CN109934391B - Intelligent scheduling method for pure electric bus - Google Patents
Intelligent scheduling method for pure electric bus Download PDFInfo
- Publication number
- CN109934391B CN109934391B CN201910139994.8A CN201910139994A CN109934391B CN 109934391 B CN109934391 B CN 109934391B CN 201910139994 A CN201910139994 A CN 201910139994A CN 109934391 B CN109934391 B CN 109934391B
- Authority
- CN
- China
- Prior art keywords
- vehicle
- time
- charging
- departure
- solution
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an intelligent scheduling method for a pure electric bus, which is used for solving the technical problems that the scheduling of the conventional electric bus is time-consuming and labor-consuming and cannot be adjusted at any time. The method adjusts the number of pure electric buses on the same line and the departure time point of the vehicles, encodes a vehicle scheduling scheme according to a designed encoding rule, utilizes a simulated annealing algorithm to perform first-stage search to generate a better candidate solution set, performs second-stage search on the candidate solution set, and adopts a heuristic adjusting method to adjust the departure time point in the scheme so as to minimize the total number of uncovered number and repeated covered number at the departure time. When the vehicle journey is generated by encoding and decoding, the vehicles needing to be charged need to be charged according to the needs, and the charging pile resources are scheduled. By adopting the method, the using number of the vehicles can be reduced, the number of the vehicle repeating covering and uncovered departure time points can be reduced, and the utilization rate of the electric buses and the charging resources can be improved.
Description
Technical Field
The invention relates to a vehicle scheduling and scheduling technology in the field of public transport, in particular to an intelligent scheduling method for a pure electric bus.
Background
At present, pure electric buses are more and more widely used, but due to the limitation of battery technology, the driving range of the electric buses is short, so that the electric buses need to meet all travel requirements by supplementing electric quantity in the course of one day. Therefore, the charging scheduling of the electric vehicles needs to be considered during the operation of the electric vehicles, and how to improve the utilization rate of the electric buses and save the charging resources is a problem to be solved, so that the scheduling problem of the electric buses needs to be researched and solved.
The scheduling problem of the electric bus scheduling is that a scheduling scheme which can cover all departure time points and minimize the operation cost of a bus company is obtained on the premise of meeting all constraint conditions according to a given departure schedule. On the basis of the traditional bus shift scheduling problem constraint condition, the electric bus shift scheduling problem needs to be added with the driving range constraint of the bus and the charging time of the bus. Moreover, the scheduling problem of the electric vehicle also needs to consider the charging process of the vehicle and the charging resource constraint. Therefore, the problems that the shift scheduling of the electric buses needs to be solved are that the driving range constraint of the electric buses is considered, the flexible charging strategy is adopted, and the charging resources are scheduled, so that the using number of the buses is optimized, the number of the vehicle repeating coverage and the number of the uncovered departure time points are reduced, and the utilization rates of the electric buses and the charging resources are improved.
At present, some researches aiming at the charging scheduling of the pure electric bus are provided. For example, chinese patent application publication No. 106991492a discloses a northern climate fast charging electric bus operation scheduling optimization method at 2017, 7/28, wherein vehicle charging is scheduled, but the scheduling of departure of the vehicle is not involved. In the prior art, the battery capacity, the number or the charging of the electric buses are mostly researched, and the research on how to optimize the use number of the vehicles, reduce the repeated coverage of the vehicles and the like is less in consideration of the specific departure time of each vehicle under the condition of determining the departure time table.
At present, the scheduling of the electric bus mainly adopts a manual scheduling mode, the manual scheduling mainly depends on the experience of a scheduling engineer, long time and more manpower are consumed, the generation period is long, and timely adjustment cannot be carried out. The scheduling problem of the electric bus is different from the traditional scheduling problem of the bus in terms of a solution method. The conventional bus shift scheduling problem solving process usually adopts a mathematical method or a heuristic algorithm, and does not consider the driving range constraint and the charging process of the bus.
The electric bus scheduling problem is relatively rarely researched. The existing method usually adopts the traditional mathematical method or the single heuristic algorithm to solve, and the single intelligent optimization method is difficult to produce better results for solving the combined optimization problem of the scale, and the solving efficiency is low. The existing research on the scheduling problem of the electric buses considers the driving range constraint of the buses, converts the scheduling problem of the electric buses into the scheduling problem of the buses with the driving range constraint, but does not consider the charging resource constraint.
In conclusion, the method in the prior art solves the problems that the characteristics of the electric buses are not fully considered, the practicability is low, and the efficiency of the generated scheduling scheme is low when the electric buses schedule.
Disclosure of Invention
The invention provides an intelligent scheduling method of a pure electric bus, aiming at solving the problems that the scheduling of the conventional electric bus depends on manual experience, consumes time and manpower and cannot be adjusted at any time.
The invention relates to an intelligent scheduling method of a pure electric bus, which executes the following steps for each bus of a public transport company:
the information of the vehicle comprises the working time and the rest time of a driver, the average running time of a train number journey of each vehicle in different time periods, the charging and discharging speed of the vehicle and the driving range of the vehicle;
setting the total departure time point number m according to a departure time table; the vehicle dispatching scheme is used for carrying out coded representation on whether vehicles are dispatched at m dispatching time points, and the coded representation is B1,B2,…,Bm(ii) a Wherein, BiShowing whether a vehicle is sent at the ith departure time point, i belongs to [1, m ]]When the ith departure moment has a vehicle to issue B i1, otherwise Bi0; when B is presentiWhen the time is 1, a travel set of the vehicle which is sent at the ith sending time point is correspondingly generated, and the sending direction and the selected several sending time points in the optional sending time interval at the time are recorded in each travel. The initial vehicle dispatching scheme randomly selects an initial departure time for N vehicles, namely randomly generates a code B containing N11,B2,…,BmAnd is combined withA set of trips for the N vehicles is generated. N is a positive integer.
And 2, decoding the codes of the initial vehicle dispatching scheme to obtain departure time points of all vehicles in the scheme. And counting the number of uncovered time points of the departure timetable and the number of repeated covered time points of departure in the scheme.
Step 3, carrying out first-stage search, and generating a candidate solution set by adopting a simulated annealing algorithm; in the simulated annealing algorithm, encoding an initial vehicle scheduling scheme as a current solution; and performing variation operation on individuals in the codes of the vehicle scheduling scheme to generate a neighborhood solution, evaluating the scheme according to the number of the vehicle uncovered and repeated covered starting time points and the using number of the vehicles in the scheme, and selecting a better solution from the neighborhood solution to add into the candidate solution set.
And 4, performing second-stage search on the candidate solution set generated in the step 3, selecting a vehicle scheduling scheme from the candidate solution set as an optimal solution, performing mutation operation on the optimal solution to generate a neighborhood solution, respectively adjusting the optimal solution and the neighborhood solution by adopting a heuristic adjusting method, evaluating the adjusted optimal solution and the adjusted neighborhood solution according to the total number of the uncovered quantity and the repeated covered quantity of the time points during vehicle sending in the scheme, selecting an optimal scheme from the optimal solution, and outputting the adjusted solution of the optimal solution as a final vehicle scheduling scheme when a set termination condition is reached.
The heuristic adjusting method is that the departure time points of all vehicles in the vehicle scheduling scheme are traversed, when a repeatedly covered departure time point is met, whether the slack time between certain vehicle travel of the departure at the time is larger than the interval time between adjacent departure time points is checked, if the slack time is larger than the interval time between the adjacent departure time points, the departure time of the vehicle is adjusted to the adjacent departure time, and then the travel of the vehicle is continuously adjusted until the condition cannot be met or a certain uncovered departure time point is met.
And 5, when the vehicle scheduling scheme is decoded and the travel of the electric bus is generated, the electric bus needing to be charged is charged as required, and the charging pile resource is scheduled.
Compared with the prior art, the electric bus dispatching method obtained by the method can reduce the number of vehicles, reduce the number of repeated vehicle coverage and uncovered vehicle departure time points, and improve the utilization rate of electric buses and charging resources. The method of the invention has the advantages and positive effects that:
(1) the method provided by the invention designs the known coding rule aiming at the scheduling problem of the electric bus to schedule the electric bus, so that convenience is provided for expressing and solving the scheduling problem of the bus. The invention adopts a mixed heuristic search method consisting of a simulated annealing algorithm and a local search strategy added with an adjusting method to search the solution space of the vehicle scheduling scheme, can avoid trapping in a local optimal solution, can more quickly and effectively find a better solution, generates a better scheduling scheme of the electric bus and improves the utilization rate of the vehicle.
(2) The invention considers the characteristics of the electric bus in the solution searching process, and adopts a more flexible and effective charging strategy and a charging resource scheduling method to solve the problems of short driving range and insufficient charging resource in the operation process of the electric bus, so that the scheme of the invention has more feasibility in implementation and simultaneously improves the utilization rate of the charging resource.
Drawings
FIG. 1 is an overall flow chart of the intelligent scheduling method of the electric bus of the present invention;
FIG. 2 is a schematic diagram of a departure time point selection according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of the encoding of a vehicle dispatch protocol according to an embodiment of the present invention;
FIG. 4 is a flowchart of a first stage simulated annealing algorithm search of the present invention;
FIG. 5 is a search flow chart of the second stage local search and heuristic tuning algorithm of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
The research object of the method is the pure electric bus on the same line, the number of strokes of each vehicle is fixed, and the number of the electric buses and the departure time point of the vehicles are optimized according to the method. Before the method of the present invention is performed, information of the departure schedule and the vehicles needs to be obtained from the public transportation company. The information of the vehicle comprises the working time and the rest time of a driver, and the average travel running time of each bus number of the vehicle in different time periods is obtained according to the statistics of the operation condition of the existing electric buses of the bus company. In addition, the charging and discharging speed of the electric bus and the driving range of the vehicle are counted according to the existing historical data record of the operation condition of the electric bus.
The method mainly comprises the steps that the departure condition of vehicles on a route is determined under the condition that a departure time table of the route is known (only the time of the starting station is determined from the starting station to the terminal station, and the time of the terminal station can be obtained through calculation), namely, each vehicle is determined to specifically depart at a certain departure time point on the departure time table, each vehicle has a plurality of departure routes one day, all vehicles form specific departure scheduling of the route, the scheduling is equivalent to the scheduling of drivers, the final aim is to cover the most departure time points with the least time, no vehicle or repeated departure (within the allowable range of a bus company) may exist at some time points, and all vehicles on the route may operate according to the scheme at a later time after the departure scheme of the route is determined.
According to the method, the average running conditions of the vehicle in different time periods are obtained according to the historical information of the vehicle, a scheduling scheme is further generated, and the residual electric quantity in different time periods of the vehicle can be obtained through calculation by performing simulation calculation on the residual electric quantity in the scheduling process.
An implementation flow of the intelligent scheduling method for the pure electric bus provided by the invention is shown in fig. 1, and implementation steps are described in detail below.
The method comprises the steps of obtaining a departure schedule of a certain line from a public transport company, obtaining statistical data of vehicles in a period of time, and obtaining information of the vehicles, wherein the information comprises working time and rest time of a driver, average running time of a trip of each vehicle in different periods of time, charging and discharging speed of the vehicles, driving range of the vehicles and the like. The invention aims to design the number of electric buses and the departure time of each electric bus for each circuit of electric buses, so that the minimum number of vehicles is used for covering all departure time points under the condition of meeting the constraint condition of the electric buses. In the method, the number of the travel of each electric bus per day is fixed.
As shown in FIG. 2, there are typically two departure stations per bus route, namely VS1And VS2And each station corresponds to one departure schedule. Each vehicle trip has its departure time and direction. For example, one vehicle will VS1As its origin, at t1At the moment of departure, t1The moment is the initial starting moment of the vehicle, and the vehicle passes through a Trip after the vehicle is started at the moment1Late arrival terminal VS2At which station a rest period T has elapsedrThen, according to the selectable departure time range T of the vehicle in the time periodWSelecting the next departure time point of the vehicle for departure in a later waiting time, wherein a plurality of departure time points may exist in the time period, so that one time point needs to be randomly selected in the range as the next departure time point of the vehicle, and then passes through the Trip2Then go back to VS1Station, vehicle past a rest period continues to select the next trip to VS2And (5) stations and so on.
For the electric bus, because the electric bus has the characteristic of short driving range, the electric quantity of the bus may not meet the travel requirement of the bus for one day, so that the electric bus needs to supplement the electric quantity to complete all the travels for one day in the operation process of one day, namely, a charging process is needed. As shown in FIG. 2, the vehicle is at t3Time slave VS1Starting from the station and passing through the Tripn-1Late arrival VS2The station, in which the remaining charge of the vehicle is not sufficient to meet the demand of the next trip, needs to be replenished with charge by a charging operation, i.e. at TcAnd supplementing the electric quantity in a time period, and continuing to select the next journey of the vehicle after supplementing a proper quantity of electric quantity according to needs until the specified number of journeys is finished, namely generating a one-day departure journey block of the electric bus. In the process of generating the departure travel block of the vehicle, the maximum working time of a driver is also considered, and when the current operation time of the vehicle exceeds the maximum working time of the driver specified by a company, the travel block is an infeasible vehicle dispatching scheme, and a new departure travel block is generated again.
When generating the departure travel block of the vehicle, determining an optional departure time interval of each vehicle in each time period according to the maximum working time and the minimum rest time of a driver, wherein the optional departure time interval generally comprises a plurality of departure time points. The constraint for each vehicle is that the departure time for the next trip must be later than the arrival time of the current trip at the station plus the minimum rest time T of the driverrWhile the maximum waiting time T of the vehicle at the station is required to be less thanw。
And setting the total departure time point number m according to the acquired departure timetable, wherein the value of m is at least equal to the total departure times in the departure timetable. The method comprises the steps of setting N electric buses initially, randomly selecting an initial starting time point for each electric bus, and then generating a departure travel block of each electric bus according to the process, wherein the departure travel blocks of the N electric buses form an initial vehicle dispatching scheme of the line. N is a positive integer and can be set empirically.
Different from the traditional bus scheduling, the electric bus has the opportunity to reach a charging station for charging every time the bus is executed for one bus number in the operation process. In the scheduling of the electric public transport vehicle, when the residual electric quantity in the running process of the vehicle cannot meet the requirement of the next journey, the vehicle needs to be charged to a charging station, and the charging time is TcAfter charging, the vehicle may continue to select the next trip. Thus, the coding of the departure times of the vehicles requires consideration of the minimumRest time TminBesides, the factors of the electric quantity of the vehicle, including the residual electric quantity of the vehicle, the starting charging time and the charging time when the vehicle needs to be charged due to insufficient electric quantity, need to be considered. The travel of all vehicles constitutes the scheduling scheme of the bus route.
The invention uses codes to represent the scheduling scheme of the electric bus, and the designed codes need to represent the scheduling scheme of all the buses. The coding scheme designed by the present invention is shown in fig. 3. Set up BiIndicating whether there is a vehicle to issue at the ith departure time point, Bi∈{0,1}，B i1 represents that the ith departure time point has a vehicle to proceed departure, Bi0 means that no vehicle is sent at the ith departure time, i belongs to [1, M ]]And M represents the total number of departure time points. When B is presentiWhen 1, the vehicle with the ith departure time point as the initial departure time point needs to be scheduled, that is, each BiCorresponding to a set of trips TSet representing the departure trips for the vehicle for one day. The TSet is made up of all the trips of the vehicle, each trip representing a trip of the vehicle from one station to another. X in FIG. 311,X12,…,X 1nThe departure distance of the vehicle which departs at the moment of departure of the 1 st departure is shown. And recording the departure direction and the selected time point of the selectable departure in each journey. Departure direction means that the vehicle is from VS1To VS2Departure or departure VS2To VS1And (5) dispatching the car. The selected first optional departure time point refers to the time point when the vehicle selects the first departure time point in the current optional departure time interval.
The departure time selected in each trip is selected and the several selectable departure times are selected. When a plurality of initial vehicle dispatching schemes are generated, namely a plurality of codes B are randomly generated1,B2,…,Bm. For each initial vehicle scheduling plan, generating B i1 corresponds to the departure vehicle's travel set.
And 2, decoding the codes of the vehicle scheduling scheme to obtain the starting time points covered by all vehicles in the vehicle scheduling scheme.
And (3) decoding the codes of the initial vehicle scheduling scheme randomly obtained in the step (1) and carrying out departure time points of each departure vehicle. In the decoding process, the maximum working time and the minimum rest time of a driver driving the vehicle need to be considered for obtaining the departure time point of each vehicle according to the constraint conditions, and the electric quantity of the vehicle needs to be judged to judge the remaining driving mileage of the vehicle and whether the vehicle needs to be charged in the running process. The method adopts partial charging in the charging process, calculates the charging time instead of charging the whole vehicle by calculating the next demand of the vehicle on the electric quantity, reduces the charging time of the vehicle in the operation process, and improves the utilization rate of the vehicle and charging resources.
The charging process mainly considers the relation between the battery State (SOC) of the vehicle and the vehicle running condition, and the vehicle has to meet the minimum electric quantity SOC if the vehicle is required to execute the vehicle-level taskminRequest that vehicle SOC be not less than SOC at departure time of the trainminAnd when the residual electric quantity of the vehicle cannot meet the requirement of the next journey, the vehicle needs to be charged, and the charging time is calculated. The charging time is calculated by mainly considering the remaining working time of the vehicle and the number of remaining number of vehicle tasks, the charging time is calculated by calculating the next possible running time of the vehicle, charging is carried out according to needs instead of charging the whole vehicle, the charging time of the vehicle in the operation process is reduced, and meanwhile, the utilization rate of the vehicle and charging resources is improved. Next, in step 5, the charging method and the charging resource scheduling of the electric bus will be described.
The method takes the coverage condition of the departure time point in the vehicle scheduling scheme as an evaluation index to judge the quality of the scheme. And counting the number of uncovered time points of the departure timetable and the number of repeated coverage departure time points in the scheme as evaluation values. The following two-stage search method is used to obtain the optimal solution.
And 3, performing first-stage search to obtain a candidate solution set of the vehicle scheduling scheme.
The first stage searching process mainly considers the covering condition of the starting time point and the number of vehicles in use. For the case of the coverage of the departure time point, in addition to the number of the departure time points which are uncovered and repeatedly covered, the balance value of the number of the departure time points which are uncovered and repeatedly covered, namely the difference value between the number of the departure time points and the number of the moment points which are repeatedly covered needs to be considered, and the smaller the difference value between the number of the departure time points and the number of the moment points which are repeatedly covered, the easier the adjustment is, so that the candidate scheme set generated in the first-stage search process is convenient for the adjustment. Meanwhile, the number of vehicles used is also considered in the calculation of the evaluation function, and the utilization rate of the vehicles is higher as the number of the vehicles is smaller. Therefore, the merit function F of the first stage search process is expressed as follows:
since the driving conditions of vehicles in different time periods may be different, the departure timetable is grouped when the coverage condition is calculated, the coverage condition of the departure time point in each group is respectively calculated, b represents the number of the departure stations, in the invention, b takes 2, two stations are VS1And VS2。niNumber of groups, u, representing departure times of the ith station grouped in orderijNumber r of uncovered time points in jth packet representing ith stationijRepresenting the number of times that the time point in the jth group of the ith station is repeatedly covered, and the number c of continuous uncovered time points appearing in the time tableiAdded as a penalty value to the objective function. c. CiRepresents the uncovered penalty value of continuous time points in the ith station time schedule, v represents the number of vehicles used, w1,w2And w3Are weights. The value of the weight can be adjusted and set according to the actual situation.
For the vehicle scheduling problem to be solved by the present invention, each vehicle scheduling scheme initialized in step 1 is a feasible solution. The first stage of the invention mainly uses a dieAnd (3) performing neighborhood search by taking the feasible solution in the step (1) as a current solution, wherein the construction of the neighborhood solution is mainly completed by performing mutation operation on individuals in the codes, and the current solution is mutated with a certain probability p, namely, for each iteration process, a random number r (0) is randomly generated<r<1) And if r is smaller than a preset mutation probability threshold value, performing mutation operation on the current solution, otherwise, not performing mutation operation. In the embodiment of the present invention, p is set to be 0.05, and the mutation probability threshold is set to be 0.05. As shown in fig. 4, a process for generating a better solution candidate for performing simulated annealing. Mutation operations are performed on the encoding of the solution, changing a certain bit or bits in the encoding to achieve the construction of the neighborhood solution. For the neighborhood solution M generated by mutation operation, Metropolis criterion is adopted to judge whether to accept the new solution. If the evaluation value of the current solution is less than or equal to F (M) and is more excellent than the current solution, adding the variant solution M into the candidate solution set, directly replacing the current solution with the variant solution, and then entering the next iteration; if F (M)>And F (S), namely the evaluation value of the variant solution is larger than that of the current solution, the current solution is more excellent, the variant solution is accepted as the current solution according to the probability p (t), and then the next iteration is carried out. Where t is the current temperature, p (t) exp ((F (M) -F (S))/t), the temperature t decreases with the increase of the number of iterations k, and the temperature t at the k-th iterationk＝t0βk，t0Beta is the annealing factor at the initial temperature. And repeating the step of constructing the domain solution to generate a neighborhood solution of the current solution, selecting to accept or abandon the variant solution until the iteration times reach the specified maximum iteration times to obtain a preferred candidate solution, adding the preferred candidate solution into the candidate solution set, and stopping searching in the first stage when the number of the candidate solutions reaches the set termination condition. The termination condition set here is set as the total number of solutions in the candidate solution set.
And 4, performing second-stage search on the candidate solution set generated by the first-stage search, adjusting solutions with the quantity of uncovered departure time points generated in the first stage being similar to the quantity of the repeated coverage departure time points, and adjusting the travel of the repeated coverage points to the uncovered departure time points. And (5) finishing the charging process of the electric buses by reducing the number of uncovered departure time points and repeated coverage of the departure time points in the scheduling scheme and combining the step 5, generating an optimal solution and outputting a scheduling scheme.
The objective function G searched in the second stage is different from that searched in the first stage, and for each vehicle dispatching scheme, the total number of uncovered quantity and repeated covered quantity at the time point of departure is mainly considered, and the smaller the total number is, the better the coverage condition is. The objective function G is represented as follows:
wherein u isiIndicating the number of uncovered time points, r, at the time of departureiRepresenting the number of repeated coverage of a moment of departure, v representing the number of vehicles, w4And w5For the set weight, the value of the weight can be adjusted and set according to the actual situation.
After the candidate solution set is generated by the simulated annealing algorithm, local search and a heuristic adjustment strategy are used for further search in the step. And searching the better solution generated by the candidate solution set through mutation by a local search method, and constructing a new solution by using the same neighborhood solution construction mode as the first stage. And reducing the number of uncovered starting time points and repeated covering starting time points in the scheduling scheme by adopting a heuristic adjusting method. The heuristic adjusting strategy mainly has the following idea that all departure time points of all vehicles in a vehicle scheduling scheme are traversed, when a repeatedly covered departure time point is met, whether the slack time between certain vehicle travel of the departure at the time is larger than the interval time between adjacent departure time points is checked, if the slack time is larger than the interval time between the adjacent departure time points, the departure time of the vehicle is adjusted to the adjacent departure time, then the vehicle travel is continuously adjusted according to the same direction, and the departure time point and the slack time of the vehicle number are modified until the condition cannot be met or a certain uncovered departure time point is met. The slack time refers to the difference between the start time of the vehicle executing the next train number task and the time the last train number task arrives at the station.
As shown in fig. 5, the second stage of the search process first selects a solution from the candidate solution set as an optimal solution T, and starts iterative search. In each iteration, firstly adjusting T to obtain an adjustment solution T ', then calculating a target function G (T') of T ', and setting the target function G (T) of T to G (T'); then, as in the first stage, performing individual mutation operation on T to generate a neighborhood solution M, adjusting M to obtain an adjusted solution M ', then calculating an objective function G (M') of M ', and setting the objective function G (M) of M to G (M'); and further judging whether the difference value between G (M) and G (T) is less than or equal to 0, if so, receiving M, updating M to be the optimal solution T, entering the next iteration, otherwise, not receiving M, and continuing to enter the next iteration search by using the current optimal solution. And repeating the iteration process until the maximum iterative times are reached, judging whether a termination condition is reached, if so, stopping the second-stage search, and outputting the current adjustment solution of the optimal solution to serve as the optimal electric bus dispatching scheme. And if the termination condition is not met, selecting one solution from the candidate solution set as the optimal solution again, and continuing the iterative search. And when the solution is adjusted, a heuristic adjusting strategy is used for adjusting the vehicle dispatching scheme. The termination condition set here refers to completing the traversal of all candidate solutions.
And 5, when the journey of the electric bus is generated, the charging requirement of the bus is also considered. When all vehicles need to be charged in a certain time period in the operation process, due to the limitation of the number of charging piles or the limitation of the voltage of the charging station, only a certain number of electric buses can be allowed to be charged simultaneously. At this time, the vehicle charging time needs to be adjusted, that is, the vehicle needing to be charged is scheduled to meet the charging resource constraint. Therefore, the charging resources, i.e. the charging piles, need to be scheduled in the decoding process of the codes of the vehicle scheduling scheme.
The charging pile scheduling process of the method mainly comprises two steps:
step A, forward adjustment based on a greedy strategy, namely, adjusting the charging time each time only considers the starting charging time and the charging time of the vehicle which has finished the scheduling coding process, and the charging time of the vehicle can be advanced properly, so that the next operation of the vehicle is not influenced, and the scheduling of charging pile resources can be finished, and the specific process is as follows:
a) if the charging pile resource is sufficient in the current vehicle charging time period, the vehicle can be directly charged;
b) if all the charging piles are occupied in the current vehicle charging starting time period, adjusting the charging starting time of the vehicle, adjusting the charging process of the vehicle forward to meet the vehicle quantity limit of the charging piles, and then generating the adjusted vehicle number task of the vehicle again;
c) and if the charging pile resource in the current vehicle charging time period is insufficient and the vehicle cannot continuously adjust the charging time period forwards due to sufficient electric quantity, scheduling and charging according to the priority of the current waiting vehicle, and executing the step B.
And step B, based on the dynamic scheduling of the priority, when the charging pile resources are insufficient and a plurality of vehicles need to be charged, the charging sequence of the vehicles is arranged by calculating the priority of each vehicle so as to complete the scheduling of the charging pile resources. The calculation of the vehicle priority is mainly based on the waiting time T of the vehicle at the charging stationwaitAnd a charging time T required for the vehiclecAnd the remaining task time T of the vehiclenext. When the vehicle charging time is the same, the higher the vehicle charging priority of the waiting time is, and similarly, for the vehicle with the same waiting time, the shorter the charging time is, the earlier the occupation of the charging pile resources can be completed, and the higher the priority is. Meanwhile, the shorter the remaining task time of the vehicle is, the earlier the vehicle can complete all the travel tasks, and the higher the priority is.
Through the two-stage searching process and the scheduling of the charging resources in each process, a solution meeting all constraint conditions, namely an optimal scheduling scheme of the electric buses can be obtained finally. And finally, selecting the vehicle to complete the scheduling driving plan according to the obtained scheduling scheme of the vehicle, thereby not only meeting the actual working requirement, but also improving the operation level and the service quality of the public transport company and improving the utilization efficiency of the electric public transport vehicle and the charging resource. The method can also be used for intelligently adjusting the vehicles running on the line according to the line bus information acquired by the bus company at regular intervals, thereby solving the problem of difficult traditional adjustment.
It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Claims (7)
1. An intelligent scheduling method for pure electric buses is characterized in that the following steps are executed for each bus of a public transport company:
step 1, obtaining information of a departure schedule and vehicles from a public transport company, and generating an initial vehicle scheduling scheme;
the information of the vehicle comprises the working time and the rest time of a driver, the number of the vehicle-class trips of each vehicle in one day, the average running time of one vehicle-class trip of each vehicle in different time periods, the charging and discharging speed of the vehicle and the driving range of the vehicle;
setting the total departure time point number m according to a departure time table; the vehicle dispatching scheme is used for carrying out coded representation on whether vehicles are dispatched at m dispatching time points, and the coded representation is B1,B2,…,Bm(ii) a Wherein, BiShowing whether a vehicle is sent at the ith departure time point, i belongs to [1, m ]]When the ith departure moment has a vehicle to issue Bi1, otherwise Bi0; when B is presentiWhen the time is 1, correspondingly generating a travel set of the vehicle which starts at the ith starting time point, and recording the starting direction and the selected several starting time points in the optional starting time interval at the time in each travel;
each initial vehicle scheduling scheme is a randomly generated code B containing N11,B2,…,BmAnd generates Bi1, corresponding to the route set of the departure vehicle; n is positiveAn integer number;
step 2, decoding the codes of the vehicle scheduling scheme to obtain the departure time points covered by all vehicles in the scheme;
step 3, carrying out first-stage search, and generating a candidate solution set by adopting a simulated annealing algorithm; in the simulated annealing algorithm, encoding an initial vehicle scheduling scheme as a current solution; generating a neighborhood solution by carrying out variation operation on individuals in the codes of the vehicle scheduling scheme, evaluating the scheme according to the number of the vehicle uncovered and repeated covered starting time points and the using number of the vehicles in the scheme, and selecting a better solution from the neighborhood solution to add into a candidate solution set;
step 4, performing second-stage search, selecting a vehicle scheduling scheme from the candidate solution set as an optimal solution, performing mutation operation on the optimal solution to generate a neighborhood solution, respectively adjusting the optimal solution and the neighborhood solution by adopting a heuristic adjusting method, evaluating the adjusted optimal solution and the adjusted neighborhood solution according to the total number of uncovered quantity and repeated covered quantity of the vehicle-sending time points in the scheme, selecting an optimal scheme from the optimal solution, and outputting the adjusted solution of the optimal solution as a final vehicle scheduling scheme when a set termination condition is reached;
the heuristic adjusting method is that the departure time points of all vehicles in the vehicle scheduling scheme are traversed, when a repeatedly covered departure time point is met, whether the slack time between certain vehicle travel of the departure at the time is larger than the interval time between adjacent departure time points is checked, if the slack time is larger than the interval time between the adjacent departure time points, the departure time of the vehicle is adjusted to the adjacent departure time, and then the travel of the vehicle is continuously adjusted until the condition cannot be met or a certain uncovered departure time point is met; the slack time refers to the difference between the starting time of the vehicle executing the next train number task and the time of the last train number task reaching the station;
step 5, when the vehicle scheduling scheme is decoded and the journey of the electric bus is generated, the electric bus needing to be charged is charged as required, and charging pile resources are scheduled; the scheduling of charging piles comprises the following steps:
step A, forward adjustment based on a greedy strategy, specifically:
the method is characterized in that the charging time is adjusted each time only by considering the starting charging time and the charging time of the vehicle which has finished the scheduling coding process, and the charging time of the vehicle can be advanced properly, so that the next operation of the vehicle is not influenced, and the scheduling of charging pile resources can be finished, and the specific process is as follows:
a) if the charging pile resource is sufficient in the current vehicle charging time period, the vehicle is directly charged;
b) if all the charging piles are occupied in the current vehicle charging starting time period, adjusting the vehicle charging starting time, adjusting the charging process of the vehicle forward to meet the resource limitation of the charging piles, and generating the adjusted vehicle number task of the vehicle again;
c) if the charging pile resource in the current vehicle charging time period is insufficient and the vehicle cannot continuously adjust the charging time period forwards due to sufficient electric quantity, executing the step B;
b, scheduling and charging according to the priority of the current waiting vehicle; the priority setting requirements of the vehicle are: when the vehicle charging time is the same, the longer the waiting time, the higher the vehicle charging priority; for vehicles with the same waiting time at the station, the shorter the charging time is, the higher the priority is; the shorter the remaining mission time of the vehicle, the higher the priority.
2. The method according to claim 1, wherein in step 1, when generating the initial vehicle dispatching scheme, the selectable departure time interval of each vehicle in each time period is determined according to the maximum working time and the minimum rest time of a driver, and the initial departure time and the departure time of the next trip are sequentially selected for each vehicle; when the vehicle journey is generated, the requirements are met: the departure time of the next journey must be later than the arrival time of the current journey at the station plus the minimum rest time T of the driverrWhile the maximum waiting time T of the vehicle at the station is required to be less thanwAnd when the residual electric quantity of the vehicle can not meet the requirement of the next journey, charging is carried out on the vehicle to a charging station, and charging is setElectric time of Tc。
3. The method according to claim 1, wherein in the step 3, the vehicle dispatching plan is evaluated by calculating an evaluation value according to the following evaluation function F:
wherein, b represents the number of the starting stations and has the value of 2; grouping departure schedules according to different time periods when calculating coverage conditions, niIs the number of groups; u. ofijRepresenting the number of uncovered departure time points in the jth group of the ith station; r isijRepresenting the number of times that the departure time point in the jth group of the ith station is repeatedly covered; c. CiRepresenting the uncovered number of the continuous time points in the departure schedule of the ith station; v represents the number of vehicles used; w is a1,w2And w3The weight value can be adjusted and set according to the actual situation.
4. The method of claim 1 or 3, wherein the step 3 of generating the better solution candidate by using the simulated annealing algorithm comprises: in the process of one iteration, the current solution S is mutated by a probability p, a neighborhood solution M is generated through mutation operation, evaluation values F (S) and F (M) of S and M are compared, if F (M) is less than or equal to F (S), M is replaced by the current solution S, and then the next iteration is carried out; if F (M) > F (S), replacing M with the current solution S by the probability p (t), and then entering the next iteration, wherein p (t) is exp ((F (M) -F (S))/t), t is the current temperature, and t is reduced along with the increase of the number of iterations; when the iteration times reach the set maximum iteration times, stopping the iteration, and adding the current solution into a candidate solution set;
and continuing to search and iterate the next initial vehicle dispatching scheme code until the number of the candidate solutions reaches the set number, and stopping searching.
5. The method of claim 1, wherein in step 4, the vehicle dispatching scheme is evaluated by calculating an evaluation value according to the following evaluation function G:
wherein u isiIndicating the number of uncovered time points, r, at the time of departureiRepresenting the number of repeated coverage of a moment of departure, v representing the number of vehicles, w4And w5Is the set weight.
6. The method of claim 1, wherein in step 4, the termination condition of the second-stage search means that the traversal of all candidate solutions in the candidate solution set is completed.
7. The method according to claim 1 or 2, wherein in the step 2, when decoding the codes of the vehicle scheduling scheme, the electric quantity of the vehicle needs to be judged, and whether the battery state of the vehicle meets the lowest electric quantity SOC is judgedminAnd the sum of the power consumption of the next vehicle number task, if the sum is not satisfied, the vehicle needs to be charged, and the charging time T is calculatedcCharging time TcThe calculation is carried out according to the remaining working time of the vehicle and the number of the remaining vehicle number tasks.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910139994.8A CN109934391B (en) | 2019-02-26 | 2019-02-26 | Intelligent scheduling method for pure electric bus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910139994.8A CN109934391B (en) | 2019-02-26 | 2019-02-26 | Intelligent scheduling method for pure electric bus |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109934391A CN109934391A (en) | 2019-06-25 |
| CN109934391B true CN109934391B (en) | 2020-11-10 |
Family
ID=66985947
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910139994.8A Active CN109934391B (en) | 2019-02-26 | 2019-02-26 | Intelligent scheduling method for pure electric bus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109934391B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112396271A (en) * | 2019-08-16 | 2021-02-23 | 顺丰科技有限公司 | Vehicle shift scheduling time reduction method and system |
| CN110581320B (en) * | 2019-09-24 | 2021-05-04 | 厦门金龙联合汽车工业有限公司 | Bus power lithium battery heat preservation mechanism management method and cloud management server |
| CN110920457A (en) * | 2019-12-24 | 2020-03-27 | 重庆国翰能源发展有限公司 | Charging module distribution method for four-pile charging pile |
| CN111325483B (en) * | 2020-03-17 | 2024-01-26 | 郑州天迈科技股份有限公司 | Electric bus scheduling method based on battery capacity prediction |
| CN113673069B (en) * | 2020-05-14 | 2023-11-03 | 南京行者易智能交通科技有限公司 | New energy bus charging model design method and device and mobile terminal equipment |
| CN112001560B (en) * | 2020-09-01 | 2023-05-16 | 湖南智慧畅行交通科技有限公司 | Two-stage bus scheduling algorithm based on iterative neighborhood search |
| CN112257999B (en) * | 2020-10-10 | 2022-11-01 | 东南大学 | Self-adaptive large-scale neighborhood searching method for large-scale pure electric bus scheduling problem |
| CN112233451B (en) * | 2020-10-13 | 2022-05-27 | 安徽交欣科技股份有限公司 | Intelligent traveling plan compiling system considering endurance mileage of pure electric bus |
| CN113256004B (en) * | 2021-05-27 | 2023-05-12 | 支付宝(杭州)信息技术有限公司 | Vehicle scheduling method, device, computer equipment and storage medium |
| CN115547052B (en) * | 2022-10-14 | 2023-09-19 | 大连海事大学 | Dynamic demand response electric bus scheduling method for improving self-adaptive large neighborhood algorithm |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102542791B (en) * | 2011-12-29 | 2014-10-01 | 北京邮电大学 | Method for dispatching public transport vehicles |
| US9168419B2 (en) * | 2012-06-22 | 2015-10-27 | Fitbit, Inc. | Use of gyroscopes in personal fitness tracking devices |
| CN103678917B (en) * | 2013-12-13 | 2016-11-23 | 杭州易和网络有限公司 | A kind of real-time arrival time Forecasting Methodology of public transport based on simulated annealing |
| US11214157B2 (en) * | 2015-12-30 | 2022-01-04 | Proterra Inc. | Electric vehicle charging to reduce utility cost |
| CN107341563A (en) * | 2017-06-09 | 2017-11-10 | 安徽富煌科技股份有限公司 | A kind of new energy bus intelligent charge dispatches combined optimization method |
-
2019
- 2019-02-26 CN CN201910139994.8A patent/CN109934391B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109934391A (en) | 2019-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109934391B (en) | Intelligent scheduling method for pure electric bus | |
| CN108199100B (en) | Electric automobile long-distance operation charging planning method in intelligent traffic | |
| CN111915176B (en) | Scheduling method and system for pure electric bus in hybrid operation mode | |
| CN108955711B (en) | Navigation method applied to intelligent charging and discharging of electric automobile | |
| CN108390421B (en) | Electric vehicle double-scale charging guiding method and system considering user satisfaction degree | |
| EP2768693B1 (en) | Method, system and charging station for charging electric vehicles | |
| CN112193116B (en) | Electric vehicle charging optimization guiding strategy considering reward mechanism | |
| CN109177802B (en) | Electric automobile ordered charging system and method based on wireless communication | |
| US20200058090A1 (en) | Method and device for determining a configuration for deployment of a public transportation system | |
| CN113283623A (en) | Electric vehicle electric quantity path planning method compatible with energy storage charging pile | |
| CN115100896B (en) | Electric demand response bus dispatching method considering opportunity charging strategy | |
| CN115547052B (en) | Dynamic demand response electric bus scheduling method for improving self-adaptive large neighborhood algorithm | |
| CN111882156B (en) | Train schedule robust optimization method for random dynamic passenger flow and energy-saving operation | |
| CN111325483A (en) | Electric bus scheduling method based on battery capacity prediction | |
| CN114444965B (en) | Single-yard multi-line electric bus collaborative scheduling method | |
| CN114742340A (en) | Optimal layout solving method for intelligent network connection sharing electric vehicle charging station in large-scale road network | |
| CN113486504A (en) | Battery management control method based on scheduling cost | |
| CN117035357A (en) | Charging and discharging scheduling method for electric taxi team in energy-traffic | |
| CN112149906A (en) | Electric vehicle trip line comprehensive optimization method considering charging time | |
| CN115663867B (en) | Electric automobile charging scheduling method based on intelligent charging network system | |
| CN113642796B (en) | Dynamic sharing electric automatic driving vehicle path planning method based on historical data | |
| CN115848196A (en) | Electric automobile ordered charging guide method based on dynamic demand and new energy consumption | |
| CN115936330A (en) | Demand response type BRT vehicle scheduling algorithm based on Lagrangian principle | |
| CN114077972A (en) | Mobile charging vehicle service scheduling method based on deep reinforcement learning | |
| CN115049272A (en) | Electric bus dispatching method for charging intermediate station based on battery exchange |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |