CN110826813A - Grid optimization method based on differential charging demands of household electric vehicle users - Google Patents

Abstract

The invention relates to a power grid optimization method based on the differentiated charging demands of household electric vehicle users, comprising: step S1: acquiring travel records of household vehicles, and acquiring temporal feature data and spatial feature data; step S2: using a vehicle use probability analysis method to perform Parameter fitting to obtain the probability function of each characteristic data, and establish the spatial transition matrix of each travel time; Step S3: simulate the travel chain of the household electric vehicle, and analyze and obtain power consumption information in the simulation process; Step S4: Divided into demand users and random users; Step S5: Determine the charging mode of demand users; Step S6: Obtain the charging probability of each random user; Step S7: Aggregate the charging needs generated in different space areas to be considered The spatial and temporal distribution map of charging load of users' charging differences is used to analyze the characteristics of charging load to optimize the power grid. Compared with the prior art, the present invention has the advantages of improving the security of the power grid and the like.

Description

Grid optimization method based on differential charging demands of household electric vehicle users

technical field

The invention relates to a power grid optimization method, in particular to a power grid optimization method based on the differential charging demands of household electric vehicle users.

Background technique

The increasing shortage of fossil fuels and the increasingly serious problems of environmental pollution make electric vehicles with the advantages of high efficiency and cleanliness have a good development prospect. Domestic and foreign auto companies are vigorously promoting the development of electric vehicles. In recent years, my country's new energy vehicle industry has developed rapidly.

With the increase in the penetration rate of electric vehicles, the impact of large-scale charging load access on the power grid is becoming more and more obvious. The superposition effect of a large number of household electric vehicle charging loads at peak load will bring difficulties to the safe operation and optimal dispatch of the power grid. Due to the large proportion of household electric vehicles and the uncertainty of users' travel and charging behavior, the charging load has unstable characteristics such as randomness, intermittentness and fluctuation in time and space. The peaks of conventional power loads are superimposed, and the peak-to-valley difference increases. If the power grid is not optimized for the charging demand of electric vehicles, it may lead to safety accidents in the power grid.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a power grid optimization method based on the differentiated charging demands of household electric vehicle users in order to overcome the above-mentioned defects in the prior art. For the research part of travel characteristics, it is proposed to use the travel chain theory as the basis, using the national family travel method. The survey (National Household Travel Survey, NHTS2017) data is used to study the daily travel rules of users, and different types of travel chain models are obtained by means of Monte Carlo simulation. Secondly, in the charging load modeling stage, a hierarchical charging decision-making model that considers the difference of users' charging is proposed, and the fuzzy reasoning method is used to study the user's random charging decision-making process considering the parking duration adequacy and time-of-use electricity price; Temporal and spatial trend of charging demand for household electric vehicles.

The object of the present invention can be realized through the following technical solutions:

A grid optimization method based on the differentiated charging demands of household electric vehicle users, including:

Step S1: Acquire the travel records of the family car, and obtain temporal feature data and spatial feature data, wherein the temporal feature data includes the start travel time, end travel time, stay time and travel length of each trip, and the spatial feature data Include the mileage and purpose of each trip;

Step S2: Based on the acquired temporal feature data and spatial feature data, use the vehicle-use probability analysis method to perform parameter fitting, obtain a probability function of each feature data, and establish a spatial transition matrix for each travel moment;

Step S3: based on the result obtained from the fitting, simulate the travel chain of the household electric vehicle, and analyze and obtain power consumption information during the simulation process;

Step S4: Divide the users into demand-type users who are determined to need charging and random-type users who may need charging according to the power consumption information;

Step S5: determining the charging mode of the demand-type user based on the parking time;

Step S6: obtaining the charging probability of each random user based on the fuzzy inference method;

Step S7: Aggregate the charging demands generated in different spatial regions, and finally obtain a charging load space-time distribution map considering the charging differences of users, and perform charging load characteristic analysis based on this to optimize the power grid.

In the step S2, the starting time of driving is fitted with a generalized extreme value distribution, the mileage is fitted with a log-normal distribution, and the parking duration is fitted with a Gaussian mixture distribution and a generalized extreme value distribution. .

The fitting method of the parking duration is as follows: the parking duration in the working area is fitted by using a Gaussian mixture distribution, and the parking duration in other areas is fitted by using a generalized extreme value distribution.

In the step S3, the process of analyzing and obtaining power consumption information in the simulation process is specifically as follows: obtaining the power consumption of the electric vehicle during the trip and updating the current state of charge of the battery through the driving mileage and the average power consumption per kilometer.

The demand-based user is a user whose current remaining power does not meet the power demand of the next trip,

The random user is a user whose current remaining power meets the power demand of the next trip.

The step S5 is specifically:

When the parking time meets the charging time required for slow charging, the charging mode of the demand-type user is determined to be slow charging; otherwise, the charging mode of the demand-type user is determined to be fast charging speed.

The step S6 specifically includes:

Step S61: Calculate the maximum saturation value that can be reached by charging within the parking duration, and convert it into the parking duration margin;

Step S62: Input the parking duration adequacy and the time-of-use electricity price as input signals into the charging decision-making model of the random user, and obtain the charging probability of the user.

The time-of-use electricity price is divided into three grades: valley, flat, and peak; the parking duration adequacy is divided into three grades: not sufficient, general sufficient and sufficient.

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

1) When studying the travel process by combining the travel chain theory and travel data, considering the interaction between temporal and spatial feature quantities, different probability functions are used to analyze the feature quantities under different operating conditions according to the distribution of actual data. , the selected analytical method is reliable.

2) According to the user's travel process, the actual energy consumption of electric vehicles is analyzed, and the user's charging load under different demand conditions is classified and studied. A hierarchical modeling method that takes into account the user's differential charging decision is proposed. scientific and rational.

3) At the same time, considering two factors that affect the user's willingness to charge: charging cost and parking adequacy, the charging load modeling is carried out, which improves the accuracy of the spatiotemporal distribution of electric vehicle charging load.

Description of drawings

Fig. 1 is the schematic flow chart of the main steps of the method of the present invention

Figure 2 is a distribution diagram of the spatiotemporal variation of the electric vehicle travel chain;

Figure 3 is a flow chart of the modeling of electric vehicle charging demand;

Figure 4 is a graph showing the distribution of charging demand for household electric vehicles;

Figure 5 is the charging load curve in the simple chain mode;

Figure 6 is the charging load curve in the complex chain mode.

Detailed ways

The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented on the premise of the technical solution of the present invention, and provides a detailed implementation manner and a specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

The analysis of the travel characteristics of electric vehicle users is the basic research of charging load modeling. The results obtained by different research methods are quite different. Therefore, it is necessary to select an appropriate method to analyze the travel process of users to improve the accuracy of charging load modeling results. sex. Most studies only conduct independent analysis on temporal feature quantities, ignoring the spatial distribution in the travel process and lack of research on the influence between feature quantities. In addition, when modeling charging load, most studies focus on considering the electricity demand due to user travel, which simplifies the research on the randomness of user charging. In fact, the user's willingness to charge is interfered by many factors. further improvement.

A grid optimization method based on the differential charging demand of household electric vehicle users, as shown in Figure 1 and Figure 3, includes:

Step S1: Acquire a travel record of the family car, and obtain temporal feature data and spatial feature data, wherein the temporal feature data includes the start travel time, end travel time, stay time and travel length of each trip, and the spatial feature data includes each trip. mileage and purpose of travel;

The present invention studies the travel characteristics of household electric vehicles by means of the travel chain theory. The time-space variation distribution in the travel chain is shown in Figure 2, which includes a time chain composed of temporal feature quantities and a space chain composed of spatial feature quantities. Two dimensions of space are used to analyze the daily travel patterns of users. The spatiotemporal feature quantities used to describe the travel chain model include:

① Number of trips, i;

②The starting point of the i-th trip, s _i ;

③ The end point of the ith trip, d _i ;

④The starting time of driving, t ₀ ;

⑤ The driving time of the i-th driving,

⑥ parking time at area d _i ,

⑦ The travel distance in the i-th trip,

Among them, the purpose of travel is divided into five categories according to functional areas, including residential area (Home, H), work area (Work, W), commercial area (Commerce, C), leisure area (Recreation, R) and related areas. Area (Other, O) for other purposes (such as picking up others, going to school, going out for medical treatment, etc.); the travel chain mode can be divided into two types: simple chain and complex chain according to the number of parking nodes. The travel patterns of are called complex chains.

Step S2: Based on the acquired temporal feature data and spatial feature data, use the vehicle-use probability analysis method to perform parameter fitting, obtain the probability function of each feature data, and establish the spatial transition matrix of each travel time, wherein the generalized driving time is used for the start time. The extreme value distribution is used for fitting, the log-normal distribution is used for the mileage, and the Gaussian mixture distribution and the generalized extreme value distribution are used for the parking duration. The fitting method of the parking duration is as follows: the parking duration in the working area is fitted by a Gaussian mixture distribution, and the parking duration in other areas is fitted by a generalized extreme value distribution.

Specifically, as follows:

Combined with the distribution of the National Travel Survey data (NHTS2017), different probability function forms are used to fit the distribution law of spatiotemporal feature quantities, using generalized extreme value distribution (as shown in Equation 1) and lognormal distribution (as shown in Equation 1) respectively. Equation 2) performs parameter estimation on the two characteristic quantities of the starting time of travel and the mileage.

Considering the influence of the spatial distribution on the parking time, the distribution characteristics of the parking time in different areas are inconsistent, and the parking time of the vehicle in the W area presents a multimodal distribution form, so the Gaussian mixture distribution (as shown in Equation 3) is used for parameter fitting. In other regions, the generalized extreme value distribution is used for fitting.

At the same time, the coupling relationship between the spatial transfer and temporal characteristics of electric vehicles is analyzed by means of the inhomogeneous Markov chain, and the spatial transfer matrix of different travel times is established.

Step S3: based on the result obtained from the fitting, simulate the travel chain of the household electric vehicle, and analyze and obtain power consumption information during the simulation process;

In step S3, in the simulation process, the process of analyzing and obtaining the power consumption information is specifically: obtaining the power consumption of the electric vehicle during the trip through the mileage and the average power consumption per kilometer and updating the current state of charge of the battery.

Ignoring the influence of external factors on power consumption, the energy consumption and mileage of electric vehicles are regarded as a linear relationship, and the power consumption during the trip can be calculated from the mileage and the average power consumption per kilometer. According to equations (4)-(6), the remaining power and state of charge of the battery when reaching the destination d _i are updated.

为完成当前行驶的总耗电量；

为到达目的地d_i时电动汽车的剩余电量；

为电动汽车电池的荷电状态；B_ev为电动汽车的电池容量。Among them, e ₀ represents the average power consumption per kilometer during the driving process of the electric vehicle;

To complete the total power consumption of the current driving;

is the remaining power of the electric vehicle when it reaches the destination d _i ;

is the state of charge of the electric vehicle battery; B _ev is the battery capacity of the electric vehicle.

Step S4: According to the power consumption information, the users are divided into demand-type users who are determined to need charging and random-type users who may need to be charged; demand-type users are users whose current remaining power does not meet the power demand for the next trip, and random-type users currently have remaining power. Users whose power meets the power requirements for the next trip.

Estimate the power consumption of the next trip, and classify users based on the estimated power. If the current remaining power does not meet the power demand for the next trip, such users are called demand users, who must be charged to ensure their normal travel; other users are called random users, and such users have charging uncertainty. The charging decision coefficient r is set as the number of decisions to control charging behavior, and the user classification basis is shown in formula (7).

为电动汽车电池的荷电状态；B_ev为电动汽车的电池容量；

为第i+1次行程中的行驶距离。Among them, e ₀ represents the average power consumption per kilometer during the driving process of the electric vehicle;

is the state of charge of the electric vehicle battery; B _ev is the battery capacity of the electric vehicle;

is the travel distance in the i+1th trip.

Step S5: Determine the charging mode of the demand-based user based on the parking time, specifically:

When the parking time meets the charging time required for slow charging, the charging mode of the demand-type user is determined to be slow charging; otherwise, the charging mode of the demand-type user is determined to be fast charging speed.

Considering the impact of parking time on the selection of demand-based user mode, when the parking time meets the charging time required for slow charging, slow charging is used to reduce the impact on battery life; if the parking time is insufficient, it is an emergency. Demand-oriented users, who urgently need to replenish the power to the expected state, need to use the fast charging mode to meet their power requirements in a short time. The charging mode selection basis is shown in formula (8).

为区域d_i中采用慢速充电时需要的充电时长；

分别为区域充电站慢充和快充功率。in, is the charging power selected by the user in the area d _i ;

is the charging time required when slow charging is used in region d _i ;

They are the slow charging and fast charging power of the regional charging stations, respectively.

Step S6: obtaining the charging probability of each random user based on the fuzzy inference method;

Step S6 specifically includes:

Step S61: Calculate the maximum saturation value that can be reached by charging within the parking duration, and convert it into the parking duration margin;

Step S62: Input the parking duration adequacy and the time-of-use electricity price as input signals into the charging decision-making model of the random user, and obtain the charging probability of the user.

The time-of-use electricity price is divided into three grades: valley, flat, and peak; the parking duration adequacy is divided into three grades: not sufficient, general sufficient and sufficient.

Considering the impact of parking time adequacy and time-of-use charging electricity price on users' willingness to charge, a charging decision model is established. With the help of fuzzy reasoning method, the user's charging psychology is analyzed, and then the charging decision is judged.

For random users, the length of parking has a certain impact on the user's willingness to charge, and users are more inclined to charge the battery to full capacity in the parking area with ample parking time. Calculate the maximum saturation value SOC _Pt that can be achieved by charging within the parking period according to formula (9), and pass the parking period The SOC _Pt that can be achieved by slow charging is used to measure the parking time adequacy AD _Pt of the electric vehicle in this area, as shown in Equation (10), where η is the charging efficiency.

Using fuzzy reasoning to take electricity price and parking time adequacy as input, a charging decision model for random users is established. The two input signals are evaluated respectively, and the time-of-use electricity price is divided into three levels: valley (level I), flat (level II), and peak (level III); the parking duration adequacy is also divided into insufficient (level I) , general abundance (level II) and abundant (level III) three levels; according to the fuzzy set from low to high, select the Z-type function suitable for expressing the low-level fuzzy process, and the bilateral Gaussian function for the intermediate-type fuzzy process. And the sigmoid function of the high-level fuzzy process is used as the membership function of three different levels, and the membership functions of different levels are shown in equations (11)-(13).

Membership function in the case of class I:

Membership function in the case of class II:

Membership function in the case of class III:

Among them, z ₁ and z ₂ are the relevant parameters in the Z-shaped function; μ and σ are the parameters in the bilateral Gaussian function (including two groups); s ₁ and s ₂ are the relevant parameters in the S-shaped function, respectively.

The result of the fuzzy inference algorithm, that is, the charging probability of the user (according to different combinations of input levels) is divided into low (level I), low (level II), medium (level III), high (level IV), high (Level V) 5 levels. And formulate fuzzy rules as shown in Table 1, where C represents the charging electricity price, AD represents the parking duration adequacy, and P represents the user's charging probability.

Table 1 Fuzzy inference rule formulation scheme

fuzzy rules C_Ⅰ C_Ⅱ C_Ⅲ AD_Ⅰ P_Ⅲ P_Ⅱ P_Ⅰ AD_Ⅱ P_Ⅳ P_Ⅲ P_Ⅱ AD_Ⅲ P_V P_Ⅳ P_Ⅲ

According to the charging decisions of different users, the charging time is calculated, and the charging period is determined, and finally the charging demand is superimposed on the charging load distribution of the corresponding area. The calculation flow of the specific charging demand is shown in Fig. 3 .

In the simulation process of this embodiment, the electricity price division form is shown in Table 2. The test values obtained by inputting the input quantities (parking time and charging electricity price) into the fuzzy logic controller in different combinations are shown in Table 3.

Table 2 Charging electricity price division table

Table 3 Test table of charging probability when different input quantities are combined

Step S7: Aggregate the charging demands generated in different spatial regions, and finally obtain a charging load space-time distribution map considering the charging differences of users, and perform charging load characteristic analysis based on this to optimize the power grid.

The temporal and spatial distribution of the charging load of household electric vehicles is finally obtained by simulating the charging process of the electric vehicle in the area set in the embodiment, as shown in Figure 4, and the distribution of the charging load generated in the simple chain and complex chain modes is shown in Figures 5 and 6. 4 to 6, the charging load in the residential area is larger than that in the commercial area. The final simulation results are analyzed, and the statistical analysis results of the distribution characteristics of the charging load in different regions are obtained, as shown in Table 4.

Table 4 Results of charging demand statistics in different regions

The above-described embodiments are only examples, and do not limit the scope of application of the present invention. These embodiments can also be implemented in various other ways, and different assumptions and substitutions can be made within the scope of not departing from the technical idea of the present invention.

Claims (8)

1. a power grid optimization method based on the charging differential demand of household electric vehicle users, is characterized in that, comprising: Step S1: Acquire the travel records of the family car, and obtain temporal feature data and spatial feature data, wherein the temporal feature data includes the start travel time, end travel time, stay time and travel length of each trip, and the spatial feature data Include the mileage and purpose of each trip; Step S2: Based on the acquired temporal feature data and spatial feature data, use the vehicle-use probability analysis method to perform parameter fitting, obtain a probability function of each feature data, and establish a spatial transition matrix for each travel moment; Step S3: based on the result obtained from the fitting, simulate the travel chain of the household electric vehicle, and analyze and obtain power consumption information during the simulation process; Step S4: Divide the users into demand-type users who are determined to need charging and random-type users who may need charging according to the power consumption information; Step S5: determining the charging mode of the demand-type user based on the parking time; Step S6: obtaining the charging probability of each random user based on the fuzzy inference method; Step S7: Aggregate the charging demands generated in different spatial regions, and finally obtain a charging load space-time distribution map considering the charging differences of users, and perform charging load characteristic analysis based on this to optimize the power grid. 2. A power grid optimization method based on the charging differential demand of household electric vehicle users according to claim 1, characterized in that, in the step S2, the starting time of driving is fitted using a generalized extreme value distribution, so The distance traveled was fitted using a lognormal distribution, and the parking duration was fitted using a Gaussian mixture distribution and a generalized extreme value distribution. 3. A power grid optimization method based on the charging differential demand of household electric vehicle users according to claim 2, wherein the fitting method of the parking duration is specifically: using a Gaussian mixture for the parking duration in the work area The distribution is fitted, and the parking duration in other areas is fitted using the generalized extreme value distribution. 4. a kind of grid optimization method based on the charging differential demand of household electric vehicle users according to claim 1, is characterized in that, in described step S3, described in the simulation process is to analyze and obtain the process of power consumption information specifically: : Obtain the power consumption of the electric vehicle during the trip and update the current state of charge of the battery through the mileage and the average power consumption per kilometer. 5. A kind of grid optimization method based on the charging differential demand of household electric vehicle users according to claim 4, is characterized in that, The demand-based user is a user whose current remaining power does not meet the power demand of the next trip, The random user is a user whose current remaining power meets the power demand of the next trip. 6. A kind of grid optimization method based on the charging difference demand of household electric vehicle users according to claim 1, is characterized in that, described step S5 is specifically: When the parking time meets the charging time required for slow charging, the charging mode of the demand-type user is determined to be slow charging; otherwise, the charging mode of the demand-type user is determined to be fast charging speed. 7. The grid optimization method based on the charging differential demand of household electric vehicle users according to claim 1, wherein the step S6 specifically comprises: Step S61: Calculate the maximum saturation value that can be reached by charging within the parking duration, and convert it into the parking duration margin; Step S62: Input the parking duration adequacy and the time-of-use electricity price as input signals into the charging decision-making model of the random user, and obtain the charging probability of the user. 8 . The grid optimization method based on the differential charging demands of household electric vehicle users according to claim 1 , wherein the time-of-use electricity price is divided into three levels: valley, flat, and peak; the parking duration The adequacy is divided into three levels: not sufficient, generally sufficient and abundant.

Images