Multi-region electric vehicle charging load prediction method and device capable of protecting privacy

Abstract

The invention discloses a privacy-protecting multi-region electric vehicle charging load prediction method and device, and belongs to the technical field of information security, and the method comprises the steps: dividing the obtained travel destination data of an electric vehicle into a plurality of regions; constructing an electric vehicle trip chain according to the plurality of divided areas; according to the constructed electric vehicle trip chain, constructing a Markov chain of a driving state; constructing a state transition probability matrix according to the constructed Markov chain; adding noise meeting the differential privacy definition to each probability value in the state transition probability matrix to obtain a state transition probability matrix after the noise is added; according to a probability value in the state transition probability matrix after the noise is added, predicting a next travel destination and a charging load of the electric vehicle; according to the method, the charging load of the electric vehicle is accurately predicted while the travel privacy of residents is protected, so that the configuration of charging facilities in different areas and the intelligent scheduling of a power distribution network can be optimized.

Description

Technical Field

[0001] The invention relates to a privacy-protected multi-region electric vehicle charging load prediction method and device, belonging to the technical field of information security. Background Art

[0002] As a representative of clean energy transportation, electric vehicles have received widespread attention in recent years. Electric vehicle charging load forecasting has become an important issue in the research of electric vehicle access analysis, distribution network dispatching and control, and orderly charging of residents. Electric vehicle charging load forecasting aims to analyze and model the charging behavior of electric vehicle users and predict the charging demand of the power grid in a certain period of time in the future. This process needs to consider multiple factors, including residents' driving destinations, charging habits, vehicle battery capacity and remaining power, distribution of charging infrastructure, and time and other external environmental variables. At present, commonly used charging load forecasting methods mainly include traditional time series analysis, machine learning algorithms, and deep learning models.

[0003] The accuracy of charging load forecasting is directly related to the stability and safety of the power system, and also affects the charging experience of electric vehicle users. However, charging load forecasting usually requires the collection and analysis of a large amount of personal data, such as users' charging habits, travel patterns, and geographic location information. These data are not only large in quantity, multi-regional, and highly complex, but also involve user privacy. How to effectively protect user privacy while performing charging load forecasting has become a difficult problem that needs to be solved urgently. Summary of the invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and to provide a privacy-preserving multi-region electric vehicle charging load prediction method and device, so as to protect the privacy of residents' travel while accurately predicting the charging load of electric vehicles, thereby optimizing the configuration of charging facilities in different regions and the intelligent scheduling of distribution networks.

[0005] To achieve the above object, the present invention is implemented by adopting the following technical solutions: In a first aspect, the present invention provides a privacy-preserving multi-region electric vehicle charging load prediction method, comprising: Obtaining electric vehicle travel destination data; Divide the acquired electric vehicle travel destination data into multiple regions; Build an electric vehicle travel chain based on the divided areas; According to the constructed electric vehicle travel chain, a Markov chain of driving status is constructed; According to the constructed Markov chain, a state transition probability matrix including the probability values ​​of transitioning from the current destination to the next destination is constructed; Add noise that satisfies the definition of differential privacy to each probability value in the state transition probability matrix to obtain the state transition probability matrix after adding noise; According to the probability values ​​in the state transition probability matrix after adding noise, the next destination and charging load of the electric vehicle are predicted.

[0006] Furthermore, the obtained electric vehicle travel destination data is divided into multiple areas, including: The acquired electric vehicle travel destination data is divided into residential area H, work area W, commercial area P and other areas O.

[0007] Furthermore, the electric vehicle travel chain is constructed according to the divided multiple areas, including: According to the residential area H, work area W, commercial area P and other areas O, an electric vehicle travel chain is constructed. …;in, Indicates the current driving location. Indicates the next travel destination, where the travel chain has no more than four destinations.

[0008] Furthermore, the Markov chain of driving state is constructed according to the constructed electric vehicle travel chain, including: Construct a Markov chain of driving states: ,in, represents the probability of transferring from the current destination to the next destination, is the state transition probability matrix.

[0009] Furthermore, according to the constructed Markov chain, a state transition probability matrix including the probability value of transferring from the current destination to the next destination is constructed, and the formula is as follows: ; in, represents the probability value of always driving in residential areas, represents the probability value of moving from a residential area to a commercial area, represents the probability value of moving from the residential area to the work area, represents the probability value of moving from the residential area to other areas, represents the probability value of moving from a commercial area to a residential area, represents the probability value of always driving in the commercial area, represents the probability value of moving from the commercial area to the work area, represents the probability value of moving from the commercial area to other areas, represents the probability value of moving from the work area to the residential area, represents the probability value of moving from the work area to the commercial area, represents the probability value of always driving in the work zone, represents the probability value of transferring from the working area to other areas, represents the probability value of moving from other areas to residential areas, represents the probability value of transferring from other areas to the working area, represents the probability value of moving from other areas to the commercial area, Indicates the probability value of always driving in other areas; in, Any element in is greater than or equal to 0, The sum of the elements in each row of is 1. Every element in is not greater than 1.

[0010] Furthermore, adding noise that satisfies the definition of differential privacy to each probability value in the state transition probability matrix to obtain the state transition probability matrix after adding noise includes: Add noise that satisfies the definition of differential privacy to the state transfer matrix, and the noise scale The calculation formula is as follows: ; in, represents the privacy budget, represents the relaxation factor; State transition matrix after adding noise , the formula is as follows: ; in, represents the probability value of always driving in residential areas after adding noise, It represents the probability value of moving from residential area to commercial area after adding noise, represents the probability value of moving from the residential area to the work area after adding noise, represents the probability value of moving from residential areas to other areas after adding noise, It represents the probability value of moving from the commercial area to the residential area after adding noise, represents the probability value of always driving in the commercial area after adding noise, represents the probability value of moving from the commercial area to the working area after adding noise, represents the probability value of moving from the commercial area to other areas after adding noise, represents the probability value of moving from the work area to the residential area after adding noise, represents the probability value of moving from the work area to the commercial area after adding noise, represents the probability value of always driving in the work area after adding noise, It represents the probability value of transferring from the working area to other areas after adding noise, represents the probability value of moving from other areas to residential areas after adding noise, It represents the probability value of transferring from other areas to the working area after adding noise, represents the probability value of moving from other areas to the commercial area after adding noise, Represents the probability value of always driving in other areas after adding noise.

[0011] Furthermore, the privacy budget The value range is 0.5+0.05*λ, where λ is a natural number from 0 to 90.

[0012] Furthermore, the method of predicting the next destination and charging load of the electric vehicle according to the probability value in the state transition probability matrix after adding noise includes: According to the state transfer matrix after adding noise The probability value of the electric vehicle is used to predict the next destination of the electric vehicle. Get the departure time of electric vehicles T 0 、The current remaining power of the electric vehicle Value, judgment Is the value less than 30%? if , it is determined that the electric vehicle needs to be charged at the current destination. The formula for calculating the charging condition is as follows: , in For electric vehicles to reach the current The remaining battery power of each destination value; is the battery capacity of the electric vehicle; if , assuming that the electric vehicle has enough power to reach the next destination, starts driving to the next destination, and the time it reaches the next destination T 1 T 0 , is the travel time; calculate T 1 The charging load of electric vehicles at the moment is as follows: , in express T 1 The charging load of electric vehicles at the moment, is the charging power, Then T1 The moment n The charging status of an electric vehicle; =1, then n An electric car is being charged. =0, then n The car is not charging.

[0013] In a second aspect, the present invention provides a privacy-protecting multi-region electric vehicle charging load prediction device, comprising: A data acquisition module, used to acquire electric vehicle travel destination data; A data partitioning module, used for partitioning the acquired electric vehicle travel destination data into multiple regions; The first building module is used to build an electric vehicle travel chain according to the divided multiple areas; The second building module is used to build a Markov chain of driving status according to the built electric vehicle travel chain; A third construction module is used to construct a state transition probability matrix including probability values ​​of transitioning from a current destination to a next destination according to the constructed Markov chain; An adding module is used to add noise that satisfies the definition of differential privacy to each probability value in the state transition probability matrix to obtain the state transition probability matrix after adding noise; The prediction module is used to predict the next destination and charging load of the electric vehicle according to the probability value in the state transition probability matrix after adding noise.

[0014] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the aforementioned methods.

[0015] In a fourth aspect, the present invention provides a computer device, comprising: Memory, for storing computer programs / instructions; A processor is used to execute the computer program / instructions to implement the steps of any of the aforementioned methods.

[0016] In a fifth aspect, the present invention provides a computer program product, comprising a computer program / instruction, which implements the steps of any one of the aforementioned methods when executed by a processor.

[0017] Compared with the prior art, the present invention has the following beneficial effects: The present invention provides a privacy-protected multi-region electric vehicle charging load prediction method and device, which uses electric vehicle travel destination data; divides the acquired electric vehicle travel destination data into multiple regions; and constructs an electric vehicle travel chain according to the divided multiple regions; According to the constructed electric vehicle travel chain, a Markov chain of driving status is constructed, and noise that meets the definition of differential privacy is added to each probability value in the state transition probability matrix to obtain the state transition probability matrix after adding noise; according to the probability values ​​in the state transition probability matrix after adding noise, the next destination and charging load of the electric vehicle are predicted; thereby, it is possible to protect the privacy of residents' travel while accurately predicting the charging load of electric vehicles, thereby optimizing the configuration of charging facilities in different regions and the intelligent scheduling of distribution networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is a flow chart of a privacy-protected multi-region electric vehicle charging load prediction method provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0020] Embodiment 1: This embodiment introduces a privacy-preserving multi-region electric vehicle charging load prediction method, including: Obtaining electric vehicle travel destination data; Divide the acquired electric vehicle travel destination data into multiple regions; Build an electric vehicle travel chain based on the divided areas; According to the constructed electric vehicle travel chain, a Markov chain of driving status is constructed; According to the constructed Markov chain, a state transition probability matrix including the probability values ​​of transitioning from the current destination to the next destination is constructed; Add noise that satisfies the definition of differential privacy to each probability value in the state transition probability matrix to obtain the state transition probability matrix after adding noise; According to the probability values ​​in the state transition probability matrix after adding noise, the next destination and charging load of the electric vehicle are predicted.

[0021] like Figure 1 As shown, the privacy-preserving multi-region electric vehicle charging load prediction method provided in this embodiment specifically involves the following steps: S11. The data preprocessing process includes steps such as data review, cleaning, coding, conversion, missing value and outlier processing, etc. The standard software Stata is used for data statistical analysis, classification and storage of data; S12. Statistical analysis of daily trip times, recording the first trip time of electric vehicles, driving time of electric vehicles, parking time of electric vehicles, The value (the ratio of remaining power to full power), the initial starting point of the electric vehicle, the destination of the electric vehicle and other characteristic data are recorded and analyzed to obtain the distribution of different electric vehicle travel data: The first trip time of electric vehicles satisfies the normal distribution, and its probability density function is: , in is the expected value of travel time, is the standard deviation of travel time.

[0022] The driving time of electric vehicles satisfies the log-normal distribution, and its probability density function is: , The expected driving time of electric vehicles in different regions is different. and standard deviation Electric vehicles from residential areas Residential Area: ; From residential area Workspace: , ; From residential area Commercial area: ; From residential area Other areas: ; From the workspace Workspace: ; From the workspace Commercial area: ; From the workspace Other areas: ;From the business district Commercial area: ; From the business district Other areas: ; From other areas Other areas: .

[0023] The parking time of electric vehicles satisfies the normal distribution, and its probability density function is: , The expected parking time of electric vehicles in different areas is different. and standard deviation For residential areas: ; For the workspace: ; For commercial areas: ; For other areas: .

[0024] The initial power of the electric vehicle satisfies the normal distribution, and its probability density function is: , in is the expected value of the initial charge, is the standard deviation of the initial power; based on the existing probability density function, a large amount of sample data can be fitted for electric vehicle charging prediction according to these characteristic distributions, thereby improving the prediction accuracy.

[0025] S13. According to the distribution of travel data characteristics, the destination areas of residents' electric vehicle travel are divided into four areas: residential area (Home, H), work area (Work, W), commercial area (Public, P) and other areas (O), and an electric vehicle travel chain is constructed so that the destination of the travel chain does not exceed four at most, such as: , The vehicle stops at different destinations, so that charging may occur at different locations. In claim 1, the parking time is estimated as the charging time of the electric vehicle.

[0026] In step S2, the data is modeled using a Markov model, and a state transition probability matrix is ​​established based on the statistical results of the electric vehicle travel data, including the following process: S21. Markov model is a random process that studies the change of state sequence. The four driving destinations of electric vehicles are regarded as a state respectively. Then the driving of electric vehicles to different destinations can be regarded as a state transfer. Then the driving state of electric vehicles is a Markov transfer process. S22. Construct the Markov chain of driving state: ,in, Indicates the current location of the electric vehicle. Indicates the next driving destination. represents the probability of transferring from the current destination to the next destination; S23, construct a state transfer matrix, and construct a driving state transfer probability matrix from all possible states of the electric vehicle from the current location to the next destination: ; Driving state transition probability matrix It has the following characteristics: (1) Any element in is greater than or equal to 0; (2) The sum of the elements in each row is 1; (3) Each element in is not greater than 1; (4) From one state to another (such as ) and the reverse state transition probability (such as )No direct connection.

[0027] In step S3, personalized privacy protection is provided. Users choose from 91 different privacy protection levels. Each user submits a personalized privacy budget. Then, according to the privacy level, Gaussian noise that meets the definition of differential privacy is added to each probability value of the state transition probability matrix. The process includes the following: S31, for the constructed state transfer matrix ,Give Each probability value in is added with noise, where the privacy budget The value range is {0.5, 0.55, 0.60, 0.65, 0.70, 0.75, ···, 4.95, 5.0}, that is, the privacy budget The value range is 0.5+0.05* λ , λ is a natural number from 0 to 90, with a total of 91 values, each of which corresponds to a privacy protection level of differential privacy. λ =0, the privacy budget Take the minimum value of 0.5, when the privacy protection effect is the strongest; λ =90, privacy budget The maximum value is 5.0, at which the privacy protection effect is the weakest. Users can choose one of the 91 different privacy budgets according to their specific situation. That is, users have 91 different privacy protection levels to choose from. After the user selects the privacy budget, the privacy budget is used to generate Gaussian noise that meets the privacy protection requirements of the corresponding level, and the state transfer matrix Add noise to each probability value in and get the perturbed state transfer matrix : ; S32, -Definition of differential privacy: Suppose there is a random function is satisfied - Differentially private, yes The set of all possible outputs, and , , if and only if for any two adjacent data sets and , then , in represents the privacy budget, The smaller it is, the higher the level of privacy protection; represents the relaxation factor, which is the minimum parameter that can violate the probability ratio of differential privacy; generally speaking, or smaller, where n is the size of the dataset; S33, Advanced Combinatorial Theorem: For all , -Differential privacy mechanism in - Satisfaction under recombination - Differential privacy, where ; S34, Gaussian mechanism definition: Given a sensitive query function , is the domain of the dataset, is the number of records in the dataset, is a set of real numbers. For the input data set , Gaussian mechanism Defined as: , in, Is a query function sensitivity, is the noise scale, The mean is 0 and the variance is Gaussian distribution (normal distribution); in order to make the Gaussian mechanism satisfy - Differential privacy requires choosing an appropriate noise standard deviation; according to the probability density function of the Gaussian distribution, we can deduce ; S35. Parallel Combination Theorem: For Random Mechanism Working on disjoint datasets , whose privacy budget is , then the combined algorithm satisfy - Differential privacy; S36. According to the Gaussian mechanism definition, when Gaussian noise is added to the Markov state transfer matrix value, since the probability value range belongs to , then the sensitivity , , . The selection of will adaptively adjust the noise generation algorithm and parameters according to the user's privacy requirements and data distribution to achieve higher data utilization and strict privacy protection.

[0028] In step S4, the next destination and charging load of the resident electric vehicle are predicted according to the state transition probability matrix value after the disturbance, which specifically includes the following process: S41, the departure time of the electric vehicle isT 0 , according to the Markov state transfer matrix after disturbance The probability value of the electric vehicle is used to predict the next destination and state of the electric vehicle. Due to the privacy budget Different, leading to The different values ​​of may change the predicted destination of each car, that is, each car may charge at different times and places, resulting in changes in charging load; S42, if , the electric vehicle is considered to have enough power to reach the next destination. Start driving to the next destination. The time when it reaches the next destination T 1 T 0 , For driving time; update electric vehicles state and start the next driving process; S43. Determine the current remaining power of the electric vehicle Is the value less than 30%? , it is determined that the electric vehicle needs to be charged at the current destination; the formula for calculating the charging condition is as follows: , in For electric vehicles to reach the current The remaining battery power of each destination value; is the battery capacity of the electric vehicle; S44. Calculate the charging load of this area at the current time. The parking time of the electric vehicle is regarded as the charging time. The formula for calculating the charging load of the electric vehicle is as follows: , in express T 1 The charging load of electric vehicles at the moment, is the charging power, Then T 1 The moment n The charging status of an electric vehicle; =1, then n An electric car is being charged. =0, then n The car was not charged; S45, Update Charging time, updated departure time Stop time and start the next driving process.

[0029] The contents involved in the above embodiment are described below in conjunction with a preferred embodiment.

[0030] The method provided in this embodiment is used to conduct experiments. The data set used in the experiment is the real US household travel survey data NHTS 2022, which includes all travel information of resident households and individuals in one day. The total sample size is 30,174 records and 85 attributes. This embodiment will perform statistical analysis based on the travel status of residents' electric vehicles in the original data, and use the Markov model based on differential privacy technology to construct a state transition probability matrix to predict the charging load of electric vehicles in residential areas, work areas, commercial areas and other areas. This embodiment sets 91 privacy protection levels in the differential privacy algorithm, and each privacy protection level corresponds to a privacy budget The values ​​are 0.5, 0.55, 0.60, 0.65, 0.70, 0.75, ···, 4.95, and 5.0. That is, the privacy budget The value range is 0.5+0.05* λ , λ is a natural number from 0 to 90, with a total of 91 values, each of which corresponds to a privacy protection level of differential privacy. λ =0, the privacy budget Take the minimum value of 0.5, when the privacy protection effect is the strongest; λ =90, privacy budget The maximum value is 5.0, at which the privacy protection effect is the weakest. Users can choose one of the 91 different privacy budgets according to their specific circumstances, that is, users have 91 different privacy protection levels to choose from.

[0031] Taking the 24-hour electricity consumption in multiple regions as the basic load, according to the attached Figure 1 The process predicts the charging load of electric vehicles in multiple regions. The battery capacity of each electric vehicle is , the car charging power is , the power consumption per kilometer is In order to measure the effect of data privacy protection, this experiment gives 0.5 and 1.0 The simulation results of measuring the accuracy of electric vehicle charging load in residential areas, work areas, commercial areas and other areas under two privacy protection levels are shown in Tables 1 and 2.

[0032] Table 1: Privacy budget The predicted value of electric vehicle charging load (kw) in different areas within 24 hours of the day Time(T) Home(H) Work(W) Public(P) Other(O) T=1 173.18 89.46 36.68 25.2 T=2 174.44 89.88 39.76 26.39 T=3 176.19 89.81 41.02 27.16 T=4 178.57 91.49 45.5 30.1 T=5 185.64 92.54 49.07 35.56 T=6 195.93 95.76 63.7 43.05 T=7 209.44 100.52 79.66 50.19 T=8 231.98 105.98 96.6 60.97 T=9 265.44 117.39 139.72 90.44 T=10 314.16 129.5 178.08 108.01 T=11 381.99 143.99 216.79 128.1 T=12 453.04 159.81 255.5 156.66 T=13 523.81 173.67 259.77 165.69 T=14 574 182.21 248.99 165.06 T=15 602.84 185.5 242.13 156.8 T=16 612.29 191.5 229.32 153.86 T=17 597.09 190.68 200.83 136.99 T=18 581.43 190.68 172.9 122.85 T=19 542.99 185.57 145.53 107.45 T=20 500.99 179.27 126.98 90.58 T=21 453.88 172.48 109.2 73.92 T=22 406.56 165.9 92.12 63.56 T=23 368.06 161.7 79.8 56.42 T=24 333.43 157.71 70.7 50.89 Table 2: Privacy budget The predicted value of electric vehicle charging load (kw) in different areas within 24 hours of the day Time(T) Home(H) Work(W) Public(P) Other(O) T=1 164.64 87.43 35.35 25.41 T=2 165.76 88.76 38.5 26.32 T=3 168.07 89.25 39.34 26.46 T=4 170.87 91.42 43.82 30.17 T=5 177.73 92.05 49.49 35.63 T=6 185.15 96.18 64.82 42.49 T=7 200.76 98.91 82.6 49.07 T=8 224.28 104.09 96.11 58.59 T=9 255.99 114.17 140.84 85.96 T=10 306.67 128.66 175.56 104.86 T=11 362.88 145.53 217.21 125.51 T=12 433.37 162.12 257.04 151.55 T=13 509.32 176.12 268.52 160.72 T=14 560.91 186.83 257.74 159.88 T=15 596.82 189.98 238.56 155.89 T=16 607.53 194.04 225.05 150.71 T=17 600.04 190.75 193.55 132.51 T=18 580.51 188.58 169.4 123.83 T=19 540.82 187.81 144.06 108.57 T=20 486.22 182.42 121.24 93.1 T=21 448.49 175.28 97.58 77.63 T=22 404.32 167.09 82.11 66.36 T=23 361.69 160.93 73.01 58.38 T=24 324.38 156.17 65.94 50.75 As shown in Tables 1 and 2, under different privacy budgets, the charging load of electric vehicles in different areas within 24 hours can be accurately predicted. It can be seen from the table that, under the premise of protecting residents' travel privacy information, the impact of privacy disturbance has little effect on the accuracy of electric vehicle charging load. In addition, the charging load of electric vehicles in multiple areas and different time periods in a day is different. Among them, the charging load in residential areas is much higher than that in work areas, commercial areas and other areas, and the charging load is higher during the time period of 11:00~23:00. The peak charging load periods in commercial areas and other areas are concentrated at 9:00~18:00 and 11:00~16:00 respectively. The peak charging load period in the work area is 11:00~18:00. According to the predicted charging load size in different time periods, the setting of charging facilities and the dispatching of distribution networks can be optimized.

[0033] Embodiment 2: This embodiment provides a privacy-protecting multi-region electric vehicle charging load prediction device, comprising: A data acquisition module, used to acquire electric vehicle travel destination data; A data partitioning module, used for partitioning the acquired electric vehicle travel destination data into multiple regions; The first building module is used to build an electric vehicle travel chain according to the divided multiple areas; The second building module is used to build a Markov chain of driving status according to the built electric vehicle travel chain; A third construction module is used to construct a state transition probability matrix including probability values ​​of transitioning from a current destination to a next destination according to the constructed Markov chain; An adding module is used to add noise that satisfies the definition of differential privacy to each probability value in the state transition probability matrix to obtain the state transition probability matrix after adding noise; The prediction module is used to predict the next destination and charging load of the electric vehicle according to the probability value in the state transition probability matrix after adding noise.

[0034] The specific functional implementation of each of the above modules can be found in the relevant contents of the method in Example 1 and will not be elaborated here.

[0035] Embodiment 3, this embodiment provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of any one of the methods described in Embodiment 1 are implemented.

[0036] Embodiment 4: This embodiment provides a computer device, including: Memory, for storing computer programs / instructions; A processor, configured to execute the computer program / instructions to implement the steps of any one of the methods described in Example 1.

[0037] Embodiment 5: This embodiment provides a computer program product, including a computer program / instruction, which implements the steps of any method described in Embodiment 1 when executed by a processor.

[0038] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the technical principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

[0039] It should be understood by those skilled in the art that the embodiments of the present disclosure may be provided as methods, systems or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present disclosure 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.

[0040] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram and 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.

[0041] 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.

[0042] 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.

[0043] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present disclosure rather than to limit its protection scope. Although the present disclosure has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that after reading the present disclosure, those skilled in the art can still make various changes, modifications or equivalent substitutions to the specific implementation methods of the invention, but these changes, modifications or equivalent substitutions are all within the protection scope of the disclosed claims to be approved.

Claims

1. A privacy-preserving multi-region electric vehicle charging load prediction method, characterized in that: include: Obtaining electric vehicle travel destination data; Divide the acquired electric vehicle travel destination data into multiple regions; Build an electric vehicle travel chain based on the divided areas; According to the constructed electric vehicle travel chain, a Markov chain of driving status is constructed; According to the constructed Markov chain, a state transition probability matrix including the probability values ​​of transitioning from the current destination to the next destination is constructed; Add noise that satisfies the definition of differential privacy to each probability value in the state transition probability matrix to obtain the state transition probability matrix after adding noise; According to the probability values ​​in the state transition probability matrix after adding noise, the next destination and charging load of the electric vehicle are predicted.

2. The privacy-preserving multi-region electric vehicle charging load prediction method according to claim 1 is characterized in that: The obtained electric vehicle travel destination data is divided into multiple areas, including: The acquired electric vehicle travel destination data is divided into residential area H, work area W, commercial area P and other areas O.

3. The privacy-preserving multi-region electric vehicle charging load prediction method according to claim 2 is characterized in that: The electric vehicle travel chain is constructed according to the divided multiple areas, including: According to the residential area H, work area W, commercial area P and other areas O, an electric vehicle travel chain is constructed. …;in, Indicates the current driving location. Indicates the next travel destination, where the travel chain has no more than four destinations.

4. The privacy-preserving multi-region electric vehicle charging load prediction method according to claim 3 is characterized in that: The method of constructing a Markov chain of driving status according to the constructed electric vehicle travel chain includes: Construct a Markov chain of driving states: ,in, represents the probability of transferring from the current destination to the next destination, is the state transition probability matrix.

5. The privacy-preserving multi-region electric vehicle charging load prediction method according to claim 4 is characterized in that: According to the constructed Markov chain, a state transition probability matrix containing the probability value of transferring from the current destination to the next destination is constructed, and the formula is as follows: ； in, represents the probability value of always driving in residential areas, represents the probability value of moving from a residential area to a commercial area, represents the probability value of moving from the residential area to the work area, represents the probability value of moving from the residential area to other areas, represents the probability value of moving from a commercial area to a residential area, represents the probability value of always driving in the commercial area, represents the probability value of moving from the commercial area to the work area, represents the probability value of moving from the commercial area to other areas, represents the probability value of moving from the work area to the residential area, represents the probability value of moving from the work area to the commercial area, represents the probability value of always driving in the work zone, represents the probability value of transferring from the working area to other areas, represents the probability value of moving from other areas to residential areas, represents the probability value of transferring from other areas to the working area, represents the probability value of moving from other areas to the commercial area, Indicates the probability value of always driving in other areas; in, Any element in is greater than or equal to 0, The sum of the elements in each row is 1. Every element in is not greater than 1.

6. The privacy-preserving multi-region electric vehicle charging load prediction method according to claim 5 is characterized in that: The step of adding noise satisfying the definition of differential privacy to each probability value in the state transition probability matrix to obtain the state transition probability matrix after adding noise includes: Add noise that satisfies the definition of differential privacy to the state transfer matrix, and the noise scale The calculation formula is as follows: ； in, represents the privacy budget, represents the relaxation factor; State transition matrix after adding noise , the formula is as follows: ; in, represents the probability value of always driving in residential areas after adding noise, It represents the probability value of moving from residential area to commercial area after adding noise, represents the probability value of moving from the residential area to the work area after adding noise, represents the probability value of moving from residential areas to other areas after adding noise, It represents the probability value of moving from the commercial area to the residential area after adding noise, represents the probability value of always driving in the commercial area after adding noise, represents the probability value of moving from the commercial area to the working area after adding noise, represents the probability value of moving from the commercial area to other areas after adding noise, represents the probability value of moving from the work area to the residential area after adding noise, represents the probability value of moving from the work area to the commercial area after adding noise, represents the probability value of always driving in the work area after adding noise, It represents the probability value of transferring from the working area to other areas after adding noise, represents the probability value of moving from other areas to residential areas after adding noise, It represents the probability value of transferring from other areas to the working area after adding noise, represents the probability value of moving from other areas to the commercial area after adding noise, Represents the probability value of always driving in other areas after adding noise.

7. The privacy-preserving multi-region electric vehicle charging load prediction method according to claim 6, characterized in that: The privacy budget The value range is 0.5+0.05*λ, where λ is a natural number from 0 to 90.

8. The privacy-preserving multi-region electric vehicle charging load prediction method according to claim 7, characterized in that: The method predicts the next destination and charging load of the electric vehicle according to the probability value in the state transition probability matrix after adding noise, including: According to the state transfer matrix after adding noise The probability value of the electric vehicle is used to predict the next destination of the electric vehicle. Get the departure time of electric vehicles T 0 、The current remaining power of the electric vehicle Value, judgment Is the value less than 30%? if , it is determined that the electric vehicle needs to be charged at the current destination. The formula for calculating the charging condition is as follows: ， in For electric vehicles to reach the current The remaining battery power of each destination value; is the battery capacity of the electric vehicle; if , assuming that the electric vehicle has enough power to reach the next destination, starts driving to the next destination, and the time it reaches the next destination T 1 T 0 , is the travel time; calculate T 1 The charging load of electric vehicles at the moment is as follows: ， in express T 1 The charging load of electric vehicles at the moment, is the charging power, Then T 1 The moment n The charging status of an electric vehicle; =1, then n An electric car is being charged. =0, then n The car is not charging.

9. A privacy-preserving multi-region electric vehicle charging load prediction device, characterized in that: include: A data acquisition module, used to acquire electric vehicle travel destination data; A data partitioning module, used for partitioning the acquired electric vehicle travel destination data into multiple regions; The first building module is used to build an electric vehicle travel chain according to the divided multiple areas; The second building module is used to build a Markov chain of driving status according to the built electric vehicle travel chain; A third construction module is used to construct a state transition probability matrix including probability values ​​of transitioning from a current destination to a next destination according to the constructed Markov chain; An adding module is used to add noise that satisfies the definition of differential privacy to each probability value in the state transition probability matrix to obtain the state transition probability matrix after adding noise; The prediction module is used to predict the next destination and charging load of the electric vehicle according to the probability value in the state transition probability matrix after adding noise.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method described in any one of claims 1 to 8 are implemented.

Images