailment loss factor. Defined as a step function, it embodies the principle of piecewise, step-wise penalty for light wastage. Based on the proportion of light wastage... The interval division uses economic means to differentiate and constrain curtailment behavior. For low curtailment situations, a lower penalty coefficient is used to ensure operational flexibility, while for high curtailment situations, the penalty weight is significantly increased.

[0228] (2) Constraints

[0229] To ensure the feasibility and security of the cooperative decision-making model for the distribution area, in addition to the constraints already considered, such as electric vehicle charging power constraints, SOC balance constraints, upper and lower limits of state of charge constraints, and grid connection and disconnection time constraints, it is also necessary to further consider constraints such as average charging time, expected state of charge, extended disconnection time, charging time window, power balance, photovoltaic output, and grid-side interactive power.

[0230] 1) Average charging time constraint:

[0231]

[0232] In the formula, The average charging time for electric vehicles; and These represent the upper and lower limits of the average charging time.

[0233] 2) Expected state of charge constraints:

[0234]

[0235] In the formula, and The first The expected state of charge and maximum state of charge of an electric vehicle; For the first The expected charge factor of an electric vehicle.

[0236] 3) Extend the offline time constraint

[0237]

[0238] In the formula, For the first The time that electric vehicles participating in the dispatch were to be off the grid was extended.

[0239] 4) Charging time window constraint

[0240]

[0241] In the formula, For the first vehicle The charging power of electric vehicles at all times; and The first The time when an electric vehicle enters or leaves the network.

[0242] 5) Power balance constraint

[0243] At any given time, the total charging demand within the distribution area must be met jointly by photovoltaic power supply and electricity purchased from the main power grid.

[0244]

[0245] In the formula, for The total charging power of all electric vehicles at any given time; for The charging power constantly comes from photovoltaics; for The charging station in the karaoke area receives electricity from the power grid.

[0246] 6) Photovoltaic output constraints

[0247] The actual photovoltaic power used for charging cannot exceed its maximum output limit:

[0248]

[0249] In the formula, for Photovoltaic output power at any given time; This represents the maximum output power of the photovoltaic system.

[0250] 7) Grid-side interactive power constraints

[0251] The concentrated charging of a large number of electric vehicles may lead to a large peak-valley difference in the power grid, necessitating the limitation of the power exchange between charging stations and the upstream distribution network.

[0252]

[0253] In the formula, for The charging power constantly supplied by the power grid; This represents the maximum peak-to-valley difference allowed by the power grid.

[0254] Step 5: For the collaborative decision-making model of the transformer substations in Step 4, an improved particle swarm optimization algorithm with constrained repair and linearly decreasing inertia weights is used to solve the problem and obtain the optimal electric vehicle charging strategy and photovoltaic power consumption scheme.

[0255] In step 5, the particle swarm optimization algorithm is selected as the basic solution framework, and an improved particle swarm optimization algorithm is obtained through targeted improvements, including:

[0256] To address the shortcomings of the standard PSO algorithm in solving the charging power optimization problem in this chapter, such as slow convergence speed, weak constraint handling capability, and poor adaptability to high dimensions, an improved Particle Swarm Optimization algorithm is designed. The core idea is to introduce a linearly decreasing inertia weight strategy and a constraint repair mechanism on the basis of the standard PSO algorithm, while optimizing the population initialization method and the pruning operation after position update, thereby improving the convergence speed and optimization accuracy, ensuring constraint feasibility, and making it more suitable for nonlinear nonconvex charging power optimization problems containing max operators and step penalty terms.

[0257] (1) Population initialization

[0258] Set the population size to ,generate Each particle represents a charging power allocation scheme. The upper bound of the variable is dynamically determined based on the user's charging and discharging factor and time window. That is, power is allocated only during the allowed charging period, and the upper bound of power is 0 during other periods. This compresses the invalid solution space and improves the feasibility and diversity of the initial population.

[0259] (2) Constraint Repair Mechanism

[0260] To ensure that all candidate solutions satisfy the constraints of time window, SOC, and energy requirement, this paper designs a three-step constraint repair mechanism to correct each particle:

[0261] 1) Time window clearing and repair: If the power allocation in a particle exceeds the allowed charging time window during a certain period, the power of that period will be set to 0 directly to eliminate invalid charging behavior;

[0262] 2) SOC upper limit constraint correction: Based on the battery SOC upper limit, if the charging power allocation scheme corresponding to a certain particle will cause the battery SOC to exceed the upper limit, the charging power during that period will be adjusted to reduce the charging rate, so that the SOC is always controlled within the specified upper limit, avoiding battery overcharging and ensuring charging safety.

[0263] 3) Energy demand replenishment and repair: If the charging power allocation scheme for the particle cannot meet the preset energy demand after the first two repair steps, the charging power for the corresponding period is appropriately increased to make up for the energy gap, under the premise of meeting the time window constraint and the SOC upper limit constraint, so as to ensure that the charging power allocation scheme is both feasible and meets the actual energy demand.

[0264] (3) Fitness evaluation

[0265] Based on the objective function constructed in this paper, the total reward corresponding to each repaired particle is calculated and used as the fitness value of the particle.

[0266] (4) Velocity and position update

[0267] To balance global exploration and local exploitation capabilities, a linearly decreasing inertia weight strategy is introduced, and particle velocity and position are updated according to standard PSO:

[0268] 1) Inertia weight decreases linearly

[0269]

[0270] In the formula, This represents the current iteration number. The maximum number of iterations, As the initial inertia weight, This refers to the inertia weights at the end of the iteration. This strategy allows the algorithm to enhance global search capabilities with larger weights in the early stages and improve local search accuracy with smaller weights in the later stages.

[0271] 2) Speed ​​Update

[0272]

[0273] In the formula, , As a learning factor, , It is a random number. This represents the optimal position in the history of an individual particle. It is the optimal position for the entire group.

[0274] 3) Update the location and trim to the feasible region.

[0275]

[0276] After the update, the particle positions will be clipped to within the preset feasible region, that is:

[0277]

[0278] At this point, it is further ensured that the variable values ​​do not exceed the physical constraints.

[0279] 4) Termination condition determination

[0280] The algorithm terminates when the maximum number of iterations is reached. Or the change in the global optimum over 20 consecutive generations is less than a preset threshold. When any condition is met, the algorithm stops iterating and outputs the global optimal solution. .

[0281] The following description, in conjunction with a preferred embodiment, illustrates the content involved in the above embodiments.

[0282] A commercial transformer substation was selected as the simulation scenario. This substation is equipped with 100 DC fast charging piles, serving 500 electric vehicle users. Each vehicle has a 60kWh battery capacity, the maximum charging power per pile is 60kW, and the charging efficiency is 0.95. A 5000kW distributed photovoltaic power station is built within the substation, with a typical daily total photovoltaic power generation of 26383.9kWh. Peak photovoltaic output is concentrated between 10:00 and 16:00. The substation also has basic loads from residential and commercial users, with a daily electricity consumption of 38017.4kWh and a peak power of 2463kW.

[0283] (1) Basic data on photovoltaics and electric vehicles are shown in Table 1.

[0284] Table 1 Basic Data on Photovoltaics and Electric Vehicles

[0285]

[0286] (2) Basic data of electric vehicle users

[0287] Table 2 shows the initial state of charge, monthly income, age, and awareness of electricity pricing policies among electric vehicle users. The consumer psychology model uses uniform parameters: dead zone inflection point, saturation zone inflection point, dead zone response rate, and saturation zone response rate are 0.05, 0.25, 0.2, and 0.9, respectively. Table 2 shows the relevant parameters of the user response rate model considering user attributes.

[0288] Table 2 Basic Data of Electric Vehicle Users

[0289]

[0290] (3) Algorithm data

[0291] Improved Particle Swarm Optimization Algorithm Particle Count The maximum number of iterations is 40. The learning factor is 120. , 2.0; Inertia weight and The values ​​are 0.9 and 0.4 respectively, and they converge to the optimal value within about 6.8 seconds.

[0292] Optimization Result Analysis

[0293] (1) Photovoltaic absorption analysis

[0294] Based on the cosine similarity between the on-demand charging load curve of electric vehicles and the photovoltaic output curve The initial value was below 0.75; after optimization, the discount factor was set to 0.32, and the dynamic electricity price discount mechanism was implemented. With the discount, and under the combined effect of the consumer psychology response model and user attribute influencing factors, 310 out of 500 electric vehicles were ultimately dispatched, with a participation rate of 62.0%.

[0295] The photovoltaic output, base load, and aggregated charging power of electric vehicles were compared between the two modes before and after optimization, and the results are as follows: Figure 5As shown, under the basic on-demand charging mode, electric vehicle charging behavior exhibits a significant morning peak characteristic influenced by user travel habits. Charging power increases sharply between 9:00 and 11:00, while photovoltaic output is still in its ramp-up phase, resulting in a mismatch between the two power curves in terms of timing. After 11:00, as vehicles gradually complete charging and enter the off-grid waiting state, electric vehicle charging power drops sharply, while photovoltaic output continues to climb to its peak between 12:00 and 14:00. During the peak photovoltaic power generation period from 13:00 to 17:00, electric vehicle charging demand is almost zero, leading to a large amount of photovoltaic power being unable to be consumed locally and resulting in curtailment. Under the optimized charging mode, the 310 electric vehicles participating in the scheduling significantly shift their charging load to the peak photovoltaic power generation period. The aggregated power of electric vehicles maintains a relatively stable and high level during this period, forming a good synergy with the photovoltaic output curve.

[0296] The photovoltaic absorption effect before and after optimization is shown in Table 3:

[0297] Table 3 Comparison of Photovoltaic Absorption Effect

[0298]

[0299] As shown in Table 3, through dynamic electricity pricing guidance and flexible user dispatch, the grid integration rate of distributed photovoltaic (PV) power in the distribution area increased from 56.9% under the immediate charging mode to 84.3% under the optimized mode, an increase of 27.4%. Correspondingly, the curtailment rate dropped significantly from 43.1% to 15.7%, and the amount of PV power consumed increased from 15019.1 kWh to 22252.1 kWh. Regarding the electric vehicle (EV) charging load itself, the total charging capacity of EVs increased from 14749.8 kWh to 16711.0 kWh after optimization. This is because the extended off-grid time of dispatching users allows for additional charging to absorb more PV power. The optimized curtailment rate decreased from 43.1% to 15.7%, with a 50% reduction in penalty per unit of curtailed PV power. The optimized charging load not only matches the peak daytime PV output but also utilizes the surplus PV power in the evening, further tapping the potential for flexible resource utilization in the distribution area and maximizing the absorption of PV output throughout the day.

[0300] (2) Economic analysis

[0301] To fully verify the comprehensive benefits of the proposed optimized strategy for orderly charging of electric vehicles oriented towards photovoltaic power consumption, the economic costs, benefits, and user charging economy before and after optimization were quantitatively analyzed. The comparative data of various indicators are shown in Table 4.

[0302] Table 4 Comparison of Economic Indicators

[0303]

[0304] Table 4 shows the following: First, the cost of electricity purchase has decreased significantly. Under the "on-demand charging" model, electric vehicles are concentrated in the morning, when the photovoltaic output is insufficient to meet the total charging load demand of electric vehicles. This necessitates purchasing electricity from the grid, resulting in a cost as high as 8713.9 yuan. After optimization, the charging power of 310 vehicles participating in the dispatch is transferred to the peak photovoltaic generation period, allowing photovoltaic power to first absorb the electric vehicle charging demand locally, reducing the electricity purchase cost to 5315.7 yuan, a relative decrease of 39.0%. Second, the amount of curtailed solar power decreased from 11364.8 kWh to 4131.8 kWh, a decrease of 63.6%. Simultaneously, the curtailment rate decreased from 43.1% to 15.7%, resulting in a decrease in the curtailment penalty coefficient from 0.30 yuan / kWh to 0.15 yuan / kWh, and a decrease in the curtailment penalty from 3409.4 yuan to 619.8 yuan, saving 2789.7 yuan. Third, charging revenue has slightly increased rather than decreased. Although dynamic pricing discounts during certain periods may lower revenue per unit of electricity, extending the off-grid time of participating users by 2 hours allows them to charge more electricity. After optimization, the total charging capacity of electric vehicles increased from 14749.8 kWh to 16711.0 kWh, resulting in an increase in charging revenue of 1049.5 yuan. Thirdly, the average charging price for users decreased. The average price actually paid by users to the charging station decreased from 1.186 yuan / kWh to 1.109 yuan / kWh, a reduction of 0.077 yuan / kWh, providing a considerable economic incentive from the electricity price perspective. In summary, the total revenue of the charging station before optimization was 5365.6 yuan, which increased to 12599.2 yuan after optimization, a net increase of 7233.6 yuan per day, representing a growth rate of 134.9%. These results demonstrate that the proposed optimization strategy not only improves the photovoltaic absorption rate and the operating characteristics of the distribution area but also brings significant economic benefits, providing economic support for flexible resource interaction in the distribution area and large-scale distributed photovoltaic grid connection.

[0305] (3) Panoramic Analysis of Cluster Electric Vehicle Charging Scheduling

[0306] The overall charging schedule for 500 electric vehicles in both the on-demand charging mode and the optimized mode is as follows: Figure 6As shown in the figure, the horizontal axis represents the 24-hour scheduling period, and the vertical axis represents the electric vehicle numbers arranged in ascending order of grid connection time. The color depth of each cell corresponds to the charging power value of the corresponding vehicle during that time period. In the "on-demand charging" mode, the bright spots of all 500 electric vehicles form a diagonal band extending from the lower left to the upper right along the Y=X direction. The width of the band is approximately 1-2 cells, or 1-2 hours, and the color depth is generally high, corresponding to the "on-demand charging" behavior of users who charge as soon as they connect to the grid and stop charging when fully charged. In the optimized mode, the bright spots of the 310 vehicles participating in the scheduling are significantly shifted to the peak photovoltaic power generation period from 10:00 to 16:00. The original diagonal band is broken into multiple rows of lighter color blocks with a width of 3-6 hours, and the color depth decreases to the 15-30kW range. Furthermore, the bright spots of each vehicle form a complementary, staggered arrangement on the time axis. The 190 vehicles not participating in the scheduling still retain the diagonal shape of "on-demand charging," but only account for 38% of the total.

[0307] The charging curves of 500 electric vehicles are overlaid and displayed one by one, and the results are as follows: Figure 7 As shown in the figure, the two sub-graphs on the left and right correspond to the instant charging mode and the optimized mode, respectively. Each sub-graph adopts a dual vertical axis structure, with the left axis representing the charging power of a single vehicle and the right axis representing the aggregated charging power of 500 vehicles. It can be seen that in the instant charging mode, the charging power of a single vehicle is distributed between 6:00 and 13:00, which is consistent with the high peak shape of the aggregated load curve between 9:00 and 11:00. In the optimized mode, the charging power of a single vehicle is distributed between 10:00 and 18:00, and the aggregated load curve shows a bell shape with obvious peak reduction and valley filling.

[0308] Example 2: This example provides a collaborative decision-making device for electric vehicle resources and photovoltaic power consumption in a power distribution area, including:

[0309] The acquisition module is used to acquire data on the driving patterns of electric vehicles, battery charging and discharging characteristics, and state of charge.

[0310] The first processing module is used to input the driving pattern data, battery charging and discharging characteristic data and state of charge data into a pre-constructed fine load model of electric vehicles to obtain the charging power and state of charge of each individual electric vehicle at each time.

[0311] The aggregation module is used to apply constraints to the charging power and state of charge of each individual electric vehicle at each time in the instant charging mode, and aggregate them to obtain the electric vehicle charging load in the instant charging mode.

[0312] The second processing module is used to calculate the matching value between photovoltaic output and total charging load based on the electric vehicle charging load under the instant charging mode and the pre-acquired photovoltaic power output data of the transformer area through cosine similarity, establish a dynamic electricity price discount model based on the matching value, and output dynamic time-of-use electricity price.

[0313] The third processing module is used to construct an electric vehicle interactive response model based on the dynamic time-of-use electricity price, combined with the consumer psychology model and the user's multi-attribute difference data, and output the electric vehicle's response decision.

[0314] The decision-making module is used to solve, under preset constraints, the optimal electric vehicle charging strategy and photovoltaic absorption scheme based on the response decision of the electric vehicle and a pre-built district collaborative decision-making model with the objectives of maximizing charging station revenue and maximizing photovoltaic absorption.

[0315] The specific functions of each module described above are explained in the relevant content of the method in Embodiment 1, and will not be repeated here.

[0316] Example 3: This example provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described in Example 1.

[0317] Example 4: This example provides an electronic device, including:

[0318] Memory, used to store computer programs / instructions;

[0319] A processor for executing the computer program / instructions to implement the steps of any of the methods described in Embodiment 1.

[0320] Example 5: This example provides a computer program product, including a computer program / instructions, which, when executed by a processor, implement the steps of the method described in any one of Examples 1.

[0321] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

[0322] Those skilled in the art will understand that embodiments of this disclosure can be provided as methods, systems, or computer program products. Therefore, this disclosure can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this disclosure can take the form of a computer program product embodied 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.

[0323] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0324] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0325] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0326] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this disclosure and not to limit its protection scope. Although this disclosure has been described in detail with reference to the above embodiments, those skilled in the art should understand that after reading this disclosure, they can still make various changes, modifications or equivalent substitutions to the specific implementation of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the pending claims.

Claims

1. A collaborative decision-making method for electric vehicle resources and photovoltaic power consumption in a transformer substation, characterized in that, include: Acquire data on the driving patterns of electric vehicles, battery charging and discharging characteristics, and state of charge. The driving pattern data, battery charging and discharging characteristic data, and state of charge data are input into a pre-constructed fine load model of electric vehicles to obtain the charging power and state of charge of each individual electric vehicle at each time. The charging power and state of charge of each individual electric vehicle at each time are constrained in the instant charging mode, and the charging load of the electric vehicle in the instant charging mode is obtained by aggregating them. Based on the electric vehicle charging load under the instant charging mode and the pre-acquired photovoltaic power output data of the transformer area, the matching value between photovoltaic power output and the total charging load is calculated by cosine similarity. Based on the matching value, a dynamic electricity price discount model is established, and a dynamic time-of-use electricity price is output. Based on the dynamic time-of-use electricity price, combined with consumer psychology models and user multi-attribute difference data, an electric vehicle interactive response model is constructed to output the electric vehicle's response decision. Based on the response decision of the electric vehicle and the pre-constructed district collaborative decision-making model with the objectives of maximizing charging station revenue and maximizing photovoltaic absorption, the optimal electric vehicle charging strategy and photovoltaic absorption scheme are output under preset constraints.

2. The method for coordinated decision-making on electric vehicle resources and photovoltaic consumption in a dynamic electricity price zone according to claim 1, characterized in that, The driving pattern data, battery charging and discharging characteristic data, and state of charge data are input into a pre-constructed fine load model of electric vehicles to obtain the charging power and state of charge of each individual electric vehicle at each time point, including: Using the electric vehicle's grid connection time, grid disconnection time, and mileage as driving pattern data, user travel behavior is fitted based on a log-normal distribution; Using the battery capacity, maximum charge / discharge power, and charging efficiency of electric vehicles as battery charge / discharge characteristic data, and combining this with the current state of charge, a charge / discharge factor model is constructed, as shown in the following formula: ； In the formula, and The first electric vehicles The charging power and discharging power at any given time; and The first The maximum charging power and maximum discharging power of a vehicle; and The first The charging factor and discharging factor of an electric vehicle; The charge / discharge factor includes a charge factor and a discharge factor, and its relationship with the state of charge is as follows: ； In the formula, For the first The current state of charge of the electric vehicle; and These are the state-of-charge reference values ​​corresponding to the charging factor and the discharging factor, respectively; and These are the coefficients of change for the charging factor and the discharging factor, respectively. Obtain the charging power of each individual electric vehicle at each time point. and state of charge .

3. The method for coordinated decision-making on electric vehicle resources and photovoltaic consumption in a dynamic electricity price zone according to claim 2, characterized in that, By applying constraints to the charging power and state of charge of each individual electric vehicle at each time point in the instant charging mode, the charging load of the electric vehicle in the instant charging mode is obtained by aggregation, including: Using the charging power and state of charge as inputs, the instant charging mode is set as follows: each electric vehicle is charged immediately at the current maximum available power after entering the network until the preset off-grid conditions are reached. The preset off-grid conditions include reaching the user-set expected state of charge or reaching the latest off-grid time. The following constraints are applied during the calculation process: State of charge equilibrium constraints: ； In the formula, For the first electric vehicles State of charge at any given moment; Improve electric vehicle charging efficiency; For the first Battery capacity of an electric vehicle; This refers to the charging time. Upper and lower limits of state of charge constraints: ； In the formula, , The first The minimum and maximum states of charge of an electric vehicle; Network entry and exit time constraints: ； In the formula, and The first The registration and deregistration times of electric vehicles; and The first The earliest registration time and the latest deregistration time of each electric vehicle; Based on the above constraints, the charging power of each electric vehicle at each time point is adjusted, and the charging load of all electric vehicles in the on-demand charging mode is obtained by summing up the data. ； in For electric vehicles in the instant charging mode Electric vehicle charging load at any given time.

4. The method for coordinated decision-making on electric vehicle resources and photovoltaic consumption in a dynamic electricity price zone according to claim 3, characterized in that, Based on the electric vehicle charging load under the instant charging mode and the pre-acquired photovoltaic power output forecast data of the substation, a matching value between photovoltaic power output and the total charging load is calculated using cosine similarity. A dynamic electricity price discount model is then established based on this matching value, outputting a dynamic time-of-use electricity price, including: The electric vehicle charging load under the instant charging mode With the obtained photovoltaic power generation forecast data of the substation area As input, cosine similarity is used to calculate the matching value between the two curves: ； In the formula, For the matching value, The cosine similarity between the two curves; Obtain peak-valley time-of-use electricity pricing from the power grid By adding electricity price discount percentages The dynamic time-of-use electricity price is obtained using the following formula: ； ； In the formula, The original peak-valley electricity price; , , These are their peak electricity price, normal electricity price, and off-peak electricity price, respectively. , , These are their corresponding time periods; Dynamic time-of-use pricing; The hyperbolic tangent function is used to describe the nonlinear mapping between the matching value deviation and the electricity price discount ratio, thus obtaining the electricity price discount ratio. The calculation formula is as follows: ； In the formula: This refers to the electricity price discount percentage; The baseline threshold for the matching value; Discount factor; This represents the rate of change in electricity price discounts.

5. The method for coordinated decision-making on electric vehicle resources and photovoltaic consumption in a dynamic electricity price zone according to claim 4, characterized in that, Based on the aforementioned dynamic time-of-use electricity pricing, combined with consumer psychology models and multi-attribute difference data of users, an electric vehicle interaction response model is constructed to output the response decision of the electric vehicle, including: Obtain the electricity price discount ratio in the dynamic time-of-use electricity price. ; Based on a consumer psychology model, a function for the average user response rate is constructed, taking the electricity price discount percentage as input, and calculating the average user response rate through a piecewise function. ； In the formula, Average user response rate; and These are the dead zone response rate and the saturation zone response rate, respectively. and These are the dead zone inflection point and the saturation zone inflection point, respectively. We obtained multi-attribute difference data on users, including economic income, age, and awareness of electricity pricing policies, and constructed economic income influencing factors for each. Age influencing factor Factors influencing awareness of electricity pricing policies The values ​​of each influencing factor are constrained to the interval (0,1); The product of the three attribute influence factors is normalized to obtain the comprehensive influence factor. : ； User calculation based on dynamic electricity price exist Base response rate at any given time: ； In the formula, For users exist Baseline response rate at any given time; For users exist Expected electricity price at any given time; The highest electricity price; By combining the baseline response rate with comprehensive influencing factors, an individual user response rate model is constructed: ； In the formula, For users exist Individual response rate at any given time; Based on the preset average user response rate and individual response rate Output the first result according to the following judgment rules. Response decision of an electric vehicle: ； In the formula, For the first Response decisions of electric vehicles; This indicates that the user is not participating; This indicates user participation.

6. The method for coordinated decision-making on electric vehicle resources and photovoltaic consumption in a dynamic electricity price zone according to claim 5, characterized in that, The objective function of the district collaborative decision-making model includes: With maximizing the total daily operating revenue of the charging stations in the area as the optimization objective, an objective function is constructed. Total revenue Including charging revenue from users Electricity purchase costs paid to the upstream main grid And the penalty losses due to insufficient solar power consumption : ； ； In the formula, Dynamic time-of-use pricing; for The total charging power of all electric vehicles at any given time; Peak-valley electricity pricing on the grid side; for The amount of electricity purchased from the power grid by the charging stations in the designated area at any given time; This refers to the amount of curtailed solar power generated by distributed photovoltaic (PV) systems, of which: ； In the formula, and They are respectively The charging power of electric vehicles that participate in and do not participate in the scheduling at all times; for Distributed photovoltaic power generation capacity at all times; Discarding loss factor: ； ； In the formula, The percentage of solar power curtailment; , , These represent low, moderate, and high wastelight loss factors, respectively. and These are low and high light curtailment ratios, respectively.

7. The method for coordinated decision-making on electric vehicle resources and photovoltaic consumption in a dynamic electricity price zone according to claim 6, characterized in that, The constraints of the collaborative decision-making model for power distribution areas include: Average charging time constraint: ； In the formula, The average charging time for electric vehicles; and These are the upper and lower limits of the average charging time. Expected state of charge constraints: ； In the formula, and The first The expected state of charge and maximum state of charge of an electric vehicle; For the first The expected charge coefficient of an electric vehicle; Extended offline time constraints: ； In the formula, For the first The off-grid time for electric vehicles participating in the dispatch was extended. For the first The extended off-grid time is allowed for electric vehicles participating in the dispatch; Charging time window constraint: ； In the formula, For the first vehicle The charging power of electric vehicles at all times; and The first The time when an electric vehicle enters or leaves the network; Electric power balance constraints: At any given time, the total charging demand within the distribution area must be met jointly by photovoltaic power supply and electricity purchased from the main power grid. ； In the formula, for The total charging power of all electric vehicles at any given time; for The charging power constantly comes from photovoltaics; for The amount of electricity purchased from the power grid by the charging stations in the designated area at any given time; Photovoltaic output constraints: The actual photovoltaic power used for charging cannot exceed its maximum output limit: ； In the formula, for Photovoltaic output power at any given time; This represents the maximum output power of the photovoltaic system. Grid-side interactive power constraints: The concentrated charging of a large number of electric vehicles may lead to a large peak-valley difference in the power grid, necessitating the limitation of the power exchange between charging stations and the upstream distribution network. ； In the formula, for The charging power constantly supplied by the power grid; This represents the maximum peak-to-valley difference allowed by the power grid.

8. The method for coordinated decision-making on electric vehicle resources and photovoltaic consumption in a dynamic electricity price zone according to claim 7, characterized in that, The solution to the transformer area collaborative decision-making model employs an improved particle swarm optimization algorithm, including: Initialize the population and set the population size to [value]. ,generate Each particle represents a charging power allocation scheme. The upper bound of the variable is dynamically determined according to the allowed charging time window. Power is allocated only during the allowed charging period, and the upper bound of the power is 0 during other periods. Perform a three-step constraint repair for each particle: Time window reset fix: Sets the power exceeding the allowed charging time window to 0; State of charge upper limit constraint correction: If the charging power causes the state of charge to exceed the limit, the power during that period will be reduced. Energy demand replenishment and repair: If the expected state of charge is not met after repair, the power will be increased to make up for the energy gap within the feasible period. The fitness value of each repaired particle is calculated based on the objective function. Introducing a linearly decreasing inertial weight strategy to update particle velocity and position: The inertia weight decreases linearly: ； In the formula, This represents the current iteration number. The maximum number of iterations, As the initial inertia weight, This refers to the inertia weight at the end of the iteration; Speed ​​updates: ； In the formula, , As a learning factor, , It is a random number. This represents the optimal position in the history of an individual particle. The optimal position for the entire group; Location update: ； After the update, the particle positions will be clipped to within the preset feasible region: ； When the maximum number of iterations is reached Or the change in the continuously preset algebraic global optimum is less than a threshold. When the iteration terminates, the global optimal solution is output. The global optimal solution is used as the optimal electric vehicle charging power allocation scheme and the corresponding photovoltaic absorption capacity.

9. A device for coordinated decision-making on electric vehicle resources and photovoltaic power consumption in a transformer substation, used to implement the coordinated decision-making method for electric vehicle resources and photovoltaic power consumption in a transformer substation as described in any one of claims 1-8, characterized in that, include: The acquisition module is used to acquire data on the driving patterns of electric vehicles, battery charging and discharging characteristics, and state of charge. The first processing module is used to input the driving pattern data, battery charging and discharging characteristic data and state of charge data into a pre-constructed fine load model of electric vehicles to obtain the charging power and state of charge of each individual electric vehicle at each time. The aggregation module is used to apply constraints to the charging power and state of charge of each individual electric vehicle at each time in the instant charging mode, and aggregate them to obtain the electric vehicle charging load in the instant charging mode. The second processing module is used to calculate the matching value between photovoltaic output and total charging load based on the electric vehicle charging load under the instant charging mode and the pre-acquired photovoltaic power output data of the transformer area through cosine similarity, establish a dynamic electricity price discount model based on the matching value, and output dynamic time-of-use electricity price. The third processing module is used to construct an electric vehicle interactive response model based on the dynamic time-of-use electricity price, combined with the consumer psychology model and the user's multi-attribute difference data, and output the electric vehicle's response decision. The decision-making module is used to solve, under preset constraints, the optimal electric vehicle charging strategy and photovoltaic absorption scheme based on the response decision of the electric vehicle and a pre-built district collaborative decision-making model with the objectives of maximizing charging station revenue and maximizing photovoltaic absorption.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When executed by a processor, the computer program implements the steps of the method according to any one of claims 1-8.

Images