Improved NSGA-II algorithm-based battery swap station charging and swap strategy optimization method

Abstract

The invention provides a charging station charging strategy optimization method based on an improved NSGA-II algorithm, and belongs to the technical field of multi-objective optimization. The problem of operation cost of an electric heavy truck battery swap station in a battery swap mode is solved; comprising the following steps: modeling an actual operation scene of the electric heavy truck in a battery swap mode, and determining the daily average battery swap demand of a battery swap station; taking the minimization of the operation cost of the battery swap station as a target, providing three target functions including the initial configuration battery number of the battery swap station, the battery charging loss and the electric power cost, and establishing a multi-target optimization model; the aspects of operator operation, parameter adjustment, crowding degree distance calculation and the like in a traditional NSGA-II algorithm are improved, and an established multi-target optimization model is solved, so that a charging and battery replacing strategy of the heavy truck in a battery replacing mode is obtained; the method is applied to the operation of the battery swap station of the electric heavy truck.

Description

Technical Field

[0001] The present invention provides a charging and swapping strategy optimization method for a battery swapping station based on an improved NSGA-II algorithm, and belongs to the technical field of mathematical modeling and multi-objective optimization. Background Art

[0002] In recent years, in order to ease environmental pressure, reducing carbon emissions has become a global consensus to address climate change, and transportation electrification, as an effective solution to reduce emissions and carbon, has become a mainstream trend. In the field of transportation, the freight industry accounts for a huge proportion of emissions, especially heavy trucks. Heavy trucks have heavy loads, long mileage, and long operating hours, resulting in a large amount of fuel consumption and greenhouse gas emissions. Electric heavy trucks have good application prospects and outstanding potential in reducing emissions and carbon.

[0003] However, the problems of electric vehicles such as range anxiety and long waiting time for charging are more serious in electric heavy trucks. Currently, the feasible solutions mainly include fast charging and battery replacement. Fast charging means that vehicles with exhausted batteries can be quickly charged at charging stations using a high C-rate. This method can indeed greatly reduce the time for recharging, but it will cause serious damage to the battery itself. Another way is to adopt a battery swap mode. At a battery swap station, heavy trucks can replace exhausted batteries with fully charged batteries within a few minutes to continue operating. The replaced exhausted batteries can be charged on the spot at the battery swap station, and can be used as batteries for swap again after being fully charged.

[0004] Due to the high initial investment and operating costs of battery swap stations, large-scale promotion and application are limited, so how to minimize the operating costs of battery swap stations becomes a key issue; many studies have proposed different planning models for battery swap station operations and applied corresponding algorithms to solve them. The optimization of battery swap station operations can be divided into three types according to different decision-making objectives: static optimization, sequential decision optimization, and multi-objective optimization.

[0005] Multi-objective optimization problem refers to a type of complex decision-making problem, which involves two or more objective functions that conflict with each other or are difficult to maximize simultaneously. It is often unrealistic to seek a single optimal solution. Multi-objective optimization algorithms are particularly suitable for such problems. Summary of the invention

[0006] In order to solve the operating cost problem of battery swap stations for electric heavy trucks in the battery swap mode, the present invention proposes an optimization method for charging and swapping strategies of battery swap stations based on an improved NSGA-Ⅱ algorithm.

[0007] The technical solution adopted by the present invention is:

[0008] The beneficial effects of the present invention compared with the prior art are as follows: a method for optimizing charging and battery swapping strategies of battery swapping stations based on an improved NSGA-Ⅱ algorithm, comprising the following steps:

[0009] Step 1: Model the actual operation scenario of electric heavy-duty trucks under the battery swap mode and determine the average daily battery swap demand at the battery swap station;

[0010] Step 2: With the goal of minimizing the operating cost of the battery swap station, three objective functions including the number of batteries initially configured at the battery swap station, battery charging loss, and electricity cost are proposed to establish a multi-objective optimization model;

[0011] Step 3: Use the improved NSGA-Ⅱ algorithm to solve the established multi-objective optimization model to obtain the charging and battery swapping strategy of heavy trucks in the battery swapping mode;

[0012] The improvements of the improved NSGA-Ⅱ algorithm are in three aspects: operation operator, parameter adjustment and congestion distance calculation;

[0013] The improvements in the operator part are as follows: random double-point crossover and sequential crossover are combined to replace single-point crossover in the crossover operation; single-point mutation and uniform mutation are used respectively according to the type of decision variables in the mutation operation, and a dynamic change step size is introduced, which is gradually reduced during the iteration process;

[0014] The improvements in the parameter adjustment part are: dynamically adjust the parameters according to the solution situation;

[0015] The improvement of the congestion distance calculation is as follows: according to the preference and importance differences between the various objective functions, a weight factor is introduced to assign a weight to each objective function, so that the algorithm can adjust the congestion distance calculation of the solution according to the importance of the objective function during the optimization process.

[0016] In step 1, it is assumed that all electric heavy trucks use the same transportation route and run at a constant speed, and the battery swap station is on the transportation route; the operation process is simply regarded as a two-way route from the starting point to the destination, and the route is divided into loaded state and unloaded state. According to the energy consumption under different load states, the number of times a battery swap can complete the route is calculated; then, according to the size of the truck fleet and the average number of routes that each truck needs to complete per day, the average number of battery swaps required by the battery swap station is obtained, and the expression is as follows:

[0017]

[0018] In the above formula: N represents the total daily demand for battery swapping at the battery swapping station, that is, the average number of rechargeable batteries required by the battery swapping station per day; K represents the size of the electric heavy-duty truck fleet; H represents the average number of routes that each heavy-duty truck needs to complete per day; R represents the rated capacity of the battery used in the battery swapping heavy-duty truck; E1 represents the energy required to complete a fixed route, and the fixed route is divided into two parts: loaded and unloaded; e1 and e2 represent the battery energy efficiency of a heavy-duty truck with a rated capacity of R under load and unload, respectively; L represents the length of the fixed route.

[0019] The first objective function in step 2 is to minimize the number of batteries required to be configured in the battery swap station, where the number of batteries required is determined by the number of available batteries in the battery swap station.

[0020] The second objective function in step 2 is to minimize the charging loss of the battery. When the health status of a battery degrades to a certain degree of rated capacity, its service life is considered to have ended. Therefore, the average number of cycles of the battery is calculated according to different charging methods. Combined with the value of a single battery, the single loss is converted into a specific cost. The battery loss cost is the sum of the losses generated during all battery charging processes.

[0021] The third objective function in step 2 is to minimize the electricity cost of battery charging. The charging electricity cost of each battery is calculated by determining the battery start charging time, the charging method and the charging power, combined with the time-of-use electricity price.

[0022] The expression of the multi-objective optimization model is as follows:

[0023]

[0024] The constraints are:

[0025]

[0026] In the above formula: B nummax (S) represents the maximum number of batteries configured in the battery swap station, B damage (S) represents the battery loss during the charging process of the overall battery, B charge (S) represents the total battery charging cost, Depl(S) represents the number of batteries configured in the battery swap station, S represents the overall decision variable, S1 represents the charging method decision variable, S2 represents the charging time decision variable, N represents the total daily battery swap demand of the battery swap station, and cyclevalue i represents the specific entity of the single loss of the i-th battery, B represents the battery set, Z represents the integer set, Indicates the time when battery i starts charging. Indicates the end time of charging of battery i, P i(t) represents the charging power of battery i at time t, Price(t) represents the electricity price at time t, and f i represents the charging mode of the ith battery, SOC represents the battery energy state, SOH represents the battery health state, and θ represents the SOH threshold when the battery reaches the end of life.

[0027]

[0028] In the above formula: cycle value i represents the specific amount of battery loss per cycle of the i-th battery, V represents the battery value, and cycletimes i represents the average number of cycles of the i-th battery charge, θ represents the SOH threshold when the battery reaches the end of life, and p i represents the capacity decay rate of the i-th battery.

[0029] The process of dynamically adjusting parameters according to the solution in the improved NSGA-Ⅱ algorithm is as follows:

[0030] First, the Pareto front coverage Cmetric is calculated based on the results of each round of iterations. g :

[0031]

[0032] In the above formula: PF is the approximate Pareto solution set after the g-th iteration, pop represents the population size; after obtaining the g-th iteration coverage Cmetric g Then dynamically adjust the crossover and mutation probabilities, and the adjustment formula is as follows:

[0033]

[0034] In the above formula: g is the current iteration number, G is the total iteration number, C g Iteration coverage Cmetric g ; is the crossover probability of the gth generation, is the mutation probability of the gth generation; Δp c is the crossover probability change step size, Δp m is the step size of mutation probability change.

[0035] The calculation formula of the crowding distance in the improved NSGA-Ⅱ algorithm is as follows:

[0036]

[0037] In the above formula: ω j represents the weight assigned to the jth objective function, f j (x i) represents the individual value of the i-th individual in the population on the j-th objective function, represents the maximum individual value of the population on the jth objective function, It represents the minimum individual value of the population on the jth objective function, i represents the individual number, and j represents the objective function number.

[0038] The beneficial effects of the present invention compared with the prior art are as follows:

[0039] (1) The method of the present invention firstly models the actual operation scenario of electric heavy-duty trucks, fully considers the operation characteristics of electric heavy-duty trucks, and classifies and discusses them according to different load conditions to obtain the average daily number of battery swaps required to be completed at the battery swap station;

[0040] (2) The method proposed in the present invention is to model the charging and battery replacement strategies of electric heavy-duty truck batteries under the battery replacement mode, with the goal of minimizing the overall operating cost of the battery replacement station, providing technical support for the operation and promotion of electric heavy-duty trucks, and providing a solution to the operation of the battery replacement station from the algorithm level; the factors considered include battery replacement demand, battery degradation and time-of-use electricity price; the number of battery inputs, battery loss and electricity cost are used as three objective functions; the constraints include full charging, battery conditions and demand restrictions, etc.; the Pareto solution set of the multi-objective optimization problem is solved by the optimization algorithm, the number of BSS battery configurations and the optimal charging method and charging time of DB are determined, and the normal operation of the battery replacement station is supported;

[0041] (3) Since the model proposed in this paper has three objective functions and some of the objective functions affect each other, they need to be balanced; the improved NSGA-Ⅱ algorithm proposed in this paper can solve the multi-objective optimization problem well; at the same time, according to the defects of the traditional NSGA-Ⅱ algorithm and the characteristics of the established model, the traditional NSGA-Ⅱ algorithm is improved in three aspects: operation operator, parameter adjustment process and congestion calculation, and an improved NSGA-Ⅱ algorithm is proposed to optimize the model, and the effect is significantly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] The present invention will be further described below in conjunction with the accompanying drawings:

[0043] Figure 1 It is a flow chart of a method for optimizing charging and swapping strategies of battery swapping stations based on an improved NSGA-Ⅱ algorithm provided by the present invention.

[0044] Figure 2 This is a multi-objective model architecture diagram of a battery swapping station charging and swapping strategy optimization method based on an improved NSGA-Ⅱ algorithm provided by the present invention.

[0045] Figure 3This is a simulation diagram of the operation scenario of a heavy-duty truck in a battery swapping mode based on a battery swapping station charging and swapping strategy optimization method provided by the present invention based on an improved NSGA-Ⅱ algorithm.

[0046] Figure 4 It is a schematic diagram of route simulation of heavy trucks in a battery swapping mode according to a battery swapping station charging and swapping strategy optimization method based on an improved NSGA-Ⅱ algorithm provided by the present invention.

[0047] Figure 5 It is a relationship diagram between the number of configured batteries and the number of available batteries in a battery swapping station charging and swapping strategy optimization method based on an improved NSGA-Ⅱ algorithm provided by the present invention.

[0048] Figure 6 It is a schematic diagram of an algorithm improvement of a battery swapping station charging and swapping strategy optimization method based on an improved NSGA-Ⅱ algorithm provided by the present invention.

[0049] Figure 7 It is an improved schematic diagram of a crossover operator and a mutation operator of a battery swapping station charging and swapping strategy optimization method based on an improved NSGA-Ⅱ algorithm provided by the present invention.

[0050] Figure 8 It is a comparison diagram of the overall population distribution of the algorithm optimization results of the battery swapping station charging and swapping strategy optimization method based on the improved NSGA-Ⅱ algorithm provided by the present invention.

[0051] Fig. 9 It is a population distribution comparison diagram in F2-F3 dimension of an algorithm optimization result of a battery swapping station charging and swapping strategy optimization method based on an improved NSGA-Ⅱ algorithm provided by the present invention.

[0052] Fig.10 It is a comparison chart of individual evaluation values ​​of the algorithm optimization results of the battery swapping station charging and swapping strategy optimization method based on the improved NSGA-Ⅱ algorithm provided by the present invention. DETAILED DESCRIPTION

[0053] like Figures 1 to 10 As shown, the present invention provides a method for optimizing the charging and battery swapping strategy of a battery swapping station based on an improved NSGA-Ⅱ algorithm. There are two main improvements. The first is to propose a multi-objective optimization model for optimizing the charging and battery swapping strategy of electric heavy truck batteries in the battery swapping mode; the second is to propose an improved NSGA-Ⅱ algorithm for optimizing and solving the proposed multi-objective optimization model. The specific steps of the optimization method are as follows:

[0054] Step 1: Model the actual operation scenario of electric heavy-duty trucks under the battery swapping mode and determine the average daily battery swapping demand at the battery swapping station (BSS). The specific process can be carried out as follows:

[0055] The specific operation process of electric heavy-duty trucks is simulated. It is assumed that all electric heavy-duty trucks use the same transportation route and run at a constant speed, and the battery swap station is on the transportation route. The operation process is simply regarded as a two-way route from the mine to the mixing station. At the same time, the route is divided into a loaded state and an unloaded state. According to the energy consumption under different load states, the number of times a battery swap can complete the route is calculated; then, according to the size of the truck fleet and the average number of routes that each truck needs to complete per day, the average number of battery swaps required by the battery swap station is obtained, and the expression is as follows:

[0056]

[0057] In the above formula: N represents the total daily demand for battery swapping at the battery swapping station, that is, the average number of rechargeable batteries required by the battery swapping station per day; K represents the size of the electric heavy-duty truck fleet; H represents the average number of routes that each electric heavy-duty truck needs to complete per day; R represents the rated capacity of the battery used in the battery swapping electric heavy-duty truck; E1 represents the energy required to complete a fixed route, and the fixed route is divided into two parts: loaded and unloaded; e1 and e2 represent the battery energy efficiency of an electric heavy-duty truck with a rated capacity of R under load and unload, respectively; L represents the length of the fixed route.

[0058] Step 2: With the goal of minimizing the operating cost of the battery swap station, three objective functions including the initial configuration of the battery swap station, battery loss and electricity cost are proposed to establish a multi-objective optimization model, as follows:

[0059] Step 2.1: Minimize the number of batteries initially configured at the battery swap station: When the battery swap station starts operating, there are no fully charged batteries in the station for swapping, and a certain number of batteries need to be configured in advance to meet the initial battery swap demand; the first objective function is to minimize the number of batteries initially configured at the battery swap station Depl(S);

[0060] The implementation principle of the first objective function is as follows: when the demand for battery swapping arrives, the battery swapping station will first check whether there are available batteries. If there are, one battery will be taken out for swapping, and the number of available batteries will be reduced accordingly; if there are no available batteries, one battery will be taken out from the configured batteries for swapping, and the number of configured batteries will be increased accordingly; when the replaced battery is fully charged and recharged, it will be added to the available battery queue for subsequent battery swapping, and the number of available batteries will be increased accordingly;

[0061] Step 2.2: Minimize battery loss. When the battery is charged at a battery swap station, the charging process will cause battery loss. Calculate the single battery loss according to the charging method used by the battery and convert it into actual value cost. The second objective function is to minimize the battery charging loss.

[0062] The implementation principle of the second objective function is as follows: when the health status of a battery degrades to a certain degree of rated capacity, its service life is considered to have ended; thus, the average number of cycles of the battery is calculated according to different charging methods; combined with the value of a single battery, the single loss is converted into a specific cost; the battery loss cost is the sum of the losses generated during the charging process of all batteries;

[0063] Step 2.3: Minimize electricity costs; When the battery is charged in the battery swap station, it consumes electricity and is affected by the time-of-use electricity price and charging power; by determining the battery start charging time, the charging method and charging power, and combining the time-of-use electricity price, the charging electricity cost of each battery is calculated; the third goal is to minimize the electricity cost of battery charging;

[0064] The implementation principle of the third objective function is as follows: the electricity cost mainly comes from the recharging process of the exhausted battery; after the exhausted battery enters the battery swap station, the algorithm determines the charging method and the starting time of charging; combined with the time-of-use electricity price, the charging cost of a single battery is equivalent to the product of the electricity price and the charging power integrated over the charging time period; the electricity cost of battery charging is the sum of the electricity costs generated by charging all batteries.

[0065] Specifically, the multi-objective optimization model includes three objective functions: BSS battery configuration quantity B nummax (S), battery loss during charging process B damage (S) and charging cost B charge (S), the specific expression is as follows:

[0066]

[0067] The constraints are:

[0068]

[0069] In the above formula: Depl(S) represents the number of batteries configured in the battery swap station, S represents the overall decision variable, S1 represents the charging method decision variable, S2 represents the charging time decision variable, N represents the total daily battery swap demand of the battery swap station, and cyclevalue i represents the specific entity of the single loss of the i-th battery, B represents the battery set, Z represents the integer set, Indicates the time when battery i starts charging. Indicates the end time of charging of battery i, P i (t) represents the charging power of battery i at time t, Price(t) represents the electricity price at time t, and f i represents the charging mode of the ith battery, SOC represents the battery energy state, SOH represents the battery health state, and θ represents the SOH threshold when the battery reaches the end of life.

[0070] The following three objective functions are modeled separately:

[0071] Minimize the number of batteries configured in the battery swap station: When the BSS is initially running, there are no fully charged batteries (Fully Charged Battery, FCB) available for exchange in the station, and a certain number of Depl (S) needs to be configured in advance to meet the battery swap demand. When the battery swap demand arrives, the BSS first checks whether the available battery Avail (t) is greater than 0: If it is greater than 0, the BSS takes out an FCB from Avail (t) for replacement, and the number of Avail (t) decreases accordingly, Avail (t) = Avail (t-1)-1; otherwise, a piece is taken out from Depl (S) for replacement, and the number of Depl (S) increases accordingly, Depl (S) = Depl (t-1) + 1. At the same time, when the replaced depleted battery (Depletded Battery, DB) is charged, it can be added to Avail (t) as an FCB for subsequent replacement, and the number of Avail (t) increases accordingly, Avail (t) = Avail (t-1) + 1. The relationship between Depl (S) and Avail (t) is as follows: Figure 5 As shown in the figure, Depl(S) increases until it reaches the peak value, and Avail(t) fluctuates with demand and FCB replenishment. Since the battery cost of battery swapping heavy trucks (BSHTs) is relatively high, the greater the investment in BSS battery configuration, the higher the initial construction cost. The present invention focuses on the maximum number of batteries required during operation, that is, the peak value B of the BSS battery configuration under the decision variable S. nummax , B nummax (S) = max(Depl(S)).

[0072] Minimize battery loss: When a battery’s state of health (SOH) degrades to θ of its rated capacity, it is considered to be at the end of its useful life (EOL), Q EOL =θ·R; Charging mode from slow to ultra-fast charging current (C-rate) C i Gradually increases, the battery capacity decay rate p i Gradually increases, combined with the SOH threshold θ, the average battery cycle times under four charging methods can be calculated i , According to the battery value V, the single loss is converted into a specific cost cycle value i , The sum of battery charging losses during operation is expressed as:

[0073]

[0074] Minimize electricity costs: Start with charging time Time to end charging The charging cost of each battery is obtained by integrating the product of the charging power and the electricity price during the battery charging cycle, and the sum is obtained as follows:

[0075]

[0076] Step 3: Use the improved NSGA-Ⅱ algorithm to solve the established multi-objective optimization model, so as to obtain the charging and battery swapping strategy of heavy trucks in the battery swapping mode. The present invention improves the traditional NSGA-Ⅱ algorithm in terms of operation operators, parameter adjustment process and congestion distance calculation.

[0077] The specific process of improving the NSGAⅡ algorithm in step 3 is as follows:

[0078] In terms of operators, the main improvements are made to the crossover operator and mutation operator; in the crossover operation, random double-point crossover and sequential crossover are combined to replace single-point crossover; in the mutation operation, single-point mutation and uniform mutation are used according to the different types of decision variables, and a dynamic change step size is introduced, which is gradually reduced during the iteration process; in terms of parameter adjustment, the traditional algorithm presets various parameters and keeps them unchanged during the iteration process. The improved NSGAⅡ algorithm proposes a strategy to dynamically adjust parameters according to the solution situation; in terms of crowding distance calculation, according to the preference and importance differences between each objective function, a weight factor is introduced to assign a weight to each objective function, so that the algorithm can adjust the crowding distance calculation of the solution according to the importance of the objective function during the optimization process.

[0079] More specific algorithm improvements are as follows:

[0080] Step 3.1: Operator improvement: Decision variables S1 and S2 are both discrete integer variables. S1 and S2 represent the DB charging mode and the start charging time, respectively. Integer codes are used to express individuals, and the chromosome structure is also divided accordingly. Since there is a time sequence before and after the arrival of the battery, in terms of crossover operations, random double-point crossover and sequential crossover are combined to replace single-point crossover to retain the partial orderliness of the chromosome and increase the diversity of offspring; in terms of mutation operations, S1 and S2 use single-point mutation and uniform mutation, respectively. At the same time, a dynamically changing step size step is introduced based on the uniform mutation of S2. A larger step is used at the beginning of the iteration to accelerate the search of the solution space. As the iteration proceeds, the step gradually decreases, and a better solution is searched instead. The operator operation is improved, such as Figure 7 shown.

[0081] Step 3.2: Dynamic parameter adjustment: In the traditional NSGA-Ⅱ algorithm, parameters such as crossover and mutation probability are preset before the iteration, and remain unchanged during the iteration process. This static parameter setting strategy cannot fully utilize the dynamic characteristics of algorithm iteration and lacks flexibility. Therefore, the present invention proposes a strategy for dynamically updating parameters in combination with iteration results during the iteration process. First, the Pareto Front coverage Cmetric is calculated based on the results of each round of iteration. g :

[0082]

[0083] In the above formula: PF is the approximate Pareto solution set after the g-th iteration, pop represents the population size; after obtaining the g-th iteration coverage Cmetric g Then dynamically adjust the crossover and mutation probabilities, and the adjustment formula is as follows:

[0084]

[0085]

[0086] In the above formula: g is the current iteration number, G is the total iteration number, C g Iteration coverage Cmetric g ; is the crossover probability of the gth generation, is the mutation probability of the gth generation; Δp c is the crossover probability change step size, Δp m Adjust the parameters for the mutation probability change step.

[0087] Step 3.3: Weighted congestion distance: The traditional NSGA-Ⅱ algorithm uses a linear method to calculate the congestion distance, as shown in the following formula:

[0088]

[0089] In order to reflect the differences in preferences and importance between different objectives, the present invention assigns a weight factor to each objective function so that the algorithm can adjust the calculation of the congestion distance of the solution according to the importance of the objective function during the optimization process. The expression is as follows:

[0090]

[0091] In the above formula: ω j represents the weight assigned to the jth objective function, f j (x i ) represents the individual value of the i-th individual in the population on the j-th objective function, represents the maximum individual value of the population on the jth objective function, represents the minimum individual value of the population on the jth objective function, i represents the individual number, and j represents the objective function number. In this embodiment, i is the battery number.

[0092] Improved NSGA-Ⅱ algorithm process Figure 6 shown.

[0093] The improved NSGA-Ⅱ algorithm of the present invention is further described below based on the specific experimental optimization results. Assume that the initial population size pop = 200, the maximum number of iterations G = 1000. The initial values ​​of the crossover probability and mutation probability pc = 0.9, pm = 0.05, and the change step size Δp c =0.01, Δp m =0.05, the crossover probability range is set to [0.5, 1], and the mutation probability range is set to [0.01, 0.3]. The population individual decision variables S1 and S2 are randomly initialized to generate a random population overall distribution, and the algorithm is used to optimize the model. After the algorithm iteration is completed, the overall population distribution comparison and the F2-F3 dimension population distribution comparison are as follows Figure 8 , Fig. 9 shown.

[0094] After the algorithm iteration is completed and the approximate Pareto front is obtained, in order to accurately compare the quality of the solution, the linear weighted method is used to process the solution set, and the evaluation value of each solution is calculated by the following formula. The smaller the individual evaluation value, the lower the overall operating cost of the BSS.

[0095]

[0096] Where N ij represents the normalized value of the jth objective function of the ith individual, ω j Indicates the weight of the jth objective function. According to the model requirements, the weights of F1, F2, and F3 are set to ω1 = 0.4, ω2 = 0.5, and ω3 = 0.1, respectively.

[0097]

[0098] Among them, F j1 and F j2 are the jth objective function values ​​of the NSGA-Ⅱ algorithm and the improved NSGA-Ⅱ algorithm respectively. Fig.10 As shown. The maximum and minimum individual evaluation values ​​of the unimproved algorithm are 0.5220 and 0.1925 respectively; the maximum and minimum individual evaluation values ​​of the improved algorithm are 0.5250 and 0.1658 respectively. In terms of mean, the mean of individual evaluation of the unimproved algorithm is 0.31845, and the mean of individual evaluation of the improved algorithm is 0.29087, which is 8.66% better than the original algorithm.

[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A charging and battery swapping strategy optimization method for battery swapping stations based on an improved NSGA-Ⅱ algorithm, characterized in that: The following steps are involved: Step 1: Model the actual operation scenario of electric heavy-duty trucks under the battery swap mode and determine the average daily battery swap demand at the battery swap station; Step 2: With the goal of minimizing the operating cost of the battery swap station, three objective functions including the number of batteries initially configured at the battery swap station, battery charging loss, and electricity cost are proposed to establish a multi-objective optimization model; Step 3: Use the improved NSGA-Ⅱ algorithm to solve the established multi-objective optimization model to obtain the charging and battery swapping strategy of heavy trucks in the battery swapping mode; The improvements of the improved NSGA-Ⅱ algorithm are in three aspects: operation operator, parameter adjustment and congestion distance calculation; The improvements in the operator part are as follows: random double-point crossover and sequential crossover are combined to replace single-point crossover in the crossover operation; single-point mutation and uniform mutation are used respectively according to the type of decision variables in the mutation operation, and a dynamic change step size is introduced, which is gradually reduced during the iteration process; The improvements in the parameter adjustment part are: dynamically adjust the parameters according to the solution situation; The improvement of the congestion distance calculation is as follows: according to the preference and importance differences between the various objective functions, a weight factor is introduced to assign a weight to each objective function, so that the algorithm can adjust the congestion distance calculation of the solution according to the importance of the objective function during the optimization process.

2. The method for optimizing charging and swapping strategies of battery swapping stations based on the improved NSGA-Ⅱ algorithm according to claim 1 is characterized in that: In step 1, it is assumed that all electric heavy trucks use the same transportation route and run at a constant speed, and the battery swap station is on the transportation route; the operation process is simply regarded as a two-way route from the starting point to the destination, and the route is divided into loaded state and unloaded state. According to the energy consumption under different load states, the number of times a battery swap can complete the route is calculated; then, according to the size of the truck fleet and the average number of routes that each truck needs to complete per day, the average number of battery swaps required by the battery swap station is obtained, and the expression is as follows: In the above formula: N represents the total daily demand for battery swapping at the battery swapping station, that is, the average number of rechargeable batteries required by the battery swapping station per day; K represents the size of the electric heavy-duty truck fleet; H represents the average number of routes that each heavy-duty truck needs to complete per day; R represents the rated capacity of the battery used in the battery swapping heavy-duty truck; E1 represents the energy required to complete a fixed route, and the fixed route is divided into two parts: loaded and unloaded; e1 and e2 represent the battery energy efficiency of a heavy-duty truck with a rated capacity of R under load and unload, respectively; L represents the length of the fixed route.

3. The method for optimizing charging and swapping strategies of battery swapping stations based on the improved NSGA-Ⅱ algorithm according to claim 2 is characterized in that: The first objective function in step 2 is to minimize the number of batteries required to be configured in the battery swap station, where the number of batteries required is determined by the number of available batteries in the battery swap station.

4. The method for optimizing charging and swapping strategies of battery swapping stations based on the improved NSGA-Ⅱ algorithm according to claim 2 is characterized in that: The second objective function in step 2 is to minimize the charging loss of the battery. When the health status of a battery degrades to a certain degree of rated capacity, its service life is considered to have ended. Therefore, the average number of cycles of the battery is calculated according to different charging methods. Combined with the value of a single battery, the single loss is converted into a specific cost. The battery loss cost is the sum of the losses generated during all battery charging processes.

5. The method for optimizing charging and swapping strategies of battery swapping stations based on the improved NSGA-Ⅱ algorithm according to claim 2 is characterized in that: The third objective function in step 2 is to minimize the electricity cost of battery charging. The charging electricity cost of each battery is calculated by determining the battery start charging time, the charging method and the charging power, combined with the time-of-use electricity price.

6. A method for optimizing charging and swapping strategies of battery swapping stations based on an improved NSGA-Ⅱ algorithm according to any one of claims 3 to 5, characterized in that: The expression of the multi-objective optimization model is as follows: The constraints are: In the above formula: B nummax (S) represents the maximum number of batteries configured in the battery swap station, B damage (S) represents the battery loss during the charging process of the overall battery, B charge (S) represents the total battery charging cost, Depl(S) represents the number of batteries configured in the battery swap station, S represents the overall decision variable, S1 represents the charging method decision variable, S2 represents the charging time decision variable, N represents the total daily battery swap demand of the battery swap station, and cyclevalue i represents the specific entity of the single loss of the i-th battery, B represents the battery set, Z represents the integer set, Indicates the time when battery i starts charging. Indicates the end time of charging of battery i, P i (t) represents the charging power of battery i at time t, Price(t) represents the electricity price at time t, and f i represents the charging mode of the ith battery, SOC represents the battery energy state, SOH represents the battery health state, and θ represents the SOH threshold when the battery reaches the end of life.

7. The method for optimizing charging and swapping strategies of battery swapping stations based on the improved NSGA-Ⅱ algorithm according to claim 6 is characterized in that: In the above formula: cycle value i represents the specific amount of battery loss per cycle of the i-th battery, V represents the battery value, and cycletimes i represents the average number of cycles of the i-th battery charge, θ represents the SOH threshold when the battery reaches the end of life, and p i represents the capacity decay rate of the i-th battery.

8. The method for optimizing charging and swapping strategies of battery swapping stations based on the improved NSGA-Ⅱ algorithm according to claim 1 is characterized in that: The process of dynamically adjusting parameters according to the solution in the improved NSGA-Ⅱ algorithm is as follows: First, the Pareto front coverage Cmetric is calculated based on the results of each round of iterations. g : In the above formula: PF is the approximate Pareto solution set after the g-th iteration, pop represents the population size; after obtaining the g-th iteration coverage Cmetric g Then dynamically adjust the crossover and mutation probabilities, and the adjustment formula is as follows: In the above formula: g is the current iteration number, G is the total iteration number, C g Iteration coverage Cmetric g ; is the crossover probability of the gth generation, is the mutation probability of the gth generation; Δp c is the crossover probability change step size, Δp m is the step size of mutation probability change.

9. The method for optimizing charging and swapping strategies of battery swapping stations based on the improved NSGA-Ⅱ algorithm according to claim 1 is characterized in that: The calculation formula of the crowding distance in the improved NSGA-Ⅱ algorithm is as follows: In the above formula: ω j represents the weight assigned to the jth objective function, f j (x i ) represents the individual value of the i-th individual in the population on the j-th objective function, represents the maximum individual value of the population on the jth objective function, It represents the minimum individual value of the population on the jth objective function, i represents the individual number, and j represents the objective function number.

Images