An electric vehicle charging situation awareness and autonomous regulation method based on multi-objective fuzzy evaluation

Abstract

The application relates to the field of electric vehicle charging scheduling, and discloses an electric vehicle charging situation awareness and autonomous regulation method based on multi-target fuzzy evaluation. Electric vehicle, parking lot and user data are collected to identify the current charging scene type; a fuzzy analytic hierarchy process (FAHP) is used to construct a hierarchical structure of user charging urgency, dynamically assign weights to the criterion layer, and calculate the real-time charging urgency coefficient of each electric vehicle by using fuzzy comprehensive evaluation (FCE); the charging urgency coefficient is used as the satisfaction weight to construct a user satisfaction and power grid load fluctuation optimization model and solve the model; model predictive control (MPC) is used to roll over the above steps in a preset time period, and a correction mechanism is introduced to update the user credit level according to the execution deviation of the last period, which is used for urgency evaluation in the next period.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle charging scheduling, and specifically to a method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation. Background Technology

[0002] With the deepening of the global energy transition and the rapid growth of electric vehicle ownership, large-scale, disorderly access to the power grid for charging of electric vehicles is occurring. In typical scenarios such as residential communities, commercial complexes, and office buildings, the peak hours of electric vehicle charging loads overlap with the peak hours of basic electricity consumption, leading to increasingly prominent problems such as increased risk of distribution transformer overload and exacerbated peak-valley differences in the power grid. To balance grid security constraints and user charging experience, existing charging scheduling methods assess charging urgency based on battery state of charge (SOC) or charging duration, and then prioritize charging. However, these methods neglect urgent vehicles with low to medium (not minimum) SOC but shorter dwell times, resulting in insufficient dimensions for evaluating charging urgency and a high risk of resource misallocation.

[0003] Existing technologies employ multi-objective optimization algorithms to balance user experience and grid safety in charging systems. However, on the user side, only "fully charged or not" is used as a Boolean constraint to establish the objective function. Achieving true Pareto optimality at the user level is therefore difficult. Summary of the Invention

[0004] Purpose of the Invention: To address the problems mentioned in the background art, this invention discloses a method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation. It constructs and weights a hierarchical structure of user charging urgency using fuzzy hierarchical analysis (FAHP), optimizes the charging priority structure, completes the user urgency dimension, and further optimizes it. MPC introduces a correction mechanism to maintain the dynamic robustness of the charging system, enabling electric vehicle charging situational awareness and autonomous control.

[0005] Technical solution:

[0006] This invention discloses a method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation, the method comprising the following steps:

[0007] S1: Collect data on electric vehicles, parking lots, and users to identify the current charging scenario type;

[0008] S2: Based on the data in S1, a hierarchical structure of user charging urgency is constructed using the fuzzy hierarchical analysis method (FAHP). This structure is dynamically weighted for the criterion layer. The real-time charging urgency coefficient for each electric vehicle is calculated using the fuzzy comprehensive evaluation (FCE).

[0009] S3: Construct and solve an optimization model of user satisfaction and power grid load fluctuation using the charging urgency coefficient as the satisfaction weight;

[0010] S4: Model predictive control (MPC) is used to execute S1 to S3 in a rolling manner with a preset time period, and a correction mechanism is introduced to update the user's credit rating based on the execution deviation of the previous period for the urgency evaluation of the next period.

[0011] Furthermore, the data collected by S1 includes vehicle status data of electric vehicles, power distribution network load data of the parking lot, and user behavior data. The vehicle status data and user behavior data include the battery state of charge (SOC), battery health status (SOH), and user's preset expected departure time for each electric vehicle. The power distribution network load data includes the real-time load rate of the distribution transformer where the charging pile is located and the regional basic load curve. Through the parking lot gate system, the current charging scenario is identified as a residential community, commercial complex, or office building, and the corresponding scenario feature library is retrieved. The collected multidimensional data is cleaned and normalized to form a standardized input dataset.

[0012] Furthermore, the hierarchical structure of user charging urgency described in S2 is constructed as follows:

[0013] The system uses user charging urgency as the target layer, battery energy state C1, time urgency C2, user credit rating C3, and grid friendliness C4 as the criteria layer, and each electric vehicle to be dispatched as the solution layer.

[0014] Furthermore, based on the aforementioned fuzzy hierarchical analysis method (FAHP), the criterion indicators are dynamically weighted as follows:

[0015] Based on the current charging scenario type, initialize the fuzzy complementary judgment matrix. ,in The importance of criterion i relative to criterion j is expressed using a scale of 0.1-0.9, and satisfies the following conditions: ; Calculate the weight vector for each criterion : Construct membership functions for battery energy state C1, time urgency C2, user credit rating C3, and grid friendliness C4 respectively.

[0016] Furthermore, based on fuzzy comprehensive evaluation (FCE), the weight vector W of each criterion and the membership value of each criterion are fuzzy synthesized to obtain the comprehensive evaluation value for each vehicle. This is the vehicle's charging urgency coefficient:

[0017]

[0018] in, Let the weight of the j-th criterion be... Let i be the membership value of vehicle i under the j-th criterion, based on the charging urgency coefficient of each vehicle. Perform priority sorting. A higher value indicates a greater urgency for charging, and thus a higher priority is given in subsequent scheduling.

[0019] Furthermore, the user satisfaction and power grid load fluctuation optimization model described in S3, under the conditions of satisfying the distribution network capacity constraints and user travel needs, obtains the charging power curve for each electric vehicle. The objective function for maximizing user satisfaction is as follows:

[0020]

[0021] Where N is the total number of electric vehicles connected to the charging system; The charging urgency coefficient for vehicle i; The initial state of charge when vehicle i arrives; The user's desired state of charge upon leaving the site; The state of charge of vehicle i when it actually leaves the site is determined by the decision variables. The result is obtained by integration;

[0022] The objective function for minimizing grid load fluctuations is as follows:

[0023]

[0024] in, The total number of time periods in the scheduling cycle is set to T=96, corresponding to a time period of 15 minutes. The total charging load of all electric vehicles during time period t; The regional base load is for time period t.

[0025] Furthermore, the user satisfaction and power grid load fluctuation optimization model constraints include distribution transformer capacity constraints, user off-site SOC constraints, and charging pile power limit constraints. The model is solved using the non-dominated sorting genetic algorithm NSGA-II to obtain the Pareto optimal solution set. From this set, the compromise solution with the highest satisfaction is selected as the final scheduling scheme, and the charging power instructions for each electric vehicle in each time period within the scheduling cycle are output. .

[0026] Furthermore, the correction mechanism described in S4 is as follows:

[0027] Rolling optimization uses a 15-minute rolling cycle. At the beginning of each cycle, S1 to S3 are re-executed to update the charging power curve for the next 24 hours.

[0028] Compare the actual results with the planned values ​​from the previous cycle to calculate the charging deviation for each vehicle. and departure time deviation Based on the execution deviation, the user credit rating C3 in the criteria layer is dynamically updated.

[0029]

[0030] in, For the current credit score, The updated credit score; , The penalty coefficient is, and ; For the planned charging amount, The planned stay is in hours; the updated credit rating will be used for the next FAHP evaluation cycle.

[0031] Beneficial effects:

[0032] 1. This invention identifies charging scenario types, initializes different FAHP fuzzy complementary judgment matrices for different scenarios, constructs a hierarchical structure of user charging urgency, dynamically assigns weights to the criterion layer, making the allocation of charging resources more in line with the real needs of users in different scenarios, reflecting the real charging urgency, and avoiding resource misallocation.

[0033] 2. This invention uses the charging urgency coefficient calculated by FAHP-FCE. As a weighting coefficient of the user satisfaction objective function in multi-objective optimization, it deeply integrates the user's continuous urgency information into the optimization model. When the power distribution capacity is limited, it can prioritize the charging needs of vehicles with the highest charging demand, while also minimizing the fluctuation of the power grid load, thus achieving Pareto optimal coordination between user experience and power grid safety.

[0034] 3. This invention employs Model Predictive Control (MPC) for periodic rolling execution optimization scheduling and introduces a credit correction mechanism based on execution deviation. This effectively addresses scheduling deviations caused by user behavior uncertainty, guides users to adjust their departure time and pick up their vehicles on time, further improves scheduling accuracy, and enhances the robustness and adaptability of the charging scheduling system. Attached Figure Description

[0035] Figure 1 This is a flowchart of the method of the present invention;

[0036] Figure 2 This is a comparison chart of the charging urgency coefficient distribution under different scenarios in embodiments of the present invention.

[0037] Figure 3 This is a comparison chart of the power grid load curves before and after dispatching according to an embodiment of the present invention;

[0038] Figure 4 This is a comparison chart of user satisfaction with different charging scheduling methods in embodiments of the present invention;

[0039] Figure 5 This is a graph showing the trend of the rolling optimization cycle and execution deviation in an embodiment of the present invention; Detailed Implementation

[0040] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0041] like Figure 1 As shown, this invention discloses a method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation. The method steps are as follows:

[0042] Step 1: Collect vehicle status data of electric vehicles, power distribution network load data of parking lots, and user behavior data, and identify the current charging scenario type;

[0043] Step 1.1: Collect the battery state of charge (SOC), battery state of health (SOH), and the expected departure time set by the user via the APP for each electric vehicle;

[0044] Step 1.2: Collect the real-time load rate and regional base load curve of the distribution transformer where the charging pile is located;

[0045] Step 1.3: Through the parking lot gate system, identify the current charging scenario as one of residential community, commercial complex, or office building, and retrieve the corresponding scenario feature database;

[0046] Step 1.4: Clean and normalize the collected multidimensional data to form a standardized input dataset.

[0047] Step 2: Based on the data collected in Step 1, the hierarchical structure of user charging urgency is constructed using the fuzzy hierarchical analysis method (FAHP), the dynamic weights of each evaluation index in the criterion layer are calculated, and the real-time charging urgency coefficient of each electric vehicle is calculated using the fuzzy comprehensive evaluation (FCE).

[0048] Step 2.1: Construct a hierarchical structure, with user charging urgency as the target layer, battery energy state C1, time urgency C2, user credit rating C3, and grid friendliness C4 as the criteria layer, and each electric vehicle to be dispatched as the scheme layer;

[0049] Step 2.2: FAHP dynamic weighting. Based on the scene type identified in Step 1.3, initialize the fuzzy complementary judgment matrix. ,in The importance of criterion i relative to criterion j is expressed using a scale of 0.1-0.9, and satisfies the following conditions: ; Calculate the weight vector for each criterion :

[0050]

[0051] in, To determine the element in the j-th row and k-th column of a matrix, The sum of all elements in column k; perform a consistency check, using the consistency index. ,in To determine the largest eigenvalue of a matrix, when the random consistency ratio... When the judgment matrix is ​​considered to have satisfactory consistency, RI is the average random consistency index.

[0052] Step 2.3: Determine the membership function

[0053] Construct linear membership functions for each criterion:

[0054] For battery state of energy criterion C1:

[0055]

[0056] in, This is the current SOC value;

[0057] For time urgency criterion C2:

[0058]

[0059] Where t is the user's expected stay time. The time required to fully charge the vehicle from its current SOC. To determine the maximum allowable dwell time threshold, take... Hour;

[0060] For user credit rating criteria C3:

[0061]

[0062] in, This represents the user's credit score, with a value range of [0, 100].

[0063] For grid friendliness criterion C4:

[0064]

[0065] in, Real-time load rate of the distribution transformer;

[0066] Step 2.4: Fuzzy Comprehensive Evaluation (FCE)

[0067] The weight vector W of each criterion is combined with the membership value of each criterion using fuzzy synthesis to obtain the comprehensive evaluation value for each vehicle. This is the vehicle's charging urgency coefficient:

[0068]

[0069] in, Let the weight of the j-th criterion be... Let be the membership value of vehicle i under the j-th criterion.

[0070] Step 2.5: Based on the charging urgency coefficient of each vehicle Perform priority sorting. A higher value indicates a greater urgency for charging, and thus a higher priority is given in subsequent scheduling.

[0071] Step 3: Using the charging urgency coefficient obtained in Step 2 As a weight for user satisfaction, the non-dominated sorting genetic algorithm NSGA-II is used to construct a multi-objective optimization model with the goal of maximizing user satisfaction and minimizing grid load fluctuations. Under the conditions of satisfying the distribution network capacity constraints and user travel needs, the charging power curve of each electric vehicle is obtained.

[0072] Step 3.1: Construct the objective function

[0073] The dual optimization objectives are to maximize user satisfaction (f1) and minimize grid load fluctuations (f2).

[0074] Objective function for maximizing user satisfaction:

[0075]

[0076] Where N is the total number of electric vehicles connected to the charging system; The charging urgency coefficient for vehicle i; The initial state of charge when vehicle i arrives; The user's desired state of charge upon leaving the site; The state of charge of vehicle i when it actually leaves the site is determined by the decision variables. The result is obtained by integration;

[0077] The objective function for minimizing grid load fluctuations is:

[0078] in, The total number of time periods in the scheduling cycle is set to T=96, corresponding to a time period of 15 minutes. The total charging load of all electric vehicles during time period t; The regional base load for time period t;

[0079] Step 3.2: Set constraints

[0080] (1) Distribution transformer capacity constraints:

[0081]

[0082] in, Rated capacity of distribution transformer;

[0083] (2) User departure SOC constraint:

[0084]

[0085] (3) Power limitation constraints of charging piles:

[0086]

[0087] in, The charging power of vehicle i during time period t. The maximum charging power allowed for vehicle i;

[0088] Step 3.3: The non-dominated sorting genetic algorithm NSGA-II is used to solve the bi-objective optimization model to obtain the Pareto optimal solution set.

[0089] Step 3.4: Select the compromise solution with the highest satisfaction as the final scheduling scheme, and output the charging power command for each electric vehicle in each time period within the scheduling cycle. .

[0090] Step 4: Model Predictive Control (MPC) is used to execute steps 1 to 3 in a rolling manner at a preset time period, and a correction mechanism is introduced to update the user's credit rating based on the execution deviation of the previous period for the urgency evaluation of the next period.

[0091] Step 4.1: Rolling optimization, with a rolling cycle of 15 minutes, at the beginning of each cycle, re-execute steps 1 to 3 to update the charging power curve for the next 24 hours;

[0092] Step 4.2: Compare the actual execution results with the planned values ​​from the previous cycle, and calculate the charging deviation for each vehicle. and departure time deviation :

[0093]

[0094] in, The actual charging amount of vehicle i in the previous cycle, in kWh; The planned charging amount for vehicle i in the previous cycle, in kWh;

[0095]

[0096] in, This represents the actual departure time of vehicle i. The user's expected departure time;

[0097] Step 4.3: Dynamically update the user's credit rating C3 based on the execution deviation.

[0098]

[0099] in, For the current credit score, The updated credit score; , The penalty coefficient is, and ; The planned stay is in hours; the updated credit rating will be used for the next FAHP evaluation cycle.

[0100] This embodiment takes a typical underground parking lot of a commercial complex in a city as an example. The parking lot is equipped with 200 AC charging piles, and the rated capacity of the distribution transformer is 1200kW. Simulation verification was carried out during the period from 11:00 to 15:00 on a certain weekday, which is the peak period for charging demand in the shopping mall.

[0101] Step 1: Multi-dimensional Situational Awareness and Data Acquisition

[0102] At t=11:00, the system collects status data from the 30 electric vehicles currently connected to the charging system, including the battery state of charge (SOC), battery health status (SOH), and the expected departure time set by the user via the app for each vehicle. Simultaneously, it collects the real-time load rate of the distribution transformer (l=65%) and the regional base load curve. The parking lot gate system identifies the current charging scenario as a "commercial complex".

[0103] The collected multidimensional data is normalized to form a standardized input dataset.

[0104] Step 2: Fuzzy evaluation of user charging urgency based on FAHP

[0105] A hierarchical structure is constructed, with user charging urgency as the target layer, battery energy state C1, time urgency C2, user credit rating C3, and grid friendliness C4 as the criteria layer, and each electric vehicle to be dispatched as the scheme layer.

[0106] FAHP Dynamic Assignment

[0107] For the commercial complex scenario, initialize the fuzzy complementary judgment matrix:

[0108]

[0109] Calculate the weights of each criterion:

[0110]

[0111] The resulting weight vector is W = [0.267, 0.420, 0.173, 0.140]. The time urgency factor C2 has the highest weight, consistent with the short user dwell time characteristic of shopping mall scenarios. The calculated CR = 0.045 < 0.1, passing the consistency test.

[0112] Determine the membership function

[0113] Taking the time urgency criterion C2 as an example, its membership function is defined as:

[0114]

[0115] Where t is the user's expected stay time. To fill the required time, Use 24 hours. The membership functions for other criteria are constructed using a similar linear form.

[0116] Step 2.4: Fuzzy Comprehensive Evaluation (FCE)

[0117] The weight vector W of each criterion is combined with the membership value of each criterion using a fuzzy synthesis operation to obtain the comprehensive evaluation value of each vehicle, which is the charging urgency coefficient of that vehicle. Taking some vehicles in this embodiment as examples, the specific calculation results are shown in Table 1: Evaluation results of charging urgency for some vehicles.

[0118] Table 1

[0119] Vehicle number SOC Expected departure time credit score Calculated EV01 25% 13:30 85 0.91 EV02 65% 14:00 92 0.72 EV03 45% 16:00 78 0.77

[0120] EV01 has the highest urgency due to its low SOC and short time to exit; EV02 has a high credit score but sufficient SOC, so its urgency is relatively low.

[0121] like Figure 2 As shown, this displays the vehicle's charging urgency coefficient. The distribution differences are evident in the underground parking lot of a commercial complex, where the weight of time urgency C2 is the highest, resulting in a significantly higher overall urgency coefficient distribution center in this scenario compared to other scenarios. This demonstrates that the present invention can automatically adjust the evaluation logic based on the feature library of different scenarios, ensuring optimal resource allocation in different social functional areas and avoiding resource misallocation.

[0122] Step 3: Autonomous Regulation Decision Based on Multi-Objective Optimization

[0123] Construct the objective function:

[0124] The goal of maximizing user satisfaction:

[0125]

[0126] The goal of minimizing power grid load fluctuations:

[0127] The scheduling cycle is 4 hours, from 11:00 to 15:00, with 15-minute intervals, and T=16.

[0128] Set constraints:

[0129] Distribution transformer capacity constraints:

[0130] User departure SOC constraint:

[0131] Charging pile power limitation constraints:

[0132] The NSGA-II algorithm was used to solve the above bi-objective optimization model with a population size of 100 and 300 iterations, yielding the Pareto optimal solution set. A compromise solution was selected as the final scheduling scheme, and charging power commands for each time period were output. Figure 3 As shown, by using the non-dominated sorting genetic algorithm NSGA-II, the system successfully guided the charging load to periods with lower base loads, reducing the peak-to-valley difference of the power grid by 23.5%. The total load curve after scheduling consistently operates smoothly below the 85% safe load threshold of the distribution transformer, further mitigating the risk of overload during peak hours.

[0133] Taking EV01 as an example, its charging power commands during certain periods are shown in Table 2:

[0134] Table 2

[0135] Time period 11:00-11:15 11:15-11:30 11:30-11:45 11:45-12:00 Power (kW) 60 60 60 60

[0136] The EV01 can be fully charged before leaving the site at 13:30, meeting user needs. For example... Figure 4 As shown, the satisfaction level of the method of the present invention compared with that of traditional disordered charging and traditional fixed-weight ordered charging in a commercial complex scenario is presented. The user satisfaction level of the method of the present invention reaches 0.83, which is about 15% higher than that of disordered charging and about 8% higher than that of traditional ordered charging.

[0137] Step four: Rolling time-domain control and feedback correction. For example... Figure 5As shown, after 3-5 scheduling cycles of iterative optimization, the average user departure time deviation decreased from the initial 23 minutes to 8 minutes; the charging quantity execution deviation rate also decreased from the initial 15% to about 6%. This demonstrates that the introduction of a credit correction mechanism can effectively suppress disturbances caused by user behavior uncertainty, and significantly enhances the scheduling accuracy and robustness of the system through feedback loop.

[0138] A new round of rolling optimization was initiated at 12:00. At this time, EV01 had been charged to 45kWh, with its SOC rising to 85%, and was expected to be fully charged before leaving the site at 13:30. Newly connected vehicles EV04 and EV05 were added to the optimization.

[0139] Calculate the execution deviation of the previous cycle. Taking EV02 as an example, its planned charging amount Q_plan = 35kWh, actual charging amount Q_actual = 32kWh, and departure time deviation Δt = 0, then:

[0140]

[0141] Update user credit rating:

[0142]

[0143] A slight decrease in credit score will affect the urgency assessment for the next cycle.

[0144] Simulation results show that, compared to traditional disordered charging methods, the method of this invention improves user satisfaction by approximately 10%-12%, reduces the peak-to-valley load difference in the power grid by 20%-25%, and consistently maintains the distribution transformer load rate within the safe threshold of 85%. Compared to traditional ordered charging methods with fixed weights, this invention employs scenario-adaptive FAHP dynamic weighting. In high-urgency scenarios such as commercial complexes, the average charging wait time for time-sensitive users is reduced by 35%; in low-urgency scenarios such as residential areas, power grid load fluctuations are reduced by 18%.

[0145] Meanwhile, based on the closed-loop feedback mechanism of MPC rolling time domain and credit correction, after 3-5 scheduling cycles of iterative optimization, the average user departure time deviation decreased from the initial 23 minutes to 8 minutes, the charging quantity execution deviation rate decreased from 15% to 6%, and the system scheduling accuracy gradually improved with the increase of the operating cycle, effectively guiding users to truthfully fill in their departure time and pick up their vehicles on time. This invention, through the deep integration of multi-objective fuzzy evaluation and autonomous control, provides an optimal decision-making scheme for electric vehicle charging scheduling that takes into account both user travel needs and grid safety constraints.

[0146] The foregoing description of the embodiments enables those skilled in the art to make or use the present invention. Various modifications to the embodiments will be readily apparent to those skilled in the art. The general principles of the invention may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention should not be limited to the embodiments shown herein, but should cover the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation, characterized in that, The method includes the following steps: S1: Collect data on electric vehicles, parking lots, and users to identify the current charging scenario type; S2: Based on the data in S1, a hierarchical structure of user charging urgency is constructed using the fuzzy hierarchical analysis method (FAHP). This structure is dynamically weighted for the criterion layer. The real-time charging urgency coefficient for each electric vehicle is calculated using the fuzzy comprehensive evaluation (FCE). S3: Construct and solve an optimization model of user satisfaction and power grid load fluctuation using the charging urgency coefficient as the satisfaction weight; S4: Model predictive control (MPC) is used to execute S1 to S3 in a rolling manner with a preset time period, and a correction mechanism is introduced to update the user's credit rating based on the execution deviation of the previous period for the urgency evaluation of the next period.

2. The method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation according to claim 1, characterized in that, The S1 data collection includes vehicle status data of electric vehicles, power distribution network load data of the parking lot, and user behavior data. The vehicle status data and user behavior data include the battery state of charge (SOC), battery health status (SOH), and the user's preset expected departure time for each electric vehicle. The power distribution network load data includes the real-time load rate of the distribution transformer where the charging pile is located and the regional basic load curve. Through the parking lot gate system, the current charging scenario is identified as a residential community, commercial complex, or office building, and the corresponding scenario feature library is retrieved. The collected multidimensional data is cleaned and normalized to form a standardized input dataset.

3. The method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation according to claim 1, characterized in that, The hierarchical structure of user charging urgency described in S2 is constructed as follows: The system uses user charging urgency as the target layer, battery energy state C1, time urgency C2, user credit rating C3, and grid friendliness C4 as the criteria layer, and each electric vehicle to be dispatched as the solution layer.

4. The method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation according to claim 3, characterized in that, Based on the aforementioned fuzzy hierarchical analysis method (FAHP), dynamic weighting of the criterion indicators is applied as follows: Based on the current charging scenario type, initialize the fuzzy complementary judgment matrix. ,in The importance of criterion i relative to criterion j is expressed using a scale of 0.1-0.9, and satisfies the following conditions: ; Calculate the weight vector for each criterion : Construct membership functions for battery energy state C1, time urgency C2, user credit rating C3, and grid friendliness C4 respectively.

5. The method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation according to claim 4, characterized in that, Based on fuzzy comprehensive evaluation (FCE), the weight vector W of each criterion and the membership value of each criterion are fuzzy synthesized to obtain the comprehensive evaluation value for each vehicle. This is the vehicle's charging urgency coefficient: ； in, Let the weight of the j-th criterion be... Let i be the membership value of vehicle i under the j-th criterion, based on the charging urgency coefficient of each vehicle. Perform priority sorting. A higher value indicates a greater urgency for charging, and thus a higher priority is given in subsequent scheduling.

6. The method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation according to claim 1, characterized in that, The user satisfaction and power grid load fluctuation optimization model described in S3, under the conditions of satisfying the distribution network capacity constraints and user travel needs, obtains the charging power curve for each electric vehicle. The objective function for maximizing user satisfaction is as follows: ； Where N is the total number of electric vehicles connected to the charging system; The charging urgency coefficient for vehicle i; The initial state of charge when vehicle i arrives; The user's desired state of charge upon leaving the field; The state of charge of vehicle i when it actually leaves the site is determined by the decision variables. The result is obtained by integration; The objective function for minimizing grid load fluctuations is as follows: ； in, The total number of time periods in the scheduling cycle is set to T=96, corresponding to a time period of 15 minutes. The total charging load of all electric vehicles during time period t; The regional base load is for time period t.

7. The method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation according to claim 6, characterized in that, The user satisfaction and power grid load fluctuation optimization model constraints include distribution transformer capacity constraints, user off-site SOC constraints, and charging pile power limit constraints. The model is solved using the non-dominated sorting genetic algorithm NSGA-II to obtain the Pareto optimal solution set. The compromise solution with the highest satisfaction is selected as the final scheduling scheme, and the charging power instructions for each electric vehicle in each time period within the scheduling cycle are output. .

8. The method for electric vehicle charging situational awareness and autonomous control based on multi-objective fuzzy evaluation according to claim 1, characterized in that, The correction mechanism described in S4 is as follows: Rolling optimization uses a 15-minute rolling cycle. At the beginning of each cycle, S1 to S3 are re-executed to update the charging power curve for the next 24 hours. Compare the actual results with the planned values ​​from the previous cycle to calculate the charging deviation for each vehicle. and departure time deviation Based on the execution deviation, the user credit rating C3 in the criteria layer is dynamically updated. ； in, For the current credit score, The updated credit score; , It is the penalty coefficient, and ; For the planned charging amount, The planned stay is in hours; the updated credit rating will be used for the next FAHP evaluation cycle.

Images