Charging facility site selection method based on space-time analysis and multiple constraint conditions

Abstract

The invention relates to a charging facility site selection method based on space-time analysis and multiple constraint conditions, and the method comprises the following steps: firstly, building a charging load space-time distribution prediction model considering dynamic transfer planning based on a real-time road condition; secondly, comprehensively considering the influence of the scale of the charging station, the reachability of user charging and the like, analyzing the selection probability of the user for different charging stations, and determining the service range of each charging station according to the selection probability; then, with the purpose of minimizing the annual total investment cost of the charging station, collaborative planning of the distributed energy storage system and the multi-power-level charging facility is carried out; and finally, establishing a charging vehicle charging station site selection model, and then solving the model by adopting an improved whale optimization algorithm based on the site selection optimization problem. The method has the advantages that the travel model is established through balanced coupling, the multi-type configuration model is constructed, two-stage locating and sizing are achieved, and the whale optimization algorithm is adopted.

Description

Technical Field

[0001] The present invention belongs to the technical field of electric vehicles, and in particular relates to a charging facility site selection method based on spatiotemporal analysis and multiple constraints. Background Art

[0002] In recent years, the sales volume of electric vehicles (EV) has exploded. It is expected that by 2035, EV will become the mainstream of the new car sales market. It is an urgent need to scientifically layout the charging infrastructure to quickly replenish EV power. At present, the research on the planning of electric vehicle charging stations (EVCS) mainly focuses on the prediction of EV charging time and space load and the site selection and capacity determination of EVCS. For the former, the existing technologies all predict the overall trend, ignoring the impact of real-time traffic conditions on power consumption, and not considering the charging selection behavior of EV users when there are multi-power charging piles. In addition, most of the existing technologies assume that each road network node is equipped with a charging station. However, considering the actual engineering significance, charging stations are often built based on economic optimization. For the latter, the existing technologies do not fully consider the access of ultra-high power charging facilities. The impact brought about is that compared with AC slow charging and general DC fast charging, the investment cost of ultra-high power charging is higher, and accurate site selection and capacity determination are the prerequisites for its efficient use; at the same time, ultra-high power charging has a greater impact on the distribution network, bringing extremely high operational safety risks and transformation costs. The contradiction between users' ultra-high power fast charging demand and distribution network safety is becoming increasingly prominent. Therefore, it is urgent to take into account users' charging needs and the safe and economic development of the distribution network, and to optimize the layout and planning of ultra-high power charging, ordinary DC fast charging, AC slow charging and other charging modes in the medium and long term; therefore, it is very necessary to provide a charging facility site selection method based on spatiotemporal analysis and multiple constraints, which can establish a travel model with balanced coupling, construct a multi-type configuration model, perform two-stage site selection and capacity determination, and adopt the whale optimization algorithm. Summary of the invention

[0003] The purpose of the present invention is to overcome the shortcomings of the prior art and to provide a charging facility site selection method based on spatiotemporal analysis and multiple constraints, which establishes a travel model through balanced coupling, constructs a multi-type configuration model, performs two-stage site selection and capacity determination, and adopts a whale optimization algorithm.

[0004] The object of the present invention is achieved by: a charging facility site selection method based on spatiotemporal analysis and multiple constraints, the method comprising the following steps:

[0005] Step 1: EV spatiotemporal distribution taking into account coupling factors: Combine the road network topology and dynamic road condition information model to calculate the charging needs of different EV owners and achieve balanced coupling between the road network and the distribution network;

[0006] Step 2: Phase 1: Establish an EV travel model based on travel chain technology, combine road congestion and temperature models to obtain the EV spatiotemporal load distribution in each period, then establish the attractiveness model of different EVCSs, and then obtain the service scope of EVCS through user selection probability;

[0007] Step 3: Phase 2: Based on the EV spatiotemporal load distribution obtained in the first phase and the service range of each candidate charging station, the second phase comprehensively considers the EVCS construction scale, the impact of EV charging power on the distribution network, and the impact of user charging accessibility factors, and constructs a multi-type charging facility planning model from the three levels of "vehicle-station-network";

[0008] Step 4: Improved whale optimization algorithm: When the demand points are known, a charging station site selection model for charging vehicles is established, and several points are selected as the site selection centers; with the minimum number of charging stations within the coverage area of ​​the charging station as the constraint condition, a function model is constructed with the goal of minimizing the sum of the product of the weight of the demand degree of each demand point within the coverage area of ​​the charging station and the distance from the charging station to the demand point, and then based on this site selection optimization problem, the improved whale optimization algorithm is used to solve the model.

[0009] The spatiotemporal distribution of EVs taking into account coupling factors in step 1 specifically includes the following steps:

[0010] Step 1.1: EV Spatiotemporal Transfer Model: EV Travel Time It obeys the normal probability distribution, and its probability density function for: , where They are The variance and mean of the function; EV travel has Markov properties, from the current node Move to the next node The probability of : , where For Node The number of all feasible options for transfer to the end point; The next transfer node in all feasible paths Number of schemes; The number of nodes that the node transfer passes through in the current solution; is the number of path choices; is the number of adjacent nodes; Respectively in and Next Segment path length; Respectively in and Next EV driving speed of the segment path;

[0011] Step 1.2: Dynamic road condition information model;

[0012] Step 1.3: Road network-distribution network balanced coupling.

[0013] The dynamic traffic information model in step 1.2 specifically includes:

[0014] Step 1.21: Congestion model: Determine the roads in each cycle based on the temporal and spatial distribution of vehicles Number of EVs in the system , the EV travel speed is determined by the interaction between the real-time number of vehicles on the road and the travel chain information Make corrections, i.e. dynamically update : , where They are The number of EVs within the unit distance of the road and on the road at the moment; is the correction constant; for The number of EVs on the road at the time; is the current road distance;

[0015] Step 1.22: Energy consumption per mile model: EV energy consumption per mile, road grade and The relationship between: , where are the energy consumption per unit mileage of expressways, trunk roads and branch roads respectively; is the EV travel speed; however, the EV energy consumption per unit mileage is different at different temperatures, and the relationship between them is expressed as: , where For temperature Energy consumption per unit mileage; is the fitting parameter; is the current ambient temperature; is a natural number.

[0016] The road network-distribution network balanced coupling in step 1.3 specifically includes:

[0017] Step 1.31: Spatial coupling: Considering the spatial overlap between the road network and the distribution network, the road network and distribution network nodes with the closest geographical locations are preferentially coupled, and the geographical location function is introduced. Characterizes the distance between the road network node and any distribution network node. This indicator is often the first indicator of coupling: , where is the node location of the distribution network; is the location of the road network node;

[0018] Step 1.32: Coupling of distribution network support margin: Since there is a capacity upper limit on the distribution network nodes, different loads will have different impacts on the distribution network. Therefore, the capacity margin index is introduced. Used to describe the impact of EV load on distribution network operation: , where For distribution network nodes Maximum rated power; for Distribution network power supply node at all times The base load; For charging stations exist The power purchased from the distribution grid at all times; is the set of distribution network nodes.

[0019] The first stage in step 2 includes the following steps:

[0020] Step 2.1: Probabilistic model of user selection of charging station: Based on the unit mileage energy consumption model, the user's real-time power consumption and resistance are calculated. The probability of a user choosing a specific charging station to charge depends on the ratio between the utility obtained by the user from the charging station and the sum of the utilities provided by all eligible charging stations in the planning area: , where For Node Users choose charging stations probability; Node Users from charging stations the utility gained; For charging stations Attractiveness to users; For Node Users go to charging stations the resistance encountered; It is a collection of all candidate charging stations in the planning area;

[0021] Step 2.2: EVCS service range: Assume that EV users charge when their battery power is less than a certain threshold. Then, the service range of the charging station is determined based on the probability of selecting different charging stations: , where For charging stations Scope of services; For Node Users choose charging stations probability; is a set of road network nodes;

[0022] Step 2.3: User charging accessibility.

[0023] The user charging accessibility in step 2.3 specifically includes:

[0024] Step 2.31: For each supply point, search for all charging demand points within the search radius and calculate the supply-demand ratio : , where They are the number of slow charging piles, fast charging piles and ultra-high power charging piles; Respectively represent their rated charging power; For Node Total EV charging load at is the distance threshold;

[0025] Step 2.32: For each demand point, search for all supply points within the search radius and sum up all supply-demand ratios to get the node Charging accessibility: , where For Node Charging accessibility for users.

[0026] The second stage in step 3 comprises the following steps:

[0027] Step 3.1: User’s parking time: EV’s parking time Basically conforms to the exponential distribution, and its probability density function for: , where is the rate parameter of the exponential distribution;

[0028] Step 3.2: The selection principle of the user's multi-power charging pile is: ① Give priority to meeting the needs within the specified time, that is, to meet the expected state of charge within the expected stay time of the owner; ② Because the electricity prices of different charging piles are different, if any charging pile can meet the owner's needs within the expected stay time, the slow charging pile with the best economy is given priority;

[0029] Step 3.3: DESS configuration: The charging pile can obtain power from the DESS or directly from the grid, while the DESS obtains power directly from the grid. The power balance relationship between them is: , where For Node exist Base load at the moment; are the charging and discharging power of DESS respectively; They are charge and discharge states respectively; For charging stations The power purchased from the distribution grid at the time;

[0030] Step 3.4: Objective function;

[0031] Step 3.5: Constraints.

[0032] The objective function in step 3.4 is specifically: comprehensively consider the interests of EVCS investors, minimize the EVCS investment cost and operating cost, and at the same time meet the user charging accessibility and distribution network basic power flow constraints to establish an EVCS site selection and capacity determination model: , where To consider the objective function of the three aspects of "vehicle-station-network"; They are the annual construction and annual operation and maintenance costs of the charging station; System network loss costs; to purchase the land; It is the operation and maintenance cost of DESS.

[0033] The constraints in step 3.5 include: user charging accessibility constraints, system power flow constraints, voltage deviation constraints, current constraints, distribution network load constraints, charging station quantity constraints, and DESS charging and discharging constraints.

[0034] The improved whale optimization algorithm in step 4 includes the following steps:

[0035] Step 4.1: Model establishment: The electric vehicle charging facility site selection model is established as follows: , the constraints are: , where the objective function It is the sum of the product of the weight of the demand degree of each demand point within the coverage area of ​​the charging station and the distance from the charging station to the demand point; for A collection of serial numbers of demand points; The maximum coverage radius of the site selection location; Indicates demand point Gathering at the selected location; Indicates the demand weight of the demand point; Indicates the distance from the demand point to the site selection location; is a variable between 0 and 1;

[0036] Step 4.2: Whale Optimization Algorithm: Surrounding the Prey: In the formula, is the number of iterations; is the current position vector; is the optimal position vector; is the correlation coefficient vector;

[0037] Step 4.3: Spiral Hunting: , where , is the displacement vector between the prey and the whale; Select probabilities for humpback whale behaviors; is the probability of surrounding the prey; is a constant used to define the spiral shape; is a random number in the interval [-1, 1];

[0038] Step 4.4: Search for prey: , where is a randomly obtained position vector;

[0039] Step 4.5: Algorithm improvement: Use the cosine decreasing method to perform nonlinear decreasing of the convergence factor. The formula is: , where is the probability of the convergence factor cosine decreasing; is the convergence factor at the current iteration number The probability of behavior selection under is a random number between (0, 1).

[0040] Beneficial effects of the present invention: The present invention is a charging facility site selection method based on spatiotemporal analysis and multiple constraints. During use, first, based on real-time road conditions, a charging load spatiotemporal distribution prediction model taking into account dynamic transfer planning is constructed; secondly, the scale of charging stations, user charging accessibility and other influences are comprehensively considered, the user's selection probability for different charging stations is analyzed and the service scope of each charging station is determined accordingly; then, with the goal of minimizing the total annual investment cost of the charging station, a distributed energy storage system (DESS) and multi-power level charging facilities are coordinated and planned; finally, a charging vehicle charging station site selection model is established, and then the improved whale optimization algorithm is used to solve the model based on this site selection optimization problem; the present invention has the advantages of establishing a travel model through balanced coupling, constructing a multi-type configuration model, two-stage site selection and capacity determination, and using a whale optimization algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a diagram of the distributed resource optimization architecture within the site of the present invention.

[0042] Figure 2 The present invention provides a flow chart for solving the two-stage site selection and capacity determination model.

[0043] Figure 3 This is a flow chart of the whale optimization algorithm of the present invention. DETAILED DESCRIPTION

[0044] The present invention will be further described below in conjunction with the accompanying drawings.

[0045] Example 1

[0046] like Figures 1-3 As shown, a charging facility site selection method based on spatiotemporal analysis and multiple constraints includes the following steps:

[0047] Step 1: EV spatiotemporal distribution taking into account coupling factors: Combine the road network topology and dynamic road condition information model to calculate the charging needs of different EV owners and achieve balanced coupling between the road network and the distribution network;

[0048] In the present invention, ① EV space-time transfer model: the transfer of EV in space is simplified to the transfer between road nodes, and the spatial transfer between nodes is the EV driving process. According to people's travel needs, the functional areas are divided into business areas (business, B), work areas (work, W), and residential areas (house, H); EV travel time It obeys the normal probability distribution, and its probability density function As shown below: , where They are The variance and mean of the function; EV travel has Markov properties, from the current node Move to the next node The probability of , as shown below: , where For Node The number of all feasible options for transfer to the end point; The next transfer node in all feasible paths Number of schemes; The number of nodes that the node transfer passes through in the current solution; is the number of path choices; is the number of adjacent nodes; Respectively in and Next Segment path length; Respectively in and Next EV driving speed on segment path.

[0049] ② Dynamic road condition information model: 1) Congestion model: First, based on the historical spatiotemporal distribution of vehicles in the region, as the data support for traffic flow correction parameters in the congestion model in the road network, the Markov decision process theory is used to determine the probability of transfer from each node to the adjacent node. Based on the transfer probability, the next transfer node is determined through Monte Carlo sampling, and the road condition in each cycle is determined according to the spatiotemporal distribution of vehicles. Number of EVs in the system , the EV travel speed is determined by the interaction between the real-time number of vehicles on the road and the travel chain information Make corrections, i.e. dynamically update , as shown below: , where They are The number of EVs within the unit distance of the road and on the road at the moment; is the correction constant; for The number of EVs on the road at the time; Current road distance; adjust road segment The road congestion can be changed; the present invention sets the same speed of EVs within the same number range, through Sure Time Road EV travel speed .

[0050] 2) Unit mileage energy consumption model: EV unit mileage energy consumption, road grade and The relationship between them is shown in the following formula: , where are the energy consumption per unit mileage of expressways, trunk roads and branch roads respectively; is the EV travel speed; however, the EV energy consumption per unit mileage is different at different temperatures, and the relationship between them is expressed as: , where For temperature Energy consumption per unit mileage; is the fitting parameter; is the current ambient temperature; is a natural number.

[0051] In summary, the comparison shows that the energy consumption is the lowest when the temperature is 20℃; when the standard temperature is set to 20℃, the energy consumption of EV is the lowest when the temperature is 20℃. Driving down Energy consumption per mile on graded roads for: , where for Energy consumption per mile on graded roads; For temperature Energy consumption per unit mileage.

[0052] ③Road network-distribution network balanced coupling: There is not only a coupling relationship between the EV road network and the distribution network in terms of geographical location, but also a coupling relationship between external policies and capacity constraints. For this reason, the coupling principles are proposed: 1) Spatial coupling: Considering the spatial overlap between the road network and the distribution network, the road network and distribution network nodes with the closest geographical location are preferentially coupled, and the geographical location degree function is introduced. Characterizes the distance between the road network node and any distribution network node. This indicator is often the first indicator of coupling: , where is the node location of the distribution network; is the location of the road network node.

[0053] 2) Distribution network support margin coupling: Since there is a capacity upper limit on the distribution network nodes, different loads will have different impacts on the distribution network. Therefore, the capacity margin index is introduced. Used to describe the impact of EV load on distribution network operation: , where For distribution network nodes Maximum rated power; for Distribution network power supply node at all times The base load; For charging stations exist The power purchased from the distribution grid at all times; is the set of distribution network nodes.

[0054] Step 2: Phase 1: Establish an EV travel model based on travel chain technology, combine road congestion and temperature models to obtain the EV spatiotemporal load distribution in each period, then establish the attractiveness model of different EVCSs, and then obtain the service scope of EVCS through user selection probability;

[0055] In the present invention, ① the probability model of user selection of charging station: the remaining power of EV when arriving at each charging and storage station can be obtained according to the traffic energy consumption model and the temperature energy consumption model. When the power threshold is less than 30% or the remaining power is insufficient to support the arrival at the next destination, a suitable EVCS is selected to stop for charging; the real-time power consumption and resistance of the user are calculated according to the unit mileage energy consumption model. The possibility of the user choosing a specific charging station for charging depends on the ratio between the utility obtained by the user from the charging station and the sum of the utilities provided by all qualified charging stations in the planning area: , where For Node Users choose charging stations probability; Node Users from charging stations the utility gained; For charging stations Attractiveness to users; For Node Users go to charging stations the resistance encountered; It is the set of all candidate charging stations in the planning area.

[0056] The resistance formula is determined by real-time traffic information, which is affected by both road congestion and temperature conditions: , where For Node From user to charging station The shortest distance; For charging stations The charging price is different in different functional areas.

[0057] , where For charging stations Attractiveness to users; is the impact coefficient of charging price; is a constant.

[0058] ②EVCS service scope: Before optimizing the configuration of EVCS, the service scope of each EVCS should be determined first. From the perspective of EV users, the EV power consumption is calculated based on the time-space transfer model and the dynamic road condition model. The present invention assumes that the EV user performs charging when the battery power is less than a certain threshold, and then determines the service scope of the charging station based on the selection probability of different charging stations: , where For charging stations Scope of services; For Node Users choose charging stations probability; is a set of road network nodes.

[0059] ③ User charging accessibility: The charging accessibility of electric vehicle users refers to the degree to which users can conveniently and efficiently find and use charging facilities in the process of using electric vehicles; the present invention uses a two-step mobile search method in ArcGIS software to obtain the charging accessibility of EV users at each node under different planning states, as follows: 1) For each supply point, search all charging demand points within the search radius and calculate the supply-demand ratio : , where They are the number of slow charging piles, fast charging piles and ultra-high power charging piles; Respectively represent their rated charging power; For Node Total EV charging load at is the distance threshold.

[0060] 2) For each demand point, search all supply points within the search radius and sum up all supply-demand ratios to get the node Charging accessibility: , where For Node Charging accessibility for users.

[0061] Step 3: Phase 2: Based on the EV spatiotemporal load distribution obtained in the first phase and the service range of each candidate charging station, the second phase comprehensively considers the EVCS construction scale, the impact of EV charging power on the distribution network, and the impact of user charging accessibility factors, and constructs a multi-type charging facility planning model from the three levels of "vehicle-station-network";

[0062] In the present invention, ① the user's parking time: the parking time of the EV Basically conforms to the exponential distribution, and its probability density function for: , where is the rate parameter of the exponential distribution.

[0063] ② User selection of multi-power charging piles: Electric car owners choose charging piles according to the following two principles:

[0064] 1) Give priority to meeting the needs within the specified time, that is, meeting the expected state of charge within the owner's expected stay time;

[0065] 2) Because different charging piles have different electricity prices, if any charging pile can meet the owner's needs within the expected stay time, the slow charging pile with the best economy should be given priority.

[0066] The present invention provides three charging piles with different powers, namely, ultra-high power (300kW), fast charging (60kW) and slow charging (7kW); EV is used to indicate the number of electric vehicles, At the node Department The charging power is : , where Respectively represent the rated power of slow, fast and ultra-high power charging piles; Indicates the battery capacity; Indicates The vehicle's state of charge.

[0067] exist Time Node Electric vehicle charging load It can be expressed as: , where For charging stations The set of EV numbers in .

[0068] ③DESS configuration: Since ultra-high power charging has a great impact on the distribution network, when large-scale electric vehicles are connected, it will bring extremely high operational safety risks and transformation costs to the distribution network. Therefore, the present invention uses DESS for energy buffering while connecting ultra-high power charging facilities to charging stations, so that DESS discharges at the peak of charging load, reducing the pressure on the distribution network power supply; Figure 1 This is a diagram of the distributed resource optimization architecture within the charging station when ultra-high power charging facilities are connected to the charging station.

[0069] Depend on Figure 1 It can be seen that the designed system consists of a DESS connected to the power grid, a charging pile, and an electric vehicle; the charging pile can obtain power from the DESS or directly from the power grid, and the DESS obtains power directly from the power grid. The power balance relationship between them is: , where For Node exist Base load at the moment; are the charging and discharging power of DESS respectively; They are charge and discharge states respectively; For charging stations The power purchased from the distribution grid at time.

[0070] ④Objective function: This invention comprehensively considers the interests of EVCS investors, takes minimizing the investment cost and operation cost of EVCS as the goal, and at the same time satisfies the constraints such as user charging accessibility and basic power flow of distribution network to establish the EVCS site selection and capacity determination model: , where To consider the objective function of the three aspects of "vehicle-station-network"; They are the annual construction and annual operation and maintenance costs of the charging station; System network loss costs; to purchase the land; It is the operation and maintenance cost of DESS.

[0071] 1) Annual construction cost of charging station: , where Respectively represent the unit investment and construction costs of slow, fast, and ultra-high power charging piles; Respectively represent the candidate charging station nodes The number of charging piles with different powers; It is a 0-1 variable, indicating whether to build a site at this node or not; is an auxiliary variable used for annualization processing, and its expression is: , where represents the discount rate; The economic life of the charging pile.

[0072] 2) Annual operation and maintenance cost of charging station: , where Respectively represent the annual operation and maintenance costs of slow, fast and super fast charging piles.

[0073] 3) System network loss costs: , where It is 96 periods; is the cost per unit network loss; for Time branch The current flowing through For branch Resistance; is the time interval, which is set to 15 minutes in the present invention; It is the collection of all branches in the distribution network.

[0074] 4) Land purchase cost: , where is the conversion ratio coefficient; is the area of ​​a single parking space; is the number of parking spaces; For the The unit price of land in the area where the EVCS is located varies depending on the functional area.

[0075] 5)DESS operation and maintenance costs: , where is the annual operation and maintenance cost per unit capacity of the DESS facility; is the maximum power of DESS.

[0076] ⑤Constraints: 1) User charging accessibility constraints: , where Charging accessibility threshold for users; is the total number of nodes.

[0077] 2) System power flow constraints: ∑ i ∈ v ( j ) [ P ij − ( P ij ) 2 + ( Q ij ) 2 ( U i ) 2 ⋅ R ij ] = ∑ l ∈ u ( j ) [ P jl + P j ] ( 27 ) , ∑ i ∈ v ( j ) [ Q ij − ( P ij ) 2 + ( Q ij ) 2 ( U i ) 2 ⋅ X ij ] = ∑ l ∈ u ( j ) [ Q jl + Q j ] ( 28 ) , ( U j ) 2 = ( U i ) 2 − 2 ( R ij P ij + X ij Q ij ) + [ ( P ij ) 2 + ( X ij ) 2 ] ( P ij ) 2 + ( Q ij ) 2 ( U i ) 2 , ∀ ij ∈ Ω L ( 29 ) , where and Branch Active and reactive power on the Node The voltage at Node Corresponding active and reactive loads; For branch The resistance of For branch Reactance; For nodes located Downstream and A collection of connected nodes; For nodes located Upstream and A collection of connected nodes.

[0078] 3) Voltage deviation constraint: , where is the voltage deviation rate; is the standard voltage; It is the maximum deviation rate allowed by the distribution network.

[0079] 4) Current Constraint: , , where For branch The maximum current allowed to flow.

[0080] 5) Distribution network load constraints: , where The upper limit of the distribution network load.

[0081] 6) Constraints on the number of charging stations: , where The number of charging stations that can be planned in the target area The maximum and minimum values ​​of .

[0082] 7) DESS charging and discharging constraints: , where They are charge and discharge states respectively; is the maximum charging power of DESS; for The power of DESS at the moment; are the charge and discharge efficiency respectively.

[0083] Step 4: Improved whale optimization algorithm: When the demand points are known, a charging station site selection model for charging vehicles is established, and several points are selected as the site selection centers; with the minimum number of charging stations within the coverage area of ​​the charging station as the constraint condition, a function model is constructed with the goal of minimizing the sum of the product of the weight of the demand degree of each demand point within the coverage area of ​​the charging station and the distance from the charging station to the demand point, and then based on this site selection optimization problem, the improved whale optimization algorithm is used to solve the model.

[0084] In summary, the present invention is a two-stage site selection and capacity determination method for EVCS with ultra-high power charging facilities. ① The proposed real-time road condition model taking into account coupling factors is more in line with the actual situation of users, and provides data support for the user's probabilistic selection of charging stations. It not only improves the accuracy and rationality of site selection and capacity determination, but also can be used to evaluate the rationality of the service scope and capacity configuration of existing charging stations in the planning area; ② The established optimization model of multi-type charging facilities in the station takes into account the multi-subject and multi-interest demands of "vehicle-station-network", takes into account the safety of the distribution network, the economy of the construction of charging facilities and user satisfaction, and can achieve a balance of interests among all subjects; ③ The proposed DESS joint planning method reduces the peak load pressure of the distribution network to a certain extent, and effectively avoids unstable factors such as voltage over-limit.

[0085] The present invention is a charging facility site selection method based on spatiotemporal analysis and multiple constraints. In use, first, in the first stage: first, balanced coupling of the road network and the distribution network is carried out, and then travel models of different EVs are established based on the travel chain technology, and finally the service range of the EVCS is obtained through the user selection probability; in the second stage: a distributed energy storage system (DESS) and three different power charging piles are selected, namely, ultra-high power (300kW), fast charging (60kW) and slow charging (7kW), and the investment economy of the charging station, the operation safety of the distribution network and the influence of the user charging accessibility factors are comprehensively considered, and a multi-type charging facility optimization configuration model is constructed from the three levels of "vehicle-station-network"; the present invention has the advantages of balanced coupling to establish a travel model, construct a multi-type configuration model, two-stage site selection and capacity determination, and adopt a whale optimization algorithm.

[0086] Example 2

[0087] like Figures 1-3 As shown, a charging facility site selection method based on spatiotemporal analysis and multiple constraints includes the following steps:

[0088] Step 1: EV spatiotemporal distribution taking into account coupling factors: Combine the road network topology and dynamic road condition information model to calculate the charging needs of different EV owners and achieve balanced coupling between the road network and the distribution network;

[0089] Step 2: Phase 1: Establish an EV travel model based on travel chain technology, combine road congestion and temperature models to obtain the EV spatiotemporal load distribution in each period, then establish the attractiveness model of different EVCSs, and then obtain the service scope of EVCS through user selection probability;

[0090] Step 3: Phase 2: Based on the EV spatiotemporal load distribution obtained in the first phase and the service range of each candidate charging station, the second phase comprehensively considers the EVCS construction scale, the impact of EV charging power on the distribution network, and the impact of user charging accessibility factors, and constructs a multi-type charging facility planning model from the three levels of "vehicle-station-network";

[0091] Step 4: Improved whale optimization algorithm: When the demand points are known, a charging station site selection model for charging vehicles is established, and several points are selected as the site selection centers; with the minimum number of charging stations within the coverage area of ​​the charging station as the constraint condition, a function model is constructed with the goal of minimizing the sum of the product of the weight of the demand degree of each demand point within the coverage area of ​​the charging station and the distance from the charging station to the demand point, and then based on this site selection optimization problem, the improved whale optimization algorithm is used to solve the model.

[0092] In the present invention, ① Model establishment: The electric vehicle charging facility site selection model is established as follows: , the constraints are: , where the objective function It is the sum of the product of the weight of the demand degree of each demand point within the coverage area of ​​the charging station and the distance from the charging station to the demand point; for A collection of serial numbers of demand points; The maximum coverage radius of the site selection location; Indicates demand point Gathering at the selected location; Indicates the demand weight of the demand point; Indicates the distance from the demand point to the site selection location; is a variable between 0 and 1. By site selection Supply, represented by 1, otherwise represented by 0.

[0093] ② Whale Optimization Algorithm: Each whale can be regarded as a particle, and each particle is a different decision variable; the implementation process of WOA mainly includes three stages: surrounding prey, spiral hunting and random search. Its mathematical model is as follows: 1) Surrounding prey: In the formula, is the number of iterations; is the current position vector; is the optimal position vector; is the correlation coefficient vector, , is the convergence factor, , is a random number between (0, 1), Linearly decreases between (0, 2); is the maximum number of iterations;

[0094] 2) Spiral Hunting: , where , is the displacement vector between the prey and the whale; is a random number between (0, 1), Select probabilities for humpback whale behaviors; is the probability of surrounding the prey; is a constant used to define the spiral shape; is a random number in the interval [-1, 1], when When the whale is closest to its prey, When , the distance is the farthest;

[0095] 3) Searching for prey: , where is a randomly obtained position vector, the algorithm defines When it is not less than 1, a leader individual is randomly selected to guide the movement of other whales in order to find better prey and enhance the global search ability of the algorithm; the process of the whale optimization algorithm is as follows Figure 3 shown.

[0096] ③Algorithm improvement: Like other meta-heuristic algorithms, the basic whale optimization algorithm is still prone to multi-peak optimization problems in the later stages of calculation, and the algorithm iteration speed needs to be improved; in the early stages of algorithm iteration, the convergence speed should be accelerated, and the convergence factor A larger value should be taken in the later stage of algorithm iteration to improve the accuracy and local search ability of the algorithm. It should take a smaller value; therefore, a method of nonlinearly decreasing the convergence factor by using cosine decreasing is proposed, and the formula is: , where is the probability of the convergence factor cosine decreasing, between (0, 1); is the convergence factor at the current iteration number The probability of behavior selection under is a random number between (0, 1).

[0097] Based on the above formula, the convergence factor can be made to decrease nonlinearly, so that the algorithm can have different decreasing rates during the iteration process, greatly improving the ability to maintain population diversity.

[0098] According to the characteristics of site selection, the present invention makes the following assumptions in establishing the model: the charging station can meet the total demand of each demand point within the coverage area, and one demand point is only supplied by one charging station; based on the above assumptions, the site selection model is established. Select from known demand points points as the site selection results.

[0099] In summary, the charging station site selection problem is a nonlinear planning model and a multi-peak optimization problem. The whale optimization algorithm is a new type of swarm intelligence optimization algorithm with the characteristics of fewer adjustment parameters, simple coding and not easy to fall into local optimality. It can also be improved in parallel, serial and embedded with other algorithms on the basic algorithm to enhance its convergence speed and search accuracy. The present invention improves the convergence factor value on the basis of the whale optimization algorithm (WOA). The improved whale optimization algorithm (IWOA) effectively improves the iteration rate, so that it can more accurately find the extreme point in the stable state during the iteration process.

[0100] The present invention is a charging facility site selection method based on spatiotemporal analysis and multiple constraints. In use, the present invention takes minimizing the number of charging stations within the coverage of the charging station as a constraint condition, and takes minimizing the sum of the product of the demand weight of each demand point and the distance from the charging station to the demand point as the goal to construct a site selection optimization model. In view of the characteristics of slow convergence speed and easy to fall into local optimality of traditional algorithms, a cosine decreasing strategy is introduced to optimize the convergence factor value selection method of the whale optimization algorithm (WOA). The optimized whale algorithm (IWOA) can calculate the site selection position more reasonably, and provide a better solution for the site selection of the charging station. The present invention has the advantages of establishing a travel model through balanced coupling, constructing a multi-type configuration model, two-stage site selection and capacity determination, and adopting a whale optimization algorithm.

Claims

1. A charging facility site selection method based on spatiotemporal analysis and multiple constraints, characterized by: The method comprises the following steps: Step 1: EV spatiotemporal distribution taking into account coupling factors: Combine the road network topology and dynamic road condition information model to calculate the charging needs of different EV owners and achieve balanced coupling between the road network and the distribution network; Step 2: Phase 1: Establish an EV travel model based on travel chain technology, combine road congestion and temperature models to obtain the EV spatiotemporal load distribution in each period, then establish the attractiveness model of different EVCSs, and then obtain the service scope of EVCS through user selection probability; Step 3: Phase 2: Based on the EV spatiotemporal load distribution obtained in the first phase and the service range of each candidate charging station, the second phase comprehensively considers the EVCS construction scale, the impact of EV charging power on the distribution network, and the impact of user charging accessibility factors, and constructs a multi-type charging facility planning model from the "vehicle-station-network" three levels; Step 4: Improved whale optimization algorithm: When the demand points are known, a charging station site selection model for charging vehicles is established, and several points are selected as the site selection centers; with the minimum number of charging stations within the coverage area of ​​the charging station as the constraint condition, a function model is constructed with the goal of minimizing the sum of the product of the weight of the demand degree of each demand point within the coverage area of ​​the charging station and the distance from the charging station to the demand point, and then based on this site selection optimization problem, the improved whale optimization algorithm is used to solve the model.

2. The charging facility site selection method based on spatiotemporal analysis and multiple constraints as claimed in claim 1, characterized in that: The spatiotemporal distribution of EVs taking into account coupling factors in step 1 specifically includes the following steps: Step 1.1: EV Spatiotemporal Transfer Model: EV Travel Time It obeys the normal probability distribution, and its probability density function for: , where They are The variance and mean of the function; EV travel has Markov properties, from the current node Move to the next node The probability of : , where For Node The number of all feasible options for transfer to the end point; The next transfer node in all feasible paths Number of schemes; The number of nodes that the node transfer passes through in the current solution; is the number of path choices; is the number of adjacent nodes; Respectively in and Next Segment path length; Respectively in and Next EV driving speed of the segment path; Step 1.2: Dynamic road condition information model; Step 1.3: Road network-distribution network balanced coupling.

3. The charging facility site selection method based on spatiotemporal analysis and multiple constraints as claimed in claim 2, characterized in that: The dynamic traffic information model in step 1.2 specifically includes: Step 1.21: Congestion model: Determine the roads in each cycle based on the temporal and spatial distribution of vehicles Number of EVs in the system , the EV travel speed is determined by the interaction between the real-time number of vehicles on the road and the travel chain information Make corrections, i.e. dynamically update : , where They are The number of EVs within the unit distance of the road and on the road at the moment; is the correction constant; for The number of EVs on the road at the time; is the current road distance; Step 1.22: Energy consumption per mile model: EV energy consumption per mile, road grade and The relationship between: , where are the energy consumption per unit mileage of expressways, trunk roads and branch roads respectively; is the EV travel speed; however, the EV energy consumption per unit mileage is different at different temperatures, and the relationship between them is expressed as: , where For temperature Energy consumption per unit mileage; is the fitting parameter; is the current ambient temperature; is a natural number.

4. The charging facility site selection method based on spatiotemporal analysis and multiple constraints as claimed in claim 2, characterized in that: The road network-distribution network balanced coupling in step 1.3 specifically includes: Step 1.31: Spatial coupling: Considering the spatial overlap between the road network and the distribution network, the road network and distribution network nodes with the closest geographical locations are preferentially coupled, and the geographical location function is introduced. Characterizes the distance between the road network node and any distribution network node. This indicator is often the first indicator of coupling: , where is the node location of the distribution network; is the location of the road network node; Step 1.32: Coupling of distribution network support margin: Since there is a capacity upper limit on the distribution network nodes, different loads will have different impacts on the distribution network. Therefore, the capacity margin index is introduced. Used to describe the impact of EV load on distribution network operation: , where For distribution network nodes Maximum rated power; for Distribution network power supply node at all times The base load; For charging stations exist The power purchased from the distribution grid at all times; is the set of distribution network nodes.

5. The charging facility site selection method based on spatiotemporal analysis and multiple constraints as claimed in claim 1, characterized in that: The first stage in step 2 includes the following steps: Step 2.1: Probabilistic model of user selection of charging station: Based on the unit mileage energy consumption model, the user's real-time power consumption and resistance are calculated. The probability of a user choosing a specific charging station to charge depends on the ratio between the utility obtained by the user from the charging station and the sum of the utilities provided by all eligible charging stations in the planning area: , where For Node Users choose charging stations The probability of Node Users from charging stations the utility gained; For charging stations Attractiveness to users; For Node Users go to charging stations the resistance encountered; It is a collection of all candidate charging stations in the planning area; Step 2.2: EVCS service range: Assume that EV users charge when their battery power is less than a certain threshold. Then, the service range of the charging station is determined based on the probability of selecting different charging stations: , where For charging stations Scope of services; For Node Users choose charging stations The probability of is a set of road network nodes; Step 2.3: User charging accessibility.

6. The charging facility site selection method based on spatiotemporal analysis and multiple constraints as claimed in claim 5, characterized in that: The user charging accessibility in step 2.3 specifically includes: Step 2.31: For each supply point, search for all charging demand points within the search radius and calculate the supply-demand ratio : , where They are the number of slow charging piles, fast charging piles and ultra-high power charging piles; Respectively represent their rated charging power; For Node Total EV charging load at is the distance threshold; Step 2.32: For each demand point, search for all supply points within the search radius and sum up all supply-demand ratios to get the node Charging accessibility: , where For Node Charging accessibility for users.

7. The charging facility site selection method based on spatiotemporal analysis and multiple constraints as claimed in claim 1, characterized in that: The second stage in step 3 comprises the following steps: Step 3.1: User’s parking time: EV’s parking time Basically conforms to the exponential distribution, and its probability density function for: , where is the rate parameter of the exponential distribution; Step 3.2: The selection principle of the user's multi-power charging pile is: ① Give priority to meeting the needs within the specified time, that is, to meet the expected state of charge within the expected stay time of the owner; ② Because the electricity prices of different charging piles are different, if any charging pile can meet the owner's needs within the expected stay time, the slow charging pile with the best economy is given priority; Step 3.3: DESS configuration: The charging pile can obtain power from the DESS or directly from the grid, while the DESS obtains power directly from the grid. The power balance relationship between them is: , where For Node exist Base load at the moment; are the charging and discharging power of DESS respectively; They are charge and discharge states respectively; For charging stations The power purchased from the distribution grid at the time; Step 3.4: Objective function; Step 3.5: Constraints.

8. The charging facility site selection method based on spatiotemporal analysis and multiple constraints as claimed in claim 7, characterized in that: The objective function in step 3.4 is specifically: comprehensively consider the interests of EVCS investors, minimize the EVCS investment cost and operating cost, and at the same time meet the user charging accessibility and distribution network basic power flow constraints to establish an EVCS site selection and capacity determination model: , where To consider the objective function of the three aspects of "vehicle-station-network"; They are the annual construction and annual operation and maintenance costs of the charging station; System network loss costs; to purchase the land; It is the operation and maintenance cost of DESS.

9. The charging facility site selection method based on spatiotemporal analysis and multiple constraints as claimed in claim 7, characterized in that: The constraints in step 3.5 include: user charging accessibility constraints, system power flow constraints, voltage deviation constraints, current constraints, distribution network load constraints, charging station quantity constraints, and DESS charging and discharging constraints.

10. The charging facility site selection method based on spatiotemporal analysis and multiple constraints as claimed in claim 1, characterized in that: The improved whale optimization algorithm in step 4 includes the following steps: Step 4.1: Model establishment: The electric vehicle charging facility site selection model is established as follows: , the constraints are: , where the objective function It is the sum of the product of the weight of the demand degree of each demand point within the coverage area of ​​the charging station and the distance from the charging station to the demand point; for A collection of serial numbers of demand points; The maximum coverage radius of the site selection location; Indicates demand point Gathering at the selected location; Indicates the demand weight of the demand point; Indicates the distance from the demand point to the site selection location; is a variable between 0 and 1; Step 4.2: Whale Optimization Algorithm: Surrounding the Prey: In the formula, is the number of iterations; is the current position vector; is the optimal position vector; is the correlation coefficient vector; Step 4.3: Spiral Hunting: , where , is the displacement vector between the prey and the whale; Select probabilities for humpback whale behaviors; is the probability of surrounding the prey; is a constant used to define the spiral shape; is a random number in the interval [-1, 1]; Step 4.4: Search for prey: , where is a randomly obtained position vector; Step 4.5: Algorithm improvement: Use the cosine decreasing method to perform nonlinear decreasing of the convergence factor. The formula is: , where is the probability of the convergence factor cosine decreasing; is the convergence factor at the current iteration number The probability of behavior selection under is a random number between (0, 1).

Images