Electric vehicle charging load space-time guiding method and device based on dynamic electricity price

Abstract

The invention relates to the technical field of power system automatic control, and particularly provides an electric vehicle charging load space-time guiding method and device based on dynamic electricity price, and the method comprises the steps: carrying out the simulation of the travel and energy consumption process of an electric vehicle in a target region through employing a Monte Carlo method based on a pre-constructed electric vehicle space-time travel model, charging events of the vehicles are counted, and a reference charging load space-time distribution simulation result of the target area is formed through aggregation; substituting the reference charging load space-time distribution simulation result of the target area into a pre-constructed double-layer optimization model and carrying out iterative solution to obtain a dynamic electricity price strategy and a charging plan of the target area; and performing charging load space-time guidance on the electric vehicle in the target area by using the dynamic electricity price strategy and the charging plan of the target area. According to the technical scheme provided by the invention, the time-space dynamic electricity price and the charging plan are formulated, so that the power grid load peak-valley difference and the user charging cost are reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automatic control technology of power system, and particularly relates to a method and device for spatiotemporal guidance of electric vehicle charging load based on dynamic electricity price. BACKGROUND

[0002] With the transformation of global energy structure and the enhancement of environmental awareness, the market penetration rate of electric vehicles is growing at an unprecedented rate due to their clean and efficient characteristics. However, the large-scale unordered access of electric vehicles to the power grid and charging, especially the concentrated charging behavior during the peak period of residential electricity, will greatly exacerbate the peak-valley load difference of the power grid, posing a serious challenge to the safe and stable operation of the distribution network. Therefore, how to effectively guide the wide and spatiotemporally mobile electric vehicle charging load so as to change it from a "load" to a "resource" that can be dispatched has become a key technical problem to be solved.

[0003] To address the above challenges, various electric vehicle charging management methods have been proposed in the prior art. One basic method is to charge freely by the user under a fixed electricity price, and the user decides the charging time and place according to his own habits. Another widely used method is to implement a time-of-use (TOU) strategy, which divides a day into several time periods (such as peak time, flat time, and valley time), and sets different fixed electricity prices for different time periods, in order to guide users to charge during the valley time when the electricity price is lower through economic leverage. In addition, there are also some guidance methods that publish power grid load state information to users through one-way communication to suggest that users stagger charging.

[0004] Although the prior art alleviates the overlap problem of charging load and peak electricity to some extent by implementing a time-of-use strategy, there are still some deficiencies:

[0005] Firstly, the guidance granularity of the existing price guidance mechanism is relatively rough, and it cannot achieve accurate smoothing of the charging load. Taking the time-of-use as an example, it only provides a few price signals at a large time scale. This leads to a large number of price-sensitive users to set the charging start time at the same time at the moment when the valley time electricity price takes effect, thereby transferring the original peak electricity to a new and severe charging load rebound peak, and the peak-valley difference problem of the distribution network has not been fundamentally solved. The root cause lies in that this fixed and non-dynamic electricity price cannot be adaptively adjusted according to the real-time load level of the power grid and the actual changes of the charging demand, and lacks closed-loop feedback regulation capability for the load.

[0006] Secondly, the existing price guiding mechanism generally lacks the guiding ability of spatial dimension. Whether it is a fixed electricity price or a time-of-use electricity price, the price signal is usually uniform in geographical space, that is, all charging facilities in the same area perform the same electricity price in the same period. This one-size-fits-all strategy in space ignores the natural differences in load characteristics and user charging demand of different functional areas (such as residential areas and working areas) in the city. For example, the power grid load of the working area is high during the day, but there is a large margin at night, and a large number of electric vehicles stay here during the day, which has charging potential. The existing technology cannot develop differentiated spatial electricity prices to exploit and utilize this time-space complementary characteristics, resulting in that the charging load is still excessively concentrated in the residential area at night, and the distributed receiving capacity of the entire urban power distribution network cannot be fully utilized.

[0007] Finally, some existing guiding methods have oversimplified user behavior and vehicle state at the modeling level. For example, when predicting charging demand, the braking energy recovery of the vehicle during actual driving is not taken into account, which may cause misjudgment of the actual required charging amount of the vehicle. At the same time, the real habit of users tending to continuous charging is also less considered, resulting in that the optimized charging strategy may be difficult to be adopted in reality due to not meeting the user's preference. These inaccuracies at the model level fundamentally weaken the accuracy of load prediction and the actual effectiveness of the guiding strategy. SUMMARY

[0008] In order to overcome the above defects, the present application provides a dynamic electricity price based electric vehicle charging load space-time guiding method and device.

[0009] In a first aspect, a dynamic electricity price based electric vehicle charging load space-time guiding method is provided, which comprises:

[0010] Based on the pre-constructed electric vehicle space-time travel model, the Monte Carlo method is used to simulate the travel and energy consumption process of electric vehicles in the target area, and the charging events of the vehicles are counted and aggregated to form the simulation results of the baseline charging load space-time distribution of the target area;

[0011] The simulation results of the baseline charging load space-time distribution of the target area are substituted into the pre-constructed double-layer optimization model and iteratively solved to obtain the dynamic electricity price strategy and charging plan of the target area;

[0012] The dynamic electricity price strategy and charging plan of the target area are used to guide the charging load space-time of electric vehicles in the target area.

[0013] Preferably, the electric vehicle space-time travel model is as follows:

[0014] L i = ((s i0 ,ti0 ),(s i1 ,t i1 ),...,(s in ,t in ))

[0015] In the above formula, L i is the trip chain of the i-th electric vehicle; i is the index of the electric vehicle number; s in is the n-th node reached by the i-th electric vehicle user in the target area; t in is the stay time of the i-th electric vehicle user after reaching the n-th node in the target area; s i0 is the starting node of the i-th electric vehicle user in the target area; t i0 is the stay time of the i-th electric vehicle user at the starting node in the target area; and n is the node sequence number in the trip chain.

[0016] Further, in the process of simulating the trip and energy consumption process of electric vehicles in the target area by the Monte Carlo method, the starting trip time of the electric vehicle user is fitted by using a mixed Gaussian model, the stay time of the electric vehicle user after reaching the node of the residential area in the target area is fitted by using a Weibull distribution model, and the stay time of the electric vehicle user after reaching the node of the work area and other areas in the target area is fitted by using a generalized extreme value distribution model.

[0017] Preferably, in the process of simulating the trip and energy consumption process of electric vehicles in the target area by the Monte Carlo method, the remaining electric quantity of the electric vehicle is updated according to the following formula:

[0018]

[0019] In the above formula, is the remaining electric quantity of the electric vehicle at the t period; B is the rated electric quantity of the electric vehicle battery; is the total electric consumption quantity of the electric vehicle at the t period; is the total recovered electric quantity of the electric vehicle at the t period;

[0020] The total recovered electric quantity of the electric vehicle at the t period is as follows:

[0021]

[0022] E f =∫δmgvdt

[0023] E w =∫AC d ρvdt

[0024] In the above formula, ω is a brake distribution factor; E ν , Ef , E w are the braking energy consumption of the electric vehicle, the energy consumption for overcoming the rolling resistance and the energy consumption for overcoming the wind resistance, respectively; v is the vehicle speed; v0 and v1 are the initial and final speeds of braking, respectively; δ and C d are the rolling resistance coefficient and the wind resistance coefficient, respectively; m is the vehicle mass; g is the gravitational acceleration; A is the vehicle frontal area; and ρ is the air density.

[0025] Preferably, in the process of simulating the travel and energy consumption process of the electric vehicle in the target area by using the Monte Carlo method, the shortest path of the electric vehicle from node n to node m in the target area is determined according to the following formula:

[0026] dist[n][m] = min(dist[n][m], dist[n][s] + dist[s][m])

[0027] In the above formula, dist[n][m] is the shortest path length from node n to node m; dist[n][s] is the shortest path length from node n to the transfer node s; dist[s][m] is the shortest path length from the transfer node s to node j; n, m and s are the index numbers of the starting node, the terminal node and the transfer node, respectively.

[0028] Preferably, the pre-constructed double-layer optimization model includes an upper-layer objective function and its corresponding upper-layer constraint condition, which take minimizing the total variance of the daily load of the power grid after considering the charging load of the electric vehicle as the optimization objective, and a lower-layer objective function and its corresponding lower-layer constraint condition, which take minimizing the total charging cost of all users as the optimization objective.

[0029] Further, the upper-layer objective function is as follows:

[0030]

[0031] In the above formula, γ is the variance of the daily load of the power grid; T is the total number of time periods in a day; P grid,t is the total load of the power grid at time period t; is the average total load per day.

[0032] Further, the upper-layer constraint condition is as follows:

[0033]

[0034] U min ≤ U t ≤ U max

[0035] (1-ζ v )V t ≤ V t≤ (1 + ζ v V t

[0036] |V t+1 -V t |≤ΔV

[0037] In the above formula, is the power of the ith electric vehicle accessing the power grid at time t; is the maximum power of the distribution network accessing at time t; N is the total number of electric vehicles; U min is the lower limit of the node voltage; U max is the upper limit of the node voltage; U t is the voltage of the node at time t; V t is the electricity price at time t; V t+1 is the electricity price at time t+1; ζ v is the price fluctuation coefficient; ΔV is the price fluctuation threshold of adjacent time periods.

[0038] Further, the lower layer objective function is as follows:

[0039] Or

[0040] In the above formula, C is the total charging cost of all electric vehicles in the region; is the charging amount of the ith electric vehicle at time t; N is the total number of electric vehicles; V t is the electricity price at time t; S is the total utility of users; S t is the utility function at time t; η t is the charging state at time t; η t-1 is the charging state at time t-1; η t+1 is the charging state at time t+1; S t (η t-1 ,η t ,η t+1 ) = V t ·η t +λ1·η t-1 ·η t +λ2·η t ·η t+1 ; V t is the electricity price at time t; λ1, λ2 are continuous charging adjustment parameters.

[0041] Further, the lower layer constraint condition is as follows:

[0042]

[0043] In the above formula, is the charging amount of the ith electric vehicle at time t; Qi(t) is the charging demand of the ith electric vehicle at time t; Δt is the duration of a single time period; Pi is the maximum charging power of the ith electric vehicle; B is the rated capacity of the battery.

[0044] In a second aspect, a dynamic price-based electric vehicle charging load space-time guidance device is provided, which comprises:

[0045] A simulation module is configured to simulate the travel and energy consumption process of electric vehicles in a target region based on a pre-constructed electric vehicle space-time travel model using a Monte Carlo method, and to count and aggregate the charging events of the vehicles to form a simulation result of the reference charging load space-time distribution of the target region.

[0046] An analysis module is configured to substitute the simulation result of the reference charging load space-time distribution of the target region into a pre-constructed double-layer optimization model and iteratively solve the model to obtain a dynamic price strategy and a charging plan for the target region.

[0047] A guidance module is configured to guide the charging load space-time of electric vehicles in the target region using the dynamic price strategy and the charging plan for the target region.

[0048] Preferably, the electric vehicle space-time travel model is as follows:

[0049] L i = ((s i0 ,t i0 ),(s i1 ,t i1 ),...,(s in ,t in ))

[0050] In the above formula, L i is the travel chain of the ith electric vehicle; i is the index of the electric vehicle; s in is the nth node reached by the user of the ith electric vehicle in the target region; t in is the stay time of the user of the ith electric vehicle after reaching the nth node in the target region; s i0 is the starting node of the user of the ith electric vehicle in the target region; t i0 is the stay time of the user of the ith electric vehicle at the starting node in the target region; and n is the node sequence number in the travel chain.

[0051] Further, in the process of simulating the travel and energy consumption of the electric vehicle in the target area by using the Monte Carlo method, the starting travel time of the electric vehicle user is fitted by using a Gaussian mixture model, the stay time of the electric vehicle user after arriving at the node of the residential area in the target area is fitted by using a Weibull distribution model, and the stay time of the electric vehicle user after arriving at the node of the working area and other areas in the target area is fitted by using a generalized extreme value distribution model.

[0052] Preferably, in the process of simulating the travel and energy consumption of the electric vehicle in the target area by using the Monte Carlo method, the residual electric quantity of the electric vehicle is updated according to the following formula:

[0053]

[0054] In the above formula, is the residual electric quantity of the electric vehicle at the t period; B is the rated electric quantity of the electric vehicle battery; is the total electric consumption of the electric vehicle at the t period; is the total recovered electric quantity of the electric vehicle at the t period;

[0055] The total recovered electric quantity of the electric vehicle at the t period is as follows:

[0056]

[0057] E f =∫δmgvdt

[0058] E w =∫AC d ρvdt

[0059] In the above formula, ω is a brake distribution factor; E ν , E f , and E w are the energy consumption of the electric vehicle braking, the energy consumption of overcoming the rolling resistance, and the energy consumption of overcoming the wind resistance, respectively; v is the vehicle speed; v0 and v1 are the initial speed and the final speed of braking, respectively; δ and C d are the rolling resistance coefficient and the wind resistance coefficient, respectively; m is the vehicle mass; g is the gravitational acceleration; A is the vehicle wind area; and ρ is the air density.

[0060] Preferably, in the process of simulating the travel and energy consumption of the electric vehicle in the target area by using the Monte Carlo method, the shortest path of the electric vehicle from the node n to the node m in the target area is determined according to the following formula:

[0061] dist[n][m]=min(dist[n][m],dist[n][s]+dist[s][m])

[0062] In the above formula, dist[n][m] is the shortest path length between node n and node m; dist[n][s] is the shortest path length between node n and transit node s; dist[s][m] is the shortest path length between transit node s and node j; n, m, s are the index numbers of the starting node, the terminal node and the transit node, respectively.

[0063] Preferably, the pre-constructed double-layer optimization model comprises: an upper-layer objective function and its corresponding upper-layer constraint condition, which take minimizing the total variance of the daily load of the power grid after considering the charging load of the electric vehicles as the optimization objective; and a lower-layer objective function and its corresponding lower-layer constraint condition, which take minimizing the total charging cost of all users as the optimization objective.

[0064] Further, the upper-layer objective function is as follows:

[0065]

[0066] In the above formula, γ is the variance of the daily load of the power grid; T is the total number of time periods in a day; P grid,t is the total load of the power grid at time period t; is the average total load per day.

[0067] Further, the upper-layer constraint condition is as follows:

[0068]

[0069] U min ≤ U t ≤ U max

[0070] (1-ζ v )V t ≤ V t ≤ (1+ζ v )V t

[0071] |V t+1 -V t |≤ΔV

[0072] In the above formula, is the power of the i-th electric vehicle connected to the power grid at time period t; is the maximum connected power of the distribution network at time period t; N is the total number of electric vehicles; U min is the lower limit of the node voltage; U max is the upper limit of the node voltage; U t is the voltage of the node at time period t; V t is the electricity price at time period t; V t+1 is the electricity price at time period t+1; ζ v is the price fluctuation coefficient; and ΔV is the fluctuation threshold of the electricity price of adjacent time periods.

[0073] Further, the lower layer objective function is as follows:

[0074] Or

[0075] In the above formula, C is the total charging cost of all electric vehicles in the region; is the charging amount of the i-th electric vehicle at time t; N is the total number of electric vehicles; V t is the electricity price at time t; S is the total user utility; S t is the utility function at time t; η t is the charging state at time t; η t-1 is the charging state at time t-1; η t+1 is the charging state at time t+1; S t (η t-1 ,η t ,η t+1 )=V t ·η t +λ1·η t-1 ·η t +λ2·η t ·η t+1 ; V t is the electricity price at time t; λ1, λ2 are continuous charging adjustment parameters.

[0076] Further, the lower layer constraint condition is as follows:

[0077]

[0078] In the above formula, is the charging amount of the i-th electric vehicle at time t; N is the total number of electric vehicles; V is the charging demand of the i-th electric vehicle at time t; Δt is the duration of a single time period; is the maximum charging power of the i-th electric vehicle; B is the rated battery capacity.

[0079] In a third aspect, a computer device is provided, comprising: one or more processors;

[0080] The processor is configured to store one or more programs;

[0081] When the one or more programs are executed by the one or more processors, the dynamic price-based electric vehicle charging load space-time guidance method is implemented.

[0082] In a fourth aspect, a computer readable storage medium is provided, having a computer program stored thereon, which, when executed, implements the dynamic price-based electric vehicle charging load space-time guidance method.

[0083] The one or more technical solutions of the present application have at least one or more of the following beneficial effects:

[0084] The present application provides a dynamic electricity price-based electric vehicle charging load space-time guidance method and device, comprising: based on a pre-constructed electric vehicle space-time travel model, the travel and energy consumption process of electric vehicles in a target area are simulated by using the Monte Carlo method, and the charging events of the vehicles are counted to form a simulation result of the reference charging load space-time distribution of the target area; the simulation result of the reference charging load space-time distribution of the target area is substituted into a pre-constructed double-layer optimization model and iteratively solved to obtain a dynamic electricity price strategy and a charging plan for the target area; and the dynamic electricity price strategy and the charging plan for the target area are used to guide the charging load space-time of electric vehicles in the target area. The technical scheme provided by the present application reduces the peak-valley difference of power grid load and the charging cost of users by formulating a space-time dynamic electricity price and a charging plan, in particular:

[0085] 1. The present application constructs an electric vehicle space-time travel model, which divides functional areas and uses a mixed Gaussian model, a Weibull distribution and a generalized extreme value distribution to probabilistically model the travel time and stay time of users, so as to more accurately predict the space-time distribution of charging demand driven by user travel purposes. Compared with traditional prediction methods that do not distinguish travel purposes, the technical scheme significantly improves the accuracy of the prediction of the reference charging load space-time distribution.

[0086] 2. The double-layer optimization model constructed by the present application minimizes the variance of power grid load in the upper layer and minimizes the charging cost of the user group in the lower layer, and solves the game model by using an iterative algorithm such as particle swarm. This structure systematically couples the stability requirements of the power grid side and the economic requirements of the user side, and the generated reference dynamic electricity price strategy can achieve "peak load shifting" of the power grid while obtaining high user acceptance due to its economic rationality, thereby improving the effectiveness of the guidance strategy. BRIEF DESCRIPTION OF DRAWINGS

[0087] Figure 1 FIG. 1 is a main step flowchart of the dynamic electricity price-based electric vehicle charging load space-time guidance method of the embodiment of the present application;

[0088] Figure 2 FIG. 4 is a double-layer optimization model framework diagram of the embodiment of the present application;

[0089] Figure 3 FIG. 6 is a dynamic electricity price optimization result diagram of each functional area of the embodiment of the present application;

[0090] Figure 4 FIG. 8 is a charging load comparison diagram before and after the charging utility function of the embodiment of the present application.

[0091] Figure 5 is a schematic diagram of electric vehicle charging demand load before and after electric vehicle braking energy recovery of an embodiment of the present application. DETAILED DESCRIPTION

[0092] The specific embodiments of the present application will be further described below in conjunction with the accompanying drawings.

[0093] To make the objectives, technical solutions, and advantages of embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0094] As disclosed in the background, with the transformation of global energy structure and the enhancement of environmental awareness, the market penetration rate of electric vehicles is growing at an unprecedented speed due to their clean and efficient characteristics. However, the large-scale unordered access of electric vehicles to the power grid and the charging behavior, especially the concentrated charging behavior in the peak period of residential electricity, will greatly exacerbate the peak-valley load difference of the power grid, posing a serious challenge to the safe and stable operation of the distribution network. Therefore, how to effectively guide the wide and space-time mobile electric vehicle charging load to change from a “load” to a dispatchable “resource” has become a key technical problem to be solved.

[0095] To address the above challenges, various electric vehicle charging management methods have been proposed in the prior art. One basic method is to charge freely by the user under a fixed electricity price, and the user decides the charging time and place according to his own habits. Another widely used method is to implement a time-of-use (TOU) strategy, which divides a day into several time periods (such as peak time, flat time, and valley time), and sets different fixed electricity prices for different time periods, in order to guide users to charge in the valley time with lower electricity prices through economic leverage. In addition, there are also some guiding methods that publish power grid load state information to users through one-way communication to suggest users to charge off-peak.

[0096] Although the prior art implements the time-of-use strategy and other strategies to alleviate the overlap problem of charging load and peak electricity to some extent, there are still some deficiencies:

[0097] Firstly, the guiding granularity of existing price guiding mechanisms is relatively rough, which cannot achieve accurate smoothing of charging load. For example, time-of-use price only provides several large time scale price signals. This leads to a large number of price-sensitive users to set the charging start time at the same time, which forms a new and severe charging load rebound peak, and the peak-valley difference problem of distribution network has not been fundamentally solved. The root cause is that this fixed and non-dynamic price cannot be adjusted according to the real-time load level of the power grid and the actual changes of charging demand, and lacks closed-loop feedback regulation ability for load.

[0098] Secondly, the existing price guiding mechanism generally lacks guiding ability in the spatial dimension. Whether it is a fixed price or a time-of-use price, the price signal is usually uniform in geographical space, that is, all charging facilities in the same area perform the same price in the same period. This one-size-fits-all strategy ignores the natural differences in load characteristics of power grids and user charging demand in different functional areas (such as residential areas and work areas) in the city. For example, the power grid load in the work area is high during the day, but there is a large margin at night, and a large number of electric vehicles stay here during the day, which has charging potential. The existing technology cannot develop differentiated spatial prices to exploit and utilize this time-space complementary characteristics, resulting in excessive concentration of charging load in residential areas at night, and failing to fully utilize the distributed accommodation capacity of the entire city distribution network.

[0099] Finally, some existing guiding methods oversimplify user behavior and vehicle state at the modeling level. For example, when predicting charging demand, the braking energy recovery of vehicles in actual driving is not considered, which will cause misjudgment of the actual required charging amount of vehicles. At the same time, the real habit of users tending to continuous charging is also less considered, which leads to the charging strategy optimized to be difficult to be adopted in reality due to not meeting user preferences. These inaccuracies at the model level fundamentally weaken the accuracy of load prediction and the actual effectiveness of guiding strategies.

[0100] In order to improve the above problems, the application provides a dynamic electricity price based electric vehicle charging load space-time guidance method and device, comprising: based on the pre-constructed electric vehicle space-time travel model, the travel and energy consumption process of the electric vehicle in the target area is simulated by using the Monte Carlo method, and the charging events of the vehicle are counted and aggregated to form the simulation results of the reference charging load space-time distribution of the target area; the simulation results of the reference charging load space-time distribution of the target area are substituted into the pre-constructed double-layer optimization model and iteratively solved to obtain the dynamic electricity price strategy and charging plan of the target area; and the dynamic electricity price strategy and charging plan of the target area are used for charging load space-time guidance of the electric vehicle in the target area. The technical scheme provided by the application reduces the peak-valley difference of the power grid load and the charging cost of the user by formulating the space-time dynamic electricity price and the charging plan.

[0101] 1. The application constructs an electric vehicle space-time travel model, which divides functional areas and uses a mixed Gaussian model, a Weibull distribution and a generalized extreme value distribution to probabilistically model the travel time and stay time of users, so that the space-time distribution of charging demand driven by user travel purposes can be more accurately predicted. Compared with traditional prediction methods that do not distinguish travel purposes, the technical scheme significantly improves the accuracy of the prediction of the reference charging load space-time distribution.

[0102] 2. The double-layer optimization model constructed by the application has a target of minimizing the load variance of the power grid in the upper layer and a target of minimizing the charging cost of the user group in the lower layer, and the game model is solved by an iterative algorithm such as a particle swarm. This structure systematically couples the stability requirement of the power grid side and the economic requirement of the user side, and the generated reference dynamic electricity price strategy can achieve the purpose of "peak load shifting" of the power grid while obtaining high user acceptance due to its economic rationality, thereby improving the effectiveness of the guidance strategy.

[0103] The above scheme will be described in detail below.

[0104] Embodiment 1

[0105] Refer to the accompanying Figure 1 , Figure 1 is the main step flowchart of the dynamic electricity price based electric vehicle charging load space-time guidance method of an embodiment of the application. As shown in Figure 1 , the dynamic electricity price based electric vehicle charging load space-time guidance method in the embodiment of the application mainly comprises the following steps:

[0106] Step S101: based on the pre-constructed electric vehicle space-time travel model, the travel and energy consumption process of the electric vehicle in the target area is simulated by using the Monte Carlo method, and the charging events of the vehicle are counted and aggregated to form the simulation results of the reference charging load space-time distribution of the target area;

[0107] Step S102: Substitute the reference charging load spatio-temporal distribution simulation result of the target region into the pre-constructed double-layer optimization model and perform iterative solving to obtain the dynamic electricity price strategy and charging plan of the target region;

[0108] Step S103: Use the dynamic electricity price strategy and charging plan of the target region to guide the charging load spatio-temporal distribution of electric vehicles in the target region.

[0109] In this embodiment, the electric vehicle spatio-temporal travel model is as follows:

[0110] L i = ((s i0 ,t i0 ), (s i1 ,t i1 ),..., (s in ,t in ))

[0111] In the above formula, L i is the travel chain of the ith electric vehicle; i is the index of the electric vehicle; s in is the nth node reached by the user of the ith electric vehicle in the target region; t in is the stay time of the user of the ith electric vehicle after reaching the nth node in the target region; s i0 is the starting node of the user of the ith electric vehicle in the target region; t i0 is the stay time of the user of the ith electric vehicle at the starting node in the target region; and n is the node sequence number in the travel chain.

[0112] This step aims to establish a mathematical model that can describe the spatio-temporal movement rules of electric vehicles in urban traffic networks. This process first establishes the traffic network topology of the target region and divides the region into multiple functional areas with different functional attributes. Then, based on historical travel statistics, the travel behavior of users, including the first travel time, the stay time in different functional areas, and other key parameters, are probabilistically modeled.

[0113] To establish a mathematical model that can accurately describe the spatio-temporal movement rules of electric vehicles, the urban traffic network needs to be topologically modeled first. By setting the boundaries of the traffic network, a real traffic network topology graph is constructed, and its mathematical model can be represented as:

[0114]

[0115] In the formula, G is the traffic network topology graph; N(G) is the set of road intersection nodes in the traffic network; L(G) is the set of roads in the traffic network; ψ G is the adjacency matrix of the roads in the traffic network; and d i, d j is the road node in the network; n is the total number of road intersection nodes in the traffic network; e ij is the connection node of d i . j is the road length of d

[0116] In one embodiment, in the process of simulating the travel and energy consumption process of electric vehicles in the target area by using the Monte Carlo method, the starting travel time of the electric vehicle user is fitted by using a Gaussian mixture model, the stay time of the electric vehicle user after arriving at the node of the residential area in the target area is fitted by using a Weibull distribution model, and the stay time of the electric vehicle user after arriving at the node of the work area and other areas in the target area is fitted by using a generalized extreme value distribution model.

[0117] In a specific embodiment, the starting travel time of the electric vehicle is modeled by using a Gaussian mixture model fitting method, and the probability density function thereof can be expressed as:

[0118]

[0119] In the formula, Φ(T) is the probability density function of the starting travel time; T is the starting travel time, in minutes; i is the index of the Gaussian component; α i is the weight coefficient of the i-th Gaussian component; κ i is the mean value of the i-th Gaussian component, in hours; ε i is the standard deviation of the i-th Gaussian component.

[0120] In addition, the stay time of the electric vehicle in different functional areas follows different probability distributions. In the residential area, the stay time is described by using a Weibull distribution, and the probability density function thereof is:

[0121]

[0122] In the formula, f r (R, λ r , θ r ) is the probability density function of the stay time in the residential area; R is the stay time of the electric vehicle in the residential area; λ r is the scale parameter of the Weibull distribution; θ r is the shape parameter of the Weibull distribution.

[0123] In the work area and other areas, the stay time is described by using a generalized extreme value (GEV) distribution, and the probability density function thereof is:

[0124]

[0125] wherein: z is a variable after standardization of the residence time; R is the residence time of the electric vehicle in the working zone or other zone; f w (z) is the probability density function of the residence time in the working zone; f o (z) is the probability density function of the residence time in the other zone; μ z is the location parameter of the generalized extreme value distribution; σ is the scale parameter of the generalized extreme value distribution; ξ w is the shape parameter of the generalized extreme value distribution for the working zone; ξ o is the shape parameter of the generalized extreme value distribution for the other zone; e is the base of the natural logarithm.

[0126] In this embodiment, in the process of simulating the travel and energy consumption process of the electric vehicle in the target zone by using the Monte Carlo method, the residual electric quantity of the electric vehicle is updated according to the following formula:

[0127]

[0128] In the above formula, is the residual electric quantity of the electric vehicle at the t period; B is the rated electric quantity of the electric vehicle battery; is the total electric consumption of the electric vehicle at the t period; is the total recovered electric quantity of the electric vehicle at the t period;

[0129] The total recovered electric quantity of the electric vehicle at the t period is as follows:

[0130]

[0131] E f =∫δmgvdt

[0132] E w =∫AC d ρvdt

[0133] In the above formula, ω is a brake distribution factor; E ν , E f , E w are respectively the energy consumption of the electric vehicle braking, the energy consumption of overcoming the rolling resistance and the energy consumption of overcoming the wind resistance; v is the vehicle speed; v0 and v1 are respectively the initial speed and the final speed of braking; δ and C d are respectively the rolling resistance coefficient and the wind resistance coefficient; m is the vehicle mass; g is the gravitational acceleration; A is the vehicle wind area; and ρ is the air density.

[0134] In this embodiment, in the process of simulating the travel and energy consumption process of the electric vehicle in the target zone by using the Monte Carlo method, the shortest path of the electric vehicle from the node n to the node m in the target zone is determined according to the following formula:

[0135] dist[n][m] = min(dist[n][m], dist[n][s] + dist[s][m])

[0136] In the above formula, dist[n][m] is the shortest path length between node n and node m; dist[n][s] is the shortest path length between node n and the transit node s; dist[s][m] is the shortest path length between the transit node s and node j; n, m, s are the index numbers of the starting node, the end node and the transit node, respectively.

[0137] In this embodiment, the Monte Carlo method is used to simulate the load space-time. The simulation process is performed on a preset number of electric vehicles one by one as follows: extracting a trip chain and the first trip time according to the constructed model; determining the destination and calculating the shortest path according to the trip chain; calculating the travel time, arrival time, consumed power, recovered power and remaining power; judging whether the charging demand is generated according to the destination stay time and the vehicle remaining power; recording the charging time and charging load of the vehicle requiring charging. By repeating the process until all vehicles complete the trip chain simulation, the charging load of all vehicles at different times and different functional area nodes is finally counted and accumulated, so as to obtain the space-time distribution of the baseline charging load.

[0138] In this embodiment, Figure 2 is a schematic diagram of a double-layer optimization model framework constructed according to an embodiment of the present application. The model constructed in this embodiment is a double-layer optimization structure based on principal-agent game. The pre-constructed double-layer optimization model includes: an upper-layer objective function with the optimization objective of minimizing the total daily load variance of the power grid after considering the charging load of electric vehicles and the corresponding upper-layer constraint condition, and a lower-layer objective function with the optimization objective of minimizing the total charging cost of all users and the corresponding lower-layer constraint condition.

[0139] In one embodiment, the upper-layer model represents the distribution network operator as the leader, and the optimization objective is to minimize the total daily load variance of the power grid after considering the charging load of electric vehicles. The upper-layer objective function is as follows:

[0140]

[0141] In the above formula, γ is the daily load variance of the power grid; T is the total number of time periods in a day; P grid,t is the total load of the power grid at time period t; is the daily average total load.

[0142] In one embodiment, the upper-layer constraint condition is as follows:

[0143]

[0144] Umin ≤U t ≤U max

[0145] (1-ζ ν )V t ≤V t ≤(1+ζ v )V t

[0146] |V t+1 -V t |≤ΔV

[0147] Specifically comprising: power constraints of distribution network, voltage constraints of distribution network, power constraints of distribution network, in the above formula, is the power of the ith electric vehicle connected to the grid at time period t; is the maximum power of the distribution network at time period t; N is the total number of electric vehicles; U min is the lower limit of the node voltage; U max is the upper limit of the node voltage; U t is the voltage of the node at time period t; V t is the electricity price at time period t; V t+1 is the electricity price at time period t+1; ζ v is the price fluctuation coefficient; ΔV is the electricity price fluctuation threshold of adjacent time periods.

[0148] In one embodiment, the lower layer model represents a group of electric vehicle users as followers, and the optimization objective is to minimize the total charging cost of all users, and the lower layer objective function is as follows:

[0149] or

[0150] In the above formula, C is the total charging cost of all electric vehicles in the region; is the charging amount of the ith electric vehicle at time period t; N is the total number of electric vehicles; V t is the electricity price at time period t; S is the total utility of users; S t is the utility function at time period t; η t is the charging state at time period t; η t-1 is the charging state at time period t-1; η t+1 is the charging state at time period t+1; S t (η t-1 ,η t ,η t+1 )=V t ·η t +λ1·η t-1 ·η t +λ2·η t ·η t+1; V t The electricity price for the time period t; λ1, λ2 are continuous charging adjustment parameters.

[0151] In one embodiment, the lower-layer constraint condition is as follows:

[0152]

[0153] Specifically, it includes user charging demand constraints, charging power constraints, and battery capacity constraints, wherein, Qi(t) is the charging amount of the i-th electric vehicle at time t; Qi(t) is the charging demand amount of the i-th electric vehicle at time t; Δt is the duration of a single time period; Pi is the maximum charging power of the i-th electric vehicle; B is the rated battery capacity.

[0154] In one specific embodiment, the pre-constructed double-layer optimization model is iteratively solved to find a solution that balances the objectives of the upper-layer model and the lower-layer model, i.e., a Stackelberg equilibrium solution.

[0155] This embodiment uses the Particle Swarm Optimization (PSO) algorithm to solve the double-layer optimization model. In the application scenario of this algorithm:

[0156] Particle: Each particle represents a complete set of dynamic electricity price strategies covering all time periods. Specifically, the position vector of a particle can be represented as V = [V1, V2,..., V t ], where V t is the electricity price for the t-th time period.

[0157] Swarm: A collection of multiple particles representing multiple sets of dynamic electricity price strategies to be evaluated.

[0158] Fitness Function: The objective function of the upper-layer optimization model, i.e., the distribution network daily load variance γ, is used as the fitness function to evaluate the quality of particles. The smaller the fitness value, the better the corresponding electricity price strategy.

[0159] The algorithm updates the position (electricity price strategy) and speed of the particles through an iterative process. Each particle adjusts its moving direction and distance based on its own historical optimal position and the global optimal position of the entire particle swarm until the preset termination condition is met.

[0160] The specific solution process of the double-layer optimization model is as follows:

[0161] Step 1: Set the particle swarm size, maximum iteration number and other algorithm parameters. In the solution space that meets the upper model's electricity price constraint, a set of initial particles, i.e. a set of initial dynamic electricity price strategies, are randomly generated.

[0162] Step 2: For each particle in the particle swarm, the dynamic electricity price strategy V represented by its position vector is passed to the lower optimization model as known input.

[0163] Step 3: After receiving the deterministic electricity price strategy V passed by the upper layer, each electric vehicle user (or user group) makes its own optimization decision based on the electricity price. That is, under the premise of meeting their own charging demand constraints, power constraints and battery capacity constraints, each user solves its charging plan in all time periods to minimize its charging cost or maximize its charging utility

[0164] Step 4: Aggregate all the charging plans determined by the users in Step 3 to obtain the total charging load space-time distribution generated by electric vehicles under the current electricity price strategy This aggregated load is fed back to the upper model as the user's response result.

[0165] Step 5: The upper model calculates the total load curve corresponding to the current electricity price strategy according to the aggregated charging load fed back by the lower layer, in combination with the basic load, and calculates its load variance γ according to the formula. This variance value is the fitness value of the current particle (electricity price strategy).

[0166] Step 6: For each particle, compare its current fitness value with its historical optimal fitness value. If the current value is better, update its historical optimal position. At the same time, compare the historical optimal fitness value of all particles with the global optimal fitness value. If a better value appears, update the global optimal position. Then, update the speed and position of each particle according to the speed and position update rules of the particle swarm algorithm.

[0167] Step 7: Check whether the maximum number of iterations is reached or the global optimal fitness value has not changed significantly for consecutive generations. If the termination condition is met, go to Step 8; otherwise, return to Step 2 and start a new round of iteration.

[0168] Step 8: When the algorithm terminates, the electricity price strategy represented by the global optimal position is the benchmark dynamic electricity price strategy. This strategy is the equilibrium solution that minimizes the load variance of the power grid after considering the users' response behavior to the price.

[0169] In a specific embodiment, in order to cope with the uncertainty factors within the operation day and improve the accuracy of the price guidance, the present application designs a dynamic electricity price rolling correction and release mechanism, which is specifically:

[0170] Before the start of the operation day, the system first performs a day-ahead optimization process. This process performs the aforementioned steps completely, i.e. constructs and solves the bi-level optimization model based on the simulation generation of the probabilistic model of user travel behavior and the benchmark load.

[0171] The output of this process is a set of benchmark dynamic pricing strategies covering the next day (e.g. 24 hours, divided into 96 15-minute periods). This strategy is formulated based on the prediction of future states as the initial plan for the operation day price guidance.

[0172] On the day of the operation day, the system enters the real-time perception phase. In this phase, the system continuously acquires actual state information through data interfaces with external information systems at a preset frequency (e.g. every 5 minutes). These information mainly include:

[0173] Real-time traffic information: data such as actual travel speed, congestion level of major road segments in the city obtained from traffic management systems or third-party map services.

[0174] Charging facility state information: data such as real-time occupancy status, idle quantity, number of queued vehicles of charging piles obtained from charging station operation systems.

[0175] Short-term load forecast correction information: updated values of ultra-short-term basic load forecast obtained from power grid dispatching systems.

[0176] Weather information: real-time weather data such as actual temperature, wind speed that may affect vehicle energy consumption is obtained.

[0177] These real-time perceived actual data will be used for comparison with day-ahead prediction data, and the deviation is the basis for triggering subsequent rolling optimization.

[0178] The price correction and release within the operation day adopts a rolling optimization mechanism. This mechanism divides the entire operation day into multiple consecutive rolling time windows (e.g. each 1 hour as a window). At the starting point of each rolling time window, the system performs a price correction calculation, the specific process is as follows:

[0179] First, update the model state and parameters: the system uses the latest information perceived in real time to update the bi-level optimization model. The update content includes: using the actual traffic information to correct the travel time of the road segments in the traffic network model; according to the actual occupancy status of the charging facilities, update the available charging selection set of users in the lower level model; use the corrected basic load forecast to update the objective function of the upper level model; and the actual load of the power grid and the actual state (position, state of charge) of the vehicle at the current time as the initial conditions for optimization solving.

[0180] Second, the system performs a rolling solution: based on the updated model and state, the system re-executes the solution process, but the time horizon of the solution is not the full 24 hours ahead, but the remaining time horizon until the end of the current day.

[0181] Third, the corrected and published price: after the rolling solution is completed, the system obtains a corrected dynamic price strategy covering the remaining time horizon. The system extracts the price corresponding to the next execution cycle (e.g., the next 15 or 30 minutes) from the strategy and publishes it to the user terminal or charging facility through the communication network.

[0182] Fourth, forward rolling window: after the current execution cycle ends, the time window of the entire system is rolled forward, and the first to third steps are repeated until the end of the operating day. Through this cycle of prediction-feedback-correction mechanism, the invention can continuously correct the price strategy to dynamically adapt to various uncertainties in the operating day.

[0183] To further illustrate the invention, the method proposed by the invention will be verified by combining specific numerical examples.

[0184] This embodiment selects a part of a city as the analysis object. First, the obtained POI (Point of Interest) data of the region is classified, and the city region is divided into three functional area types of residential area, working area and other area according to the kernel density estimation method, so as to obtain the traffic network structure of the region and the functional area division result of each road node.

[0185] In this embodiment, the number of simulated electric vehicles is set to 5000, the rated capacity of the single vehicle battery is set to 50kWh, the energy consumption rate is set to 15kWh per 100km, and the charging power is set to 10kW. When the particle swarm optimization algorithm is used for solution, the number of particle swarms is set to 10, and the maximum number of iterations is set to 200.

[0186] According to the published 2017 NHTS (National Household Travel Survey) travel survey data set, the proportion of each type of travel chain is determined as shown in Table 1.

[0187] Table 1

[0188] Travel chain type Proportion Residential area-work area-residential area 0.47 Residential area-other area-residential area 0.15 Residential area-work area-other area-residential area 0.12 Residential area-other area-work area-residential area 0.13 Other area-residential area / work area-other area 0.03 Work area-residential area / other area-work area 0.03

[0189] To verify the effect of the technical solution of the invention, three different comparison scenarios are set:

[0190] Scenario 1 (fixed price): a fixed charging price is set for all functional areas, and the value is 0.428 yuan / kWh.

[0191] Scenario 2 (Time-of-Use Pricing): A uniform time-of-use pricing is set for all functional areas, with a peak price of 0.568 yuan / kWh from 8:00 to 22:00 and an off-peak price of 0.288 yuan / kWh from 22:00 to 8:00 the next day.

[0192] Scenario 3 (Dynamic Electricity Price): Using the method proposed in this invention, dynamic electricity price optimization is performed for different functional areas.

[0193] The three scenarios were optimized and solved separately to obtain the total load variation of the region within a day under different scenarios. Table 2 shows a comparison of the peak-valley difference of the distribution network load and the total charging cost for users under the three scenarios.

[0194] Table 2

[0195] Scenario Power distribution network load peak-valley difference (kW) Total cost of user charging (yuan) Scenario one 19638.25 65907.52 Scenario two 9272.95 50029.13 Scenario three 7714.96 45315.26

[0196] As shown in Table 2, compared to Scenario 1 (fixed electricity price) and Scenario 2 (time-of-use pricing), Scenario 3 (the method of this invention) corresponds to the lowest peak-to-valley load difference in the distribution network. Specifically, the peak-to-valley load difference in Scenario 3 is reduced by 60.71% compared to Scenario 1 and by 16.80% compared to Scenario 2. Simultaneously, the total user charging cost corresponding to Scenario 3 is also the lowest, decreasing by 31.24% compared to Scenario 1 and by 9.42% compared to Scenario 2.

[0197] To analyze the spatiotemporal guidance capability of the method of this invention, the dynamic electricity price optimization results generated for different functional areas in scenario three are shown in the appendix. Figure 3 As shown.

[0198] According to the appendix Figure 3 The electricity pricing strategy shown alters the charging load distribution across different functional areas. Specifically, the charging load in residential areas is redirected to the off-peak electricity hours from 8:00 PM to 6:00 AM the following day; the charging load in work areas is concentrated between 4:00 PM and 8:00 PM; and the charging load in other areas is mainly distributed between 12:00 PM and 11:00 PM. Through this spatiotemporal guidance, a significant portion of the charging load originally concentrated in residential areas is transferred to work areas and other areas. Statistics show that the latter two account for 36% of the total charging load.

[0199] To verify the impact of the charging utility function on the model accuracy, a comparison was made in Scenario 3 of the total regional charging load before and after considering the utility function, as shown in the attached figure. Figure 4The smoothness of the total charging load curve is 0.65 before considering the charging utility function, and the smoothness of the total charging load curve is 0.70 after considering the charging utility function. The increase in the smoothness value indicates that the model can better reflect the real charging behavior after considering the continuous charging preference of the user, thereby improving the accuracy of the charging load distribution prediction.

[0200] To verify the influence of the brake energy recovery model on the model accuracy, the total charging demand load before and after considering the brake energy recovery is compared, as shown in FIG. 6. Figure 5 In this example, all electric vehicles in the region save 8781.07 kWh of electricity during the whole day of travel after considering the brake energy recovery, and the saved amount accounts for 11.14% of the total power consumption of the vehicle.

[0201] In particular, during the travel peak period from 7:00 to 17:00, the amount of electricity recovered by the brake energy recovery accounts for 26.41% of the total power consumption during the period. This result shows that considering the brake energy recovery can more accurately determine the real-time power of the electric vehicle, and further more accurately determine the charging demand of the user, thereby improving the accuracy of the entire model.

[0202] Embodiment 2

[0203] Based on the same inventive concept, the application also provides a dynamic price-based electric vehicle charging load space-time guiding device, which comprises:

[0204] A simulation module is configured to simulate the travel and energy consumption process of electric vehicles in a target region by using a Monte Carlo method based on a pre-constructed electric vehicle space-time travel model, and to count and aggregate the charging events of the vehicles to form a simulation result of the benchmark charging load space-time distribution of the target region.

[0205] An analysis module is configured to substitute the simulation result of the benchmark charging load space-time distribution of the target region into a pre-constructed double-layer optimization model and iteratively solve the model to obtain a dynamic price strategy and a charging plan of the target region.

[0206] A guiding module is configured to guide the charging load space-time of electric vehicles in the target region by using the dynamic price strategy and the charging plan of the target region.

[0207] Preferably, the electric vehicle space-time travel model is as follows:

[0208] L i = ((s i0 ,t i0 ), (s i1 ,t i1 ),..., (s in ,tin ))

[0209] In the above formula, L i is the trip chain of the i-th electric vehicle; i is the index of the electric vehicle; s in is the n-th node reached by the i-th electric vehicle user in the target area; t in is the stay time of the i-th electric vehicle user after reaching the n-th node in the target area; s i0 is the starting node of the i-th electric vehicle user in the target area; t i0 is the stay time of the i-th electric vehicle user at the starting node in the target area; n is the node number in the trip chain.

[0210] Further, in the process of simulating the travel and energy consumption process of electric vehicles in the target area by the Monte Carlo method, the starting travel time of the electric vehicle user is fitted by using a Gaussian mixture model, the stay time of the electric vehicle user after reaching the node of the residential area in the target area is fitted by using a Weibull distribution model, and the stay time of the electric vehicle user after reaching the node of the work area and other areas in the target area is fitted by using a generalized extreme value distribution model.

[0211] Preferably, in the process of simulating the travel and energy consumption process of electric vehicles in the target area by the Monte Carlo method, the remaining electric quantity of the electric vehicle is updated according to the following formula:

[0212]

[0213] In the above formula, is the remaining electric quantity of the electric vehicle at t period; B is the rated electric quantity of the electric vehicle battery; is the total electric consumption of the electric vehicle at t period; is the total recovered electric quantity of the electric vehicle at t period;

[0214] Wherein, the total recovered electric quantity of the electric vehicle at t period is as follows:

[0215]

[0216] E f =∫δmgvdt

[0217] E w =∫AC d ρvdt

[0218] In the above formula, ω is a brake distribution factor; E ν , E f , E wrespectively are the energy consumption of electric vehicle braking, energy consumption of overcoming rolling resistance and energy consumption of overcoming wind resistance; v is the vehicle speed; v0, v1 are the initial speed and final speed of braking respectively; δ and C d respectively are the rolling resistance coefficient and wind resistance coefficient; m is the vehicle mass; g is the gravity acceleration; A is the vehicle frontal area; ρ is the air density.

[0219] Preferably, in the process of simulating the travel and energy consumption process of the electric vehicle in the target area by using the Monte Carlo method, the shortest path of the electric vehicle from node n to node m in the target area is determined according to the following formula:

[0220] dist[n][m] = min (dist[n][m], dist[n][s] + dist[s][m])

[0221] In the above formula, dist[n][m] is the shortest path length from node n to node m; dist[n][s] is the shortest path length from node n to the transfer node s; dist[s][m] is the shortest path length from the transfer node s to node j; n, m, s are the index numbers of the starting node, the terminal node and the transfer node respectively.

[0222] Preferably, the pre-constructed double-layer optimization model includes: an upper layer objective function and its corresponding upper layer constraint condition, which take minimizing the total variance of the power grid daily load after considering the charging load of the electric vehicle as the optimization objective; and a lower layer objective function and its corresponding lower layer constraint condition, which take minimizing the total charging cost of all users as the optimization objective.

[0223] Further, the upper layer objective function is as follows:

[0224]

[0225] In the above formula, γ is the variance of the power grid daily load; T is the total number of time periods in a day; P grid,t is the total load of the power grid at time period t; is the average total load per day.

[0226] Further, the upper layer constraint condition is as follows:

[0227]

[0228] U min ≤ U t ≤ U max

[0229] (1-ζ v )V t ≤ V t ≤ (1+ζ v )Vt

[0230] |V t+1 -V t |≤ΔV

[0231] In the above formula, is the power of the ith electric vehicle accessing the power grid at time t; is the maximum power of the distribution network accessing at time t; N is the total number of electric vehicles; U min is the lower limit of the node voltage; U max is the upper limit of the node voltage; U t is the voltage of the node at time t; V t is the electricity price at time t; V t+1 is the electricity price at time t+1; ζ ν is the price fluctuation coefficient; ΔV is the price fluctuation threshold of adjacent time periods.

[0232] Further, the lower layer objective function is as follows:

[0233] Or

[0234] In the above formula, C is the total charging cost of all electric vehicles in the region; is the charging amount of the ith electric vehicle at time t; N is the total number of electric vehicles; V t is the electricity price at time t; S is the total utility of users; S t is the utility function at time t; η t is the charging state at time t; η t-1 is the charging state at time t-1; η t+1 is the charging state at time t+1; S t (η t-1 ,η t ,η t+1 )=V t ·η t +λ1·η t-1 ·η t +λ2·η t ·η t+1 ; V t is the electricity price at time t; λ1, λ2 are continuous charging adjustment parameters.

[0235] Further, the lower layer constraint condition is as follows:

[0236]

[0237] In the above formula, is the charging amount of the ith electric vehicle at time t; Qi(t) is the charging demand of the ith electric vehicle at time t; Δt is the duration of a single time period; Qi(t) is the charging demand of the ith electric vehicle at time t; Δt is the duration of a single time period;

[0238] Embodiment 3

[0239] Based on the same inventive concept, the present application further provides a computer device, which comprises a processor and a memory, the memory is used to store a computer program, the computer program comprises program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc., which are the computing core and control core of the terminal, and are suitable for implementing one or more instructions, and are specifically suitable for loading and executing one or more instructions in the computer storage medium to implement a corresponding method flow or a corresponding function, so as to implement the steps of the electric vehicle charging load space-time guidance method based on dynamic electricity price in the above embodiment.

[0240] Embodiment 4

[0241] Based on the same inventive concept, the present application further provides a storage medium, specifically a computer readable storage medium (Memory), which is a memory device in the computer device, and is used to store programs and data. It can be understood that the computer readable storage medium herein can include the built-in storage medium in the computer device, and of course can also include the expansion storage medium supported by the computer device. The computer readable storage medium provides a storage space, and the storage space stores the operating system of the terminal. Moreover, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and the instructions can be one or more computer programs (including program codes). It should be noted that the computer readable storage medium herein can be a high-speed RAM memory, or a non-volatile memory such as at least one disk memory. One or more instructions stored in the computer readable storage medium can be loaded and executed by the processor to implement the steps of the electric vehicle charging load space-time guidance method based on dynamic electricity price in the above embodiment.

[0242] Those skilled in the art will appreciate that embodiments of the present application can be readily used as a method, apparatus such as a system, or computer program product. Accordingly, the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer readable program code.

[0243] The present application is described in reference to the drawings, which are as follows. Figure one one or more processes and / or blocks Figure one means for performing the functions specified in the one or more processes and / or blocks

[0244] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the Figure one one or more processes and / or blocks Figure one one or more blocks

[0245] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the Figure one one or more processes and / or blocks Figure one one or more blocks

[0246] Finally, it should be noted that the above-described embodiments are merely intended for describing and illustrating, not limiting, the technical solution of the present application. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, and any modification or replacement should be covered within the protection scope of the claims of the present application.

Claims

1. A dynamic electricity price-based spatiotemporal guiding method for electric vehicle charging load, characterized in that, The method comprises: Based on the pre-constructed electric vehicle space-time travel model, the Monte Carlo method is used to simulate the travel and energy consumption process of electric vehicles in the target area, and the charging events of the vehicles are counted to form the simulation results of the reference charging load space-time distribution of the target area; The simulation results of the reference charging load space-time distribution of the target area are substituted into the pre-constructed double-layer optimization model and iteratively solved to obtain the dynamic electricity price strategy and charging plan of the target area; The dynamic electricity price strategy and charging plan of the target area are used to guide the charging load space-time of electric vehicles in the target area.

2. The method of claim 1, wherein, The electric vehicle space-time travel model is as follows: L i = ((s i0 ,t i0 ), (s i1 ,t i1 ),..., (s in ,t in )) In the above formula, L i is the trip chain of the i-th electric vehicle; i is the index of the electric vehicle number; s in is the n-th node reached by the i-th electric vehicle user in the target area; t in is the stay time of the i-th electric vehicle user after reaching the n-th node in the target area; s i0 is the starting node for the i-th electric vehicle user in the target area; t i0 is the dwell time of the ith electric vehicle user at the origin node in the target area; n is the node sequence number in the trip chain.

3. The method of claim 2, wherein, In the process of simulating the travel and energy consumption process of electric vehicles in the target area by the Monte Carlo method, the starting travel time of electric vehicle users is fitted by using a mixed Gaussian model, the stay time of electric vehicle users after arriving at the node of the residential area in the target area is fitted by using a Weibull distribution model, and the stay time of electric vehicle users after arriving at the node of the work area and other areas in the target area is fitted by using a generalized extreme value distribution model.

4. The method of claim 1, wherein, In the process of simulating the travel and energy consumption process of electric vehicles in the target area by the Monte Carlo method, the remaining electric quantity of the electric vehicle is updated according to the following formula: In the above formula, B is the rated capacity of the battery of the electric vehicle; is the total power consumption of the electric vehicle during the t period of travel; is the total recovered power of the electric vehicle during the t period. The total recovered electric quantity of the electric vehicle in the t period is as follows: E f = ∫δmgvdt E w = ∫AC d ρvdt In the above formula, ω is a braking distribution factor; E ν , E f , E w are energy consumption of electric vehicle braking, energy consumption of overcoming rolling resistance and energy consumption of overcoming wind resistance, respectively; v is vehicle speed; v0 and v1 are initial speed and final speed of braking, respectively; δ and C d are rolling resistance coefficient and wind resistance coefficient, respectively; m is vehicle mass; g is gravitational acceleration; A is vehicle frontal area; and ρ is air density.

5. The method of claim 1, wherein, In the process of simulating the travel and energy consumption process of electric vehicles in the target area by the Monte Carlo method, the shortest path between node n and node m in the target area is determined according to the following formula: dist[n][m]=min(dist[n][m],dist[n][s]+dist[s][m]) In the above formula, dist[n][m] is the shortest path length between node n and node m; dist[n][s] is the shortest path length between node n and the transfer node s; dist[s][m] is the shortest path length between the transfer node s and node j; n, m and s are the number indexes of the starting node, the end node and the transfer node respectively.

6. The method of claim 1, wherein, The pre-constructed double-layer optimization model comprises: an upper layer objective function and its corresponding upper layer constraint condition, which take minimizing the total variance of the power grid daily load after considering the charging load of electric vehicles as the optimization target; and a lower layer objective function and its corresponding lower layer constraint condition, which take minimizing the total charging cost of all users as the optimization target.

7. The method of claim 6, wherein, The upper layer objective function is as follows: In the above formula, γ is the variance of the daily load of the power grid; T is the total number of time periods in a day; P grid,t is the total load of the power grid for the t time period; is the average daily total load.

8. The method of claim 7, wherein, The upper layer constraint condition is as follows: U min ≤U t ≤U max (1 - ζ v V t ≤ V t ≤ (1 + ζ v V t |V t+1 -V t |≤ΔV In the above formula, is the grid access power of the ith electric vehicle at time period t; is the maximum grid access power of the distribution network at time period t; N is the total number of electric vehicles; U min is the lower limit of the node voltage; U max is the upper limit of the node voltage; U t Vtis the voltage of the node at time period t; V t Vtis the price of electricity at time period t; V t+1 Vt+1is the price of electricity at time period t+1; ζ v ζ is the price fluctuation coefficient; ΔV is the price fluctuation threshold of adjacent time periods.

9. The method of claim 7, wherein, The lower layer objective function is as follows: or In the above formula, C is the total charging cost of all electric vehicles in the area; Qi(t) is the charging amount of the ith electric vehicle at the t period; N is the total number of electric vehicles; V t P(t) is the electricity price at the t period; S is the total utility of the user; S t U(t) is the utility function at the t period; η t SoC(t) is the state of charge at the t period; η t-1 State of charge for the period t-1; η t+1 is the state of charge for the t+1 period; S t (η t-1 ,η t ,η t+1 ) = V t ·η t + λ1·η t-1 ·η t + λ2·η t ·η t+1 ; V t is the electricity price for the t period; λ1, λ2 are continuous charging adjustment parameters.

10. The method of claim 9, wherein, The lower layer constraint condition is as follows: In the above formula, Qi(t) is the charging amount of the i-th electric vehicle at time t; Qi(t) is the charging amount of the i-th electric vehicle at time t; Δt is the duration of a single time period; Qi(t) is the charging amount of the i-th electric vehicle at time t; Δt is the duration of a single time period; Qi(t) is the charging amount of the i-th electric vehicle at time t; Δt is the duration of a single time period; 11. A dynamic electricity price-based electric vehicle charging load spatiotemporal guidance device, characterized by, The device comprises: A simulation module is configured to simulate the travel and energy consumption process of electric vehicles in the target area by using the Monte Carlo method based on the pre-constructed electric vehicle space-time travel model, and count the charging events of the vehicles to form the simulation results of the reference charging load space-time distribution of the target area; An analysis module is configured to substitute the simulation results of the reference charging load space-time distribution of the target area into the pre-constructed double-layer optimization model and iteratively solve the model to obtain the dynamic electricity price strategy and charging plan of the target area; The dynamic electricity price strategy and charging plan of the target area are used to guide the charging load space-time of electric vehicles in the target area. A guiding module is configured to guide the charging load of the electric vehicles in the target region in time and space by using the dynamic electricity price strategy and the charging plan of the target region.

12. The apparatus of claim 11, wherein, The electric vehicle time-space travel model is as follows: L i = ((s i0 ,t i0 ), (s i1 ,t i1 ),..., (s in ,t in )) In the above formula, L i is the trip chain of the i-th electric vehicle; i is the index of the electric vehicle number; s in is the n-th node reached by the i-th electric vehicle user in the target area; t in is the stay time of the i-th electric vehicle user after reaching the n-th node in the target area; s i0 is the starting node for the i-th electric vehicle user in the target area; t i0 is the dwell time of the ith electric vehicle user at the origin node in the target area; n is the node sequence number in the trip chain.

13. The apparatus of claim 12, wherein, In the process of simulating the travel and energy consumption of the electric vehicles in the target region by using the Monte Carlo method, the starting travel time of the electric vehicle users is fitted by using a Gaussian mixture model, the stay time of the electric vehicle users after arriving at the nodes of the residential areas in the target region is fitted by using a Weibull distribution model, and the stay time of the electric vehicle users after arriving at the nodes of the work areas and other areas in the target region is fitted by using a generalized extreme value distribution model.

14. The apparatus of claim 11, wherein, In the process of simulating the travel and energy consumption of the electric vehicles in the target region by using the Monte Carlo method, the residual electricity quantity of the electric vehicles is updated according to the following formula: In the above formula, B is the rated capacity of the battery of the electric vehicle; is the total power consumption of the electric vehicle during the t period of time; is the total recovered power of the electric vehicle during the t period of time; In the process of simulating the travel and energy consumption of the electric vehicles in the target region by using the Monte Carlo method, the shortest path between the node n and the node m in the target region is determined according to the following formula: E f = ∫δmgvdt E w = ∫AC d ρvdt In the above formula, ω is a braking distribution factor; E ν , E f , E w are energy consumption of electric vehicle braking, energy consumption of overcoming rolling resistance and energy consumption of overcoming wind resistance, respectively; v is vehicle speed; v0 and v1 are initial speed and final speed of braking, respectively; δ and C d are rolling resistance coefficient and wind resistance coefficient, respectively; m is vehicle mass; g is gravitational acceleration; A is vehicle frontal area; and ρ is air density.

15. The apparatus of claim 11, wherein, dist[n][m]=min(dist[n][m],dist[n][s]+dist[s][m]) In the above formula, dist[n][m] is the shortest path length between the node n and the node m; dist[n][s] is the shortest path length between the node n and the transfer node s; dist[s][m] is the shortest path length between the transfer node s and the node j; n, m and s are the index numbers of the starting node, the terminal node and the transfer node respectively. The pre-constructed double-layer optimization model comprises an upper-layer target function and a corresponding upper-layer constraint condition, and a lower-layer target function and a corresponding lower-layer constraint condition.

16. The apparatus of claim 11, wherein, The upper-layer target function is as follows:

17. The apparatus of claim 16, wherein, The upper-layer constraint condition is as follows: In the above formula, γ is the variance of the daily load of the power grid; T is the total number of time periods in a day; P grid,t is the total load of the power grid for the t time period; is the average daily total load.

18. The apparatus of claim 17, wherein, The lower-layer target function is as follows: U min ≤U t ≤U max (1 - ζ ν V t ≤ V t ≤ (1 + ζ v V t |V t+1 -V t |≤ΔV In the above formula, is the grid access power of the ith electric vehicle at time period t; is the maximum grid access power of the distribution network at time period t; N is the total number of electric vehicles; U min is the lower limit of the node voltage; U max is the upper limit of the node voltage U t Vtis the voltage of the node at time period t; V t Vtis the price of electricity at time period t; V t+1 Vt+1is the price of electricity at time period t+1; ζ v ζ is the price fluctuation coefficient; ΔV is the price fluctuation threshold of adjacent time periods.

19. The apparatus of claim 17, wherein, The lower-layer constraint condition is as follows: or In the above formula, C is the total charging cost of all electric vehicles in the area; Qi(t) is the charging amount of the ith electric vehicle at the t period; N is the total number of electric vehicles; V t P(t) is the electricity price at the t period; S is the total utility of the user; S t U(t) is the utility function at the t period; η t SoC(t) is the state of charge at the t period; η t-1 State of charge for the period t-1. η t+1 is the state of charge for the t+1 period; S t (η t-1 ,η t ,η t+1 ) = V t ·η t + λ1·η t-1 ·η t + λ2·η t ·η t+1 ; V t is the electricity price for the t period; λ1, λ2 are continuous charging adjustment parameters.

20. The apparatus of claim 19, wherein, comprises: In the above formula, Qi(t) is the charging amount of the i-th electric vehicle at time t; Qi(t) is the charging amount of the i-th electric vehicle at time t; Δt is the duration of a single time period; Qi(t) is the charging amount of the i-th electric vehicle at time t; Δt is the duration of a single time period; Qi(t) is the charging amount of the i-th electric vehicle at time t; Δt is the duration of a single time period; 21. A computer device, comprising: one or more processors; the processor is configured to execute one or more programs; when the one or more programs are executed by the one or more processors, the dynamic electricity price based electric vehicle charging load time-space guiding method according to any one of claims 1 to 10 is implemented. a computer program is stored thereon, and the computer program is executed to implement the dynamic electricity price based electric vehicle charging load time-space guiding method according to any one of claims 1 to 10.

22. A computer-readable storage medium, characterized in that, ​

Images