Two-stage charging scheduling method and system considering behavior preference and uncertainty

Abstract

The invention provides a two-stage charging scheduling method and system considering behavior preference and uncertainty, and the method comprises the steps: determining a scheduling time period, and constructing an electric vehicle charging scheduling multi-objective optimization model; analyzing the user charging behavior data, and extracting user charging behavior preferences; classifying the users according to the charging behavior preferences of the users; the similarity of environmental factors is considered, historical optimization scheduling results are selected, and reference charging scheduling modes of different types of users are determined; mapping a reference charging scheduling mode to a user individual scheduling scheme based on deep learning to obtain a day-ahead charging scheduling scheme after collaborative optimization; and considering the uncertainty of the charging time of the user, adopting a time-ahead charging scheduling method, and presetting time to adjust the day-ahead charging scheduling scheme after collaborative optimization to obtain a final charging scheduling scheme. According to the invention, the minimization of the charging cost, the improvement of the operation and maintenance efficiency and the balance of the power grid load can be realized.

Description

Technical Field

[0001] The present invention belongs to the field of charging scheduling, and in particular relates to a two-stage charging scheduling method and system taking into account behavioral preference and uncertainty. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] With the global energy transformation and the improvement of environmental awareness, electric vehicles, as representatives of new energy vehicles, are gradually becoming an important choice for future transportation. However, the popularity of electric vehicles has also brought new challenges, especially in densely populated areas such as urban residential areas. How to efficiently and economically manage the charging needs of a large number of electric vehicles has become an urgent problem to be solved.

[0004] In addition, grid load management, the complexity of user charging behavior, the limitations of charging facilities, economic and environmental impacts, demand-side response, and the integration of new energy are all key factors that need to be considered in charging scheduling. Existing charging scheduling technologies often rely on simplified models and algorithms, which are difficult to adapt to complex and changing actual needs. Summary of the invention

[0005] In order to solve the above problems, the present invention proposes a two-stage charging scheduling method and system taking into account behavioral preferences and uncertainties. The present invention optimizes the charging time distribution and charging amount distribution of electric vehicles by introducing charging scheduling in the two stages of day-ahead and day-ahead, combining deep learning with genetic algorithms, thereby minimizing the charging cost, improving the operation and maintenance efficiency, and balancing the grid load.

[0006] According to some embodiments, the present invention adopts the following technical solutions:

[0007] A two-stage charging scheduling method taking into account behavioral preference and uncertainty includes the following steps:

[0008] Determine the scheduling time period and build a multi-objective optimization model for electric vehicle charging scheduling;

[0009] Analyze user charging behavior data and extract user charging behavior preferences;

[0010] Categorize users based on their charging behavior preferences;

[0011] Considering the similarity of environmental factors, historical optimization scheduling results are selected to determine the benchmark charging scheduling mode for different categories of users;

[0012] Based on deep learning, the benchmark charging scheduling mode is mapped to the user's individual scheduling plan to obtain the collaboratively optimized day-ahead charging scheduling plan;

[0013] Taking into account the uncertainty of user charging time, the method of time-ahead charging scheduling is adopted. The time is set in advance to adjust the collaboratively optimized day-ahead charging scheduling plan to obtain the final charging scheduling plan.

[0014] As an optional implementation, the process of determining the scheduling time period is to determine the scheduling time period based on the overall power demand forecast result of the electric vehicle and the grid demand side response requirement.

[0015] As an optional implementation, the process of constructing a multi-objective optimization model for electric vehicle charging scheduling includes: the objective function of the multi-objective optimization model includes minimizing charging operation and maintenance costs, minimizing peak loads, minimizing carbon emissions, maximizing the amount of renewable energy power generation consumed, and maximizing the amount of overall load reduction during the demand response period;

[0016] The constraints of the model include electric vehicle charging capacity constraints, residential charging load upper limit constraints, user latest departure time constraints and demand side response requirement constraints.

[0017] As an optional implementation, the user charging behavior data includes arrival time, charging time, power outage time, departure time and charging amount data.

[0018] As an optional implementation, the scheduling mean of different categories of users is selected as the benchmark charging scheduling mode for the users in that category.

[0019] As an optional implementation, the process of mapping the benchmark charging scheduling mode to the user's individual scheduling plan based on deep learning includes:

[0020] Based on the deep learning algorithm, the charging behavior of individual users is predicted, and the initial population for charging scheduling optimization of different categories of users is obtained;

[0021] Determine whether the initial population meets the requirements of the multi-objective optimization model. If so, output the results of all categories of populations as the residential area electric vehicle scheduling optimization results, otherwise proceed to the next step;

[0022] Select, crossover and mutate different types of populations to obtain new populations;

[0023] Determine whether the new population meets the requirements of the multi-objective optimization function. If so, use the results of all categories of populations as the optimization results of electric vehicle scheduling in residential areas. Otherwise, return to the previous step.

[0024] As an optional implementation, when the uncertainty of the user's charging time is taken into account, the following users are targeted: users who return home early and are forced to charge in advance, users who return home early but are not forced to charge in advance, and users who postpone charging.

[0025] As a further limited implementation method, for users who return home early and are forced to charge in advance, their charging requirements are taken into consideration and their charging scheduling plan is replanned;

[0026] For users who return home early and are not required to charge in advance, the charging scheduling plan remains unchanged;

[0027] For users who postpone charging, the scheduling plan after the home time remains unchanged. The unmet power demand in the time period from the predicted home time to the actual home time is statistically calculated, and charging scheduling for the unmet power demand is carried out in the next scheduling time period.

[0028] As a further limited implementation method, the forward charging scheduling optimization model taking into account the uncertainty of the user's charging time is:

[0029]

[0030] Where n0, n1, …, n k ,…,n K To postpone the dispatch of the car number, t k0 ,t k1 ,…,t kh ,…t kH is the charging time period of electric vehicle k, s0, s1, …, s m ,…,s M To force users to charge in advance, t m0 ,t m1 ,…,t me ,t mE is the charging time period of the electric vehicle q, θ is the set of schedulable time periods;

[0031] The model is proposed to be solved by methods such as second-order cone programming.

[0032] As a further limited implementation, the constraints of the forward charging scheduling optimization model taking into account the uncertainty of the user's charging time include:

[0033]

[0034] In the formula, To postpone charging, the charging schedule of electric vehicle i has met the required power, P j To force the electric vehicle j to charge in advance, is the charging power of car i that is delayed for charging during time period t, is the charging power of vehicle j that is forced to charge ahead of schedule during time period t, is the maximum power allowed by the charging circuit in time period t, The charging power occupied by the charging line during time period t.

[0035] A two-stage charging scheduling system considering behavioral preference and uncertainty, comprising:

[0036] a multi-objective optimization module configured to determine a scheduling time period and construct a multi-objective optimization model for electric vehicle charging scheduling;

[0037] A behavior preference extraction module is configured to analyze the user's charging behavior data, extract the user's charging behavior preferences, and classify the users according to their charging behavior preferences;

[0038] A charging scheduling mode determination module is configured to consider the similarity of environmental factors, select historical optimization scheduling results, and determine the benchmark charging scheduling mode for different categories of users;

[0039] The day-ahead collaborative optimization module is configured to map the benchmark charging scheduling mode to the user's individual scheduling plan based on deep learning, and obtain the collaboratively optimized day-ahead charging scheduling plan;

[0040] The pre-time optimization scheduling module is configured to take into account the uncertainty of user charging time, adopt the pre-time charging scheduling method, pre-set the time to adjust the collaboratively optimized day-ahead charging scheduling plan, and obtain the final charging scheduling plan.

[0041] A computer-readable storage medium is used to store computer instructions, and when the computer instructions are executed by a processor, the steps in the above method are completed.

[0042] An electronic device comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps in the above method are completed.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] 1. The present invention adopts a two-stage charging scheduling strategy, namely, the day-ahead and the hour-ahead, to effectively deal with the uncertainty of the user's return home and charging time, and improve the flexibility and reliability of charging scheduling.

[0045] 2. The present invention achieves comprehensive optimization of charging scheduling for electric vehicles in residential areas by establishing an optimization model with multiple objectives, including charging operation and maintenance costs, peak load reduction, carbon emission reduction, new energy power consumption, and demand response.

[0046] 3. The present invention uses deep learning methods to map charging scheduling modes of different charging behavior preference types to individual charging scheduling plans, which improves the solution efficiency and reduces the randomness of the initial solution domain; combined with genetic algorithms to achieve collaborative optimization solutions among different categories of populations, further improving the solution efficiency of charging scheduling.

[0047] 4. The present invention uses a pre-time charging scheduling method to adjust the charging scheduling plan in real time every half an hour or one hour to adapt to the user's return home and charging time changes.

[0048] 5. While taking into account the uncertainty of user charging time, the present invention mainly considers economic goals, and realizes the maximization of cost-effectiveness by optimizing models and constraints. It adopts efficient algorithms such as second-order cone programming to solve the optimization scheduling model, ensuring the accuracy and efficiency of the model solution. In addition, in the optimization model, it comprehensively considers constraints such as the charging amount of electric vehicles, the upper limit of residential charging load, the latest departure time of users, and the response requirements on the demand side, ensuring the feasibility of the charging scheduling plan.

[0049] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0051] Figure 1 is a schematic diagram of a two-stage charging scheduling process taking into account behavioral preference and uncertainty in an embodiment;

[0052] Figure 2 The present invention is a schematic diagram of a multi-category collaborative optimization day-ahead charging scheduling process for electric vehicles considering charging behavior preferences according to an embodiment. DETAILED DESCRIPTION

[0053] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0054] It should be noted that the following detailed descriptions are all illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0055] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.

[0056] In the absence of conflict, the embodiments in this application and the features in the embodiments may be combined with each other.

[0057] Embodiment 1

[0058] A two-stage charging scheduling method taking into account behavioral preference and uncertainty includes the following steps:

[0059] Determine the scheduling time period and build a multi-objective optimization model for electric vehicle charging scheduling;

[0060] Analyze user charging behavior data and extract user charging behavior preferences;

[0061] Categorize users based on their charging behavior preferences;

[0062] Considering the similarity of environmental factors, historical optimization scheduling results are selected to determine the benchmark charging scheduling mode for different categories of users;

[0063] Based on deep learning, the benchmark charging scheduling mode is mapped to the user's individual scheduling plan to obtain the collaboratively optimized day-ahead charging scheduling plan;

[0064] Taking into account the uncertainty of user charging time, the method of time-ahead charging scheduling is adopted. The time is set in advance to adjust the collaboratively optimized day-ahead charging scheduling plan to obtain the final charging scheduling plan.

[0065] A detailed introduction is given below.

[0066] First, if Figure 1 As shown in the figure, the charging preferences of group users often show diversity, while the charging behavior preferences of single users show uncertainty. In order to realize the joint analysis and modeling of user charging behaviors of groups and individuals, the unsupervised model (clustering) and the supervised model (deep neural network) are combined to realize the joint analysis and modeling of user charging behaviors of groups and individuals.

[0067] Firstly, the multi-category coordinated day-ahead charging scheduling optimization of electric vehicles is carried out considering charging behavior preferences.

[0068] The optimization of the day-ahead charging scheduling of electric vehicles in residential areas should, on the one hand, fully consider the various needs and constraints of residential areas and establish a multi-objective optimization model for charging scheduling; on the other hand, a fast solution method for the scheduling optimization model should be constructed.

[0069] The charging behaviors of users in residential communities have group similarities. The user charging behavior preferences are integrated into the solution of the multi-objective optimization model of charging scheduling in residential communities. Charging scheduling modes with different charging behavior preference types are defined based on experience / historical data. Deep learning methods are used to map different charging scheduling modes to individual charging scheduling plans, generate initial solution domains of different categories, and solve the problem of high randomness of the initial solution domains. Genetic algorithms are further combined to achieve collaborative optimization solutions between different categories of populations (solution domains), ultimately achieving the goal of improving solution efficiency.

[0070] The optimization process of multi-category coordinated day-ahead charging scheduling of electric vehicles considering charging behavior preferences is as follows: Figure 2 shown.

[0071] The specific steps include:

[0072] (1) Constructing a multi-objective optimization model for short-term charging scheduling of electric vehicles

[0073] Before conducting short-term charging scheduling for electric vehicles, the scheduling time period must be determined. The definition of this time period is determined based on the overall power demand forecast results of electric vehicles and the demand-side response requirements of the power grid.

[0074] Minimizing the overall charging operation and maintenance cost is one of the goals of short-term charging scheduling for electric vehicles in residential areas. Its calculation mainly considers factors such as charging electricity charges and other operation and maintenance costs. In order to obtain the lowest charging operation and maintenance cost, this embodiment uses 30 minutes as the charging interval and establishes the charging operation and maintenance cost objective function as follows:

[0075]

[0076] Where i is the car number; t is the 15-minute interval number, t=1 is the first 30-minute scheduling time period, and t=2 is the second scheduling time period; P i t is the power demand of car i at time t; C t is the electricity price at time t, O t is other operation and maintenance cost at time t.

[0077] Furthermore, we hope to minimize the peak load and establish the objective function as follows:

[0078]

[0079] Where P t is the total power demand in time period t after scheduling optimization; P is the peak load threshold (P t When it is greater than P, it is defined as peak load).

[0080] Expected minimum carbon emissions, the objective function is:

[0081]

[0082] In the formula, Q t is the carbon emission in time period t.

[0083] Considering absorbing as much renewable energy power generation as possible, the objective function is:

[0084]

[0085] Where Wt is the amount of renewable energy electricity consumed by electric vehicles in time period t.

[0086] It is expected that the overall load will be reduced as much as possible during the demand response period. The objective function is:

[0087]

[0088] In the formula, is the predicted power demand of electric vehicles in time period t.

[0089] Other objective functions will be further studied and summarized during the implementation of the project.

[0090] Electric vehicle charging in residential areas must meet multiple constraints, such as the amount of electric vehicle charging, the upper limit of residential charging load, the latest time for users to leave, and demand-side response requirements. The constraints are as follows:

[0091]

[0092] In the formula, is the demand side response, t s -t p The minimum load reduction required during the period, P i is the minimum power requirement of electric vehicle i, is the maximum load limit of electric vehicles in the residential area during time period t.

[0093] (2) Collaborative optimization solution combining charging behavior preference, deep learning and genetic algorithm

[0094] Define charging scheduling modes for different categories of users, use deep learning to map charging scheduling modes to individual user charging plans, generate the initial solution domain (initial population) for optimal scheduling, and further combine genetic algorithms to achieve multi-category collaborative optimization solutions for charging scheduling of electric vehicles in residential areas. The specific technical routes are as follows:

[0095] Step 1: Classify users according to their charging behavior preferences;

[0096] Step 2: Considering the similarity of weather, working days, etc., select the historical optimization scheduling results and take the scheduling mean of different categories of users as the benchmark charging scheduling mode s for this category of users j ;

[0097] Step 3: Based on deep learning, the benchmark charging scheduling mode is mapped to the user's individual scheduling plan, and the charging behavior of individual users is predicted based on the deep learning algorithm. The prediction process is as follows:

[0098]

[0099] In the formula, is the charging scheduling plan for car i of the jth category, g() is the mapping method from the deep learning-based benchmark charging mode to the individual charging scheduling plan for the migration application. Through this step, the initial population for charging scheduling optimization of different categories of users is obtained;

[0100] Step 5: Determine whether the initial population meets the requirements of the multi-objective optimization model. If so, output the results of all categories of populations as the residential area electric vehicle scheduling optimization results. Otherwise, proceed to step 6.

[0101] Step 6: Select, crossover and / or mutate different types of populations to obtain new populations;

[0102] Step 7: Determine whether the new population meets the requirements of the multi-objective optimization function. If so, output the results of all categories of populations as the residential area electric vehicle scheduling optimization results. Otherwise, proceed to step 6.

[0103] Some embodiments also include a comparative analysis of the convergence and solution efficiency of the optimization results of the proposed method and traditional genetic algorithms, particle swarm optimization methods, etc.

[0104] Next, forward charging scheduling is performed taking into account the uncertainty of user charging time.

[0105] Considering the uncertainty of the user's return home and charging time, the method of pre-time charging scheduling is adopted to make real-time adjustments to the charging scheduling plan every half an hour or an hour, and adjust the charging scheduling plan through local fine-tuning. Three types of users are mainly considered: users who return home early and are forced to charge in advance, users who return home early but are not forced to charge in advance, and users who postpone charging. The detailed plan is as follows:

[0106] For users who return home early and are forced to charge in advance, their charging requirements will be taken into consideration and their charging scheduling will be re-planned;

[0107] For users who return home early and are not required to charge in advance, the charging scheduling plan remains unchanged;

[0108] For users who postpone charging, the scheduling plan after the home time remains unchanged. The unmet power demand in the time period from the predicted home time to the actual home time is statistically calculated, and charging scheduling for the unmet power demand is carried out in the next scheduling time period.

[0109] For the uncertain user charging scheduling, the economic goal is mainly considered. The forward charging scheduling optimization model taking into account the uncertainty of user charging time is as follows:

[0110]

[0111] Where n0, n1, …, n k ,…,n KTo postpone the dispatch of the car number, t k0 ,t k1 ,…,t kh ,…t kH is the charging time period of electric vehicle k, s0, s1, …, s m ,…,s M To force users to charge in advance, t m0 ,t m1 ,…,t me ,t mE is the charging time period of electric vehicle q, θ is the set of schedulable time periods.

[0112] The above optimization scheduling model should satisfy the following constraints:

[0113]

[0114] In the formula, To postpone charging, the charging schedule of electric vehicle i has met the required power, P j To force the electric vehicle j to charge in advance, is the charging power of car i that is delayed for charging during time period t, is the charging power of vehicle j that is forced to charge ahead of schedule during time period t, is the maximum power allowed by the charging circuit in time period t, The charging power occupied by the charging line during time period t.

[0115] The above model is proposed to be solved by methods such as second-order cone programming.

[0116] Embodiment 2

[0117] A two-stage charging scheduling system considering behavioral preference and uncertainty, comprising:

[0118] a multi-objective optimization module configured to determine a scheduling time period and construct a multi-objective optimization model for electric vehicle charging scheduling;

[0119] A behavior preference extraction module is configured to analyze the user's charging behavior data, extract the user's charging behavior preferences, and classify the users according to their charging behavior preferences;

[0120] A charging scheduling mode determination module is configured to consider the similarity of environmental factors, select historical optimization scheduling results, and determine the benchmark charging scheduling mode for different categories of users;

[0121] The day-ahead collaborative optimization module is configured to map the benchmark charging scheduling mode to the user's individual scheduling plan based on deep learning, and obtain the collaboratively optimized day-ahead charging scheduling plan;

[0122] The pre-time optimization scheduling module is configured to take into account the uncertainty of user charging time, adopt the pre-time charging scheduling method, pre-set the time to adjust the collaboratively optimized day-ahead charging scheduling plan, and obtain the final charging scheduling plan.

[0123] It can be understood that the above-mentioned units / modules can be separately or completely combined into one or several other units / modules, or one (some) of the units can be further divided into multiple functionally smaller units, which can achieve the same operation without affecting the realization of the technical effects of the embodiments of the present application.

[0124] The above modules of the system are divided based on logical functions. In practical applications, the function of one module can also be implemented by multiple modules, or the functions of multiple modules can be implemented by one module. For example, the day-ahead collaborative optimization module of this embodiment may include:

[0125] The individual prediction module is configured to predict the charging behavior of individual users based on a deep learning algorithm to obtain the initial population for optimizing charging scheduling for different categories of users;

[0126] The first judgment module is configured to judge whether the initial population meets the requirements of the multi-objective optimization model. If so, the results of all categories of populations are output as the residential area electric vehicle scheduling optimization results, otherwise, the next module is entered;

[0127] The population update module is configured to select, cross and mutate different categories of populations to obtain new populations;

[0128] The second judgment module is configured to judge whether the new population meets the requirements of the multi-objective optimization function. If so, the results of all categories of populations are used as the residential area electric vehicle scheduling optimization results, otherwise, it returns to the previous module.

[0129] Similarly, in other embodiments of the present application, the system may also include other units / modules. In actual applications, these functions may also be implemented with the assistance of other units, and may be implemented by collaboration of multiple units.

[0130] According to another embodiment of the present application, the system described in this embodiment can be constructed and the method of Embodiment 1 can be implemented by running a computer program (including program code) capable of executing the steps involved in the corresponding method described in Embodiment 1 on a general-purpose computing device such as a computer, which includes processing elements and storage elements such as a central processing unit (CPU), random access memory (RAM), and read-only memory (ROM). The computer program can be recorded on, for example, a computer-readable recording medium, and loaded into the above-mentioned computing device through the computer-readable recording medium and run therein.

[0131] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0132] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0133] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0134] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0135] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principle of the present invention without creative labor shall be included in the protection scope of the present invention.

Claims

1. A two-stage charging scheduling method taking into account behavioral preferences and uncertainties, characterized in that: The following steps are involved: Determine the scheduling time period and build a multi-objective optimization model for electric vehicle charging scheduling; Analyze user charging behavior data and extract user charging behavior preferences; Categorize users based on their charging behavior preferences; Considering the similarity of environmental factors, historical optimization scheduling results are selected to determine the benchmark charging scheduling mode for different categories of users; Based on deep learning, the benchmark charging scheduling mode is mapped to the user's individual scheduling plan to obtain the collaboratively optimized day-ahead charging scheduling plan; Taking into account the uncertainty of user charging time, the method of day-ahead charging scheduling is adopted. The time is set in advance to adjust the collaboratively optimized day-ahead charging scheduling plan to obtain the final charging scheduling plan.

2. A two-stage charging scheduling method taking into account behavioral preference and uncertainty as claimed in claim 1, characterized in that: The process of determining the dispatching time period is to determine the dispatching time period based on the overall power demand forecast results of electric vehicles and the grid demand side response requirements.

3. A two-stage charging scheduling method taking into account behavioral preference and uncertainty as claimed in claim 1, characterized in that: The process of constructing a multi-objective optimization model for electric vehicle charging scheduling includes: the objective functions of the multi-objective optimization model include minimizing charging operation and maintenance costs, minimizing peak loads, minimizing carbon emissions, maximizing the amount of renewable energy power generation consumed, and maximizing the amount of overall load reduction during the demand response period; The constraints of the model include electric vehicle charging capacity constraints, residential charging load upper limit constraints, user latest departure time constraints and demand side response requirement constraints.

4. A two-stage charging scheduling method taking into account behavioral preference and uncertainty as claimed in claim 1, characterized in that: The user charging behavior data includes arrival time, charging time, power outage time, departure time and charging amount data.

5. A two-stage charging scheduling method taking into account behavioral preference and uncertainty as claimed in claim 1, characterized in that: The dispatch mean of different categories of users is selected as the benchmark charging dispatch mode for this category of users.

6. A two-stage charging scheduling method taking into account behavioral preference and uncertainty as claimed in claim 1, characterized in that: The process of mapping the benchmark charging scheduling mode to the user's individual scheduling plan based on deep learning includes: Based on the deep learning algorithm, the charging behavior of individual users is predicted, and the initial population for charging scheduling optimization of different categories of users is obtained; Determine whether the initial population meets the requirements of the multi-objective optimization model. If so, output the results of all categories of populations as the residential area electric vehicle scheduling optimization results, otherwise proceed to the next step; Select, crossover and mutate different types of populations to obtain new populations; Determine whether the new population meets the requirements of the multi-objective optimization function. If so, use the results of all categories of populations as the optimization results of electric vehicle scheduling in residential areas. Otherwise, return to the previous step.

7. A two-stage charging scheduling method taking into account behavioral preference and uncertainty as claimed in claim 1, characterized in that: When considering the uncertainty of user charging time, the following users are targeted: users who return home early and are forced to charge in advance, users who return home early but are not forced to charge in advance, and users who postpone charging; For users who return home early and are forced to charge in advance, their charging requirements are considered and their charging scheduling plan is re-planned; For users who return home early and are not required to charge in advance, the charging scheduling plan remains unchanged; For users who postpone charging, the scheduling plan after the home time remains unchanged. The unmet power demand in the time period from the predicted home time to the actual home time is statistically calculated, and charging scheduling for the unmet power demand is carried out in the next scheduling time period.

8. A two-stage charging scheduling method taking into account behavioral preference and uncertainty as claimed in claim 1, characterized in that: The forward charging scheduling optimization model considering the uncertainty of user charging time is: Where n0, n1, …, n k ,…,n K To postpone the dispatch of the car number, t k0 ,t k1 ,…,t kh ,…t kH is the charging time period of electric vehicle k, s0, s1, …, s m ,…,s M To force users to charge in advance, t m0 ,t m1 ,…,t me ,t mE is the charging time period of the electric vehicle q, θ is the set of schedulable time periods; The model is proposed to be solved by methods such as second-order cone programming.

9. A two-stage charging scheduling method taking into account behavioral preference and uncertainty as claimed in claim 8, characterized in that: The constraints of the forward charging scheduling optimization model taking into account the uncertainty of user charging time include: In the formula, To postpone charging, the charging schedule of electric vehicle i has met the required power, P j To force the electric vehicle j to charge in advance, is the charging power of car i that is delayed for charging during time period t, is the charging power of vehicle j that is forced to charge ahead of schedule during time period t, is the maximum power allowed by the charging circuit in time period t, The charging power occupied by the charging line during time period t.

10. A two-stage charging scheduling system taking into account behavioral preferences and uncertainties, characterized in that: include: a multi-objective optimization module configured to determine a scheduling time period and construct a multi-objective optimization model for electric vehicle charging scheduling; A behavior preference extraction module is configured to analyze the user's charging behavior data and extract the user's charging behavior preference; Categorize users based on their charging behavior preferences; A charging scheduling mode determination module is configured to consider the similarity of environmental factors, select historical optimization scheduling results, and determine the benchmark charging scheduling mode for different categories of users; The day-ahead collaborative optimization module is configured to map the benchmark charging scheduling mode to the user's individual scheduling plan based on deep learning, and obtain the collaboratively optimized day-ahead charging scheduling plan; The pre-time optimization scheduling module is configured to take into account the uncertainty of user charging time, adopt the pre-time charging scheduling method, pre-set the time to adjust the collaboratively optimized day-ahead charging scheduling plan, and obtain the final charging scheduling plan.

Images