Electric vehicle ordered charging optimization method based on three-phase imbalance and deviation treatment

Abstract

The invention discloses an electric vehicle ordered charging optimization method based on three-phase imbalance and deviation treatment. The method comprises the following steps: constructing a two-dimensional excitation model based on charging excitation and charging duration; segmenting 24 hours of a day, and introducing a time-phased coefficient to correct the two-dimensional excitation model to form a time-phased user response rate model; obtaining the current three-phase unbalance degree and the voltage deviation of the power distribution network before and after implementation of each candidate charging scheme, and calculating the excitation income in combination with an operator excitation mechanism to obtain an operator net profit index; and according to the operator net profit index, with the operator net profit maximization as the target, a multi-constraint ordered charging optimization model is established by integrating the user response rate and the power quality excitation profit, and a genetic algorithm and a simulated annealing algorithm are adopted for joint solving to obtain an optimal charging period and a power distribution scheme. According to the invention, the participation enthusiasm of the user and the operator can be improved, the three-phase unbalance degree and the voltage deviation are obviously reduced, and the electric energy quality and economy are improved.

Description

Technical Field

[0001] This invention belongs to the field of electric vehicle charging management and distribution network operation optimization technology, and relates to electric vehicle orderly charging optimization technology, specifically to an electric vehicle orderly charging optimization method based on three-phase imbalance and deviation management. Background Technology

[0002] In recent years, the new energy vehicle industry has experienced rapid development, with production, sales, and ownership of new energy vehicles all showing a high-speed growth trend. However, the large-scale connection of EV loads to the power distribution network, especially the slow charging method via single-phase lines to low-voltage distribution networks, poses a severe challenge to the safe and stable operation of the power system. If these charging loads are not controlled, when their charging time coincides with the peak period of the distribution network's base load, it will cause a "peak-on-peak" phenomenon, leading to insufficient capacity or even damage to power distribution equipment. In addition, EVs can also cause power quality problems such as voltage deviation and three-phase imbalance during charging, increasing grid losses and shortening the lifespan of electrical equipment. Therefore, it is urgent to study orderly charging strategies that can effectively reduce the impact of EV charging on the distribution network, in order to improve the grid's EV acceptance capacity and ensure the safe and stable operation of the grid.

[0003] Currently, research on orderly charging strategies mainly focuses on optimizing peak shaving and valley filling, smoothing load fluctuations, reducing network losses, and lowering user charging costs. However, the large-scale access of EVs for disorderly charging has many impacts on the distribution network. In addition to causing a surge in grid load and resulting in "peak-on-peak" phenomena, it also causes power quality issues such as three-phase imbalance and voltage deviation.

[0004] Therefore, a new technological solution is needed to address these issues. Summary of the Invention

[0005] Purpose of the invention: To address the problems of insufficient user participation behavior modeling, inadequate consideration of power quality factors, lack of efficient optimization algorithms, and lack of real-time execution and dynamic adjustment mechanisms in existing orderly charging strategies, this invention provides an orderly charging optimization method for electric vehicles that considers the management of three-phase imbalance and voltage deviation. This method is simple to implement, low in cost, and fast in response. It can improve the power quality of the distribution network and the economic benefits of operators while meeting the travel needs of users.

[0006] Technical Solution: To achieve the above objectives, this invention provides an optimized method for orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation, comprising the following steps:

[0007] S1: Construct a two-dimensional excitation model based on charging excitation and charging duration;

[0008] S2: Divide the 24 hours of a day into segments and introduce a time-segment coefficient γ to correct the two-dimensional incentive model, forming a time-segmented user response rate model;

[0009] S3: Obtain the three-phase current imbalance and voltage deviation of the distribution network before and after the implementation of each candidate charging scheme, calculate the incentive revenue in combination with the operator's incentive mechanism, and obtain the operator's net profit index.

[0010] S4: Based on the operator's net profit index, with the goal of maximizing the operator's net profit, and taking into account user response rate and power quality incentive benefits, a multi-constraint ordered charging optimization model is established. The model is then solved by combining genetic algorithm and simulated annealing algorithm to obtain the optimal charging time period and power allocation scheme.

[0011] Furthermore, the two-dimensional excitation model in step S1 is expressed as follows:

[0012]

[0013] Where: α and β are the influencing factors of charging incentive level and charging time, respectively, which are related to EV users’ consumption concepts and travel plans, and can be obtained through historical data statistical analysis in practice; d represents charging incentive, m represents actual charging time, m0 represents planned charging time, and Ψ is the ratio of charging incentive to charging time, which can be adjusted by charging operators based on their own revenue.

[0014] Furthermore, the user response rate model in step S2 is expressed as follows:

[0015]

[0016] Where γ is the time-segmentation coefficient.

[0017] Furthermore, the incentive reward in step S3 is calculated as follows:

[0018]

[0019] Where: κ1 is the three-phase unbalanced excitation coefficient, κ2 is the voltage deviation excitation coefficient, and ε I 0 ε I These represent the maximum three-phase current imbalance in the low-voltage output branch of the transformer before and after the ordered charging strategy, respectively, during the day. 0 and e represent the maximum deviations of the voltage amplitude of all nodes in the system from the rated voltage before and after the ordered charging strategy, respectively. The calculation method is as follows:

[0020]

[0021]

[0022] Where: t is a specific point in time within a day, usually a discrete scheduling time; T represents the set of all discrete time points of the day; n is the nth node number in the distribution network of the transformer area; A={1,…,n} is the set of all nodes in the transformer area.

[0023] Furthermore, the formula for maximizing the operator's net profit in step S4 is as follows:

[0024]

[0025] Where: π is the net profit of the charging operator, R is the EV charging service fee revenue, r is the power quality incentive, and C is the user's discount fee. The calculation method is as follows:

[0026]

[0027]

[0028]

[0029] Where: N is the total number of EVs charged at the charging station in one day; x i,t Let x be the charging state of the i-th EV at time t, when x i,t When x = 1, the EV is in a charging state. i,t When P = 0, the EV is in a non-charging state; i c1 is the rated charging power of the i-th EV; c1 is the charging service fee set by the charging operator; Δt is the time interval length. For the response rate function; C i ev d represents the discount amount incurred by the i-th EV due to accepting dispatch; d represents the discount given to the EV user by the charging operator; t represents the discount amount. i s The moment when the i-th EV begins charging; l i e The extended charging time for the i-th EV to accept the operator's orderly charging strategy.

[0030] Furthermore, the multi-constraint ordered charging optimization model in step S4 includes charging time extension constraints, user state of charge constraints, node voltage constraints, and power balance constraints.

[0031] Furthermore, in the multi-constraint ordered charging optimization model of step S4:

[0032] The constraint on extended charging time is expressed as follows:

[0033] While providing users with charging incentives (d), charging operators also require users to extend their total charging time to:

[0034]

[0035] Where m is the charging time required by the charging operator, and l i c The calculation method is as follows: This refers to the charging duration during which the EV does not accept scheduling or the charging incentive is 0.

[0036]

[0037] Among them, SOC i + and SOC i - These represent the initial and target states of charge of the EV, respectively; E i c For the battery capacity of the EV; P i The rated charging power for electric vehicles;

[0038] The user's state of charge constraint is expressed as follows:

[0039] Each electric vehicle's state of charge (SOC) should reach the target state of charge when leaving the charging station, subject to the following constraints:

[0040]

[0041] The node voltage constraint is expressed as:

[0042]

[0043] Among them: U min U max These are the minimum and maximum voltages specified by national standards, respectively.

[0044] The power balance constraint is expressed as:

[0045]

[0046] Where i and j represent the node numbers in the distribution network, and φ∈{a,b,c} represents the three-phase phases; and These represent the active power and reactive power injected into phase φ by node i, respectively; and These represent the active and reactive power of the load at node i in phase φ, respectively. and These are the voltage magnitudes at nodes i and j on phase φ, respectively; and Let φ be the voltage phase angles at nodes i and j respectively; The voltage phase angle difference between nodes i and j at phase φ; and These are the real and imaginary parts of the equivalent admittance between nodes i and j under phase φ, respectively. and φ represents the active and reactive power imbalance of node i on phase φ, respectively; m is the total number of nodes in the distribution network.

[0047] Furthermore, the process of jointly solving the problem using a genetic algorithm and a simulated annealing algorithm in step S4 includes:

[0048] A1: Input the total number of electric vehicles N, the charging excitation d, and the charging time m, and initialize the genetic generation counter g and the annealing counter n;

[0049] A2: Based on the scheduling problem, chromosome coding is performed, and the charging status of each electric vehicle at each scheduling time is represented as binary code "1" or "0", forming a chromosome structure with the number of electric vehicles and the number of scheduling time periods throughout the day as the matrix dimension;

[0050] A3: Set the initial temperature T0 and the final temperature T f And the cooling coefficient θ;

[0051] A4: Set the population size SIZE and the maximum number of iterations GEN, randomly generate SIZE charging schemes as the initial population, among which EVs willing to accept scheduling are within the allowed time window [t]. i s , t i s +l i e Within [t], the charging state is randomly generated; EVs that do not accept scheduling will have their charging status randomly generated within [t]. i s , t i s +l i c The charging status remains constant at 1 throughout the time period;

[0052] A5: Calculate the fitness value of each individual based on the operator's net profit objective function, and include electric vehicles that do not participate in orderly charging in the fitness calculation;

[0053] A6: Use the roulette wheel selection method to select parent and parent individuals, randomly set crossover points to perform gene crossover operations; under a given mutation probability, perform even-number bit inversion mutation on the charging state within the schedulable time period;

[0054] A7: Compare the fitness of the parent and offspring according to the Metropolis criterion. If the offspring is better than the parent or accepts a worse solution with a specified probability, the offspring is retained and enters the new population. The individual with the highest current fitness is always retained.

[0055] A8: Perform cooling operation based on initial temperature, final temperature, and cooling coefficient;

[0056] A9: Repeat steps A5 to A8 until the maximum number of iterations is reached or the fitness converges, and output the electric vehicle orderly charging scheduling scheme with the optimal fitness value.

[0057] Further, step A6 includes:

[0058] Based on the fitness value calculated in step A5, a parent generation charging scheduling scheme is selected from the current population using a selection mechanism based on fitness ratio.

[0059] After the parent generation selection is completed, the selected parent generation scheduling scheme is subjected to gene recombination operation under a given crossover probability. By randomly setting the crossover position, the charging state of different electric vehicles at each scheduling time is recombined to generate a new offspring scheduling scheme. Subsequently, under a given mutation probability, the charging state variables within the schedulable time window are perturbed and updated, and the charging state in some time periods is reversed and adjusted.

[0060] This invention considers the impact of charging incentives and charging time on user response behavior, as well as the operator incentive benefits brought about by the improvement of three-phase imbalance and voltage deviation. It combines the advantages of genetic algorithms and simulated annealing algorithms to construct an efficient optimization model. Under the premise of meeting users' travel needs, it achieves coordinated optimization of economy and power quality, and has the ability to execute in real time and adjust dynamically.

[0061] Beneficial Effects: Compared with existing technologies, this invention improves traditional consumer psychology models by introducing responsiveness and responsiveness deviation, establishing a user response rate model that considers both charging incentives and charging time, thus achieving refined modeling of user participation behavior; it proposes a charging operator incentive mechanism that takes into account three-phase imbalance and voltage deviation improvement, quantifies the benefits of power quality governance and incorporates them into the optimization objective, achieving synergistic optimization of economic efficiency and power quality; it adopts a hybrid optimization method combining genetic algorithms and simulated annealing algorithms, which has strong global search capabilities and can utilize the Metropolis criterion to escape local optima, accelerating convergence speed and improving solution accuracy; it constructs an optimization model that includes multiple constraints such as voltage, current, user SOC, and scheduling time, ensuring the engineering feasibility of the optimization results and significantly improving the adaptability and stability of the scheduling scheme in actual operation. Attached Figure Description

[0062] Figure 1 This is a schematic diagram illustrating the implementation process of the method of the present invention;

[0063] Figure 2 This is a low-voltage power distribution network topology diagram for a certain region;

[0064] Figure 3 This is a diagram illustrating the three-phase current imbalance in the low-voltage outlet branch of a transformer.

[0065] Figure 4 This is a diagram of the voltage deviation at the end node under disordered charging.

[0066] Figure 5 This is a comparison chart of the convergence curves of different algorithms;

[0067] Figure 6 This is a diagram illustrating the three-phase current imbalance in the low-voltage outlet branch of the transformer before and after adopting the ordered charging strategy.

[0068] Figure 7 This is a diagram showing the voltage deviation of the end nodes before and after adopting the ordered charging strategy. Detailed Implementation

[0069] The present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. After reading this invention, any modifications of the invention in various equivalent forms by those skilled in the art will fall within the scope defined by the appended claims.

[0070] Example 1:

[0071] like Figure 1 As shown, this embodiment provides an optimized method for orderly charging of electric vehicles based on three-phase imbalance and deviation management, including the following steps:

[0072] S1: Construct a two-dimensional excitation model based on charging excitation and charging duration;

[0073] The two-dimensional excitation model is expressed as follows:

[0074]

[0075] Where: α and β are the influencing factors of charging incentive level and charging time, respectively, which are related to EV users’ consumption concepts and travel plans, and can be obtained through historical data statistical analysis in practice; d represents charging incentive, m represents actual charging time, m0 represents planned charging time, and Ψ is the ratio of charging incentive to charging time, which can be adjusted by charging operators based on their own revenue.

[0076] S2: Divide the 24 hours of a day into segments and introduce a time-segment coefficient γ to correct the two-dimensional incentive model, forming a time-segmented user response rate model;

[0077] The user response rate model is expressed as follows:

[0078]

[0079] Where γ is the time-segmentation coefficient.

[0080] In this embodiment, the time-segmentation coefficient γ is selected as follows:

[0081] During the period from 00:00 to 10:00, users are not sensitive to the charging completion time and have the strongest response to incentives. The time-segmentation coefficient γ is taken as 0.9.

[0082] During the period from 10:00 to 14:00, the demand for charging is high at noon, and users are unwilling to adjust their charging schedules. The time-segment coefficient γ is set to 0.3.

[0083] During the period from 14:00 to 19:00, the afternoon gradually enters the evening peak, user plans are stable, and the response rate is low. The time period coefficient γ is set to 0.6.

[0084] During the period from 19:00 to 00:00, the evening is the main peak charging period, and users are most sensitive to delayed charging. The time period coefficient γ is set to 0.2.

[0085] S3: Obtain the three-phase current imbalance and voltage deviation of the distribution network before and after the implementation of each candidate charging scheme, calculate the incentive revenue in combination with the operator's incentive mechanism, and obtain the operator's net profit index.

[0086] The power grid company determines the three-phase current imbalance ε before and after the charging operator implements the orderly charging strategy. I 0 ε I and voltage deviation degree e 0 The changes in e will be used to provide charging operators with corresponding incentive revenue r;

[0087] The incentive payout r is calculated as follows:

[0088]

[0089] Wherein: κ1 is the three-phase unbalance excitation coefficient, κ2 is the voltage deviation excitation coefficient, which is set by the power grid company according to power quality management requirements; ε I 0 ε I These represent the maximum three-phase current imbalance in the low-voltage output branch of the transformer before and after the ordered charging strategy, respectively, during the day. 0 e and e represent the maximum deviations of the voltage amplitude of all nodes in the system from the rated voltage before and after the ordered charging strategy.

[0090] Three-phase current imbalance ε I The calculation method for voltage deviation e is as follows:

[0091]

[0092]

[0093]

[0094]

[0095] Among them: I t a Let I be the effective value of phase A current in the low-voltage outlet branch of the transformer substation at time t. t b Let I be the effective value of phase B current in the low-voltage outlet branch of the transformer substation at time t. t c Let I be the effective value of the C-phase current in the low-voltage outlet branch of the transformer substation at time t; t avg U is the average three-phase current of the low-voltage outlet branch at time t; t is a specific point in time within a day, usually a discretized scheduling time; T represents the set of all discrete time points of the day; n is the nth node number in the distribution network of the transformer substation; A={1,…,n} is the set of all nodes in the transformer substation; U n , t Let n be the voltage at node n at time t; This is the node's rated voltage.

[0096] The above calculations reveal the effectiveness of each charging scheme in mitigating three-phase imbalance and voltage deviation, allowing for the determination of corresponding incentive benefits. This incentive cost will serve as a positive revenue term in the optimization objective function, encouraging charging operators to actively participate in load regulation to reduce power quality issues in the distribution network.

[0097] S4: Based on the operator's net profit index, with the goal of maximizing the operator's net profit, and taking into account user response rate and power quality incentive benefits, a multi-constraint ordered charging optimization model is established, which includes voltage, current, user SOC and scheduling time. The model is then solved by combining genetic algorithm and simulated annealing algorithm to obtain the optimal charging time period and power allocation scheme.

[0098] When establishing an optimization model for the orderly charging strategy of electric vehicles, the optimization objective is to maximize the net profit of charging operators. The net profit of charging operators can be calculated as the difference between their revenue and expenditure. Specifically, the net profit of charging operators... It can be calculated based on the difference between its revenue and expenses. Therefore, the formula for maximizing the operator's net profit is:

[0099]

[0100] Where: π is the net profit of the charging operator, R is the EV charging service fee revenue, r is the power quality incentive, and C is the discount fee paid by the charging operator to encourage users to participate in dispatching. The calculation method is as follows:

[0101]

[0102]

[0103]

[0104] Where: N is the total number of EVs charged at the charging station in one day; x i,t Let x be the charging state of the i-th EV at time t, when x i,t When x = 1, the EV is in a charging state. i,t When P = 0, the EV is in a non-charging state; i c1 is the rated charging power of the i-th EV; c1 is the charging service fee set by the charging operator; Δt is the time interval length. For the response rate function; C i ev d represents the discount amount incurred by the i-th EV due to accepting dispatch; d represents the charging incentive; t represents the discount amount. i s The moment when the i-th EV begins charging; l i e The extended charging time for the i-th EV to accept the operator's orderly charging strategy.

[0105] The multi-constraint ordered charging optimization model includes constraints on charging time extension, user state of charge, node voltage, and power balance, specifically:

[0106] The constraint on extended charging time is expressed as follows:

[0107] While providing users with charging incentives (d), charging operators also require users to extend their total charging time to:

[0108]

[0109] Where m represents the charging time required by the charging operator. In this embodiment, it is considered that the charging operator requires the EV charging time to be extended by a maximum of 1 time, i.e. ;l i c The calculation method is as follows: This refers to the charging duration during which the EV does not accept scheduling or the charging incentive is 0.

[0110]

[0111] Among them, SOC i + and SOC i - These represent the initial and target states of charge of the EV, respectively; E i cFor the battery capacity of the EV; P i The rated charging power for electric vehicles;

[0112] The user's state of charge constraint is expressed as follows:

[0113] Charging operators must schedule EV charging times while ensuring the travel needs of EV users are met; that is, the EV's State of Charge (SOC) should reach the target state of charge when the user leaves the charging station. The constraint is:

[0114]

[0115] The node voltage constraint is expressed as:

[0116]

[0117] Among them: U min U max These are the minimum and maximum voltages specified by national standards, which are set to 0.9 and 1.07 respectively in this embodiment.

[0118] The power balance constraint is expressed as:

[0119]

[0120] Where i and j represent the node numbers in the distribution network, and φ∈{a,b,c} represents the three-phase phases; and These represent the active power and reactive power injected into phase φ by node i, respectively; and These represent the active and reactive power of the load at node i in phase φ, respectively. and These are the voltage magnitudes at nodes i and j on phase φ, respectively; and Let φ be the voltage phase angles at nodes i and j respectively; The voltage phase angle difference between nodes i and j at phase φ; and These are the real and imaginary parts of the equivalent admittance between nodes i and j under phase φ, respectively. and φ represents the active and reactive power imbalance of node i on phase φ, respectively; m is the total number of nodes in the distribution network.

[0121] Under the constraints described above, an optimization method combining genetic algorithm and simulated annealing algorithm (GA-SA) is used to iteratively solve the optimization model and obtain the optimal charging period and power allocation scheme for each electric vehicle that meets the constraints.

[0122] An optimization method based on the fusion of genetic algorithm and simulated annealing algorithm (GA-SA) is used to iteratively solve the optimization model. The solution process includes:

[0123] A1: Input the total number of electric vehicles N, the charging excitation d, and the charging time m, and initialize the genetic generation counter g and the annealing counter n;

[0124] A2: Based on the scheduling problem, chromosome coding is performed, and the charging status of each electric vehicle at each scheduling time is represented as binary code "1" or "0", forming a chromosome structure with the number of electric vehicles and the number of scheduling time periods throughout the day as the matrix dimension;

[0125] A3: Set the initial temperature T0 and the final temperature T f And the cooling coefficient θ;

[0126] A4: Set the population size SIZE and the maximum number of iterations GEN, randomly generate SIZE charging schemes as the initial population, among which EVs willing to accept scheduling are within the allowed time window [t]. i s , t i s +l i e Within [t], the charging state is randomly generated; EVs that do not accept scheduling will have their charging status randomly generated within [t]. i s , t i s +l i c The charging status remains constant at 1 throughout the time period;

[0127] A5: Based on the charging operator's net profit objective function, evaluate each charging scheduling scheme in the current population and calculate the fitness value corresponding to each scheduling scheme; the fitness value is evaluated using the charging operator's net profit as the evaluation index, and its calculation method is as follows:

[0128]

[0129] Where F is the fitness value of the charging dispatch scheme, π is the net profit of the charging operator, R is the EV charging service fee revenue, r is the power quality incentive, and C is the discount fee paid by the charging operator to encourage users to participate in the dispatch. When calculating the fitness value, all electric vehicles participating in charging are uniformly included in the evaluation scope. Among them, electric vehicles that do not participate in the orderly charging dispatch are only included in the charging service fee revenue and do not generate incentive discount costs or power quality incentive revenue.

[0130] A6: Based on the fitness value calculated in step A5, a parent charging scheduler is selected from the current population using a fitness-based selection mechanism. The fitness value is used as the selection weight, prioritizing schedulers with higher net profits for subsequent evolutionary processes. Specifically, the probability of the k-th scheduler being selected as the parent is expressed as:

[0131]

[0132] In the formula, p k F represents the probability that the k-th individual is selected as the parent; k is the fitness value of the k-th individual; SIZE is the population size; For the first The fitness value corresponding to each scheduling scheme.

[0133] After the parent generation selection is completed, a gene recombination operation is performed on the selected parent generation scheduling scheme under a given crossover probability. By randomly setting the crossover position, the charging state of different electric vehicles at each scheduling time is recombined to generate a new offspring scheduling scheme. Subsequently, under a given mutation probability, the charging state variable within the schedulable time window is perturbed and updated. By reversing the charging state in some time periods, the diversity of the scheduling scheme is enhanced and local optima are avoided.

[0134] A7: Compare the fitness of the parent and offspring according to the Metropolis criterion. If the offspring is better than the parent or accepts a worse solution with a specified probability, the offspring is retained and enters the new population. The individual with the highest current fitness is always retained.

[0135] A8: Perform cooling operation based on initial temperature, final temperature, and cooling coefficient;

[0136] A9: Repeat steps A5 to A8 until the maximum number of iterations is reached or the fitness converges, and output the electric vehicle orderly charging scheduling scheme with the optimal fitness value.

[0137] This embodiment employs a hybrid optimization method (GA-SA) combining a genetic algorithm (GA) and a simulated annealing (SA) algorithm to solve the ordered charging strategy for electric vehicles. The genetic algorithm is a heuristic search algorithm that mimics natural selection and biological evolution. Through operations such as population initialization, selection, crossover, and mutation, it gradually evolves the population to a better solution space, possessing strong global search capabilities. However, traditional GA algorithms are prone to getting trapped in local optima and have relatively weak local search capabilities. The simulated annealing algorithm, by mimicking the solid-state annealing process, adopts a probabilistic search strategy. It allows for accepting poor new solutions at high initial temperatures and gradually reduces the probability of accepting poor solutions as the temperature decreases, thus possessing the ability to escape local optima. The genetic annealing algorithm (GA-SA), formed by combining the two algorithms, can balance the global search capability of GA and the local escape capability of SA, improving the convergence speed and global optimization performance.

[0138] This embodiment utilizes a multi-dimensional parameter set consisting of charging excitation level and charging duration to generate an initial population, providing a diverse optimization search space for the Genetic Annealing Algorithm (GA-SA), enabling the algorithm to fully explore the differences between different charging strategies. By setting the population size to 400, the maximum number of iterations to 500, the crossover rate to 0.8, the mutation rate to 0.05, the initial temperature to 100, the termination temperature to 0.1, and the cooling coefficient to 0.8, the algorithm achieves a good balance between global search and local escape capabilities, thereby effectively avoiding getting trapped in local optima.

[0139] Through iterative search and evaluation of a large number of scheme combinations, the GA-SA algorithm can quickly converge to the global optimum while ensuring the stability and feasibility of the solution, providing efficient optimization support for the orderly charging scheduling of electric vehicles.

[0140] In summary, this embodiment first acquires charging demand data from electric vehicle users and power distribution network operation status data. After data cleaning, feature filtering, and normalization, a user response rate feature set and a power quality feature set are constructed, including characteristic parameters such as charging incentive level, charging duration, node voltage, three-phase current, and voltage deviation rate. Subsequently, an optimization objective function considering charging operator profits, power quality incentive revenue, and user response rate is established, and multiple constraints such as charging time, SOC, voltage range, and node power balance are introduced to form a complete ordered charging strategy optimization model. During the optimization process, a genetic annealing algorithm (GA-SA) combining genetic algorithm and simulated annealing algorithm is used as the optimization tool. Through population initialization, fitness calculation, crossover mutation, and iterative search using annealing criteria, it quickly converges to the global optimum, obtaining the optimal charging power allocation scheme for each electric vehicle at different time periods.

[0141] Example 2:

[0142] To verify the effectiveness and effect of the method of the present invention, this embodiment applies the method of the present invention to a specific scenario, as follows:

[0143] 1) Constructing a simulated power distribution network model and charging load access methods

[0144] This embodiment uses, as follows: Figure 2 The diagram shows a 12-node three-phase four-wire low-voltage distribution network topology with four electric vehicle (EV) charging stations. All charging loads are randomly connected to the low-voltage distribution area via single-phase lines. To verify the effectiveness of the proposed strategy, the three-phase charging load data is proportionally distributed to the four charging stations. The charging service fee is assumed to be 1.44 yuan / kWh. The relevant parameters of the EV user response rate model and the charging operator incentive mechanism are shown in Table 1. The main parameters of the genetic annealing algorithm include: population size 400, maximum number of iterations 500, crossover rate 0.8, mutation rate 0.05, initial temperature 100, termination temperature 0.1, and cooling coefficient 0.8.

[0145] 2) Impact assessment of disordered charging on power quality of distribution network

[0146] (1) Three-phase imbalance

[0147] This embodiment adopts Figure 2 The 12-node three-phase four-wire low-voltage distribution network topology shown is used as a simulation model, including 4 electric vehicle charging stations and a total of 3000 electric vehicles. It is assumed that the single-phase access of electric vehicles is uniformly distributed. The three-phase charging load is evenly distributed across the 4 charging stations, and power flow calculations are performed over 96 scheduling periods (each period lasting 15 minutes). The charging service fee is set at 1.44 yuan / kWh, the charging power is set at 7kW (slow charging) and 20kW (fast charging), the charging efficiency is 0.9, and the electric vehicle battery capacity is 35kWh.

[0148] Connect the three-phase electric vehicle charging load Figure 2 In the low-voltage distribution network shown, three-phase power flow calculations were performed at 96 time points to obtain the three-phase imbalance at each time point. Figure 3 The three-phase current imbalance in the low-voltage side outlet branch of the transformer before and after the unordered charging load of the EV is connected. From Figure 3As can be seen, after the EV charging load was connected, the three-phase current imbalance of the transformer's low-voltage outlet branch increased between 00:00 and 3:00 and between 18:30 and 24:00, and decreased around 13:30, 16:00, and 17:30. This indicates that the disordered EV charging load can either increase or decrease the three-phase current imbalance. Table 1 shows the changes in the maximum three-phase current imbalance of the transformer's low-voltage outlet branch throughout the day before and after the disordered EV charging load was connected. As can be seen from Table 1, the maximum three-phase current imbalance of the low-voltage outlet branch increased from 27.5% to 47% after the disordered EV charging load was connected.

[0149] Further analysis reveals that whether EV charging increases the three-phase current imbalance depends on whether the phase with the largest charging current deviation and the phase with the largest base load current deviation are the same. The three-phase current imbalance increases between 18:30 and 24:00 because during this period, the phase with the largest charging current deviation is phase C, which is also the phase with the largest base load current deviation. The charging current exacerbates the difference in current between phases, thus increasing the three-phase imbalance. The three-phase current imbalance decreases around 13:30 because at this time, the two phases with the largest deviations are not the same, and the charging current cancels out the difference in current between phases, thus reducing the three-phase current imbalance.

[0150] Table 1 Maximum Three-Phase Current Unbalance in Low-Voltage Outlet Branch

[0151]

[0152] (2) Voltage deviation

[0153] Figure 4 This demonstrates the voltage deviation at the end node of the low-voltage distribution area within one day after the unordered charging load of EVs was connected. Figure 4 It can be seen that during the period from 15:30 in the afternoon to 24:00 at night, the voltage of the end nodes of the low-voltage distribution network generally exceeded the limit, and the voltage deviation reached the maximum of -19.8% at 19:00, 20:30 and 22:15 at night.

[0154] 3) Establish an optimization model for orderly charging of electric vehicles.

[0155] With the goal of maximizing the net revenue of charging operators, an optimization model is established and constraints are set, taking into account charging service fee revenue, user discount expenditure and power quality incentive revenue.

[0156] 4) The charging scheduling scheme is optimized using the Genetic Annealing Algorithm (GA-SA).

[0157] During population initialization, multiple charging strategies are randomly generated, and each EV's charging status is represented by a matrix encoding within an acceptable scheduling period. The fitness function is calculated using the charging operator's net revenue. A new population is generated through selection, crossover, and mutation operations using a genetic algorithm, and the Metropolis criterion of simulated annealing is used to determine whether to accept offspring schemes. This algorithm can escape local optima and accelerate global convergence in multiple iterations. Figure 5 The convergence curves of the GA-SA algorithm and the traditional GA and SA algorithms are compared. It can be seen that the method of the present invention converges faster and can escape local optima multiple times.

[0158] Table 2 Maximum Three-Phase Current Imbalance Before and After Adopting the Ordered Charging Strategy

[0159]

[0160] Figure 6 This demonstrates the effectiveness of implementing an ordered charging strategy in mitigating the three-phase current imbalance in the low-voltage side outlet branch of the transformer. From... Figure 6 It can be observed that after adopting the ordered charging strategy, the overall three-phase current imbalance of the transformer's low-voltage outlet branch decreases, but it slightly increases at certain times such as 4:00, 5:00, and 11:00 compared to before the optimized charging strategy. Table 2 shows that the maximum three-phase current imbalance of the transformer's low-voltage outlet branch decreased from 47% before optimization to 9.9%. The optimized charging strategy reduces the three-phase current imbalance by transferring the power of the phase with the largest current deviation to other charging times.

[0161] Figure 7 This demonstrates the voltage deviation of the low-voltage distribution area's end node within a day before and after orderly charging, from... Figure 7 As can be seen, the node voltage increased significantly from 16:00 to 24:00 after adopting the ordered charging strategy, but decreased significantly from 02:00 to 07:00. This is because, after adopting the ordered strategy, some of the charging load was shifted from the evening and night to the morning.

[0162] Table 3 Maximum voltage deviation before and after adopting the ordered charging strategy

[0163]

[0164] As shown in Table 3, after adopting the orderly charging strategy, the maximum voltage deviation at the end node of the low-voltage distribution area decreased from -19.8% during disordered charging to -10%, a significant reduction in voltage deviation. Although the voltage dropped slightly from 02:00 to 07:00 due to the shift of some charging load to the early morning period, the overall voltage level was still significantly better than that under disordered charging conditions. In particular, it effectively improved the low voltage problem during the high-load periods in the evening and at night, enhancing the voltage quality and operational safety of the distribution area.

Claims

1. A method for optimizing orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation, characterized in that, Includes the following steps: S1: Construct a two-dimensional excitation model based on charging excitation and charging duration; S2: Divide the 24 hours of a day into segments and introduce a time-segment coefficient γ to correct the two-dimensional incentive model, forming a time-segmented user response rate model; S3: Obtain the three-phase current imbalance and voltage deviation of the distribution network before and after the implementation of each candidate charging scheme, calculate the incentive revenue in combination with the operator's incentive mechanism, and obtain the operator's net profit index. S4: Based on the operator's net profit index, with the goal of maximizing the operator's net profit, and taking into account user response rate and power quality incentive benefits, a multi-constraint ordered charging optimization model is established. The model is then solved by combining genetic algorithm and simulated annealing algorithm to obtain the optimal charging time period and power allocation scheme.

2. The optimized method for orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation as described in claim 1, characterized in that, The two-dimensional excitation model in step S1 is expressed as follows: ； Where: α and β are the influencing factors of charging incentive level and charging duration, respectively; d represents charging incentive, m represents actual charging duration, m0 represents planned charging duration, and Ψ represents the ratio between charging incentive and charging duration.

3. The method for optimizing orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation as described in claim 2, characterized in that, The user response rate model in step S2 is expressed as follows: ； Where γ is the time-segmentation coefficient.

4. The method for optimizing orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation as described in claim 3, characterized in that, The incentive payout in step S3 is calculated as follows: ； Where: κ1 is the three-phase unbalanced excitation coefficient, κ2 is the voltage deviation excitation coefficient, and ε I 0 ε I These represent the maximum three-phase current imbalance in the low-voltage output branch of the transformer before and after the ordered charging strategy, respectively, throughout the day. 0 and e represent the maximum deviations of the voltage amplitude of all nodes in the system from the rated voltage before and after the ordered charging strategy, respectively. The calculation method is as follows: ； ； Where: t is a specific point in time within a day; T represents the set of all discrete points in time on that day; n is the nth node number in the distribution network of the transformer area; A={1,…,n} is the set of all nodes in the transformer area.

5. The method for optimizing orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation according to claim 4, characterized in that, The formula for maximizing the operator's net profit in step S4 is as follows: ； Where: π is the net profit of the charging operator, R is the EV charging service fee revenue, r is the power quality incentive, and C is the user's discount fee. The calculation method is as follows: ； ； ； Where: N is the total number of EVs charged at the charging station in one day; x i,t Let x be the charging state of the i-th EV at time t, when x i,t When x = 1, the EV is in a charging state. i,t When P = 0, the EV is in a non-charging state; i c1 is the rated charging power of the i-th EV; c1 is the charging service fee set by the charging operator; Δt is the time interval length. For the response rate function; C i ev d represents the discount amount incurred by the i-th EV due to accepting dispatch; d represents the discount given to the EV user by the charging operator; t represents the discount amount. i s The moment when the i-th EV begins charging; i e The extended charging time for the i-th EV to accept the operator's orderly charging strategy.

6. The optimized method for orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation as described in claim 5, characterized in that, The multi-constraint ordered charging optimization model in step S4 includes charging time extension constraints, user state of charge constraints, node voltage constraints, and power balance constraints.

7. The method for optimizing orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation as described in claim 6, characterized in that, In the multi-constraint ordered charging optimization model of step S4: The constraint on extended charging time is expressed as follows: While providing users with charging incentives (d), charging operators also require users to extend their total charging time to: ； Where m is the charging time required by the charging operator, and l i c The calculation method is as follows: This refers to the charging duration during which the EV does not accept scheduling or the charging incentive is 0. ； Among them, SOC i + and SOC i - These represent the initial and target states of charge of the EV, respectively; E i c For the battery capacity of the EV; P i The rated charging power for electric vehicles; The user's state of charge constraint is expressed as follows: Each electric vehicle's state of charge (SOC) should reach the target state of charge when leaving the charging station, subject to the following constraints: ； The node voltage constraint is expressed as: ； Among them: U min U max These are the minimum and maximum voltages specified by national standards, respectively. The power balance constraint is expressed as: ； ； Where i and j represent the node numbers in the distribution network, and φ∈{a,b,c} represents the three-phase phases; and These represent the active power and reactive power injected into phase φ by node i, respectively; and These represent the active and reactive power of the load at node i in phase φ, respectively. and These are the voltage magnitudes at nodes i and j on phase φ, respectively; and Let φ be the voltage phase angles at nodes i and j respectively; The voltage phase angle difference between nodes i and j at phase φ; and These are the real and imaginary parts of the equivalent admittance between nodes i and j under phase φ, respectively. and φ represents the active and reactive power imbalance of node i on phase φ, respectively; m is the total number of nodes in the distribution network.

8. The method for optimizing orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation as described in claim 7, characterized in that, The process of using a combination of genetic algorithm and simulated annealing algorithm to solve the problem in step S4 includes: A1: Input the total number of electric vehicles N, the charging excitation d, and the charging time m, and initialize the genetic generation counter g and the annealing counter n; A2: Based on the scheduling problem, chromosome coding is performed, and the charging status of each electric vehicle at each scheduling time is represented as binary code "1" or "0", forming a chromosome structure with the number of electric vehicles and the number of scheduling time periods throughout the day as the matrix dimension; A3: Set the initial temperature T0 and the final temperature T f And the cooling coefficient θ; A4: Set the population size SIZE and the maximum number of iterations GEN, randomly generate SIZE charging schemes as the initial population, among which EVs willing to accept scheduling are within the allowed time window [t]. i s , t i s +l i e Within [t], the charging state is randomly generated; EVs that do not accept scheduling will have their charging status randomly generated within [t]. i s , t i s +l i c The charging status remains constant at 1 throughout the time period. A5: Calculate the fitness value of each individual based on the operator's net profit objective function, and include electric vehicles that do not participate in orderly charging in the fitness calculation; A6: Use the roulette wheel selection method to select parent and parent individuals, randomly set crossover points to perform gene crossover operations; under a given mutation probability, perform even-number bit inversion mutation on the charging state within the schedulable time period; A7: Compare the fitness of the parent and offspring according to the Metropolis criterion. If the offspring is better than the parent or accepts a worse solution with a specified probability, the offspring is retained and enters the new population. The individual with the highest current fitness is always retained. A8: Perform cooling operation based on initial temperature, final temperature, and cooling coefficient; A9: Repeat steps A5 to A8 until the maximum number of iterations is reached or the fitness converges, and output the electric vehicle orderly charging scheduling scheme with the optimal fitness value.

9. The method for optimizing orderly charging of electric vehicles based on three-phase imbalance and deviation mitigation as described in claim 8, characterized in that, Step A6 includes: Based on the fitness value calculated in step A5, a parent generation charging scheduling scheme is selected from the current population using a selection mechanism based on fitness ratio. After the parent generation selection is completed, the selected parent generation scheduling scheme is subjected to gene recombination operation under a given crossover probability. By randomly setting the crossover position, the charging state of different electric vehicles at each scheduling time is recombined to generate a new offspring scheduling scheme. Subsequently, under a given mutation probability, the charging state variables within the schedulable time window are perturbed and updated, and the charging state in some time periods is reversed and adjusted.

Images