Electric vehicle charging scheduling method based on improved red mouth blue magpie algorithm and related equipment

Abstract

The invention provides an electric vehicle charging scheduling method based on an improved red-mouth blue-pie algorithm and related equipment, and relates to the technical field of electric vehicle charging scheduling. The method comprises the following steps: initializing a population, wherein an individual in the population represents a charging decision vector of an electric vehicle; taking the peak-valley load difference of the micro-grid as a fitness function; updating the search capability control parameters, and controlling the individuals to select corresponding group behaviors according to the search capability control parameters so as to perform position updating; carrying out boundary processing and constraint processing on each updated individual; the fitness value of the current population is updated, the individuals in the current population are divided into elite individuals, secondary elite individuals and non-elite individuals according to the fitness value, and therefore different disturbance strategies are adopted to conduct disturbance on each kind of individuals so as to conduct position updating; and the process is repeated until a preset iteration stop condition is met, and at the moment, the position of each individual in the population is the optimal charging decision of the corresponding electric vehicle.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle charging scheduling, and in particular, to an electric vehicle charging scheduling method and related devices based on an improved red-billed blue magpie algorithm. Background Art

[0002] With the rapid development of the electric vehicle industry, the charging demand of electric vehicles has been increasing year by year, bringing great pressure to the power grid. Especially during peak hours, the large-scale charging of electric vehicles may cause the microgrid load to overload, and even lead to the stability problems of the power system. Therefore, how to reasonably schedule the charging process of electric vehicles and reduce the load fluctuation of the microgrid has become an important research direction for the current power system optimization.

[0003] Traditional electric vehicle charging scheduling methods mostly rely on simple heuristic algorithms, such as genetic algorithm (GA), particle swarm optimization (PSO), etc. However, when facing complex constraint conditions and nonlinear optimization problems, these methods often face problems such as local optimal solutions and slow convergence speed. Summary of the Invention

[0004] In order to reduce the load fluctuation of the microgrid caused by the large-scale charging of electric vehicles, the present invention provides an electric vehicle charging scheduling method and related devices based on an improved red-billed blue magpie algorithm. By minimizing the peak-valley load difference of the microgrid, it can effectively optimize the charging power scheduling of electric vehicles, balance the charging demand of electric vehicles and the grid load, and effectively improve the stability and resource utilization efficiency of the power grid.

[0005] In the first aspect, the present invention provides an electric vehicle charging scheduling method based on an improved red-billed blue magpie algorithm, including:

[0006] Step 1: Define the scheduling period and the minimum unit charging duration;

[0007] Step 2: Initialize the population and algorithm parameters; wherein, an individual in the population represents a charging decision vector of an electric vehicle, and the value of an element of the charging decision vector represents the charging power of the corresponding unit charging time period. The dimension of the charging decision vector is determined by the scheduling period and the minimum unit charging duration; the algorithm parameters include the number of electric vehicles, the charging power constraint condition, and the microgrid load constraint condition;

[0008] Step 3: Use the peak-valley load difference of the microgrid as the fitness function, and update the fitness value of the current population according to the fitness function;

[0009] Step 4: Update the search ability control parameter, and for each individual, control the individual to select the corresponding group behavior according to the search ability control parameter to perform position update; the group behavior includes large group behavior and small group behavior;

[0010] Step 5: For each updated individual, perform boundary processing and constraint processing so that the updated population satisfies the charging power constraint condition and the microgrid load constraint condition;

[0011] Step 6: Update the fitness values of the current population, and divide the individuals in the current population into elite individuals, sub-elite individuals, and non-elite individuals according to the fitness values;

[0012] Step 7: For elite individuals, sub-elite individuals, and non-elite individuals, use different perturbation strategies to perturb each type of individual for position update; and compare the fitness values of the individuals before and after the update, and record the better fitness value and the corresponding better solution;

[0013] Step 8: Repeat Steps 4 to 7 until a preset iteration stop condition is reached. At this time, the position of each individual in the population is the optimal charging decision for the corresponding electric vehicle.

[0014] Further, the updated search ability control parameter specifically includes:

[0015]

[0016] where ∈ is the search ability control parameter, t is the current iteration number, and T is the maximum iteration number.

[0017] Further, for each individual, control the individual to select the corresponding group behavior for position update according to the search ability control parameter, specifically including:

[0018] Generate a random number within (0, 1). If the random number is less than the search ability control parameter, select the small group behavior; otherwise, select the large group behavior.

[0019] Further, in Step 7, for elite individuals, sub-elite individuals, and non-elite individuals, use different perturbation strategies to perturb each type of individual for position update, specifically including:

[0020] Add a large perturbation to the elite individuals; add a small perturbation to the sub-elite individuals; for the non-elite individuals, select an elite individual for linear combination of positions and then add a random perturbation;

[0021] For all individuals, compare the fitness values of the individuals before and after the perturbation, and retain the position of the individual corresponding to the smaller fitness value.

[0022] Further, before comparing the fitness values of individuals before and after perturbation for all individuals, it further includes: generating a random number within (0, 1), and if the random number is less than a preset threshold, re-perturbing the non-elite individuals after perturbation according to the global optimal solution of the population and the average position of the small group, so that the non-elite individuals approach the global optimal solution.

[0023] Further, re-perturbing the non-elite individuals after perturbation according to the global optimal solution of the population and the average position of the small group specifically includes:

[0024]

[0025] where ω represents the weight, t is the current iteration number, T is the maximum iteration number, represents the position after perturbation, X i represents the position before perturbation, rand1 and rand2 represent random numbers generated within (0, 1), X food represents the global optimal solution of the population, X pmean represents the average position of the small group.

[0026] Further, in step 2, the population is initialized using the Sobol sequence.

[0027] On the other hand, the present invention provides an electric vehicle charging scheduling device based on an improved red-billed blue magpie algorithm, including:

[0028] A setting module for defining a scheduling period and a minimum unit charging duration;

[0029] An initialization module for initializing the population and algorithm parameters; wherein, an individual in the population represents a charging decision vector of an electric vehicle, and the element values of the charging decision vector represent the charging power corresponding to the unit charging time period, and the dimension of the charging decision vector is determined by the scheduling period and the minimum unit charging duration; the algorithm parameters include the number of electric vehicles, charging power constraint conditions, and microgrid load constraint conditions;

[0030] A fitness calculation module for using the peak-valley load difference of the microgrid as a fitness function and updating the fitness values of the current population according to the fitness function;

[0031] A position update module, configured to update search ability control parameters, and for each individual, control the individual to select a corresponding group behavior for position update according to the search ability control parameters; the group behaviors include large-group behaviors and small-group behaviors; for each updated individual, perform boundary processing and constraint processing to make the updated population meet the charging power constraint condition and the microgrid load constraint condition; divide the individuals in the current population into elite individuals, sub-elite individuals, and non-elite individuals according to fitness values; and for elite individuals, sub-elite individuals, and non-elite individuals, respectively adopt different perturbation strategies to perturb each type of individual for position update; and compare the fitness values of the individuals before and after the update, and record the better fitness value and the corresponding better solution.

[0032] A judgment module, configured to judge whether a preset iteration stop condition is reached, and if so, use the position of each individual in the population at this time as the optimal charging decision for the corresponding electric vehicle.

[0033] In a third aspect, the present invention provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the method described in the first aspect is implemented.

[0034] In a fourth aspect, the present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the method described in the first aspect is implemented.

[0035] The beneficial effects of the present invention are as follows:

[0036] (1) By optimizing the charging scheduling strategy of electric vehicles, the present invention can effectively reduce the load fluctuation of the microgrid, especially the peak-valley difference of the load. This makes the load of the microgrid more balanced during the peak charging period of electric vehicles, avoiding the risk of overload of the power grid when the charging load is concentrated, thereby improving the operation stability of the microgrid.

[0037] (2) Based on the improved red-billed blue magpie algorithm (MRBMO), the present invention combines the Sobol sequence and global and local search strategies in the optimization process, and dynamically adjusts the search trade-off parameters, ensuring that both the breadth of global search and the accuracy of local search can be maintained during the charging scheduling process, so as to find the optimal solution more efficiently. At the same time, the random perturbation mechanism and particle swarm update mechanism of the elite individuals and the second-echelon individuals of the algorithm can further accelerate convergence and improve the solution accuracy.

[0038] (3) By optimizing the charging time period of electric vehicles, the present invention can effectively avoid the burden on the power grid during peak hours and reduce the response cost of power suppliers to peak loads. In addition, since the adjustment of the charging time period can better align with the production time period of renewable energy, it helps to reduce the use of traditional fossil energy and carbon emissions, having good environmental benefits.

[0039] (5) The charging scheduling method and related equipment of the present invention have strong adaptability and can perform scheduling optimization according to different electric vehicle charging demands, microgrid load characteristics, and external conditions. Whether in urban areas, rural power grids, or large-scale electric vehicle charging scenarios, it can effectively optimize the impact of electric vehicle charging on the microgrid. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a flowchart showing a method for scheduling electric vehicle charging based on an improved red-billed blue magpie algorithm provided by an embodiment of the present invention;

[0041] Figure 2 It is a comparison chart of sample distribution situations using random initialization of the population and Sobol sequence initialization of the population respectively provided by an embodiment of the present invention;

[0042] Figure 3 It is a structural schematic diagram of a device for scheduling electric vehicle charging based on an improved red-billed blue magpie algorithm provided by an embodiment of the present invention;

[0043] Figure 4 It is a structural block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0044] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0045] As a newly emerging nature-inspired optimization algorithm, the red-billed blue magpie algorithm (RBMO) has been widely applied to various engineering optimization problems due to its strong global search ability and good optimization performance. However, the traditional red-billed blue magpie algorithm still has deficiencies such as slow convergence speed and insufficient population diversity. Therefore, the present invention improves the traditional red-billed blue magpie algorithm and proposes a method for scheduling electric vehicle charging based on the improved red-billed blue magpie algorithm, including the following steps:

[0046] S101: Define the scheduling period and the minimum unit charging duration;

[0047] Specifically, the scheduling period refers to the time interval from the start of the charging scheduling task designed by the present invention to the next start. In this embodiment, the scheduling period is set to 24 hours, and the minimum unit charging duration is set to 15 minutes. It can be understood that the scheduling period and the minimum unit charging duration can be set as needed.

[0048] S102: Initialize the population and algorithm parameters; where an individual in the population represents a charging decision vector of an electric vehicle, the element values of the charging decision vector represent the charging power in the corresponding unit charging time period, and the dimension of the charging decision vector is determined by the scheduling period and the minimum unit charging duration; the algorithm parameters include the number of electric vehicles, the charging power constraint condition, and the microgrid load constraint condition;

[0049] Specifically, taking the scheduling period of 24 hours and the minimum unit charging duration of 15 minutes as an example, 24 hours can be evenly divided into 96 time slices, each with a duration of 15 minutes. Thus, the charging decisions of the entire population can be represented as an N×96 matrix X, where N represents the number of electric vehicles, and each row of this matrix represents the charging decision vector of an individual. The specific representation is as follows:

[0050]

[0051] X(i,:) = [P i,1 ,P i,2 ,...,P i,t ,P i,96 , i = 1,2,...,N

[0052] where X(i,:) is a 96 - dimensional vector representing the charging decisions of electric vehicle i in 96 time periods; P i,t represents the charging power of electric vehicle i in the t - th time period.

[0053] Generally speaking, the charging modes of electric vehicles are divided into fast - charging mode and slow - charging mode. Correspondingly, the charging power of electric vehicles must also be selected between the charging powers corresponding to the fast - charging mode and the slow - charging mode. This is the constraint condition that the charging power needs to satisfy, which can be specifically expressed as follows:

[0054]

[0055] At the same time, within a scheduling period, the sum of the charging powers of all electric vehicles and the basic load of the microgrid shall not exceed the maximum load capacity of the microgrid. This is the microgrid load constraint condition that should be satisfied, that is:

[0056]

[0057] Among them, L t is the basic load of the microgrid (household appliances, lighting equipment, heating equipment, etc. always maintain a certain electricity demand, and these demands form the basic load of the microgrid), and P max is the maximum load of the microgrid.

[0058] S103: Use the peak-valley load difference of the microgrid as the fitness function, and update the fitness value of the current population according to the fitness function;

[0059] Specifically, the electric vehicle charging scheduling method designed by the present invention aims to minimize the peak-valley load difference of the microgrid, so as to achieve the balance of power demand and energy optimization during the electric vehicle charging process. Therefore, in the embodiments of the present invention, the peak-valley load difference ΔL of the microgrid, which is the peak-valley difference of the microgrid load, is directly used as the fitness function, that is, according to the charging power decision variable of the current time period, the load situation of the microgrid is calculated, and the formula is expressed as follows:

[0060]

[0061] It can be understood that in practical applications, the maximum tolerance value ΔL of the microgrid load fluctuation can be preset in advance max , correspondingly, the peak-valley difference should not exceed this maximum tolerance value ΔL max , that is:

[0062]

[0063] It should be noted that before the start of the first iteration process, the fitness value is initialized to infinity; in this way, for the first iteration process, it can be ensured that the calculated fitness value must be less than "infinity", thereby triggering "update", that is, recording the current minimum fitness value (also called the optimal fitness value BestValue), and the global optimal solution X corresponding to BestValue food ;

[0064] S104: Update the search ability control parameters. For each individual, control the individual to select the corresponding group behavior according to the search ability control parameters to perform position update; the group behavior includes large group behavior and small group behavior;

[0065] Specifically, in the Red-billed Blue Magpie algorithm, 2 to 5 individuals are randomly selected to form a small group, and 10 to N individuals are randomly selected to form a large group. The behavior of the small group can quickly adapt to environmental changes and conduct a detailed search in a local area. However, due to the relatively small scale of the small group, its search range is relatively limited and may not be able to effectively cover the entire solution space. The behavior of the large group can cover a wider search space and, through collective cooperation, can more effectively utilize global information and improve the search efficiency. However, the flexibility of the large group behavior is relatively low, and the search of the local area is not detailed enough. Therefore, the present invention controls the behavior of individuals in the population by introducing a search ability control parameter, so that the two group behaviors can be reasonably combined throughout the scheduling process, balancing exploration and exploitation, and improving the performance of the entire charging scheduling method.

[0066] S105: For each updated individual, perform boundary processing and constraint processing to make the updated population meet the charging power constraint condition and the microgrid load constraint condition;

[0067] Specifically, boundary processing means that after each position update, for individuals that cross the boundary, truncation processing needs to be adopted, and the out-of-bounds position is set to the upper and lower bound values. The formula is as follows:

[0068]

[0069] where X min and X max respectively represent the minimum and maximum values of the charging power.

[0070] Constraint processing means that after each position update, it is necessary to check whether each dimension of the individual (i.e., the power of each unit charging time period of each electric vehicle) meets the constraint conditions; and the sum of the charging powers of all electric vehicles and the sum of the basic loads of the microgrid shall not exceed the maximum load of the microgrid.

[0071] S106: Update the fitness value of the current population, and divide the individuals in the current population into elite individuals, sub-elite individuals, and non-elite individuals according to the fitness value;

[0072] S107: For elite individuals, sub-elite individuals, and non-elite individuals, different perturbation strategies are adopted to perturb each type of individual for position update; and compare the fitness values of the individuals before and after the update, and record the better fitness value and the corresponding better solution;

[0073] Specifically, by clustering the population individuals according to the fitness value, different perturbation strategies can be adopted for each type, which can improve the diversity, randomness, and flexibility of the population individuals, thereby avoiding falling into local optimal solutions. After the position update, a historical best preservation mechanism is introduced (i.e., save the best BestValue and X food) That is, if the fitness value of the individuals in the current generation deteriorates, the positions and fitness values of the previous generation are restored.

[0074] S108: Repeat steps S104 to S107 until a preset iteration stop condition is reached (for example, when the maximum iteration number T is reached or the load fluctuation meets the preset value). At this time, the position of each individual in the population is the optimal charging decision for the corresponding electric vehicle.

[0075] The electric vehicle charging scheduling method provided by the embodiments of the present invention can effectively reduce the load fluctuation of the microgrid, especially the peak-valley difference of the load, by optimizing the charging scheduling strategy of the electric vehicle. This makes the load of the microgrid more balanced during the peak charging period of the electric vehicle, avoiding the risk of overload of the power grid when the charging load is concentrated, thereby improving the operation stability of the microgrid. During the charging scheduling process, the global and local search strategies are combined, and by dynamically adjusting the search ability control parameters, it is ensured that both the breadth of the global search and the accuracy of the local search can be maintained during the charging scheduling process, so as to find the optimal solution more efficiently.

[0076] In one embodiment, the embodiments of the present invention adopt the following method to update the search ability control parameters, which specifically include:

[0077]

[0078] where ∈ is the search ability control parameter, t is the current iteration number, and T is the maximum iteration number.

[0079] Specifically, using the exponential decay formula to update the search ability control parameter can achieve a smooth transition from global search in the initial stage to local development in the later stage.

[0080] In one embodiment, for each individual, according to the search ability control parameter, control the individual to select the corresponding group behavior for position update, which specifically includes: generating a random number within (0, 1). If the random number is less than the search ability control parameter, select the small group behavior; otherwise, select the large group behavior.

[0081] Specifically, the position updates corresponding to the small group behavior and the large group behavior include two stages: (1) The search stage, and the position update formula is as follows:

[0082]

[0083] where X r is the position of a randomly selected individual in the population, rand1, rand2, and rand() are all random numbers within (0, 1), X pmean is the average position or the central position of the small group, X qmeanis the average position or central position of the large group, X i represents the updated individual position, represents the updated individual position after update.

[0084] (2) In the development stage, the position update formula is as follows:

[0085]

[0086] where X food is the global optimal solution of the population, CF is the fitness control factor, N(0,1) is the random perturbation term of the standard normal distribution, and the meanings of other parameters are the same as above, which will not be elaborated here.

[0087] In one embodiment, for elite individuals, sub-elite individuals and non-elite individuals, different perturbation strategies are respectively adopted to perturb each type of individual for position update, specifically including: adding a larger amplitude perturbation to elite individuals; adding a smaller amplitude perturbation to sub-elite individuals; for non-elite individuals, an elite individual is selected for linear combination of positions and then a random perturbation is added; for all individuals, the fitness values of the individuals before and after perturbation are compared, and the individual position corresponding to the smaller fitness value is retained.

[0088] Specifically, according to the ascending order of fitness values, the population is divided into elite individuals (e.g., the top 10%), sub-elite individuals (e.g., the top 20%) and non-elite individuals (also called ordinary individuals). A larger amplitude perturbation (e.g., the perturbation intensity factor is 0.05) is added to elite individuals to enhance the ability to jump out of the local optimum. A smaller amplitude perturbation (e.g., the perturbation intensity factor is 0.01) is added to sub-elite individuals to improve the population diversity. For ordinary individuals, an elite individual is randomly selected as a partner, and after linear combination with its position, a random perturbation is added to improve the diversity. The above perturbation process can be expressed by the formula as follows:

[0089]

[0090] where X j is the randomly selected elite individual, and N(0,1) represents a random variable subject to the standard normal distribution (mean is 0, variance is 1).

[0091] In order to further improve the population quality, in one embodiment, non-elite individuals are also perturbed again with a certain probability (e.g., rand() < 0.2) using the particle swarm update mechanism. The perturbation principle is: according to the global optimal solution of the population and the average position of the small group, the perturbed non-elite individuals are perturbed again so that the non-elite individuals approach the global optimal solution. As an implementable way, the position update formula is:

[0092]

[0093] Among them, ω represents the weight, t is the current iteration number, T is the maximum iteration number, represents the position after perturbation, X i represents the position before perturbation, rand1 and rand2 represent random numbers generated within (0, 1), X food represents the global optimal solution of the population, X pmean represents the average position of the small group.

[0094] Specifically, in the initial stage, the weight is relatively large (ω≈0.9), and the step size during individual update is relatively large, which helps to explore the entire search space and avoid falling into local optimal solutions; as the iteration progresses, the weight gradually decreases, and the step size during individual update becomes smaller, which is more suitable for fine search near the global optimum; the update method of the weight ω makes ω decrease linearly in the first half of the iteration process and further decrease exponentially in the second half of the iteration process, which can improve the convergence ability of the charging scheduling method in the later stage; in the above position update method, ordinary individuals are affected by both the global optimal solution and the small group, enabling them to converge quickly to the optimal region but not completely concentrate on a single point. This can enhance the balance ability of the algorithm between exploration and exploitation, ensure the robustness of the algorithm, make it perform more stably in multi-modal functions and complex constraint optimization problems, and further enhance the population diversity and local development ability.

[0095] In one embodiment, the charging power of each electric vehicle is regarded as a decision variable, and its value is determined within each time period. The value of each decision variable is restricted within the charging power range of the electric vehicle. For example, the power upper limit in the fast charging mode is 10kW, and the power upper limit in the slow charging mode is 3.5kW; in order to improve the initial coverage uniformity of the population in the solution space, the Sobol sequence is used for population initialization in this embodiment.

[0096] Specifically, compared with random initialization, the Sobol sequence is a low-discrepancy sequence that can generate uniformly distributed sample points in a multi-dimensional space, ensuring that the population has a more comprehensive global exploration ability at the initial stage of the search. This initialization method effectively reduces the risk of the algorithm falling into local optima, especially prominent in high-dimensional and complex solution spaces. The population sample distributions generated by using the Sobol sequence and the random method are as Figure 2 shown, and it is obvious that the sample distribution generated by this method is more uniform.

[0097] Based on the same inventive concept, as Figure 3As shown in the figure, an embodiment of the present invention further provides an electric vehicle charging scheduling device based on an improved red-billed blue magpie algorithm, including a setting module, an initialization module, a fitness calculation module, a position update module, and a judgment module.

[0098] Specifically, the setting module is used to define the scheduling period and the minimum unit charging duration. The initialization module is used to initialize the population and algorithm parameters; among them, an individual in the population represents a charging decision vector of an electric vehicle, and the element value of the charging decision vector represents the charging power of the corresponding unit charging time period, and the dimension of the charging decision vector is determined by the scheduling period and the minimum unit charging duration; the algorithm parameters include the number of electric vehicles, the charging power constraint condition, and the microgrid load constraint condition. The fitness calculation module is used to use the peak-valley load difference of the microgrid as the fitness function and update the fitness value of the current population according to the fitness function. The position update module is used to update the search ability control parameter. For each individual, according to the search ability control parameter, control the individual to select the corresponding group behavior for position update; the group behavior includes large group behavior and small group behavior; for each updated individual, perform boundary processing and constraint processing to make the updated population meet the charging power constraint condition and the microgrid load constraint condition; divide the individuals in the current population into elite individuals, sub-elite individuals, and non-elite individuals according to the fitness value; and for elite individuals, sub-elite individuals, and non-elite individuals, respectively use different perturbation strategies to perturb each type of individual for position update; and compare the individual fitness values before and after the update, and record the better fitness value and the corresponding better solution. The judgment module is used to judge whether the preset iteration stop condition is reached. If so, use the position of each individual in the population at this time as the optimal charging decision of the corresponding electric vehicle, specifically including the charging power of each electric vehicle in each time period and the corresponding charging mode (fast charging or slow charging).

[0099] The electric vehicle charging scheduling device provided by the present invention can effectively reduce the load fluctuation of the microgrid, especially the peak-valley difference of the load, by optimizing the charging scheduling strategy of the electric vehicle. This makes the load of the microgrid more balanced during the peak charging period of electric vehicles, avoiding the risk of grid overload when the charging load is concentrated, thereby improving the operation stability of the microgrid. In the charging scheduling process, a global and local search strategy is combined. By dynamically adjusting the search ability control parameter, it is ensured that both the breadth of global search and the accuracy of local search can be maintained during the charging scheduling process, so as to find the optimal solution more efficiently.

[0100] Figure 4 An example of the entity structure diagram of an electronic device is shown in Figure 4As shown in the figure, the electronic device may include: a processor 401, a communications interface 402, a memory 403, and a communication bus 404. Among them, the processor 401, the communications interface 402, and the memory 403 complete communication with each other through the communication bus 404. The processor 401 can call the logical instructions in the memory 403 to execute an electric vehicle charging scheduling method based on an improved red-billed blue magpie algorithm. The method includes: Step 1: Define a scheduling period and a minimum unit charging duration; Step 2: Initialize the population and algorithm parameters. Among them, an individual in the population represents a charging decision vector of an electric vehicle. The element value of the charging decision vector represents the charging power of the corresponding unit charging time period. The dimension of the charging decision vector is determined by the scheduling period and the minimum unit charging duration. The algorithm parameters include the number of electric vehicles, charging power constraint conditions, and microgrid load constraint conditions; Step 3: Use the peak-valley load difference of the microgrid as the fitness function, and update the fitness value of the current population according to the fitness function; Step 4: Update the search ability control parameters. For each individual, control the individual to select the corresponding group behavior for position update according to the search ability control parameters. The group behavior includes large group behavior and small group behavior; Step 5: For each updated individual, perform boundary processing and constraint processing to make the updated population meet the charging power constraint conditions and the microgrid load constraint conditions; Step 6: Update the fitness value of the current population, and divide the individuals in the current population into elite individuals, sub-elite individuals, and non-elite individuals according to the fitness value; Step 7: For elite individuals, sub-elite individuals, and non-elite individuals, respectively use different perturbation strategies to perturb each type of individual for position update; and compare the individual fitness values before and after the update, and record the better fitness value and the corresponding better solution; Step 8: Repeat Steps 4 to 7 until a preset iteration stop condition is reached. At this time, the position of each individual in the population is the optimal charging decision of the corresponding electric vehicle.

[0101] In addition, when the logic instructions in the above-mentioned memory 403 are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0102] An embodiment of the present invention further provides a computer program product. The computer program product includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the electric vehicle charging scheduling method based on the improved red-billed blue magpie algorithm provided by each of the above method embodiments.

[0103] An embodiment of the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the electric vehicle charging scheduling method based on the improved red-billed blue magpie algorithm provided by each of the above method embodiments.

[0104] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on such an understanding, the technical solution, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disc, etc., and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0105] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of each embodiment of the present invention.

Claims

1. An electric vehicle charging scheduling method based on an improved red-billed blue magpie algorithm, characterized in that: include: Step 1: Define the scheduling cycle and minimum unit charging time; Step 2: Initialize the population and algorithm parameters; wherein an individual in the population represents a charging decision vector of an electric vehicle, the element value of the charging decision vector represents the charging power of the corresponding unit charging time period, and the dimension of the charging decision vector is determined by the scheduling cycle and the minimum unit charging time; the algorithm parameters include the number of electric vehicles, charging power constraints and microgrid load constraints; Step 3: Taking the peak-valley load difference of the microgrid as a fitness function, and updating the fitness value of the current population according to the fitness function; Step 4: Update the search capability control parameter, and for each individual, control the individual to select a corresponding group behavior to perform location update according to the search capability control parameter; the group behavior includes a large group behavior and a small group behavior; Step 5: For each updated individual, boundary processing and constraint processing are performed so that the updated population satisfies the charging power constraint condition and the microgrid load constraint condition; Step 6: Update the fitness value of the current population, and divide the individuals in the current population into elite individuals, sub-elite individuals and non-elite individuals according to the fitness value; Step 7: For elite individuals, sub-elite individuals, and non-elite individuals, use different perturbation strategies to perturb each type of individuals to update their positions; compare the individual fitness values ​​before and after the update, and record the better fitness value and the corresponding better solution; Step 8: Repeat steps 4 to 7 until the preset iteration stop condition is reached. At this time, each individual position in the population is the optimal charging decision for the corresponding electric vehicle.

2. The electric vehicle charging scheduling method based on the improved red-billed blue magpie algorithm according to claim 1 is characterized in that: The update search capability control parameters specifically include: Among them, ∈ is the search capability control parameter, t is the current iteration number, and T is the maximum iteration number.

3. The electric vehicle charging scheduling method based on the improved red-billed blue magpie algorithm according to claim 2 is characterized in that: For each individual, the search capability control parameter is used to control the individual to select a corresponding group behavior to perform location update, specifically including: A random number is generated within (0,1). If the random number is smaller than the search capability control parameter, a small group behavior is selected; otherwise, a large group behavior is selected.

4. The electric vehicle charging scheduling method based on the improved red-billed blue magpie algorithm according to claim 1 is characterized in that: In step 7, different perturbation strategies are used to perturb elite individuals, sub-elite individuals, and non-elite individuals to update their positions, including: Add a larger disturbance to the elite individuals; add a smaller disturbance to the sub-elite individuals; for non-elite individuals, select an elite individual for position linear combination and then add random disturbance; For all individuals, compare the fitness values ​​of the individuals before and after the disturbance, and retain the individual positions corresponding to the smaller fitness values.

5. The electric vehicle charging scheduling method based on the improved red-billed blue magpie algorithm according to claim 4 is characterized in that: Before comparing the fitness values ​​of individuals before and after the disturbance for all individuals, the method also includes: generating a random number in (0,1); if the random number is less than a preset threshold, the non-elite individuals after the disturbance are disturbed again according to the global optimal solution of the population and the average position of the small group, so that the non-elite individuals move closer to the global optimal solution.

6. The electric vehicle charging scheduling method based on the improved red-billed blue magpie algorithm according to claim 5 is characterized in that: According to the global optimal solution of the population and the average position of the small group, the non-elite individuals after disturbance are disturbed again, including: Among them, ω represents the weight, t is the current iteration number, T is the maximum iteration number, represents the position after disturbance, X i represents the position before the disturbance, rand1 and rand2 represent the generated random numbers in (0,1), and X food represents the global optimal solution of the population, X pmean Represents the average position of the small group.

7. The electric vehicle charging scheduling method based on the improved red-billed blue magpie algorithm according to any one of claims 1 to 6, characterized in that: In step 2, the Sobol sequence is used to initialize the population.

8. An electric vehicle charging scheduling device based on an improved red-billed blue magpie algorithm, characterized in that: include: The setting module is used to define the scheduling cycle and the minimum unit charging time; An initialization module is used to initialize the population and algorithm parameters; wherein an individual in the population represents a charging decision vector of an electric vehicle, the element value of the charging decision vector represents the charging power of the corresponding unit charging time period, and the dimension of the charging decision vector is determined by the scheduling cycle and the minimum unit charging time; the algorithm parameters include the number of electric vehicles, charging power constraints and microgrid load constraints; A fitness calculation module, used to take the peak-valley load difference of the microgrid as a fitness function, and update the fitness value of the current population according to the fitness function; A location update module is used to update the search capability control parameters, and for each individual, the individual is controlled to select the corresponding group behavior for location update according to the search capability control parameters; the group behavior includes large group behavior and small group behavior; for each individual after update, boundary processing and constraint processing are performed so that the updated population satisfies the charging power constraint condition and the microgrid load constraint condition; the individuals in the current population are divided into elite individuals, sub-elite individuals and non-elite individuals according to the fitness value; and for elite individuals, sub-elite individuals and non-elite individuals, different perturbation strategies are used to perturb each type of individual for location update; and the individual fitness values ​​before and after the update are compared, and the better fitness value and the corresponding better solution are recorded; The judgment module is used to judge whether the preset iteration stop condition is reached. If so, the position of each individual in the population at this time is used as the optimal charging decision of the corresponding electric vehicle.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method according to any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images