Electric vehicle rapid charging pile load prediction method considering traffic flow estimation

Abstract

The invention discloses an electric vehicle rapid charging pile load prediction method considering traffic flow estimation. The method comprises the following steps: constructing a model for searching a feasible path of an electric vehicle and a fuel vehicle in a traffic network; by analyzing the energy consumption behavior of the electric vehicle in the travel process, an electric vehicle battery electric quantity change model is constructed, and a traffic flow distribution model including the electric vehicle and the fuel vehicle is established with the goal of reaching the user equilibrium state to the maximum extent. And proposing three sub-problems and obtaining the load prediction level of the fast-charging type charging pile of the electric vehicle through a prediction algorithm. According to the method, the electric vehicle and the fuel vehicle are comprehensively considered, the load level of the electric vehicle fast-charging type charging pile can be predicted, the adverse effect of the uncertainty and volatility of the charging load of the electric vehicle on a power grid is reduced, a basis is provided for a power grid dispatching center to formulate a dispatching plan, and the operation efficiency of the power grid is optimized.

Description

Technical Field

[0001] The present invention relates to a technology in the field of fast-charging charging piles, and specifically to a method for predicting the load of electric vehicle fast-charging piles taking into account traffic flow estimation. Background Art

[0002] The existing prediction technology for fast-charging electric vehicle charging piles uses probabilistic methods and artificial intelligence-based methods. Probabilistic methods include Markov chain, Monte Carlo, etc. The reliability parameters of the probabilistic method system are calculated by mathematical formulas, so it has unique advantages in prediction speed and accuracy. Artificial intelligence-based methods are often used for short-term load forecasting. In the analysis process, it is not necessary to establish an accurate model of the object and can represent complex nonlinear problems. However, most of the current prediction technologies based on deep learning are based on historical data and do not take into account the theory of transportation networks. Summary of the invention

[0003] In view of the problem that the load level of existing fast-charging electric vehicle charging piles is highly uncertain and difficult to predict, the present invention proposes a load prediction method for electric vehicle fast-charging piles taking into account traffic flow estimation. By comprehensively considering electric vehicles and fuel vehicles, the load level of electric vehicle fast-charging piles can be predicted, and the adverse effects of the uncertainty and volatility of electric vehicle charging load on the power grid can be reduced, providing a basis for the power grid dispatching center to formulate a dispatching plan, and optimizing the power grid operation efficiency.

[0004] The present invention is achieved through the following technical solutions:

[0005] The present invention relates to a method for predicting the load of a fast charging pile for electric vehicles taking into account traffic flow estimation. A model for electric vehicles and fuel vehicles to search for feasible paths in a traffic network is constructed. By analyzing the energy consumption behavior of electric vehicles during travel, a model for the change in battery power of electric vehicles is constructed. A traffic flow distribution model including electric vehicles and fuel vehicles is established with the goal of maximizing the user equilibrium state. Three sub-problems are proposed and the load prediction level of a fast charging pile for electric vehicles is obtained through a prediction algorithm.

[0006] The model of electric vehicles and fuel vehicles searching for feasible paths in the traffic network refers to: constructing an optimal path selection problem based on the traffic network, including: a node set θ consisting of a starting point, a garage, a parking lot, etc. as an end point, and a junction, etc. as a relay point N and the set of lines connecting the points Θ A The attributes of the participants in the operation of the traffic network include: the starting point (Origination, o), the destination (Destination, d), and the selected path plan (path, q). All traffic flows with the same starting point and destination are an od pair, and the electric vehicle charging stations are located at the intersections of the roads.

[0007] In the model, each od pair of cars starts from the starting point and reaches the end point, and the constraints of its path include:

[0008] ① A set of OD path solutions for traffic flow selection must be based on existing roads in: is a binary variable, when When , it means that the traffic flow of this od pair passes through the road between node i and node j, and is from node i to node j; when When , it means that the traffic flow of this group of od pairs did not go from node i to node j through the road between nodes i and j. as auxiliary binary variables. is a constant. When there is a road from node i to node j, otherwise

[0009] ② Each road intersection (including the path starting point and the path end point) is either not passed or only passed once in: is a binary variable, when When , it means that the traffic flow of this od pair passes through the road between node i and node j, and is from node i to node j; when When , it means that the traffic flow of this group of od pairs did not go from node i to node j through the road between nodes i and j.

[0010] ③ Each road intersection (except the path starting point and path end point) must be the starting point of a road the same number of times as the end point of another road Among them, k o and k d They are the starting point and the end point respectively.

[0011] ④The path start point (path end point) must be the start point (end point) of a certain road. in: is a binary variable, when When , the traffic flow of this od pair passes through the road between node i and node j, and is from node i to node j; when When , the traffic flow of this group of od pairs does not go from node i to node j through the road between nodes i and j; is an auxiliary binary variable; is a constant. When there is a road from node i to node j, otherwise k o and k dThey are the starting point and the end point respectively.

[0012] The electric vehicle battery power variation model includes:

[0013] 1) When the electric vehicle chooses the q path at point i(j), the power consumption satisfies: in: The power of the electric vehicle when it chooses path q at point i(j), D ij is the energy consumption of the electric vehicle on the road section (i, j), is the charge amount of the electric vehicle when it chooses the q path at point j, is the auxiliary variable, M is an infinite number, is the length of the road from node i to node j, and is the power consumption per unit length.

[0014] 2) The power of the electric vehicle at the starting point is equal to the sum of the initial power of the electric vehicle and the charging power at the starting point in: is the initial charge of the electric vehicle.

[0015] 3) The upper and lower limits of electric vehicle battery capacity meet the following requirements: Where: E min is the lower limit of battery capacity, E max The upper limit of battery capacity.

[0016] 4) Considering the mileage anxiety of the car owner, that is, the worry caused by the low battery when driving an electric car, the difference between the battery power of the electric car when the electric car chooses the q path at point i(j) and the energy consumption of the electric car on the road section (i, j) satisfies: in: This is the range anxiety for electric car drivers, which means that the driver must ensure that the battery level is greater than this value.

[0017] 5) The relationship between the charging capacity of the electric vehicle and the charging pile satisfies: in: is a binary variable. When there is a charging pile at point i, is 1, otherwise is 0.

[0018] In the electric vehicle battery power change model, when the electric vehicle chooses to charge, the battery will be fully charged: in: is a binary variable. When the electric vehicle decides to charge at node m, is 1, otherwise is 0, Θ evcs It is the collection of nodes where electric vehicle charging piles are located.

[0019] The traffic flow distribution model including electric vehicles and fuel vehicles includes:

[0020] in: is the traffic flow of the electric vehicle choosing the q path for this group of rs, W rs is the total traffic volume of electric vehicles in this group of rs, x a is the traffic volume of section a, x m is the traffic volume of charging pile m, Θ evcs It is the collection of nodes where electric vehicle charging piles are located.

[0021] Where: t a is the time to pass through road a, is the free travel time, u a is the road capacity; the quartic term in the expression is eliminated by piecewise linearization; Where: t m is the time taken by the electric vehicle to charge at the charging pile at node m, is the charging rated power, is the fixed service time, u m is a constant coefficient used to characterize the congestion effect.

[0022] in: is the charging load of charging station m, in: The maximum capacity of the charging pile.

[0023] in: is a binary variable, is a slack variable, rs is a set of od pairs, is the traffic flow of path q in the path set of the pair rs. is the travel cost of the electric vehicle choosing path q for this group of od, c rs is the minimum travel cost of electric vehicles for this group of od, Θ rs is the path set of the od pair of rs, Θ OD is a set of od pairs.

[0024] in: is the cost of choosing path q, k is the coefficient of converting time cost into money cost, ρ average is the average electricity price of all charging piles, is the charging electricity price of the charging pile at node m.

[0025] The decomposition refers to decomposing the entire problem into a simplified main problem, an optimal path generation subproblem, and a feasible path generation subproblem, specifically:

[0026] a) Simplified main problem: For all od pairs of traffic, after the optimal path generation subproblem and the feasible path generation subproblem generate a path set and corresponding charging options for each od pair of traffic, the traffic flow of each path is calculated for all od pairs of traffic according to the approximate state of user equilibrium. The objective function is to minimize the slack variable, specifically: Restrictions include:

[0027] b) Optimal path generation sub-problem: For a pair of od pairs of traffic, find the optimal path with the lowest comprehensive cost. Specifically: on the basis of satisfying all the restrictions of the original problem Where: t′ a and t′ m are obtained by solving the simplified main problem. a and t m , which are known constants for the optimal path generation subproblem.

[0028] c) Feasible path generation sub-problem: For a pair of od pairs of traffic, in order to overcome the capacity limitation of charging piles, the simplified main problem has no feasible solution, and find a feasible path with the lowest comprehensive cost. Specifically, on the basis of satisfying all the restrictions of the original problem and in: is the set of charging choices for the paths found in the previous calculation of the traffic flow, γ q，before，i is a vector, representing all charging options for a path in the charging option set of the path found in the previous calculation of the traffic flow by this set of od, is a binary variable, which is the charging choice of one of the charging piles for a certain path in the traffic path set of this set of OD.

[0029] The prediction algorithm specifically includes:

[0030] Step 1) Set the relevant parameters of the charging pile and the transportation network and initialize t′ a and t′ m , where: the superscript ' represents the known quantity in the optimal path generation subproblem and the feasible path generation subproblem, and the non-superscript is the result obtained after solving the subproblem;

[0031] Step 2) Solve the optimal path generation sub-problem for each set of od pairs of vehicles, and update the path set for each set of od pairs;

[0032] Step 3) When all ODs for the traffic flow are not updated after step 2), the algorithm process is terminated, and the calculation results such as the charging load of the charging pile are obtained;

[0033] Step 4) Solve the simplified main problem;

[0034] Step 5) When the simplified main problem obtains a feasible solution, we can obtain t a and t m , and use t a and t m Update t′ a and t m Then return to step 2) to continue;

[0035] Step 6) When the simplified main problem does not obtain a feasible solution, solve the feasible path generation sub-problem for each set of od pairs of traffic flow, and update the path set of each set of od pairs and return to step 4) to continue. Technical Effects

[0036] The present invention combines traffic network theory, considers the impact of traffic flow on the load level of fast charging piles, estimates traffic flow through traffic network theory, and then predicts the charging load of fast charging piles. Compared with the existing technology, the prediction result is more reasonable. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flow chart of the present invention;

[0038] Figure 2 It is the flow chart of prediction algorithm;

[0039] Figure 3 This is an example effect diagram. DETAILED DESCRIPTION

[0040] like Figure 1 As shown, a method for predicting the load of an electric vehicle fast charging pile taking into account traffic flow estimation involved in this embodiment includes:

[0041] Step 1: The traffic network operation topology adopted is based on the road network within the second ring road of Xi'an City, Shaanxi Province, with a total of 41 traffic network nodes and 72 two-way roads. The road network in the example is designed to simulate a real-world city. This method is suitable for large-scale road networks, and its operation model is adapted to the highway system. The example sets 15 electric vehicle charging piles, located at nodes 1, 3, 4, 7, 10, 13, 17, 19, 20, 24, 28, 32, 33, 35, and 40. The traffic demand of electric vehicle drivers is simulated by 50 groups of od pairs, and the traffic flow of each group of od pairs of electric vehicles is normally distributed on the time scale. The coefficient k for converting time cost into money cost is 3, the battery capacity of the electric vehicle is set to 32kwh, the constant coefficient used to characterize the congestion effect is set to 0.5, and it is assumed that the upper limit of the battery power is set to 32kwh, the lower limit of the power is set to 3.2kwh, and the power consumption per unit distance is set to 0.15kwh / km.

[0042] Step 2: Build a model for electric vehicles and fuel vehicles to search for feasible paths in the traffic network;

[0043] Step 3: Establish a model of electric vehicle power changes;

[0044] Step 4: Establish a vehicle traffic network model;

[0045] Step 5: Decompose the entire problem into three sub-problems, namely: the simplified main problem, the optimal path generation sub-problem and the feasible path generation sub-problem.

[0046] Step 6: Use the prediction algorithm to solve the three sub-problems obtained in step 5, including:

[0047] 6.1) Set the relevant parameters of charging piles and transportation network, and initialize t ′ a and t ′ m ;

[0048] 6.2) Solve the optimal path generation subproblem for each set of od pairs of vehicles, and update the path set for each set of od pairs;

[0049] 6.3) When all ODs for traffic flow are not updated after 6.2), the algorithm process is terminated and the calculation results such as charging pile charging load are obtained;

[0050] 6.4) Solve the simplified main problem;

[0051] 6.5) When the simplified main problem obtains a feasible solution, we can obtain t a and t m , and use t a and t m Update′ a and t m Then return to 6.2) to continue;

[0052] 6.6) When the simplified main problem does not yield a feasible solution, solve the feasible path generation subproblem for each set of od pairs of traffic flow, and update the path set for each set of od pairs and return to 6.4) to continue.

[0053] After specific practical experiments, the experimental data that can be obtained under the specific environment settings of calling the commercial solver gurobi in python are as follows: The load level of the charging station within one hour of the simulation experiment is shown in Table 1 and Figure 3 As shown:

[0054] Table 1 Charging station node number Charging station load / (kwh) 1 545.733 3 637.5413 4 1056.773 7 698.1378 10 925.0046 13 1143.072 17 1999.968 19 1191.068 20 1245.29 24 1122.154 28 2000 32 1183.725 33 1028.535 35 736.2856 40 1011.845

[0055] Compared with the existing technology, the prediction technology of this method takes into account the influence of traffic network factors and introduces Wardrop user equilibrium theory.

[0056] The above-mentioned specific implementation can be partially adjusted in different ways by those skilled in the art without departing from the principle and purpose of the present invention. The protection scope of the present invention shall be based on the claims and shall not be limited by the above-mentioned specific implementation. Each implementation scheme within its scope shall be subject to the constraints of the present invention.

Claims

1. A method for predicting the load of electric vehicle fast charging piles taking into account traffic flow estimation, characterized in that: Construct a model for electric vehicles and fuel vehicles to search for feasible paths in the traffic network; by analyzing the energy consumption behavior of electric vehicles during travel, build an electric vehicle battery power change model, and establish a traffic flow distribution model including electric vehicles and fuel vehicles with the goal of maximizing the user equilibrium state. Three sub-problems are proposed and the load prediction level of electric vehicle fast charging piles is obtained through prediction algorithms; The model of electric vehicles and fuel vehicles searching for feasible paths in the traffic network refers to: constructing an optimal path selection problem based on the traffic network, including: a node set θ consisting of a starting point, a garage, a parking lot, etc. as an end point, and a junction, etc. as a relay point N and the set of lines connecting the points Θ A The attributes of the participants in the operation of the traffic network include: the starting point (Origination, o), the destination (Destination, d), and the selected path plan (path, q). All traffic flows with the same starting point and destination are an od pair, and the electric vehicle charging stations are located at the intersections of the roads.

2. The method for predicting the load of electric vehicle fast charging piles taking into account traffic flow estimation according to claim 1 is characterized in that: In the model, each od pair of cars starts from the starting point and reaches the end point, and the constraints of its path include: ① A set of OD path solutions for traffic flow selection must be based on existing roads in: is a binary variable, when When , it means that the traffic flow of this od pair passes through the road between node i and node j, and is from node i to node j; when When , it means that the traffic flow of this group of od pairs did not go from node i to node j through the road between nodes i and j. As auxiliary binary variables, is a constant. When there is a road from node i to node j, otherwise ② Each road intersection (including the path starting point and the path end point) is either not passed or only passed once in: is a binary variable, when When , it means that the traffic flow of this od pair passes through the road between node i and node j, and is from node i to node j; when When , it means that the traffic flow of this group of od pairs does not go from node i to node j through the road between nodes i and j; ③ Each road intersection (except the path starting point and path end point) must be the starting point of a road the same number of times as the end point of another road ④The path start point (path end point) must be the start point (end point) of a certain road. in: is a binary variable, when When , the traffic flow of this od pair passes through the road between node i and node j, and is from node i to node j; when When , the traffic flow of this group of od pairs does not go from node i to node j through the road between nodes i and j; is an auxiliary binary variable; is a constant. When there is a road from node i to node j, otherwise k o and k d They are the starting point and the end point respectively.

3. The method for predicting the load of electric vehicle fast charging piles taking into account traffic flow estimation according to claim 1 is characterized in that: The electric vehicle battery power variation model includes: 1) When the electric vehicle chooses the q path at point i(j), the power consumption satisfies: in: The power of the electric vehicle when it chooses path q at point i(j), D ij is the energy consumption of the electric vehicle on the road section (i, j), is the charge amount of the electric vehicle when it chooses the q path at point j, is the auxiliary variable, M is an infinite number, is the length of the road from node i to node j, and is the power consumption per unit length; 2) The power of the electric vehicle at the starting point is equal to the sum of the initial power of the electric vehicle and the charging power at the starting point in: is the initial power of the electric vehicle; 3) The upper and lower limits of electric vehicle battery capacity meet the following requirements: Where: E min is the lower limit of battery capacity, E max is the upper limit of battery capacity; 4) Considering the mileage anxiety of the car owner, that is, the worry caused by the low battery when driving an electric car, the difference between the battery power of the electric car when the electric car chooses the q path at point i(j) and the energy consumption of the electric car on the road section (i, j) satisfies: in: The range anxiety of electric car drivers, that is, the driver must ensure that the power is greater than this value; 5) The relationship between the charging capacity of the electric vehicle and the charging pile satisfies: in: is a binary variable. When there is a charging pile at point i, is 1, otherwise is 0.

4. The method for predicting the load of electric vehicle fast charging piles taking into account traffic flow estimation according to claim 1 is characterized in that: In the electric vehicle battery power change model, when the electric vehicle chooses to charge, the battery will be fully charged: in: is a binary variable. When the electric vehicle decides to charge at node m, is 1, otherwise is 0, Θ evcs It is the collection of nodes where electric vehicle charging piles are located.

5. The method for predicting the load of electric vehicle fast charging piles taking into account traffic flow estimation according to claim 1 is characterized in that: The traffic flow distribution model including electric vehicles and fuel vehicles includes: in: is the traffic flow of the electric vehicle choosing the q path for this group of rs, W rs is the total traffic volume of electric vehicles in this group of rs, x a is the traffic volume of section a, x m is the traffic volume of charging pile m, Θ evcs It is the collection of nodes where the electric vehicle charging piles are located; ii) Where: t a is the time to pass through road a, is the free travel time, u a is the road capacity; the quartic term in the expression is eliminated by piecewise linearization method; Where: t m is the time taken by the electric vehicle to charge at the charging pile at node m, is the charging rated power, is the fixed service time, u m is a constant coefficient used to characterize the congestion effect; iii) in: is the charging load of charging station m, in: is the maximum capacity of the charging pile; iv) in: is a binary variable, is a slack variable, rs is a set of od pairs, is the traffic flow of path q in the path set of the pair rs. is the travel cost of the electric vehicle choosing path q for this group of od, c rs is the minimum travel cost of electric vehicles for this group of od, Θ rs is the path set of the od pair of rs, Θ OD is the set of od pairs; v) in: is the cost of choosing path q, k is the coefficient of converting time cost into money cost, ρ average is the average electricity price of all charging piles, is the charging electricity price of the charging pile at node m.

6. The method for predicting the load of electric vehicle fast charging piles taking into account traffic flow estimation according to claim 1, characterized in that: The decomposition refers to decomposing the entire problem into a simplified main problem, an optimal path generation subproblem, and a feasible path generation subproblem, specifically: a) Simplified main problem: For all od pairs of traffic, after the optimal path generation subproblem and the feasible path generation subproblem generate a path set and corresponding charging options for each od pair of traffic, the traffic flow of each path is calculated for all od pairs of traffic according to the approximate state of user equilibrium. The objective function is to minimize the slack variable, specifically: Restrictions include: b) Optimal path generation sub-problem: For a pair of od pairs of traffic, find the optimal path with the lowest comprehensive cost. Specifically: on the basis of satisfying all the restrictions of the original problem Where: t′ a and t′ m are obtained by solving the simplified main problem. a and t m , which are known constants for the optimal path generation subproblem; c) Feasible path generation sub-problem: For a pair of od pairs of traffic, in order to overcome the capacity limitation of charging piles, the simplified main problem has no feasible solution, and find a feasible path with the lowest comprehensive cost. Specifically, on the basis of satisfying all the restrictions of the original problem and in: is the set of charging choices for all paths of the traffic flow by this group of OD, γ q，before，i is a vector, representing all charging choices of this set of od for a certain path in the traffic path set, is a binary variable, which is the charging choice of one of the charging piles for a certain path in the traffic path set of this set of OD.

7. The method for predicting the load of electric vehicle fast charging piles taking into account traffic flow estimation according to any one of claims 1 to 6, characterized in that: The prediction algorithm specifically includes: Step 1) Set the relevant parameters of the charging pile and the transportation network and initialize t′ a and t′ m , where: the superscript ' represents the known quantity in the optimal path generation subproblem and the feasible path generation subproblem, and the non-superscript is the result obtained after solving the subproblem; Step 2) Solve the optimal path generation sub-problem for each set of od pairs of vehicles, and update the path set for each set of od pairs; Step 3) When all ODs for the traffic flow are not updated after step 2), the algorithm process is terminated, and the calculation results such as the charging load of the charging pile are obtained; Step 4) Solve the simplified main problem; Step 5) When the simplified main problem obtains a feasible solution, we can obtain t a and t m , and use t a and t m Update t′ a and t m Then return to step 2) to continue; Step 6) When the simplified main problem does not obtain a feasible solution, solve the feasible path generation sub-problem for each set of od pairs of traffic flow, and update the path set of each set of od pairs and return to step 4) to continue.

Images