Bus timetable synchronization and cross-line scheduling method for oil-electricity hybrid motorcade

Abstract

The invention relates to a bus timetable synchronization and cross-line scheduling method for a fuel-electric hybrid fleet, and the method comprises the steps: constructing a data set according to bus network information, including lines, transfer stations, passenger arrival rates, driving time and power consumption; constructing a multi-target integer programming model which minimizes the operation cost of a bus company and the waiting time of passengers and maximizes the number of convenient transfer times; an NSGAII-SA-PSO nested heuristic algorithm is designed, and the most suitable algorithm is allocated to decision variables with different characteristics, so that the quality of the solution is improved. According to the invention, through timetable synchronization and vehicle cross-line coordination, the operation efficiency of buses is significantly improved, and the travel experience of passengers is effectively improved.

Description

TECHNICAL FIELD

[0001] The application provides a bus schedule synchronization and cross-line scheduling method for a hybrid electric bus fleet, and belongs to the technical field of urban public transportation. BACKGROUND

[0002] A large number of bus systems have long been plagued by the contradiction between insufficient service efficiency and low utilization rate of transport capacity. At present, most bus lines still use the operation mode of fixed route, fixed schedule and fixed vehicle formation, which is difficult to adjust flexibly according to different degrees of passenger demand changes. This mode not only cannot effectively cope with the service pressure of concentrated travel during peak hours, but also leads to low resource utilization efficiency due to excess transport capacity during off-peak hours; at the same time, there is a lack of coordination between different line schedules, which exacerbates the waiting of passengers during transfer. In addition, the promotion of the concept of green travel to electric vehicles is in sharp conflict with the fact that a large number of existing fuel vehicles, so building a hybrid electric bus fleet has become an important direction to improve the bus system.

[0003] The bus schedule synchronization and cross-line scheduling for a hybrid electric bus fleet involves multiple interrelated optimization links: (1) schedule design: the departure times of each line should be optimized in coordination, and the headway should be reasonably arranged to minimize passenger transfer and waiting time; (2) vehicle scheduling: in a multi-line and multi-period operation environment, the dynamic matching scheme of vehicles and lines needs to be determined to reduce the size of the fleet while ensuring service levels; (3) charging planning: for electric vehicles in the hybrid fleet, an efficient charging strategy is developed to minimize the interference of charging behavior on the operation trip while ensuring sufficient battery power. However, existing research usually handles schedule synchronization, hybrid electric bus fleet and vehicle scheduling separately, and lacks a method for schedule synchronization and cross-line scheduling for a hybrid electric bus fleet in the same framework. Therefore, how to realize the coordinated optimization of the hybrid bus fleet while considering passenger demand and vehicle transport capacity has become a technical problem that needs to be solved. SUMMARY

[0004] The application solves the problem: In response to the low efficiency of current bus services and the demand for hybrid electric bus fleets, the regional vehicle coordination mode is introduced to change the fixed single-line operation mode, and a bus schedule synchronization and cross-line scheduling method for a hybrid electric bus fleet is proposed. The method can adjust the allocation of transport capacity on different lines by using the cross-line scheduling method of buses, and synchronize the schedules of different lines to reduce transfer time, thereby improving the service quality and efficiency of bus lines, reducing bus operating costs, and improving passenger travel experience.

[0005] The technical solution of the present application: facing the demand of passengers to reduce travel time, introducing cross-line cooperative scheduling mode, synchronizing the time table of each line. Construct a multi-objective optimization model that minimizes the public transport operating cost and passenger waiting time, maximizes the number of convenient transfers. And design NSGAII-SA-PSO nested heuristic algorithm, so as to generate high-quality, oil-electric hybrid vehicle fleet-oriented public transport timetable synchronization and cross-line scheduling scheme.

[0006] In order to achieve the above object, the technical scheme adopted by the present application is as follows: a kind of bus timetable synchronization and cross-line scheduling method for oil-electric hybrid vehicle fleet, comprising the following steps: First step, information processing: according to the information of public transport network, determine the line, starting and ending point, station and transfer station, calculate the travel time and power consumption between nodes, based on historical passenger flow data to obtain the passenger arrival rate of each line starting point; Second step, construct mathematical model: minimize the public transport operating cost and passenger waiting time, maximize the number of convenient transfers as the goal, meet the constraints of vehicle scheduling and charging scheme feasible, vehicle head distance within the predetermined threshold, all bus trips are executed, establish mathematical optimization model; Third step, algorithm solving: the multi-objective integer programming model obtained is solved by using NSGAII-SA-PSO nested heuristic algorithm, and the Pareto frontier containing the following decision variables is obtained: (1) line timetable: The departure time of each class of the line ; (2) vehicle type and scheduling scheme: the number of hired oil cars And the number of electric cars , the oil-electric type corresponding to each car , which line and which class each car serves in turn;(3) vehicle charging scheme: The charging amount of the first task of the vehicle after th time, the public transport operating cost , passenger waiting time and the number of convenient transfers are calculated.

[0007] Compared with the prior art, the present application has the following beneficial effects: For the complex problem of regional vehicle coordination and timetable optimization in the public transportation system, the existing research often considers vehicle cross-line scheduling, timetable synchronization and hybrid electric vehicle fleet in isolation, or considers public transportation operating cost and passenger waiting and transfer cost in isolation, which is difficult to meet the actual demand. The present application builds a multi-objective optimization model, with the goal of minimizing public transportation operating cost and passenger waiting time, and maximizing the number of convenient transfers, to realize the coordinated optimization of departure timetable, vehicle charging scheme and scheduling scheme, significantly improving the overall performance of the system, and having important practical significance.

[0008] The present application proposes a bus system optimization framework based on NSGAII-SA-PSO nested heuristic algorithm, which uses the operator characteristics and iteration characteristics of different heuristic algorithms to improve the adaptability of algorithm types and decision variable types; high-quality initial solutions and auxiliary variables (such as offsets) are designed to further improve the overall quality and diversity of the solution set. BRIEF DESCRIPTION OF DRAWINGS

[0009] Figure 1 is a flow chart of a bus timetable synchronization and cross-line scheduling method for a hybrid electric vehicle fleet; Figure 2 is a schematic diagram of a regional public transportation network. DETAILED DESCRIPTION

[0010] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other. In order to achieve the above-mentioned purpose, the present application adopts the following technical scheme.

[0011] The present application proposes a bus timetable synchronization and cross-line scheduling method for a hybrid electric vehicle fleet, as shown in Figure 1 , including the following steps: First step, information processing: according to the public transportation network information, determine the line, starting and ending point, station and transfer station situation, calculate the travel time and power consumption between nodes, and get the passenger arrival rate at the starting point of each line based on historical passenger flow data; Second step, build mathematical model: take minimizing public transportation company operating cost and passenger waiting time, and maximizing the number of convenient transfers as the goal, meet the constraints of vehicle scheduling and charging scheme feasibility, headway within the predetermined threshold, and all public transportation trips are executed, establish a mathematical optimization model; the symbols used are summarized in the following table: The mathematical model is as follows: First, the operating cost of the bus company is minimized according to the value coefficient and the amount of each cost. Equation (1) sequentially calculates the total fixed cost of the gasoline bus, the total fixed cost of the electric bus, the total cost of the gasoline bus carrying passengers, the total cost of the gasoline bus empty running (inter-route scheduling, the first bus leaving the station, the last bus entering the station), the total cost of the electric bus carrying passengers, and the total cost of the electric bus empty running (inter-route scheduling, charging in and out of the station, the first bus leaving the station, and the last bus entering the station): (1) wherein, set: - the set of all bus vehicles, , wherein represents the total number of vehicles; - the set of all tasks executed by the vehicle , , wherein represents the total number of daily tasks executed by the vehicle ; parameters: - the fixed cost of the gasoline bus ; - the empty running cost of the gasoline bus, 10,000 yuan ; - the passenger carrying running cost of the gasoline bus, 10,000 yuan ; - the empty running carbon emission conversion cost of the gasoline bus, 10,000 yuan ; - the passenger carrying running carbon emission conversion cost of the gasoline bus, 10,000 yuan ; - the fixed cost of the electric bus ; - the empty running cost of the electric bus, 10,000 yuan ; - the passenger carrying running cost of the electric bus, 10,000 yuan ; - the empty running time from station to station ; - the passenger carrying running time from station to station ; Index: - sequence number of bus line, ; - sequence number of bus vehicle, ; - sequence number of vehicle task, (e.g. for a vehicle); ; - start point of line ; - end point of line ; - depot; Decision variables: - binary variable, taking value 1 if vehicle is an electric bus, otherwise 0; Auxiliary variables: - integer variable, number of oil buses; - integer variable, number of electric buses, with ; - binary variable, taking value 1 if vehicle is charged after its th task, here ; - integer variable, vehicle is working on line after its th task.

[0012] Equation (2) represents the minimization of the total waiting time of passengers on each line. For a bus service , when passengers arrive satisfying a Poisson distribution, the waiting time of a passenger between bus service and bus service on line is .

[0013] (2) where, Set: - set of all bus lines, ​​​​where denotes the total number of lines; - the bus line , where denotes the total number of daily departures of the line ; Parameters: - the bus departure number, for example for the line ; - the passenger arrival rate for the line ; Auxiliary variables: - integer variable, headway between the departure of the line and the departure of the line .

[0014] Equation (3) represents the maximization of the number of convenient transfers per transfer station. For a transfer station , this includes the total number of convenient transfers from the line to the line , the total number of convenient transfers from the line to the line .

[0015] (3) where Set: - the set of all transfer stations; Index: - the transfer station of the line and the line , where (if the line is not transferable to , then is not set) ; Auxiliary variables: - binary variable, which takes the value if the departure of the line can be conveniently transferred to the line .

[0016] Equation (4) represents that each departure of each line is performed and performed once. These departures can be the first mission of a bus (by the variable (Note: This could also be a non-first-time task for a particular bus (determined by variables)). illustrate).

[0017] st : (4) in, Decision variables: - A binary variable, taking hour vehicle The first mission was for the line train schedule ; - A binary variable, taking hour vehicle Executed line train schedule Then it will be executed. line Flight schedule.

[0018] Equation (5) indicates that the vehicle After completing a task (using lines) Train schedule For example, at most one task can be executed at a time (there may be no next task to be executed).

[0019] (5) Equation (6) indicates that each vehicle Each has a first task (by a variable) illustrate).

[0020] (6) Equation (7) defines the auxiliary variable. If the line Train schedule Not by vehicle Execute, then Set to 0; otherwise, it indicates a circuit. Train schedule It is a vehicle The Secondary task.

[0021] (7) in, Auxiliary variables: -Integer variables, lines train schedule It is a vehicle The Secondary task.

[0022] Equation (8) defines the auxiliary variable. , hour Indicates vehicle The Which line does this task operate on; otherwise Also take 0, to represent a vehicle. The This task is not a route. .

[0023] (8) In equation (9), because the line train schedule It is a vehicle The This task, therefore the vehicle The Departure time for this mission , equal to line train schedule Departure time .

[0024] (9) in, Decision variables: - Integer variable, bus route train schedule Departure time in minutes; Auxiliary variables: - Integer variable, vehicle The The next mission will depart at [time]. Equation (10) indicates that the first train on each route departs at 0:00; (10) Equation (11) indicates that the last bus on each route departs at [time missing]. time; (11) in, parameter: -Bus operating hours; Equation (12) indicates that the vehicle Task Before starting, the following needs to be completed: departure from the depot, all work tasks, charging time for the trolleybus (if needed), empty vehicle travel time (through the depot) for the trolleybus (if not needed), empty vehicle travel time (not through the depot) for the oil bus.

[0025] (12) where parameters: - charging speed (% / minute); decision variables: - integer variable, charging amount after the nth task of the vehicle Here satisfies .

[0026] Equation (13) defines the headway between the shift of the line and the shift of the line .

[0027] (13) Equation (14) indicates that the headway has an upper limit and a lower limit .

[0028] (14) where parameters: - upper limit of the headway; - lower limit of the headway.

[0029] Because the electric amount of the trolleybus should be within the interval , the charging amount of the trolleybus also has an upper limit and a lower limit each time it is charged: (15) where auxiliary variables: - integer variable, lower limit of the charging amount after the nth task of the vehicle - integer variable, upper limit of the charging amount after the nth task of the vehicle ​​​​​​​​​​Maximum charge amount after the end of the task .

[0030] Equation (16) indicates that the vehicle The The amount of charge after the end of the task must be sufficient for the vehicle to complete the next task and be sufficient to return to the depot for recharging. The left side of equation (16) represents the lower limit of charging plus the total battery capacity (100%) minus the vehicle's charge level. The Total power consumption after this task Subtract the first The energy consumption for returning to the depot after completing the task. The left side of the equation equals the total battery capacity after charging, assuming minimal charging. This is sufficient to satisfy the right side of the equation, representing the total energy required to travel from the depot to the starting point, complete the task, and return to the depot.

[0031] (16) in, parameter: -from Station Electricity consumption of empty trains running at the station ; -from Station Electricity consumption for passenger transport at the station ; Auxiliary variables: - Integer variable, vehicle The The vehicle battery had consumed a significant amount of power after the mission ended. , Here satisfy .

[0032] The right side of equation (17) represents the vehicle. The When the mission ends and the battery returns to the depot for charging, this is the amount of charge the battery has consumed. The charging limit is equal to this value, which is exactly enough to fully charge the battery.

[0033] (17) Equation (18) defines the auxiliary variable. , indicating vehicle The Does the device charge after the task ends? When taking 0 Also take 0, when When not 0 Take 1.

[0034] (18) in, - A very large positive number.

[0035] Equation (19) defines the auxiliary variable. It indicates a vehicle. The The amount of electricity consumed by the vehicle's battery after the mission ends. This includes: the electricity consumed during the initial mission's departure from the depot, the electricity consumed during all passenger-carrying missions, the electricity consumed by empty vehicles dispatched without charging (not passing through the depot), the electricity consumed by empty vehicles dispatched while charging (passing through the depot), and minus all charging amounts.

[0036] (19) Equation (20) defines the auxiliary variable. He indicated the route train schedule Arrive at the transfer station The time is equal to the departure time plus the passenger travel time.

[0037] (20) in, -Integer variables, lines train schedule Arrive at the transfer station The time (minutes).

[0038] Equation (21) defines the auxiliary variable. He indicated the route train schedule Is it convenient to transfer to the line? When he takes 1, there exists at least one line. train schedule On the line train schedule Arrive after And the difference in arrival time is within the threshold. Within this time, the line train schedule Convenient transfer to the line Otherwise, it will take 0, indicating a line. train schedule Inconvenient transfer to the line .

[0039] ; (twenty one) wherein, Parameters: - upper limit of the convenient transfer threshold; - lower limit of the convenient transfer threshold.

[0040] Equation (22) defines an auxiliary variable which represents the total number of tasks of the vehicle . For a specific , is equal to the number of all and and.

[0041] (22) Equation (23) defines a set .

[0042] (23) Equation (24) defines an auxiliary variable which is equal to the number of 0 in .

[0043] (24) Equation (25) defines an auxiliary variable which is equal to the number of 1 in .

[0044] (25) Equations (26)-(28) define the value range of each auxiliary variable and decision variable.

[0045] (26) (27) (28) Third step, using algorithm to solve: using NSGAII-SA-PSO nested heuristic algorithm to solve the obtained multi-objective integer programming model, to obtain the Pareto frontier containing the following decision variables: (1) line timetable: the departure time of each class of the line ; (2) vehicle type and scheduling scheme: the number of hired oil vehicles and the number of electric vehicles , the oil-electric type corresponding to each vehicle , which class of which line each vehicle serves in turn; (3) vehicle charging scheme: the charging amount of the vehicle after the th task is completed , the operating cost of the bus company , the waiting time of passengers , and the number of convenient transfers .

[0046] For the Pareto front obtained by the algorithm, the weights of the three objective functions are set by linear weighting, and the most suitable solution is obtained from the Pareto front to obtain the final timetable, oil-electric type selection, vehicle scheduling and charging scheme.

[0047] In the third step, the multi-objective integer programming model obtained is solved by using the NSGAII-SA-PSO nested heuristic algorithm, and the objective functions are reasonably weighted to obtain the vehicle scheduling scheme: the number of hired oil and electric vehicles, and which line and which class each vehicle serves in turn. Specifically as follows: (1) The multi-objective integer programming model obtained is solved by using the NSGAII-SA-PSO nested heuristic algorithm, and the objective functions are reasonably weighted to obtain the value of the decision variable ; (2) If , the first task of the th vehicle is the th class of the th line; if , the first task of the th vehicle is not the th class of the th line; (3) If , the th vehicle will execute the th class of the th line after executing the th class of the th line; if , the th vehicle will not execute the th class of the th line after executing the th class of the th line.

[0048] Step 3, design the NSGAII-SA-PSO nested heuristic algorithm: (a) Use NSGA-II algorithm to solve the timetable: the chromosome structure, crossover and mutation operation of NSGA-II algorithm are adapted to the bus timetable; (b) Use simulated annealing algorithm (SA) to solve the vehicle oil-electric type: this decision variable is relatively simple, and SA such single factor search algorithm can be used to save the overall solving time; (c) Use particle swarm optimization (PSO) to solve the vehicle scheduling and charging scheme: the particle coordinate structure and offset of PSO , adapted.

[0049] In summary, the NSGAII-SA-PSO nested heuristic algorithm is used to solve the multi-objective integer programming model obtained, and the timetable, vehicle type, vehicle scheduling and charging scheme are obtained. The variables used in the algorithm are as follows:

[0050] (3.1) In NSGA-II, generate the initial timetable information of each individual in the initial population: for each individual, construct a zero matrix with a width of and a height of , and replace the tail column with . Then arrange the departure time for each line and ensure that the headway is within .

[0051] At the same time, a special initial solution will be introduced in the population: in order to make the passenger waiting time as short as possible, add a timetable individual with all equal headways to the initial population.

[0052] (3.2) Calculate the objective function values and of each individual in the initial timetable population.

[0053] (3.3) In SA, generate the initial vehicle type table. Generate two vehicle oil-electric type tables: a zero matrix with sufficient width and a height of , indicating that all oil vehicles are used; another column with sufficient width and a height of , whose elements are all , indicating that all electric vehicles are used.

[0054] (3.4) In PSO, generate the vehicle task table and charging plan table information of each particle in the initial particle swarm: for each particle, construct a matrix with a width of and a height of , and each element in the matrix is an ordered number pair with a shape of . For each class of each line, the algorithm can achieve the following goals: (a) Calculate the number of vehicles that can arrive on time, and use it as a threshold to randomly generate an offset ; (b) Calculate the charging amount that meets the upper and lower limit constraints, and use it as a threshold to randomly generate an offset ; (c) Record and to the matrix respectively.

[0055] (3.4.1) Task selection offset These are auxiliary variables used to help describe the vehicle's mission. In determining the route... Train schedule offset Before that, it's necessary to confirm whether I can make it to his departure time. Vehicles that arrive at the starting point before the start. If Then choose to be in The last vehicle to arrive. If If the number of vehicles exceeds the reachable number, the earliest arriving vehicle will be selected. If the number of vehicles does not exceed the reachable number, then select the last one. Late arrival vehicles. If no vehicle can arrive on time, then... Take 0 and dispatch new vehicles.

[0056] (3.4.2) Charge offset It is an auxiliary variable used to help describe the vehicle's charging amount. In determining the vehicle... No. Offset of the secondary task Before that, it is necessary to determine his current feasible charging limit. and lower limit .like Then the charging amount equals .like Then the charging amount equals Otherwise, the charging amount is equal to... .

[0057] In summary, by maintaining and Non-negative values ​​can prevent selecting vehicles that cannot arrive in time or whose charging capacity exceeds the feasible range, thus avoiding infeasible solutions. and The values ​​are entered into the matrix and used as particle coordinates for iteration.

[0058] At the same time, a special initial solution will be introduced into the particle swarm optimization. To minimize the time wasted by the vehicle and maximize charging when possible, a solution is added where all values ​​are... The offset matrix is ​​added to the initial particle swarm.

[0059] (3.5) Iteration of PSO: (29) (3.5.1) PSO's first Generation: Calculates the value of each particle in the current generation of the particle swarm. The value is then used to calculate the velocity of each particle in the initial swarm using the velocity formula (Equation (29)). The coordinates of the next generation particle swarm are then obtained.

[0060] (3.5.2) PSO's first Generation to the first Generation: Calculates the value of each particle in the current generation of the particle swarm. Values. Select the worst few particles and randomly generate their coordinates in the next generation. The coordinates of the remaining particles in the next generation are calculated using the velocity formula (Equation (29)).

[0061] Finish After each iteration, the coordinates of the entire particle swarm and the coordinates of each particle are output to the SA. value.

[0062] (3.6) SA iteration: Each iteration of SA requires obtaining the optimal vehicle task table, charging schedule, and corresponding data obtained after the iteration from PSO. value.

[0063] (3.6.1) SA's first Substitute: The two initial oil-electric type tables (all oil vehicles, all electric vehicles) generated in step (3.3) have their corresponding values ​​calculated in PSO. Value. Choose the better one to name. One is the target of the exchange. Randomly select one... A portion is exchanged with the target, and obtained .Will Input to PSO.

[0064] (30) (3.6.2) SA's first generation : Obtained from the previous generation , and The selection is calculated using equation (30). As probability .

[0065] (a) When, set The probability is ,set up The probability is .

[0066] (b) When, set .

[0067] For two The configuration methods generate new swap targets respectively: (a) Settings Time: If Best in history The historical optimal is taken as the exchange target. The historical optimal , and the number of oil vehicles in the , the oil vehicles are exchanged for electric vehicles, thereby generating an exchange target; if there is no such oil vehicle, a random exchange target is generated.

[0068] (b) Set : a random exchange target is generated.

[0069] After iterations, the NSGA-II outputs and the corresponding value.

[0070] (3.7) Iteration of NSGA-II: each iteration of NSGA-II needs to obtain the optimal vehicle task table and charging plan table obtained after iteration from PSO, and the optimal oil-electric type table and corresponding value obtained after iteration from SA.

[0071] For each individual in the parent population, based on the schedule, calculate their objective function values and , and obtain a value suitable for the schedule through SA and PSO. Perform crowding degree calculation and non-dominated degree sorting on each individual in the parent population. Select the best ones directly into the next generation, the worst ones for mutation, and the rest for crossover. Obtain the offspring and continue the cycle until iterations are completed.

[0072] (3.8) The weights of the three objective functions are set by using linear weighting, and the most suitable solution is obtained from the Pareto front to obtain the final schedule, oil-electric type selection, vehicle scheduling and charging scheme.

[0073] More specifically, the current bus route generally uses a single-line circulation mode, and the bus vehicle circulates to serve the same line. Due to the busy public transportation demand, a large number of vehicles are often invested to meet the operation requirements. In view of this problem, the present application proposes a cross-line scheduling operation mechanism, which allows vehicles to be flexibly allocated between different lines to balance the demand between different lines. The present application aims to guarantee the passenger demand of the public transport network, and schedule the line served by the vehicle to save the public transport operation cost.

[0074] (1) First, the line, starting and ending points, station and transfer station conditions are determined, the travel time and power consumption between nodes are calculated, and the passenger arrival rate at the starting point of each line is predicted based on historical passenger flow data.

[0075] (2) Then, build the mathematical model.

[0076] (3) Finally, use the NSGAII-SA-PSO nested heuristic algorithm to solve the model. Then select a reasonable weighting method to get the bus schedule, task allocation of each vehicle, and the electric vehicle charging scheme.

[0077] The specific case is as follows, for example, Figure 2 The basic parameters can be referred to the following table:

[0078] The following are the decision variables before optimization: (1) Line schedule:

[0079] (2) Vehicle task allocation table: Here each vehicle only serves a single line.

[0080] ; ;

[0081] (3) Vehicle type, charging plan table: Here each vehicle will be fully charged when charging.

[0082] ; ;

[0083] After calculation: The total operating cost of the bus company is 410,081 yuan, the total waiting time of passengers is 1,253,156 minutes, and the total number of convenient transfers is 17 times.

[0084] For this bus network, without changing the total passenger demand of the bus network, the schedule is redesigned, the oil-electric type of the vehicle is selected, the vehicle is allowed to cross the line, and the vehicle task allocation and charging scheme table is redesigned. The specific steps are as follows: First, information processing: According to the bus network information, determine the line, starting and ending point, station and transfer station situation, calculate the travel time and power consumption between nodes, and get the passenger arrival rate at each line starting point based on historical passenger flow data.

[0085] Second, build the mathematical model: Take minimizing the operating cost of the bus company and passenger waiting time, maximizing the number of convenient transfers as the goal, meet the constraints of vehicle scheduling and charging scheme feasibility, vehicle headway within the predetermined threshold, all bus trips are executed, etc. Establish a mathematical optimization model; Thirdly, the NSGAII-SA-PSO nested heuristic algorithm is used to solve the obtained multi-objective integer programming model, and the line timetable, vehicle scheduling scheme and vehicle charging scheme are obtained. The excellent solution after optimization is selected and shown as follows: (1) Line timetable:

[0086] (2) Vehicle task allocation table: each vehicle here can run across lines.

[0087]

[0088] (3) Vehicle type and charging plan table: each vehicle here will be charged according to the plan.

[0089]

[0090] After calculation, it is obtained that: The total operation cost of the bus company is 228,710 yuan, the total waiting time of passengers is 1,184,020 minutes, and the total number of convenient transfer is 22 times.

[0091] The above example shows the effectiveness and superiority of the present application, and it is obtained that the bus timetable synchronization and cross-line scheduling method for the hybrid electric vehicle fleet can reduce the operation cost of the bus company, save the waiting and transfer time of passengers, and effectively improve the efficiency and benefit of the public transportation system.​​

Claims

1. A method for bus schedule synchronization and cross-line scheduling for a hybrid electric fleet, characterized in that, Includes the following steps: The first step is information processing: Based on the public transportation network information, determine the routes, origins and destinations, depots and transfer stations, calculate the travel time and power consumption between each node, and obtain the passenger arrival rate at the origin of each route based on historical passenger flow data. The second step is to construct a mathematical model: taking the minimization of the bus company's operating costs and passenger waiting time and the maximization of convenient transfers as the objectives, and satisfying the constraints of vehicle scheduling and charging schemes being feasible, vehicle headway being within a predetermined threshold, and all bus services being executed, a mathematical optimization model is established. Third, using algorithm solution: using NSGAII-SA-PSO nested heuristic algorithm to solve the multi-objective integer programming model, get the following decision variables containing the Pareto frontier: (1) line schedule: The departure time of each class of line ; (2) vehicle type and scheduling scheme: the number of hired oil vehicles and the number of electric vehicles , the oil-electric type corresponding to each vehicle , which class of which line each vehicle serves in turn; (3) vehicle charging scheme: The charging amount of each vehicle after the end of the th task , the operation cost of the bus company , the passenger waiting time and the number of convenient transfer .

2. The method of bus schedule synchronization and cross-line dispatching for a hybrid fleet of buses according to claim 1, wherein: For the Pareto front obtained by the algorithm, the weights of the three objective functions are set by a linear weighting method. The most suitable solution is obtained from the Pareto front, resulting in the final timetable, selection of hybrid vehicle type, vehicle scheduling and charging scheme.

3. The method of bus schedule synchronization and cross-line dispatching for hybrid electric fleet of claim 1, wherein: The mathematical model is as follows: First, based on the value coefficients and usage of each cost, the operating cost of the bus company is minimized. Formula (1) is used to calculate the total fixed cost of gasoline buses, the total fixed cost of electric buses, the total cost of gasoline buses with passengers, the total cost of gasoline buses with empty buses including inter-line scheduling, the first bus leaving the depot, and the last bus entering the depot, the total cost of electric buses with passengers, and the total cost of electric buses with empty buses including inter-line scheduling, entering and leaving the depot while charging, the first bus leaving the depot, and the last bus entering the depot: （1） in, gather: -A collection of all buses ,in Indicates the total number of vehicles; -vehicle The set of all task sequence numbers executed. ,in Indicates vehicle Total number of daily tasks executed; parameter: - Fixed costs of gasoline buses; - Cost of running an empty gasoline bus; - Passenger operating costs for gasoline-powered buses; - The carbon emission cost of gasoline buses running empty; - The carbon emission cost of gasoline-powered buses carrying passengers; - Fixed costs of electric buses; - Cost of running an empty electric bus; - Operating costs for electric buses carrying passengers; -from Station Station empty train travel time ; -from Station Station passenger travel time ; index: - Bus route number. ; - Bus vehicle serial number ; - Vehicle mission number. (by vehicle) (For example) -line starting point; -line end; -Station; Decision variables: - A binary variable, taking hour vehicle It must be an electric vehicle; otherwise, it must be a gasoline vehicle. Auxiliary variables: - An integer variable representing the number of fuel vehicles; - Integer variable, number of trams, has ; - A binary variable, taking Vehicle The It will recharge after the mission is completed. satisfy ; - Integer variable, vehicle The This task operates on the line. ; Equation (2) represents minimizing the total waiting time for passengers on each route; for each route train schedule When passengers arrive at a location satisfying a Poisson distribution, the passengers are on the route. train schedule With train schedule The waiting time is ; （2） in, gather: -A collection of all bus routes, in Indicates the total number of lines; -Bus routes The set of all class numbers in the middle. in Indicates the line The total number of daily flights; parameter: -Bus route number. (by line) (For example) -line Passenger arrival rate; Auxiliary variables: -Integer variables, lines train schedule With train schedule The distance between the front ends of the cars; Equation (3) represents maximizing the number of convenient transfers at each transfer station; for transfer stations This includes from the line Convenient transfer to the line Total number of times, from the line Convenient transfer to the line Total number of times; （3） in, gather: -A collection of all transfer stations; index: -line With the line The transfer stations, among which If the line and If there is no transfer, then no system will be set up. ; Auxiliary variables: - A binary variable, taking Time Line train schedule It allows for convenient transfers to the line. ; Equation (4) indicates that each trip on each route is executed once; These trips are the first missions of a certain bus; determined by variables. This indicates that the task may be a non-first-time task for a particular bus, determined by variables. illustrate; s.t. : (4) in, Decision variables: - A binary variable, taking hour vehicle The first mission was for the line train schedule ; - A binary variable, taking hour vehicle Executed line train schedule Then it will be executed. line Schedule; Equation (5) indicates that the vehicle After completing a task (using lines) Train schedule For example, at most one task can be executed at a time. （5） Equation (6) indicates that each vehicle Each has an initial task, determined by variables. illustrate; （6） Equation (7) defines the auxiliary variable. If the line Train schedule Not by vehicle Execute, then Set to 0; otherwise, it indicates a circuit. Train schedule It is a vehicle The Sub-task; （7） in, Auxiliary variables: -Integer variables, lines train schedule yes vehicle The Sub-task; Equation (8) defines auxiliary variables , hour Indicates vehicle The Which line does this task operate on; otherwise Also take 0, to represent a vehicle. The This task is not a route. ; （8） In equation (9), because the line train schedule yes vehicle The This task, therefore the vehicle The Departure time for this mission , equal to line train schedule Departure time ; （9） in, Decision variables: - Integer variable, bus route train schedule Departure time; Auxiliary variables: - Integer variable, vehicle The The next mission will depart at the specified time; Equation (10) indicates that the first train on each route departs at 0:00; （10） Equation (11) indicates that the last bus on each route departs at [time missing]. time; （11） in, parameter: -Bus operating hours; Equation (12) indicates that the vehicle Task Before starting, the following needs to be completed: all work tasks from the depot, the charging time of the trolley, the empty running time of the trolley and the empty running time of the fuel truck; （12） in, parameter: -Charging speed (% / minute); Decision variables: - Integer variable, vehicle The Charge amount after the mission ends. , Here satisfy ; Equation (13) defines the line train schedule With train schedule The distance between the front of the car ; （13） Equation (14) indicates that the distance between the front of the vehicle and the front of the vehicle is... There is an upper limit and lower limit ; （14） in, parameter: - Maximum distance between the front of the vehicle; -Lower limit of vehicle frontage; Because the electric vehicle's battery level should be within the specified range. Therefore, the amount of electricity a trolley can charge each time it is charged also has upper and lower limits: （15） in, Auxiliary variables: - Integer variable, vehicle The The minimum charge level after the end of the task; - Integer variable, vehicle The The maximum amount of charge after the end of the task; Equation (16) indicates that the vehicle The The amount of charge after the end of the task must be at least enough for the vehicle to complete the next task, and enough to return to the depot for charging after completion; the left side of equation (16) represents the lower limit of charging plus the total battery capacity of 100%, minus the vehicle's... The Total power consumption after this task Subtract the first The power consumption of returning to the depot after the mission; the left side of the equation equals the total battery capacity after charging is completed with the minimum charging amount; and this is sufficient to satisfy the right side of the equation, that is, the total power demand for traveling from the depot to the starting point, completing the work task, and traveling from the end point to the depot. （16） in, parameter: -from Station Electricity consumption of empty trains running at the station ; -from Station Electricity consumption for passenger transport at the station ; Auxiliary variables: - Integer variable, vehicle The The vehicle battery had consumed a significant amount of power after the mission ended. , Here satisfy ; Equation (17) indicates that the vehicle The After the task is completed, the battery charge should not exceed 100%; the right side of equation (17) represents the vehicle The When the mission ends and the battery returns to the station to recharge, the amount of power that the battery has consumed is the maximum amount that can be charged. （17） Equation (18) defines the auxiliary variable. , indicating vehicle The Whether to charge after the task ends; when When taking 0 Also take 0, when When not 0 Take 1; （18） in, -A very large positive number; Equation (19) defines the auxiliary variable. , indicating vehicle The The amount of electricity consumed by the vehicle battery after the mission is completed; including: the electricity consumed when the vehicle leaves the depot for the first mission, the electricity consumed when the vehicle is carrying passengers for all work missions, the electricity consumed when the vehicle is dispatched without charging and does not pass through the depot, the electricity consumed when the vehicle is dispatched with charging and passes through the depot, minus all the charging amount. （19） Equation (20) defines the auxiliary variable. ; indicates a line train schedule Arrive at the transfer station The time is equal to the departure time plus the passenger travel time; （20） in, -Integer variables, lines train schedule Arrive at the transfer station The moment; Equation (21) defines the auxiliary variable. ; indicates a line train schedule Is it convenient to transfer to the line? When the value is 1, there exists at least one line. train schedule On the line train schedule Arrive after And the difference in arrival time is within the threshold. Within this time, the line train schedule Convenient transfer to the line Otherwise, it will take 0, indicating a line. train schedule Inconvenient transfer to the line ; ； （21） in, parameter: - Maximum threshold for convenient transfers; -Lower limit of convenient transfer threshold; Equation (22) defines the auxiliary variable. He said the vehicle Total number of tasks; for a specific , Equal to all and and; （22） Equation (23) defines the set ; （23） Equation (24) defines the auxiliary variable. ,equal The number of 0s; （24） Equation (25) defines the auxiliary variable. ,equal The number of 1s in the middle; （25） Equations (26)-(28) define the range of values ​​for each auxiliary variable and decision variable; （26） （27） （28）。 4. The method for synchronizing bus timetables and cross-line scheduling for hybrid electric bus fleets according to claim 1, characterized in that: In the third step, the NSGAII-SA-PSO nested heuristic algorithm is used to solve the obtained multi-objective integer programming model, and the objective function is appropriately weighted to obtain the vehicle scheduling scheme: the number of hired gasoline and electric vehicles, and which route and which shift each vehicle will serve in sequence; as detailed below: (1) The multi-objective integer programming model is solved by using the NSGAII-SA-PSO nested heuristic algorithm, and the objective function is reasonably weighted to obtain the decision variables. The value; (2) If ,but The car's first mission was The line Schedule; if ,but The car's first mission was not The line Schedule; (3) If ,but The car has performed line The shift will continue after the first shift. line Schedule; if ,but The car has performed line This will not be implemented after the shift. line Flight schedule.

5. The method for synchronizing bus timetables and cross-line scheduling for hybrid electric bus fleets according to claim 4, characterized in that: The NSGAII-SA-PSO nested heuristic algorithm includes the following steps: (3.1) In NSGA-II, generate the initial timetable information for each individual in the initial population: construct a wide array for each individual. Gao Wei The zero matrix, then using Replace its last train; then schedule departure times for each line, ensuring the headway between trains is within [specified range]. Inside; At the same time, a special initial solution will be introduced into the population: in order to minimize the passenger waiting time, add an individual with all train head-to-head distances to the initial population. (3.2) Calculate the objective function value for each individual in the initial timetable population. and ; (3.3) In SA, generate the initial vehicle type table; generate two vehicle hybrid type tables: one column with sufficient width and height. A zero matrix represents the use of all fuel vehicles; the other column has sufficient width and height. A matrix whose elements are all 0 This indicates that all vehicles will be trolleys; (3.4) In PSO, generate vehicle task table and charging schedule information for each particle in the initial particle swarm: construct a wide array for each particle. Gao Wei A matrix, where each element is of shape ... The algorithm obtains ordered pairs of numbers; for each train on each route, the algorithm achieves the following objectives: (a) Calculate the number of vehicles that can arrive in time, and use this as a threshold to randomly generate offsets. ; (b) Calculate the charging amount that satisfies the upper and lower charging limit constraints, and use this as a threshold to randomly generate an offset. ; (c) will and Record them separately in the matrix; (3.5) Iteration of PSO: （29） (3.6) SA iteration: Each iteration of SA requires obtaining the optimal vehicle task table, charging schedule, and corresponding data obtained after the iteration from PSO. value; (3.7) NSGA-II Iteration: Each iteration of NSGA-II requires obtaining the optimal vehicle task table and charging plan table obtained after the iteration from PSO, and the optimal hybrid type table and corresponding hybrid type table obtained after the iteration from SA. value; (3.8) The weights of the three objective functions are set by linear weighting, and the most suitable solution is obtained from the Pareto front, resulting in the final timetable, selection of oil and electric vehicle type, vehicle scheduling and charging scheme.

6. The method for synchronizing bus timetables and cross-line scheduling for hybrid electric bus fleets according to claim 5, characterized in that: Step 3.4 includes the following steps: (3.4.1) Task selection offset These are auxiliary variables used to help describe the vehicle's tasks; in determining the route... Train schedule offset Before that, it's necessary to confirm whether I can make it to his departure time. Vehicles that arrive at the starting point before the destination; if Then choose to be in The latest vehicle to arrive; if If the number of vehicles exceeds the reachable number, the earliest arriving vehicle will be selected; if... If the number of vehicles does not exceed the reachable number, then select the last one. Late arrival vehicles; if no vehicles can arrive on time, then... Take 0 and dispatch new vehicles; (3.4.2) Charge offset It is an auxiliary variable used to help describe the vehicle's charging amount; in determining the vehicle No. Offset of the secondary task Before that, it is necessary to determine his current feasible charging limit. and lower limit ;like Then the charging amount equals ;like Then the charging amount equals Otherwise, the charging amount equals .

7. The method for synchronizing bus timetables and cross-line scheduling for hybrid electric bus fleets according to claim 5, characterized in that: Step 3.5 includes the following steps: (3.5.1) PSO's first Generation: Calculates the value of each particle in the current generation of the particle swarm. The value is then used to calculate the velocity of each particle in the initial population using the velocity formula (Equation (29)); the coordinates of the next generation particle swarm are obtained. (3.5.2) PSO's first Generation to the first Generation: Calculates the value of each particle in the current generation of the particle swarm. Value; select the worst few particles and randomly generate their coordinates in the next generation; calculate the coordinates of the next generation for the remaining particles using the velocity formula (Equation (29)); Finish After the iteration, output the coordinates of the entire particle swarm and the coordinates of each particle to SA. value.

8. The method for synchronizing bus timetables and cross-line scheduling for hybrid electric bus fleets according to claim 5, characterized in that: Step 3.6 includes the following steps: (3.6.1) SA's first Substitute: The two initial oil-electric type tables generated in step (3.3), including all oil vehicles and all electric vehicles, have their corresponding values ​​calculated in PSO. Value; choose the better one to name. The other is used as the target of the exchange; randomly select A portion is exchanged with the target, and obtained ,Will Input to PSO; （30） (3.6.2) SA's first generation : Obtained from the previous generation , and ; Calculate the selection using equation (30) As probability ; (a) When, set The probability is ,set up The probability is ; (b) When, set .

9. The method for synchronizing bus timetables and cross-line scheduling for hybrid electric bus fleets according to claim 8, characterized in that: For two The configuration methods generate new swap targets respectively: (a) Settings Time: If historical best The best in history As the target of the exchange; if historical best ,and Number of tasks involving oil vehicles If these oil tankers are replaced with electric cars, then an exchange target is generated; if no such oil tankers are available, a random exchange target is generated. (b) Settings Time: Generates random swap targets; Finish After the next iteration, output to NSGA-II With the corresponding value; For each individual in the parent population, their objective function value is calculated based on the timetable. and Then, an appropriate timetable is obtained through SA and PSO. For each individual in the parent generation, calculate crowding degree and rank them by non-dominance degree; select the best batch to directly enter the next generation, the worst batch for mutation, and the rest for crossover; obtain offspring and continue the cycle until complete. The next iteration.

Images