A dynamic scheduling optimization method for shared electric bicycle battery transportation

Abstract

A method for dynamic scheduling optimization of shared electric motorcycle battery transportation is first designed according to the actual situation of shared electric motorcycles. Then, a shared electric motorcycle battery transportation dynamic scheduling optimization problem model is established, with the goal of maximizing the number of replaceable batteries in the vehicle within the specified time, including the prediction of the number of low-battery electric motorcycles in the parking spot, the addition of electric motorcycles within the cycle update time, and the change in the number of low-battery electric motorcycles in the unserviced parking spot. Finally, a hybrid water wave optimization algorithm based on differential evolution is designed to solve the maximum number of replacement batteries for the vehicle. The present invention solves the vehicle path planning problem for operators to replace batteries for low-battery electric motorcycles in regional shared electric motorcycle parking spots.

Description

Technical Field

[0001] The present invention belongs to the field of shared electric motorcycle battery transportation scheduling, and involves a dynamic scheduling optimization method for maximizing the number of replacement electric motorcycle batteries. The problem is solved based on a hybrid water wave optimization algorithm of differential evolution. Background Art

[0002] Shared electric motorcycles are a product of the sharing economy. Compared with shared bicycles, they are more time-saving and labor-saving because they are powered by electricity. Shared electric motorcycles require users to park at fixed parking spots for use and return, and the electric motorcycles in the parking spots cannot be charged autonomously, so operators must use manual operations to provide battery replacement services for low-power electric motorcycles. As time goes by, due to the mobility of electric motorcycles in the region, the number of low-power electric motorcycles in the parking spots will change accordingly, and the number of requests for battery replacement is uncertain. Therefore, designing a reasonable battery transportation scheduling method to replace low-power electric motorcycles can improve user satisfaction and reduce operator costs, which is a very valuable research issue. Summary of the invention

[0003] In response to the above problems, the present invention designs a hybrid water wave optimization algorithm based on differential evolution, and proposes a dynamic scheduling optimization method for shared electric motorcycle battery transportation of multiple vehicles. The model divides the working time into several periodic update times, which are used to update the changes in the number of low-battery electric motorcycles in the parking spots and the new parking spots that need replacement services, and allocates parking spots that need battery replacement services to vehicles, so as to obtain the vehicle service path that maximizes the number of batteries that can be replaced within the specified working time.

[0004] In order to achieve the above object, the present invention provides the following inventive steps:

[0005] A method for dynamic scheduling optimization of shared electric motorcycle battery transportation, the method comprising the following steps:

[0006] Step 1: Use t' to represent the current time, initially t' = 0; use t * Indicates the periodic update time, that is, every time t * Get the motorcycle information in the parking spot and re-update the vehicle path; use t to represent the current cycle time, that is, the time after several update cycles. In the initial stage, t = 0, and S = {s1, s2, ..., s n} means there are n shared motorcycle parking spots in the area. Represented as the current cycle time t parking point s i (s i ∈S), that is, the number of low-battery motorcycles that need to be replaced; d represents the battery warehouse; V = {v1,v2,…,v k} represents k homogeneous transport vehicles parked in the warehouse, where the maximum number of batteries that vehicle k can load is b k,max The average speed of each vehicle is represented as v, and the number of battery replacement requests at each parking spot meets There are m battery stations for vehicles to recharge their batteries during service, H = {h1,h2,…,h m} indicates that the distance between the warehouse and the parking point is D1(d,s i ), the distance between parking points is D2(s i ,s j ), the distance between the battery station and the parking point is D3(h i ,s i ). The average time taken by maintenance personnel to replace the battery of a shared electric motorcycle is μ, the average time taken to load and unload the battery at the battery station is θ, and the maximum working time of the vehicle is expressed as

[0007] Step 2: Obtain the vehicle information at the current cycle update time, including the vehicle's current location, the time the vehicle arrives at the location, the time the vehicle completes the location task, the number of remaining replaceable batteries in the vehicle, and the number of batteries the vehicle has replaced; add the parking spots waiting for service to the service pool.

[0008] Step 3: Take the current location of the vehicle as the starting point, and take the parking point information and vehicle information in the waiting service pool as input parameters. After calculation in steps 19-34, the vehicle service path is obtained. Assuming that there are g parking points in the service pool waiting for the assigned path, then at the current cycle time t, vehicle v k The service path can be expressed as Among them, n k Denoted as assigned to vehicle v k The number of parking spots, Denote that vehicle k is the jth (1≤j≤n k ) parking spots provide replacement services,

[0009] Step 4: Use Represented as vehicle v k The time to arrive at point j, when vehicle k arrives at the first parking point The driving time is:

[0010] Step 5: Use Indicates vehicle v k The time to complete the task at point j, vehicle v k Change the first stop The times for low to medium battery are: Then the vehicle vk Complete the first parking spot The service task time is

[0011] Step 6: Use Indicates that in T * k,j Vehicle v k The number of batteries that can be replaced is the remaining number. Then the vehicle v k The number of replaceable batteries is:

[0012] Step 7: Vehicle v k Go to the next parking spot If The vehicle will then go to a nearby suitable battery station to recharge the battery and then serve the next parking spot. Otherwise, skip to step 10.

[0013] Vehicle k At the parking spot Choose the most suitable battery station for replenishing batteries for:

[0014]

[0015] Step 8: Vehicle v k Arrival at the parking spot The time is:

[0016]

[0017] Step 9: Vehicle v k At the parking spot The time to complete the replacement task is:

[0018]

[0019] Step 10: Vehicle v k Go directly to the parking spot Arrival at the parking spot The time is:

[0020]

[0021] Step 11: Vehicle v k Go directly to the parking spot The time to complete the point replacement service task is:

[0022]

[0023] Step 12: When the vehicle v k Service parking spot After that, the number of remaining replaceable batteries on the vehicle is updated to:

[0024] Step 13: Vehicle v k The number of motorcycle batteries replaced during this cycle time t is:

[0025]

[0026] Among them, round() means round down, c k is the vehicle v at the end of the current cycle time t k the number of stops served on its route;

[0027] Step 14: When the current time t′≤t, go to step 7; otherwise, go to step 15;

[0028] Step 15: If the vehicle is at a parking point, wait for it to complete the task at that point, and the current time when it is completed is t'; if the vehicle is on the way from parking point i to parking point j, wait for the vehicle to complete the task at point j, and the current time when it is completed is t'; if the vehicle is at a battery station, wait for it to complete the task at that point, and the current time when it is completed is t';

[0029] Step 16: Save the current time t′ vehicle v k Status information, including the vehicle's current location, the time it will arrive at the location, the time it will complete the location, the number of batteries remaining to be replaced, and the number of batteries that have been replaced;

[0030] Step 17: Set the cycle time t = t′ + t * , get the periodic update time t * The information of low-battery motorcycles in the parking spots in the area, including newly added parking spots that meet the replacement conditions and parking spots that have not been served. Since the number of low-battery motorcycles may change on the way to the parking spots, the number of requested battery replacements will be calculated based on the current situation in the parking spots. Increasing δ, the estimated value is And order That is to make a maximum estimate of the number of motorcycles in the parking lot.

[0031] Step 18: Find the maximum number of low-battery motorcycles that can be replaced by all vehicles, and combine the mathematical model of steps 10, 11, and 13 to express it as:

[0032]

[0033]

[0034] Step 19: Design a fitness function based on the discrete water wave optimization algorithm based on differential evolution. Define the fitness function according to the target problem and constraints in step 18. in For vehicle v j The number of motorcycle batteries replaced during the period at time t;

[0035] Step 20: Encode using the sorting sequence method to form a sorting sequence P i :

[0036] P i =(π1,π2,..,π j ..,π D ), where D = n + k - 1, P i Represents the i-th sorted sequence. π j represents the sequence number, π j =j(j=1,2,..,D);

[0037] Step 21: Randomly generate water waves with a population size of NP, denoted as P i =(π1,π2,..,π j ..,π D ), (i=1,2,..,N) and find the optimal solution P in the population i * , whose fitness is f(P i * );

[0038] Step 22: If the algorithm reaches the termination condition, the optimal solution is output; otherwise, jump to step 31;

[0039] Step 23: Propagation operation, for each water wave P i Cycle execution wavelength Second-rate;

[0040] Step 24: If f(P′ i )>f(P i ), then use the water wave P′ i Replace P i , otherwise jump to step 28;

[0041] Step 25: If f(x′)>f(P i * ), then replace P with water wave x′ i * ;

[0042] Step 26: Breaking wave operation, according to the original optimal fitness f max and the fitness of the new optimal solution to determine the solitary wave where kmax To generate the maximum value of an independent wave, ∈ is a very small positive number to avoid anomalies caused by the denominator being zero;

[0043] Step 27: Randomly generate a probability number r less than 1. If r>c, randomly select water wave P i Otherwise, randomly select a dimension d=rand(1,D-1) and swap the left and right dimensions of d.

[0044] Step 28: Calculate the population size NP(g) of the current generation according to the following formula:

[0045]

[0046] Among them, g is the current iteration number, g max is the maximum number of iterations set, NP max and NP min Respectively represent the upper and lower limits of the population size;

[0047] Step 29: If NP>NP(g), NP=NP-NP(g);

[0048] Step 30: Calculate the wavelength of the water wave according to the following formula where f max and f min They are respectively represented as the maximum fitness and the minimum fitness in the current population, α is the wavelength attenuation coefficient, ∈ is a very small positive number to avoid anomalies caused by the denominator being zero, and go to step 23;

[0049] Step 31: If Then go to the next step; otherwise jump to step 2;

[0050] Step 32: Output the final solution and obtain the maximum number of batteries replaced in the vehicle.

[0051] Further, the step 23 includes the following operations:

[0052] Step 23-1: Among them, rand(0,1) is a randomly generated value between [0,1], F is the mutation rate, and Swap(X) is a random swap function that randomly selects two dimensions in the water wave for swapping;

[0053] Step 23-2: Among them, rand(0,1) is a randomly generated value between [0,1], CR is the crossover rate, PMX(P m ,P i ) is a partial matching crossover strategy function;

[0054] Step 23-3: Conflict detection: establish a mapping relationship based on the two sets of sequences exchanged to ensure that the new subsequence formed has no conflict;

[0055] Step 23-4: According to this formula, the one with the best fitness value is selected as the new water wave.

[0056] Beneficial effects of the present invention: This paper establishes a dynamic scheduling optimization problem model for shared electric motorcycle battery transportation of multiple vehicles, and uses a hybrid water wave optimization algorithm based on differential evolution to solve it, thereby achieving vehicle path optimization that maximizes the replacement of low-power electric motorcycle batteries at parking spots within the maximum working time, improving the battery replacement rate, thereby reducing the operating cost of the entire electric motorcycle and improving user satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 is a schematic diagram of the initial path of multi-vehicle service;

[0058] Figure 2 It is the new path obtained after periodic updating. DETAILED DESCRIPTION

[0059] The present invention will be further described below in conjunction with the accompanying drawings.

[0060] Combination Figure 1 and Figure 2 , a shared electric motorcycle battery transportation dynamic scheduling optimization method, comprising the following steps:

[0061] Step 1: If Figure 1 , initially t,t′=0; t * =1, that is, the service route is updated every 1 hour. In the figure, there is a warehouse (D), two battery stations (H1, H2), two transport vehicles (V1, V2) and three parking spots to be served (S1, S2, S3). The maximum amount of batteries that a vehicle can carry is represented by curly brackets, the number of low-power motorcycles requesting battery replacement at each parking spot is represented by small brackets, and the number on the dotted line between the points represents the average time (in hours) for the vehicle to travel between the two points; in addition, the number of batteries in the given warehouse and battery station is sufficient for the vehicle in this problem. Assuming that the maximum number of batteries that the vehicle can carry is 100, the average time to replace the battery for each motorcycle is 1 minute, and the average time spent loading and unloading batteries at the battery station is 10 minutes. The maximum working time of the vehicle is expressed as

[0062] Step 2: Obtain the information of the vehicle at the current cycle update time, including the current location of the vehicle, the time when the vehicle arrives at the location, the time when the vehicle completes the location task, the number of remaining replaceable batteries in the vehicle, and the number of replaced batteries in the vehicle; add the parking spots waiting for service to the service pool;

[0063] Step 3: Take the current location of the vehicle as the starting point, and use the parking point information and vehicle information in the waiting service pool as input parameters, and obtain the vehicle service path after calculation in step 20.

[0064] After periodic update Figure 2

[0065] Step 4: Use Represented as vehicle v k The time when the vehicle arrives at point j is t = 0. The travel time when the vehicle v1 arrives at the first parking point s1 is: minute, minute;

[0066] Step 5: Use Indicates vehicle v k The time to complete the task at point j, and the time for vehicle v1 to replace the low-power battery in the first parking point s1 is: minutes, then the time for vehicle v1 to complete the first parking point service task is minutes; the time for vehicle v2 to replace the low-power battery in the first parking point s3 was 47 minutes, and the task at the completion point was 77 minutes;

[0067] Step 6: Use Indicates that in T * k,j Vehicle v k The number of remaining replaceable batteries, then the number of replaceable batteries for vehicle v1 is: The number of replaceable batteries in vehicle v1 is 53;

[0068] Step 7: Vehicle v1 goes to the next parking station s2 to serve it. If Then the vehicle goes to a suitable nearby battery station to recharge the battery and then serves the next parking point s2, otherwise jumps to step 10;

[0069] Vehicle k At the parking spot Choose the most suitable battery station for replenishing batteries for:

[0070]

[0071] Step 8: Vehicle v k Arrival at the parking spot The time is:

[0072]

[0073] Step 9: Vehicle v k At the parking spot The time to complete the replacement task is:

[0074]

[0075] Step 10: Vehicle v1 goes directly to parking point s2. The time when it arrives at parking point s2 is:

[0076] minute;

[0077] Step 11: Vehicle v k Go directly to the parking spot The time to complete the point replacement service task is:

[0078] The time for vehicle v1 to complete task s2 is 119 minutes;

[0079] Step 12: When the vehicle v k Service parking spot After that, the number of remaining replaceable batteries on the vehicle is updated to: The remaining replaceable battery of vehicle v1 after serving parking spot s2 is 11;

[0080] Step 13: Vehicle v k The number of motorcycle batteries replaced during this cycle time t is:

[0081]

[0082] Among them, round() means round down, c k is the vehicle v at the end of the current cycle time t k The number of stops served on its route. The maximum number of battery replacements for vehicle v1 is 89, and the maximum number of battery replacements for vehicle v2 is 47;

[0083] Step 14: When the current time t′≤t, go to step 7; otherwise, go to the next step;

[0084] Step 15: If the vehicle is at a parking point, wait for it to complete the task at that point, and the current time when it is completed is t'; if the vehicle is on the way from parking point i to parking point j, wait for the vehicle to complete the task at point j, and the current time when it is completed is t'; if the vehicle is at a battery station, wait for it to complete the task at that point, and the current time when it is completed is t';

[0085] Step 16: Save the current time t′ vehicle v kStatus information, including the vehicle's current location, the time it will arrive at the location, the time it will complete the location, the number of batteries remaining to be replaced, and the number of batteries that have been replaced;

[0086] Step 17: Set the cycle time t = t′ + t * , get the periodic update time t * The information of low-battery motorcycles in the parking spots in the area, including newly added parking spots that meet the replacement conditions and parking spots that have not been served. Since the number of low-battery motorcycles may change on the way to the parking spots, the number of requested battery replacements will be calculated based on the current situation in the parking spots. Increasing δ, the estimated value is And order That is, make a maximum estimate of the number of motorcycles in the parking spot, such as Figure 2 ,When the cycle time is 60 minutes, the replanned vehicle service path is obtained;

[0087] Step 18: Find the maximum number of low-battery motorcycles that can be replaced by all vehicles. Combined with step 13, the mathematical model of this problem is expressed as:

[0088]

[0089]

[0090] In this case, the solution value calculated by this objective function is 129, which is the maximum number of replaced batteries;

[0091] Step 19: Design a fitness function based on the discrete water wave optimization algorithm based on differential evolution. Define the fitness function according to the target problem and constraints in step 15. in For vehicle v j The number of motorcycle batteries replaced during the period at time t;

[0092] Step 20: Encode using the sorting sequence method to form a sorting sequence P i :

[0093] P i =(π1,π2,..,π j ..,π D ), where D = n + k - 1, P i Represents the i-th sorted sequence. π j represents the sequence number, π j =j(j=1,2,..,D);

[0094] Step 21: Randomly generate water waves with a population size of NP, denoted as P i=(π1,π2,..,π j ..,π D ), (i=1,2,..,N) and find the optimal solution P in the population i * , whose fitness is f(P i * );

[0095] Step 22: If the algorithm reaches the termination condition, output the current optimal solution and jump to step 32;

[0096] Step 23: Propagation operation, for each water wave P i Perform the following operations;

[0097] Step 23-1: Among them, rand(0,1) is a randomly generated value between [0,1], F is the mutation rate, and Swap(X) is a random swap function that randomly selects two dimensions in the water wave for swapping;

[0098] Step 23-2: Among them, rand(0,1) is a randomly generated value between [0,1], CR is the crossover rate, PMX(P m ,P i ) is a partial matching crossover strategy function;

[0099] Step 23-3: Conflict detection: establish a mapping relationship based on the two sets of sequences exchanged to ensure that the new subsequence formed has no conflict;

[0100] Step 23-4: According to the formula, the one with the best fitness value is selected as the new water wave;

[0101] Step 24: If f(P′ i )>f(P i ), then use the water wave P′ i Replace P i , otherwise jump to step 28;

[0102] Step 25: If f(x′)>f(P i * ), then replace P with water wave x′ i * ;

[0103] Step 26: Breaking wave operation, according to the original optimal fitness f max and the fitness of the new optimal solution to determine the solitary wave where k max To generate the maximum value of an independent wave, ∈ is a very small positive number to avoid anomalies caused by the denominator being zero;

[0104] Step 27: Randomly generate a probability number r less than 1. If r>c, randomly select water wave P i Otherwise, randomly select a dimension d=rand(1,D-1), swap the left and right dimensions of d, and c=0.3;

[0105] Step 28: Calculate the population size NP(g) of the current generation according to the following formula:

[0106]

[0107] Among them, g is the current iteration number, g max The maximum number of iterations set. NP max and NP min Respectively represent the upper and lower limits of the population size;

[0108] Step 29: If NP>NP(g), NP=NP-NP(g);

[0109] Step 30: Calculate the wavelength of the water wave according to the following formula where f max and f min They are respectively represented as the maximum fitness and the minimum fitness in the current population, α is the wavelength attenuation coefficient, ∈ is a very small positive number to avoid anomalies caused by the denominator being zero, and go to step 23;

[0110] Step 31: If Then go to the next step; otherwise, jump to step 2;

[0111] Step 32: Output the final number and obtain the maximum number of batteries replaced for the vehicle.

Claims

1. A method for dynamic scheduling optimization of shared electric motorcycle battery transportation, characterized in that: The method comprises the following steps: Step 1: Use t′ to represent the current time, initially t = 0; use t * Indicates the periodic update time, that is, every time t * Obtain the motorcycle information in the parking spot and re-update the vehicle path; t represents the current cycle time, that is, the time after several update cycles, and t = 0 in the initial stage; S = {s1, s2, ..., s n } means there are n shared motorcycle parking spots in the area. Represented as the current cycle time t parking point s i Number of batteries to be replaced within s i ∈S, that is, a low-battery motorcycle; d represents the battery warehouse; V = {v1, v2, …, v k } represents k homogeneous transport vehicles parked in the warehouse, where the maximum number of batteries that vehicle k can load is b k,max The average speed of each vehicle is represented as v, and the number of battery replacement requests at each parking spot meets There are m battery stations for vehicles to recharge their batteries during service, H = {h1,h2,…,h m } indicates that the distance between the warehouse and the parking point is D1(d,s i ), the distance between parking points is D2(s i ,s j ), the distance between the battery station and the parking point is D3(h i ,s i ), the average time taken by maintenance personnel to replace the battery of a shared electric motorcycle is μ, the average time taken to load and unload the battery at the battery station is θ, and the maximum working time of the vehicle is expressed as Step 2: Obtain the information of the vehicle at the current cycle update time, including the current location of the vehicle, the time when the vehicle arrives at the location, the time when the vehicle completes the location task, the number of remaining replaceable batteries in the vehicle, and the number of replaced batteries in the vehicle; add the parking spots waiting for service to the service pool; Step 3: Take the current location of the vehicle as the starting point, and take the parking point information and vehicle information in the waiting service pool as input parameters. After calculation in steps 19-34, the vehicle service path is obtained. Assuming that there are g parking points in the service pool waiting for the assigned path, then at the current cycle time t, vehicle v k The service path can be expressed as Among them, n k Represented as assigned to vehicle v k The number of parking spots, s′ k,j ′ represents the replacement service provided by vehicle k to the jth parking point in its route, 1≤j≤n k , Step 4: Use Represented as vehicle v k The time to arrive at point j, when vehicle k arrives at the first parking point The driving time is: Step 5: Use Indicates vehicle v k The time to complete the task at point j, vehicle v k Change the first parking spot The times for low to medium battery are: Then the vehicle v k Complete the first parking spot The service task time is Step 6: Use Indicates that in T * k,j Vehicle v k The number of batteries that can be replaced is the remaining number of batteries. k The number of replaceable batteries is: Step 7: Vehicle v k Go to the next parking spot To serve it, 1 <j≤n k ,if The vehicle will then go to a nearby suitable battery station to recharge the battery and then serve the next parking spot. Otherwise, skip to step 10; Vehicle k At the parking spot Choose the most suitable battery station for replenishing batteries for: Step 8: Vehicle v k Arrival at the parking spot The time is: Step 9: Vehicle v k At the parking spot The time to complete the replacement task is: Step 10: Vehicle v k Go directly to the parking spot Arrival at the parking spot The time is: Step 11: Vehicle v k Go directly to the parking spot The time to complete the point replacement service task is: Step 12: When the vehicle v k Service parking spot After that, the number of remaining replaceable batteries on the vehicle is updated to: Step 13: Vehicle v k The number of motorcycle batteries replaced during this cycle time t is: Among them, round() means round down, c k is the vehicle v at the end of the current cycle time t k the number of stops served on its route; Step 14: Current time t ′ ≤t, go to step 7; otherwise, go to the next step; Step 15: If the vehicle is at the parking point, wait for it to complete the task at that point. The current time when it is completed is t ′ ; If the vehicle is on the way from parking point i to parking point j at this time, wait for the vehicle to complete the task at point j. The current time when the task is completed is t ′ ; If the vehicle is at the battery station at this time, wait for it to complete the task at that point. The current time when it is completed is t ′ ; Step 16: Save the current time t′ vehicle v k Status information, including the vehicle's current location, the time the vehicle arrived at the location, the time the vehicle completed the location task, the number of batteries remaining in the vehicle that can be replaced, and the number of batteries that have been replaced; Step 17: Set the cycle time t = t′ + t * , get the periodic update time t * The information of low-battery motorcycles in the parking spots in the area where the vehicle is located includes new parking spots that meet the replacement conditions and parking spots that have not been served yet. Since the number of low-battery motorcycles may change on the way to the parking spots, the number of requested battery replacements will be increased according to the current situation in the parking spots. Increasing δ, the estimated value is And order That is, make a maximum estimate of the number of motorcycles in the parking spot; Step 18: Find the maximum number of low-power motorcycles that can be replaced by all vehicles. Combining steps 10, 11, and 13, the mathematical model of this problem is expressed as: Step 19: Design a fitness function based on the discrete water wave optimization algorithm based on differential evolution. Define the fitness function according to the target problem and constraints in step 18. in For vehicle v j The number of motorcycle batteries replaced during the period at time t; Step 20: Encode using the sorting sequence method to form a sorting sequence P i :P i =(π1,π2,..,π j ..,π D ), where D = n + k - 1, P i represents the i-th sorting sequence, π j represents the sequence number, π j =j(j=1,2,..,D); Step 21: Randomly generate water waves with a population size of NP, denoted as P i =(π1,π2,..,π j ..,π D ), (i=1,2,..,N) and find the optimal solution P in the population i * , whose fitness is f(P i * ); Step 22: If the algorithm reaches the termination condition, the optimal solution is output; otherwise, jump to step 31; Step 23: Propagation operation, for each water wave P i Cycle execution wavelength Second-rate; Step 24: If f(P′ i )>f(P i ), then use the water wave P′ i Replace P i , otherwise jump to step 28; Step 25: If f(x′)>f(P i * ), then replace P with water wave x′ i * ; Step 26: Breaking wave operation, according to the original optimal fitness f max and the fitness of the new optimal solution to determine the solitary wave where k max To generate the maximum value of an independent wave, ∈ is a very small positive number to avoid anomalies caused by the denominator being zero; Step 27: Randomly generate a probability number r less than 1. If r>c, randomly select water wave P i Otherwise, randomly select a dimension d=rand(1,D-1) and swap the left and right dimensions of d. Step 28: Calculate the population size NP(g) of the current generation according to the following formula: Among them, g is the current iteration number, g max is the maximum number of iterations set, NP max and NP min Respectively represent the upper and lower limits of the population size; Step 29: If NP>NP(g), NP=NP-NP(g); Step 30: Calculate the wavelength of the water wave according to the following formula where f max and f min They are respectively represented as the maximum fitness and the minimum fitness in the current population, α is the wavelength attenuation coefficient, ∈ is a very small positive number to avoid anomalies caused by the denominator being zero, and go to step 23; Step 31: If Then go to the next step; otherwise jump to step 2; Step 32: Output the final solution and obtain the maximum number of batteries replaced in the vehicle.

2. A method for dynamic scheduling optimization of shared electric motorcycle battery transportation as claimed in claim 1, characterized in that: The step 23 includes the following operations: Step 23-1: Among them, rand(0,1) is a randomly generated value between [0,1], F is the mutation rate, and Swap(X) is a random swap function that randomly selects two dimensions in the water wave for swapping; Step 23-2: Among them, rand(0,1) is a randomly generated value between [0,1], CR is the crossover rate, PMX(P m ,P i ) is a partial matching crossover strategy function; Step 23-3: Conflict detection: establish a mapping relationship based on the two sets of sequences exchanged to ensure that the new subsequence formed has no conflict; Step 23-4: According to this formula, the one with the best fitness value is selected as the new water wave.

Images