A method for designing and scheduling vehicles for a hierarchical line of customized urban and rural passenger transport
By designing tiered routes between urban and rural areas and optimizing vehicle scheduling, the problems of operating costs and punctuality in urban and rural passenger transport have been solved, achieving efficient customized passenger transport services and meeting personalized travel needs.
Patent Information
- Application Number
- CN202611269621.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-20
- Publication Date
- 2026-11-06
AI Technical Summary
Traditional integrated urban and rural passenger transport service networks cannot meet personalized travel needs, especially in balancing operating costs and punctuality between towns and cities. Existing customized passenger transport route designs suffer from low occupancy rates and increased operating distances when traveling across regions.
A hierarchical collaborative design is adopted, combining urban-rural cross-regional main lines with town and village connecting branch lines. A timetable optimization model and a vehicle scheduling collaborative optimization model are constructed. A non-dominated sorting genetic algorithm is used to generate customized passenger transport routes and timetable operation plans. Supply and demand matching rate, travel time and operating costs are optimized through direct main line and branch line connecting modes.
It has achieved broader demand coverage, on-time arrival, and reduced operating costs, ensuring the service quality and economy of customized passenger transport, and providing decision support for actual operation plans.
Smart Images
Figure CN123050781A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of route design and vehicle scheduling technology, and in particular to a method for designing and scheduling tiered routes for customized urban and rural passenger transport. Background Technology
[0002] The demand for diversified travel between urban and rural areas continues to increase, while the "fixed-point, fixed-route" operation model of the traditional integrated urban and rural passenger transport service network can no longer meet the personalized travel needs with higher requirements for on-time delivery, such as commuting, going to school, or traveling to transportation hubs. As an important supplement to the conventional passenger transport service network, customized passenger transport is an important way to meet the diversified travel needs of rural residents.
[0003] Against the backdrop of the transformation of urban and rural passenger transport routes into public transport, the route design method for customized passenger transport for cross-regional travel mainly follows the urban customized bus or demand-responsive bus service network design method, that is, adopting the route form of "multi-point to single point" or "multi-point to multi-point" direct access. Based on travel demand, a composite design process has been formed, including station site selection, route design, timetable compilation and driving plan.
[0004] Depending on the research scenario, for medium- to long-distance travel in high-density, peak-hour passenger corridors, the design concepts of customized commuter buses and airport express lines can be referenced. For example, existing technical solutions propose a customized bus network planning model based on a multi-point to single-point service model, after clustering the origin and destination areas. Existing technical solutions also propose a customized bus route design method considering passenger walking accessibility for long-distance travel. Existing technical solutions also consider passenger preference factors and construct a two-stage customized bus network design model. For areas with less travel demand or more dispersed travel needs, demand-responsive or flexible bus design concepts can be referenced. For example, existing technical solutions construct a demand-responsive connecting bus route and timetable optimization model considering passenger travel time windows. Existing technical solutions consider the dispersed travel demand of residents in rural areas and construct a demand-responsive bus route design model. Existing technical solutions consider the spatiotemporal dispersion of travel demand and construct a customized passenger transport route and vehicle joint optimization model based on dynamic boarding and alighting points. Existing technical solutions take the connection between areas with low bus coverage and transportation hubs as the research scenario and construct a combined scheduling model.
[0005] If the existing route design methods based on the "one-stop direct" model are directly applied to urban and rural travel scenarios, it will be difficult to balance operating costs and travel efficiency under specific demand distribution characteristics. Taking the morning peak hours as an example, the direction of cross-regional passenger flow between urban and rural areas is usually unidirectional, with the main passenger flow directions being: administrative village to town, and town to urban area. However, due to the spatial structure of the town area, the passenger flow concentration in the central town is high, while the travel demand of the surrounding administrative villages is more dispersed in space. Therefore, if a multi-terminal direct route operation model from administrative village to urban area is adopted, the low occupancy rate will not be able to effectively control operating costs for long-distance cross-regional travel. At the same time, compared with urban built-up areas, the road conditions in towns and administrative villages are more complex. If a long-distance route model of administrative village to town to urban area is directly deployed, the increased detour distance of operating vehicles will also affect the punctuality, a key service indicator of customized passenger transport.
[0006] Therefore, how to provide a design method for customized urban and rural passenger transport routes that balances service quality and operating costs is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention
[0007] To address the problems existing in the prior art, the purpose of this invention is to provide a method for designing and scheduling customized passenger transport routes in urban and rural areas using a hub-and-spoke hierarchical approach. This method breaks down the traditional "one-stop direct" route format of customized routes and adopts a hierarchical collaborative design approach that combines urban and rural cross-regional main lines with town and village connecting branch lines to achieve customized services. Furthermore, it constructs a timetable optimization model for the main line mode, a route design model for the branch line connecting mode, a timetable and vehicle scheduling collaborative optimization model, and designs a model solving algorithm based on intelligent optimization to output customized passenger transport routes and timetable operation plans.
[0008] To achieve the above objectives, the present invention provides the following solution: A method for designing and scheduling customized urban and rural passenger transport routes, comprising: Obtain booking demand data for customized passenger transport; We set up a mainline cross-regional direct mode and a branch line connection mode, and constructed a mainline timetable optimization model with supply and demand matching rate and travel time as objective functions. At the same time, we constructed a collaborative optimization model for branch line route design, timetable and vehicle scheduling oriented towards the branch line connection mode with supply and demand matching rate, travel time cost and operating cost as objective functions. Based on the reservation travel demand data, a non-dominated sorting genetic algorithm that integrates genetic operators and elite strategies is used to solve the main line timetable optimization model and the collaborative optimization model, and output a customized passenger transport route operation plan; during the solution of the branch line model, vehicle scheduling operations are performed to optimize the fleet size and reduce operating costs.
[0009] Optionally, constructing the main timetable optimization model includes: The reservation travel demand data is divided into two categories: departure from townships and departure from administrative villages. The boarding and alighting points of passengers departing from townships are located in townships and urban areas, respectively. The travel trajectories of passengers departing from administrative villages are set sequentially as administrative villages, townships, and urban areas. If the reservation travel demand data originates from townships, then from a timetable compilation perspective, the departure times of direct route vehicles from intermediate stops and their arrival times at urban destinations must meet the time window of the reservation demand. If the reservation travel demand data originates from administrative villages, shuttle vehicles must transport passengers to intermediate stops on the direct route. Simultaneously, the departure times of mainline direct route vehicles from these intermediate stops must not be later than the target time window, and shuttle vehicles serving the demand originating from these administrative villages must arrive at these intermediate stops before the mainline vehicles depart. Therefore, the mainline timetable optimization model is constructed with supply-demand matching rate and travel time as the objective functions. ; ; ; in, To maximize the supply and demand matching rate, To minimize travel time costs, For station Corresponding vehicle set, A collection of needs originating from townships. This is a collection of connection needs originating from administrative villages. In the main storyline, the passengers being served The travel time cost indicator Indicates if travel demand Can be from the station Vehicles dispatched If the service is active, its value is 1; otherwise, its value is 0. The demand originating from townships, This represents the number of connections required originating from administrative villages. To improve the supply and demand matching rate, For travel time costs.
[0010] Optionally, the method further includes: constraints on the main timetable optimization model: Vehicle task allocation constraints: ; Time window constraint: ; ; Timetable collaborative compilation constraints: ; Vehicle capacity constraints: ; Runtime constraints: ; ; ; Variable constraints: ; in, Main line vehicles Arrival at the intermediate station Time, Main line vehicles k Arrival time at the destination in the city center For demand Expected arrival time at the destination Main line vehicles At the intermediate station Service hours, In the main line, two stations and The distance between, The speed of vehicles traveling on the main line. For vehicle occupancy rate, The capacity of vehicles operating on the main line. For demand The corresponding number of reservations The earliest departure time within the main line's operating hours. The latest departure time within the main line's operating hours. The main train schedule is as follows. Main line vehicles Arrival at the intermediate station Time, Main line vehicles Leaving the intermediate station Time, In the main line, two stations and Runtime between For station Each train dispatch Departure time, , For each passenger's expected departure time, a coefficient is introduced. Define requirements The travel time window is adjusted to increase the flexibility of the operational plan, taking into account each passenger's desired departure time. Define coefficients Thus, it can meet the demand The corresponding departure station defines its travel time window. ,coefficient for The corresponding time window parameters.
[0011] Optionally, constructing the collaborative optimization model for the branch line connection mode includes: ; ; ; ; ; in, To improve the supply and demand matching rate, For operating distance costs, For travel time cost, The number of operating vehicles required to complete all train trips. To maximize the supply and demand matching rate, To minimize operating distance costs, To minimize travel time costs, For the station assembly, For the connecting area, the station The corresponding set of vehicles Indicates if travel demand Can be from the station Vehicles dispatched If the service is active, its value is 1; otherwise, its value is 0. For site collection, For feeder lines, the operating cost per unit distance. As a branch line, there are two stations. and The distance between, Indicates if the station Mainline vehicles dispatched From demand points Drive to the point of demand If it is 1, its value is 1; otherwise, its value is 0. As a branch line, the demand The travel time cost indicator This represents the number of connections required originating from administrative villages. For station d A certain vehicle was dispatched. For stationd The corresponding set of vehicles The cost of dispatching each vehicle. For station A certain train that was dispatched, For station Collection of departing trains, Indicates if train number Vehicles If the value is 1, it is used; otherwise, it is used for 0.
[0012] Optionally, the method further includes: constraints on the cooperative optimization model for branch connection modes: Vehicle task allocation constraints: ; ; Time window constraint: ; Coordination constraints of timetables for the two types of route modes: ; Vehicle capacity and occupancy constraints: ; Runtime constraints for consecutive nodes: ; ; Variable constraints: ; ; ; in, This is a collection of connection needs originating from administrative villages. For connecting line vehicles Arrival at the intermediate station Time, For connecting line vehicles Arrive at the intermediate station of the corresponding direct route Time, For the connecting area, the station The corresponding set of vehicles For the capacity of branch line operating vehicles, For branch line vehicles At the intermediate station Service hours, Two demand points in the branch line and Runtime between For connecting line vehicles Arrival at the intermediate station Time, For a sufficiently large number, For station The connection between the dispatched vehicle and the first stop along the way. For station The dispatched vehicles return to the depot after completing their service. For passengers' expected departure time, for The corresponding time window parameters, Main line vehicles Leaving the intermediate station Time, This refers to the occupancy rate.
[0013] Optionally, using the non-dominated sorting genetic algorithm to output the main line cross-regional direct mode and branch line connecting mode customized passenger transport route operation plan includes: Step 1: Generate the initial population according to the integer encoding rules; Step 2: Select the initial population using a tournament strategy, perform crossover and mutation operations on the initial population using an adaptive genetic operator to generate offspring populations, and calculate the objective function values corresponding to the offspring populations. Step 3: Merge the parent population of the previous generation with the offspring population to obtain a new population; Step 4: Use the fast nondominated sorting method to output the nondominated frontiers at multiple different levels corresponding to the new population; Step 5: Iteratively assess the non-dominant frontier using an elite strategy to generate a new parent population: If ,but until The crowding distance of each individual in the non-dominated front is calculated and sorted from high to low. The individual with the highest sorting value is selected. The individual corresponding to the given position is incorporated into the population, resulting in a new parent population; among them... For population, As a non-dominant frontier, Population size; Step 6: Repeat steps 2-5 until the termination condition of the iteration state is met. Merge the population and offspring population in the last iteration and output the Pareto optimal solution corresponding to the merged population using the fast non-dominated sorting method. This can be regarded as a customized passenger transport route operation plan, and solutions that do not meet the occupancy rate constraint are eliminated. In the generated branch line connection mode route operation plan, vehicle scheduling operations are further performed to optimize the fleet size and reduce operating costs.
[0014] Optionally, the initial population of the non-dominated sorting genetic algorithm includes: The initial population of the main timetable optimization model: ; The initial population of the branch cooperative optimization model: ; in, The departure timetable or departure time sequence corresponding to each individual. For the first time during the operating period The departure time of a vehicle or train. For vehicles In executing the line The order of station accesses when corresponding to a train number task. For the line The site number, For the line The number of stations visited.
[0015] Optionally, the genetic operators for the adaptive mechanism in the non-dominated sorting genetic algorithm include: ; ; ; ; in, For adaptive crossover probability, The number of objective functions, , , These are the th in the current population. The maximum, minimum, and average fitness values of each objective function. The larger of the fitness values of the two individuals to be crossed is given. The fitness value of the individual to be mutated. For adaptive mutation probability, These are parameters used in the crossover probability calculation process. These are parameters used in the calculation of the probability of mutation. , , For other parameters, satisfy ; , , For other parameters, satisfy .
[0016] Optionally, the vehicle dispatching operation includes: Step 1: Set the maximum number of iterations to 1. Arrange the trains in the set of trains departing from the station in ascending order of departure time, and set up a temporary set; Step 2: Starting from the first train in the train set, iterate through the elements in the train set in sequence to determine whether there is a target train set among the elements in the train set excluding the train in the current iteration round, so that the trains in the target train set can be connected according to time and space conditions. Step 3: If the target train set exists, store the generated train chain in the temporary set, delete the train elements in the target train set from the train set, and determine whether the elements in the train set have been traversed. If the traversal is complete, merge the train set and the temporary set to generate a new train chain set. If the traversal is not complete, re-determine whether the target train set exists in the current train set excluding the trains in the next iteration round, until the traversal is complete. Step 4: If the target train set does not exist, the train set remains unchanged, and it is determined whether the elements in the train set have been traversed. If they have, the train set and the temporary set are merged to generate a new train chain set. If the traversal is not completed, it is determined again whether the target train set exists in the current train set excluding the trains in the next iteration round, until the traversal is completed.
[0017] The beneficial effects of this invention are as follows: (1) The customized passenger transport “hub-and-spoke” layered route design mode adopted in this invention fully considers the uneven distribution of travel demand in the spatial area of the township area, changes the existing “one-stop direct” design concept of customized passenger transport or customized bus between multiple origins and destinations, and forms a two-level service network mode of “direct main line and branch line connection”, so that the demand coverage is wider.
[0018] (2) The model constructed in this invention considers both the spatial distribution of travel demand and the travel time characteristics of users. By optimizing the operation plan of different route types and adopting a strategy of coordinating the compilation of timetables for direct and connecting routes, it aims to meet the travel time window and comfort (one person, one seat) of each user, ensure that passengers served by the two types of route modes can arrive at their destination on time, and consider vehicle scheduling issues to reduce operating costs.
[0019] (3) The hub-and-spoke hierarchical route design adopted in this invention takes into account indicators such as supply and demand matching rate, travel time and operating cost, avoiding the problem of excessively high operating distance costs caused by directly deploying long-distance direct routes from "administrative villages to townships to urban areas", while ensuring punctuality, a key service characteristic of customized passenger transport. In addition, this invention designs a solution algorithm based on intelligent optimization according to the model characteristics to generate a coordinated operation plan for customized passenger transport route design, timetable and vehicle scheduling. The results can provide decision support for the formulation of actual operation plans for customized passenger transport. Attached Figure Description
[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0021] Figure 1 This is a schematic diagram of a method for designing and scheduling customized urban and rural passenger transport routes according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a scenario according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the algorithm encoding in an embodiment of the present invention; Figure 4 This is a schematic diagram of the algorithm crossover operator in an embodiment of the present invention; Figure 5 This is a schematic diagram of the algorithm mutation operator in an embodiment of the present invention; Figure 6 This is a schematic diagram of an embodiment of the present invention; Figure 7 This is a schematic diagram of the main line of an embodiment of the present invention; Figure 8 This is a flowchart illustrating a method for designing and scheduling customized urban and rural passenger transport routes according to an embodiment of the present invention. Detailed Implementation
[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0023] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0024] like Figure 8 As shown in the figure, this embodiment discloses a method for designing and scheduling tiered routes for customized urban and rural passenger transport, including the following steps: The first step is to obtain booking demand data for customized passenger transport. This data can be obtained through travel surveys or by customized passenger transport operators through mobile applications. The travel information for each booking request is extracted, specifically including the passenger's departure point, destination, and travel time window information such as expected departure and arrival times.
[0025] The second step involves developing timetable plans for the main cross-regional routes between urban and rural areas (passenger flow direction: towns to urban areas), constructing a timetable optimization model with supply-demand matching rate and travel time as the objective functions. The main routes connect major towns and central urban hubs within the operating area and adopt a multi-point-to-single-point direct route mode, enabling this mode to cover customized needs in areas with large passenger flow.
[0026] The third step involves designing routes and timetables for town and village shuttle feeder lines (passenger flow direction: administrative village to township). Under this route model, for travel demands not covered by the main line (such as administrative village to township to urban area direction), based on the arrival and departure information of main line vehicles at intermediate stops, a shuttle route model is used to transport these demands to the main line's intermediate stops. While meeting the travel time windows of each booked user, a collaborative timetable compilation strategy for the main line and feeder lines is adopted. Using supply-demand matching rate, travel time cost, and operating cost as objective functions, a collaborative optimization model for route design, timetables, and vehicle scheduling is constructed. By designing timetable connection constraints for the two route models, collaborative timetable compilation for the main line and feeder lines is achieved, while simultaneously considering indicators such as supply-demand matching rate, travel time cost, and operating cost.
[0027] The fourth step involves using an intelligent optimization-based method to solve the model. This algorithm is based on the Non-dominated Sorting Genetic Algorithms (NSGA-II) framework, which integrates the adaptive mechanism of genetic operators and elite strategies to solve the two types of models. It also introduces the vehicle chain theory to solve the vehicle scheduling problem.
[0028] Finally, by solving two types of mathematical optimization models, a customized passenger transport route operation plan is output, including the stations along each route, timetable information such as the arrival and departure times of operating vehicles at each station, and the allocation relationship between vehicles and trains.
[0029] Specifically, the logic flowchart of the urban-rural customized passenger transport tiered route design and vehicle scheduling method provided in this embodiment is as follows: Figure 1 As shown, the specific steps include: The first step is to obtain booking demand data for customized passenger transport. This data can be obtained through travel surveys or by customized passenger transport operators through mobile applications. The travel information for each booking request is extracted, specifically including the passenger's departure point, destination, and travel time window information such as expected departure time and expected arrival time.
[0030] The second step involves designing customized routes for cross-regional travel scenarios from rural areas (including administrative villages and townships) to urban areas, such as... Figure 2 As shown. Taking the morning rush hour as an example, the origin and destination points of a set of travel demands are located in townships (departure areas) and urban areas (destination areas), respectively. Within the departure areas, passenger travel demand in townships is more concentrated during peak hours, while the spatial distribution of travel demand in administrative villages is more dispersed. Considering the travel purposes of residents from townships and administrative villages traveling to their workplaces in urban areas, transportation hubs (bus stations, train stations), etc., based on a hub-and-spoke hierarchical design approach, two route design methods are considered: ①The first type is the main line cross-regional direct mode, where the boarding and alighting stations are located in towns and urban areas respectively. Considering the long travel distances between urban and rural areas, no intermediate stations are set up between the two areas to achieve rapid direct travel. This route mode connects major towns with central urban areas where demand is concentrated and travel volume is high.
[0031] ②The second type is the feeder connection mode, which uses intermediate stops on direct routes within the origin region as the destination of connecting routes. This connects users to needs that cannot be directly covered by the direct route mode (such as a user's travel path: village-township-urban area) to these intermediate stops for transfer to the main line. The hub-and-spoke design approach, through timetable coordination between different levels of routes, avoids the excessively high operating costs associated with directly connecting villages and urban areas.
[0032] The third step involves using customized passenger booking demands as input information for the route design and timetable optimization model. Each person's travel demand... Including: place of departure ,destination Expected departure time Expected arrival time Information such as these. Simultaneously, by defining requirements. This allows for more flexible operational plans by defining travel time windows. For example, based on demand... Expected departure time Introducing parameters ,exist The corresponding departure station defines its departure time window. .
[0033] The fourth step is to construct a timetable optimization model for the first type of route (main line cross-regional direct mode) described in the second step.
[0034] During the modeling process, based on time window constraints, feeder line vehicles should arrive at the corresponding intermediate stop (transfer point) before mainline vehicles leave the township area, ensuring that feeder line passengers have sufficient time to transfer at that intermediate stop. When compiling the mainline timetable, in addition to meeting the time window constraints corresponding to travel demand originating from townships, it is necessary to also consider the timetable information of feeder line vehicles arriving at the mainline intermediate stops, ensuring that passengers originating from administrative villages can transfer to the mainline and arrive at their urban destination on time. The specific modeling steps are as follows: (1) Demands originating from administrative villages The time window for arriving at the destination in the city (e.g., the latest arrival time) ) Reverse the calculation to the intermediate stops of the corresponding direct line when the passenger transfers to the main line. The time window can be determined based on the destination within the city. With the site Distance between Vehicle speed Calculations are performed. From a timetable creation perspective, the requirements... If customized passenger transport services are available (travel route: administrative village—township—urban area), the following two conditions must be met: ①Direct route operating vehicles Leaving this intermediate station No later than (Unit: minutes) to ensure vehicle In demand Arrive at the destination in the city before the latest expected arrival time; ② Serving the needs shuttle vehicles Mainline vehicles are required Leave the station The train arrived at this intermediate stop earlier to ensure passengers could transfer to their desired vehicle. .
[0035] (2) Assume the main line is The starting point of the operating vehicles is recorded as (Main line departure depot); Assuming this main line The number of intermediate stops is Then the line It can be represented as a set of station sequences, denoted as The final stop of this route in the urban destination area is denoted as Therefore, the model's output is the station. Each train dispatch Departure time Mainline vehicles At each intermediate stop Arrival time and departure time Among them, the vehicle collection Each car corresponds to a set of train numbers. One of the train-related tasks.
[0036] (3) The definitions of symbols, parameters and variables in the model are shown in Table 1.
[0037] Table 1. Model Symbols and Parameter Definitions (4) Establish a timetable optimization model for the mainline direct route mode. This model considers the following two objective functions: (1); (2); Function (1) aims to maximize the supply-demand matching rate. The calculation of function (1) consists of two parts: demand originating from townships and demand originating from administrative villages. During the model construction process, a certain connection demand is determined. ( Can the main line vehicles be selected based on one of the conditions for customized passenger transport services? arrive Does the transfer time at intermediate stations meet the calculated arrival time window? Meanwhile, demand... Whether the service can be accepted ultimately depends on the timetable information of the branch line. Therefore, the value of function (1) is the maximum value under the main line timetable optimization level, and its actual value needs to be analyzed based on the results of the collaborative optimization model of line design, timetable and vehicle scheduling for branch line connection mode described in step 5.
[0038] Function (2) aims to minimize travel time costs, which are calculated as follows: the number of passengers served by the main line (passengers departing from towns and administrative villages), The expected time deviation to reach the destination in the urban area. Mainline vehicles Arrival time at the destination in the city (need Actual arrival time and demand Expected arrival time Perform the calculation: (3); The timetable optimization problem for the mainline direct route is transformed into a multi-objective optimization problem, and the mainline timetable optimization model can then be expressed as: (4); The main timetable optimization model includes the following constraints: (5); (6); (7); (8); (9); (10); (11); (12); (13); Constraint (5) assigns constraints to vehicle tasks, with each requirement... Service can be provided by a maximum of one commercial vehicle.
[0039] Constraints (6) and (7) are time window constraints. One of the conditions for users to choose customized passenger transport is that the timetable meets their travel time window. For demand starting from a town, the vehicle arrives at a midway station on the main line. and urban destinations This condition must be met at that time.
[0040] Constraint (8) indicates the arrival and departure times of vehicles at intermediate stations during the mainline timetable compilation process. The timing also needs to take into account the needs of those departing from administrative villages, so as to achieve the goal of coordinating the compilation of the main and branch timetables.
[0041] Constraint (9) is a vehicle capacity constraint, meaning the number of passengers on board should be less than the vehicle capacity to meet the "one person, one seat" operational characteristic. Simultaneously, considering operating costs and revenues, the vehicle occupancy rate... It should also be within a certain range.
[0042] Constraints (10)-(12) are operating time constraints. Constraint (10) indicates that the departure time of each train on the main line should be within the preset operating time; Constraint (11) is the time constraint for the operation of the trains on the main line. At two consecutive stations and The travel time relationship between them, i.e., vehicles arrive The moment it is before leaving the preceding station Time and and The running time between them is related. Constraint (12) indicates that the vehicle On the site The departure time is calculated from the arrival time and the service time (or stop time).
[0043] Constraint (13) is a variable constraint.
[0044] The fifth step is to construct a collaborative optimization model for the design, timetable, and vehicle scheduling of village-township branch line connection types.
[0045] (1) This model describes the travel process from administrative villages to townships. Specifically, within an administrative village, shuttle vehicles arrive at the passenger's pre-booked demand point within the passenger's travel time window and transport the passenger to a stop on the main route within the township area. The route information output by the model can be described by the sequence of demand points (or stops). To ensure passengers arrive on time in the shuttle area, the demand... ( The actual departure point is the intermediate stop on the connecting route. The timetable information for each route includes: the number of trains operated. Departure time ,vehicle At each intermediate stop on the connecting route Arrival time Departure and departure times Arrive at the corresponding intermediate station on the main line. ( (time) Among them, without considering vehicle scheduling factors, the vehicle set Each car corresponds to a set of train numbers. One of the train-related tasks.
[0046] (2) Let For the connecting network in the customized passenger transport service network, among which Represents a set of nodes (sites), including a set of facilities. Demand points (travel origins) collection Collection of intermediate stations on the main line (The final stop of each connecting route), i.e. For each demand point ( This corresponds to a specific reservation travel demand. ( Meanwhile, demand This corresponds to a travel time window. Let represent the set of arcs between nodes in the network, and satisfy ... .
[0047] The definitions of symbols, parameters, and variables in the model are shown in Table 1 in step four.
[0048] (3) Establish a collaborative optimization model for route design, timetable, and vehicle scheduling for feeder line connection mode. This model first considers the following three objective functions: (14); (15); (16); Function (14) aims to maximize the supply-demand matching rate. In the scenario of coordinated operation of main and branch lines, the output of function (4) includes the potential travel demand within the connecting area that can be served by customized passenger transport, which is used as the demand input for the branch line coordinated optimization model. This optimization objective enables customized passenger transport to cover more potential travel demand within the connecting area.
[0049] Function (15) minimizes the operating distance cost, which is measured by the distance traveled by the operating vehicle through the arc in the network.
[0050] Function (16) minimizes the travel time cost, where Defined in two parts: the first part is serving the needs. shuttle vehicles Arrive at the intermediate station on the main line Time and Main Line Direct Trains (need You need to transfer to a vehicle )arrive Time deviation (vehicle) arrive The timeframe is the model output result in step four); the second part is the requirements. On the site Transfer to Time: (17); To further optimize operating costs, based on the concept of trip chains in public transportation planning and operation theory, this invention considers the fleet sizing optimization (FSO) problem in customized passenger transport operations in connecting areas, i.e., minimizing the number of operating vehicles; the goal of this problem is to find a set of trip chains to serve passenger demand. In this invention, a trip chain refers to a set of trips that can be performed by the same operating vehicle. The input to solving the fleet sizing problem is a set of branch lines / trips in the connecting area. The connecting area consists of stations... d Each train dispatch It includes four basic attributes, namely the departure station (or depot) of the train. Departure time The train's arrival station (or depot). Arrival time Where h is the train number.
[0051] Therefore, if there is a set of consecutive train numbers If constraint (18) is satisfied, it means that this group of trains can be executed sequentially by a single train. This group of trains is then called a train sequence. (18); in, This indicates the total number of trains in the train number chain. Constraint (18) indicates two train numbers. and It can be done by a vehicle in order of train number ( , This can be executed by the station. Furthermore, it can be carried out by the site. d A certain car was dispatched ( A set of train services executed can be represented as: (19); in Indicates by vehicle The train number chain to be executed. Therefore, once all train numbers are given, a set of train number chains is defined. ,in This represents the number of vehicles required to execute all train services within the connecting area (i.e., the number of train service chains). Within this set, a group of train services completed by a single vehicle is also referred to as that vehicle's driving schedule. This can also be viewed as a driving plan for all vehicles. Each vehicle trip in this invention... It can only be carried out by one vehicle, and each trip... It can only be within one train chain.
[0052] Furthermore, the branch line collaborative optimization model needs to consider the target number of operating vehicles (vehicle scheduling problem or fleet size problem). Reducing the total number of operating vehicles can significantly reduce the overall operating cost. Therefore, based on the above concepts, compared with the aforementioned operating distance cost and travel time cost, this invention defines the fixed cost as the required number of operating vehicles: (20); in Describes the vehicle The number of train services that can be operated (train service chain); and then This indicates the total number of vehicles required to serve all train services. The cost of dispatching each vehicle.
[0053] The collaborative optimization model for route design, timetable, and vehicle scheduling for feeder connection modes is described as a multi-objective optimization model. Therefore, the feeder collaborative optimization model can be expressed as: (twenty one); The constructed multi-objective optimization model includes the following constraints: (twenty two); (twenty three); (twenty four); (25); (26); (27); (28); (29); (30); (31); Constraints (22) and (23) assign constraints to vehicle tasks. Each requirement... Service is provided by a maximum of one operating vehicle (or one route). Each vehicle From the station Depart, complete the corresponding train tasks for each route, and then return to the depot.
[0054] Constraint (24) is a time window constraint, i.e., the number of connecting line vehicles. Reaching potential demand Departure point (demand point) The time should be within its departure time window.
[0055] Constraint (25) is a timetable coordination constraint for the two types of route modes. This is to ensure the demand in the connecting areas. Arrive at the destination in the city on time, and, based on constraint (24), connect the shuttle bus. Arrive at the destination The time (i.e., the time corresponding to the intermediate station on the main line) should be within the operating time of a certain main line vehicle. leave Previous arrivals, i.e., potential demand in the connecting areas One of the conditions for being serviceable is that the timetable information for both types of line modes must simultaneously satisfy constraints (8), (24) and (25).
[0056] Constraint (26) is the vehicle capacity and occupancy rate ( )constraint.
[0057] Constraints (27) and (28) indicate that the vehicle is at two consecutive nodes. and The runtime relationship between them, where It is a sufficiently large number.
[0058] Constraints (29), (30) and (31) are variable constraints.
[0059] The sixth step involves solving the two types of multi-objective optimization models constructed in steps four and five using an intelligent optimization algorithm to output customized passenger transport route design and timetable compilation schemes. Based on the model characteristics, this algorithm is grounded in the Non-dominated Sorting Genetic Algorithm II (NSGA-II), solving the model by designing an adaptive mechanism for genetic operators and introducing an elite strategy. The main steps are as follows: (1) Encoding process (population initialization). The basic idea of the algorithm is to use individuals (chromosomes) to describe the route design and timetable compilation scheme. Based on the initial population, the population is updated through multiple generations of genetic operations, and finally the corresponding Pareto solution is output, that is, multiple operation schemes.
[0060] First, an initial population (initial solution) is generated through an encoding process. , For the main timetable optimization problem in step four, an integer encoding method is adopted, based on the number of departures within the operating period. (Number of vehicles) is used as the coding length for each chromosome, such as Figure 3 As shown. Therefore, the value of each gene locus in the chromosome can be determined by a randomly generated integer during the operating period, and thus... Each individual It can be described as: (32); In the formula: It can also be viewed as a departure timetable or departure time sequence corresponding to each individual; For the first time during the operating period The departure time of each vehicle (trip); furthermore, the departure time of each operating vehicle in the mainline mode. The arrival time at each intermediate station can be determined by the departure time and the travel time between stations.
[0061] Regarding the collaborative optimization problem of branch line design, timetable, and vehicle scheduling in step five, individual... It should include the station Information such as the stop order and departure times for connecting routes. According to the definition of the feeder service model, each route... It should include departure stations, intermediate stops on connecting routes (demand points), and destination stations (intermediate stops on the main line). Therefore, an integer encoding method, such as... Figure 3 As shown, for a certain connecting line ,definition: (33); In the formula: For the line The station number sequence, For the line The number of stations visited; Describes the vehicle In executing the line The order of station visits corresponds to the train schedule; "0" represents the departure station in the administrative village area; "1" represents the destination station of the connecting route. Simultaneously, define... For the operation of the line vehicle The departure time is determined by station "0". Similarly, the vehicle... The arrival and departure times at each intermediate station can be determined by the time constraints between stations.
[0062] Based on the above integer encoding rules, the initial population and initial solutions for the two types of problems are generated, and their population sizes are preset as follows: and .exist In this context, the length of each individual is defined as... ,in For the number of departures; in In this context, the length of each individual is defined as... ,in For each route, there is a corresponding bus trip and departure time. Therefore, each individual route defines an operational plan. Furthermore, since connecting routes need to be designed based on reservation demand, each... ( ), needs to be Departure time window and vehicles corresponding to demand points Arrival times are matched to determine whether the point is a vehicle. The actual stops along the route are visited, and operations such as updating the chromosomes and calculating the corresponding objective function values are performed. In addition, the vehicle... exist Arrive at the desired point before the earliest departure time, without affecting Based on the timely departure and arrival of other demand points, vehicles Available in Waiting for the earliest departure time at the demand point to increase operational flexibility and improve the supply-demand matching rate.
[0063] (2) Genetic operators. During the iteration process of the algorithm, the offspring population needs to be generated from the previous generation population through genetic operators such as crossover and mutation. The operation steps of the two operators are as follows.
[0064] ① Crossover operator. For example... Figure 4 As shown, two individuals are selected from the population. For the selected individuals, within the interval... Generate random numbers Based on crossover probability ,like If the crossover occurs, a crossover operation is performed; otherwise, no operator operation is performed. The crossover strategy involves exchanging gene segments of the same length based on the crossover position, resulting in two new individuals.
[0065] ② Mutation operators. For example... Figure 5 As shown, a single individual is selected from the population, within the interval Generate random numbers Based on the mutation probability ,like If the mutation operation is successful, then a mutation operation is performed; otherwise, no operator operation is performed. This strategy uses a swapping approach; for example, in the multi-objective optimization problem in step five, the numbers at two gene positions are randomly swapped on the selected individuals to generate new individuals.
[0066] ③ Adaptive mechanism of genetic operators. Based on the model characteristics, an adaptive adjustment mechanism for crossover and mutation probabilities is introduced to dynamically update the two parameters according to the population state. The specific calculation process is as follows: (34); (35); (36); (37); In the above calculation process, the corresponding fitness values need to be calculated based on the objective functions in steps four and five, and then the calculation... and Then calculate all and The average value is used to obtain the crossover probability of the current iteration. With the probability of mutation In the formula: The number of objective functions, , , These are the th in the current population. The maximum, minimum, and average fitness values of each objective function. The larger of the fitness values of the two individuals to be crossed is given. Let be the fitness value of the individual to be mutated. Furthermore, the other parameters in the formula satisfy... , .
[0067] (3) Vehicle scheduling process. For solving the branch line cooperative optimization model, this invention considers the fleet size problem. Based on generating a set of feasible solutions (vehicle set), this invention designs an operator to optimize the fleet size by connecting different vehicle sets to further optimize the number of operating vehicles. The input of this operator is a set of feasible vehicle sets, in which each vehicle set includes the aforementioned four basic attribute information. According to the definition of vehicle chain, each vehicle can serve only one vehicle set; or it can connect vehicle sets through constraint (18) to perform multiple vehicle tasks, thereby reducing the number of operating vehicles. The specific steps of this optimization operator are as follows: Algorithm input: Feasible train schedules .
[0068] Algorithm output: Train number chain set .
[0069] Step 1: Set the maximum number of iterations to 1. That is, from the station Set of dispatched trains The number of trains in the collection. At the same time, the collection... The train numbers are arranged in ascending order of departure time, denoted as... , This represents the number of trains in the set. Simultaneously, a temporary set is defined. From the set The first train in Initially, the elements in the set are iterated and evaluated sequentially, proceeding to steps 2-5. Then, the set... After traversing all elements and completing the iteration judgment, proceed to step 6.
[0070] Step 2: Based on Step 1, consider the train number selected in the current iteration round. And determine the current set Remove elements Does a set of train numbers exist? This allows the train numbers in the set to be connected according to time and space conditions (i.e., constraint (18)). If they exist, proceed to step 3; otherwise, proceed to step 4.
[0071] Step 3: Generate the train number chain Store to temporary collection At the same time, this set The train number element in the set Delete it and proceed to step 5.
[0072] Step 4: Set Leave it unchanged and proceed to step 5.
[0073] Step 5: Determine the set Have all elements in the list been traversed? If so, proceed to step 6; otherwise, select... For the next train in the sequence, proceed to step 2 for judgment.
[0074] Step 6: Merge sets and Generate a new set of train number chains In this step, the set The remaining trains do not belong to any train number chain, that is... Each remaining train trip requires a separate vehicle to perform its task. Meanwhile, the train trip set generated in step 2 can have all its tasks completed by a single vehicle. Finally, the generated set... It describes the number of vehicles required to complete all train missions.
[0075] (4) Solving the algorithm iteration process. The core steps of the algorithm include encoding (population initialization), fast non-dominated sorting, selection operation, fitness evaluation, crowding distance calculation, and genetic operator (crossover and mutation) operations. The fast non-dominated sorting, selection operation, fitness evaluation, and crowding distance calculation processes refer to existing research. Combining steps (1) and (2), the specific operation steps of the algorithm are as follows, where steps 1-3 are the initialization stage and steps 4-10 are the algorithm iteration stage.
[0076] The solution for the two types of models is adopted in stages. First, the main line timetable model is solved, and the solution results (such as the potential travel demand information of the filtered branch line connection areas and the timetable information of the main line) are used as the input of the branch line collaborative optimization model. Then, the branch line model is solved, and finally the route operation plan of the main line and the branch line is output.
[0077] Step 1: Algorithm Initialization. Input relevant parameters, such as population size. Individual length, number of iterations wait.
[0078] Step 2: Population Initialization. An initial population is generated according to the integer encoding rules. ( ).
[0079] Step 3: Put As the parent population, Select, crossover, and mutation operations are performed to generate offspring populations. and calculate The objective function values are used to determine dominance relationships and generate non-dominated fronts in fast non-dominated sorting. The selection process employs a tournament strategy commonly used in current literature. Proceed to step 4; update the iteration state. .
[0080] Step 4: Algorithm Iteration Phase If the algorithm termination condition is met, proceed to step 10; otherwise, proceed to steps 5-9 for algorithm iteration.
[0081] Step 5: Transfer the parent population of the previous generation With offspring population The population will be merged, with a size of 2. A new population is obtained, denoted as .
[0082] Step 6: Output using the fast non-dominated sort method. Corresponding Non-dominated frontiers at different levels and satisfy , .in," " indicates a dominant relationship, for example, " "express Dominate .
[0083] Step 7: Employ an elite strategy to generate the first... New parent population Assume the population size is The steps are as follows: First, let Secondly, order ,right Perform iterations, if ,but Otherwise, if Then calculate Calculate the crowding distance of each individual and sort them (from high to low), then select the one with the highest sorting value. The individual corresponding to the given position is incorporated into the new parent population. This keeps the size of the new parent population at a certain level. .
[0084] Step 8: Population Select, crossover, and mutation operations are performed to generate offspring populations. and calculate The corresponding objective function values (used for determining dominance relationships and generating non-dominated fronts in fast non-dominated sorting).
[0085] Step 9: Update the iteration state Return to step 4.
[0086] Step 10: Transfer the population from the last iteration. , The algorithm merges solutions, performs non-dominated sorting, and outputs the Pareto optimal solution, at which point it terminates. When outputting the results, solutions that do not meet the occupancy rate constraint need to be removed. Simultaneously, in the generated branch line connection mode route operation plan, further vehicle scheduling operations are performed to optimize fleet size and reduce operating costs.
[0087] In one possible implementation, the scenario is urban-rural travel to a transportation hub, which includes five township areas and one destination (the urban transportation hub). Each township area contains several demands originating from that area itself, as well as several demands originating from nearby villages. The study area also includes one mainline direct line depot and four branch line connecting line depots. The compilation environment for this embodiment is Matlab software.
[0088] The key parameters of the models and algorithms involved in the specific implementation are shown in Table 2.
[0089] Table 2. Values of key parameters in the model and algorithm. Based on the research scenario, reservation demands are categorized by origin: villages or townships. If originating from a village, the travel route is: village → township → urban area; if originating from a township, the travel route is: township → urban area. Information such as stations, facilities, and demand points is edited into coordinates to output the spatial layout of the scenario, such as... Figure 6 As shown.
[0090] This embodiment selects a travel scenario where residents travel to the city's train station. The train departure time for this scenario is set at 8:59. Considering that passengers need to complete necessary procedures such as security checks, and that trains typically begin ticket checking 15-20 minutes before departure, the latest arrival time at the destination is set at 8:29. According to the description in the specific implementation, the requirement for direct route vehicles to arrive at the train station before this time is one of the constraints for users selecting customized passenger transport. Table 3 shows the travel demand information of booked passengers and examples of corresponding service areas. This embodiment includes a total of 59 booking demand points.
[0091] Table 3 Example of Appointment Request Information Operational plan for the mainline direct route. All intermediate stations on the mainline direct route are pre-set and can be represented by station numbers: Om →Z1→Z2→Z3→Z4→Z5→ Dm ;in Om For the departure point (direct route depot). Dm For the destination (city railway station), such as Figure 7 As shown. The main timetable model is solved using an intelligent optimization algorithm. The algorithm is executed 10 times in the example and the optimal value is obtained. In the Pareto solution set of the output, the timetable information of the scheme with the highest supply and demand matching rate is shown in Table 4. For example, the vehicle of a certain train departs from the depot at 7:22:44 and arrives at the city railway station at 8:27:00, which meets the constraint of the latest arrival time. Under this operation scheme, the average travel time deviation index corresponding to the objective function (2) is about 9 minutes, and the travel efficiency is good.
[0092] Table 4. Timetable Compilation Scheme for Mainline Direct Routes Feeder Line Operation Plan. Based on the main line operation plan, the potential demand in the feeder line connection area is updated according to the original travel demand, and the demand point information is used as the data input for the feeder line model. In the selected case data, there are four feeder line depots: J1, J2, J3, and J4. For vehicles departing from each depot, the intermediate stops of the final direct route have the following correspondence: J1→Z5; J2→Z4; J2→Z3; J3→Z2; J4→Z1. Figure 6 The scenario includes five connecting areas, and intelligent optimization algorithms are applied to solve the branch line model for each, outputting the operation plan for each connecting area. Table 5 shows an example of the route operation plan corresponding to the maximum supply-demand matching rate in the Pareto solution output for each connecting area. The startup cost per trip is 5 RMB. Taking the route plan J4→59→57→54→56→58→Z1 in connecting area 5 of Table 5 as an example, the vehicle departs from station J4, serving passengers corresponding to demand points 59, 57, 54, 56, and 58 in sequence, and finally transporting passengers to the intermediate station Z1 of the direct route. In this example, the vehicle arrives at Z1 at 7:26:08; as shown in Table 4, the main line operating vehicle leaves Z1 at 7:28:02, indicating that when the connecting line vehicle arrives at Z1, the main line vehicle has not yet left Z1, satisfying the timetable coordination constraint requirement and realizing the coordinated compilation of the timetable for the main line and branch line.
[0093] Table 5. Examples of Line Operation Schemes under Optimal Supply-Demand Matching Rate Comparative Analysis of Different Operating Modes. This embodiment compares the differences in operating costs between the "hub-and-spoke" route design mode and the "direct" route mode, as shown in Table 6. The distance cost calculation method for the direct route mode in Table 6 is: the operating distance corresponding to the "village-to-city" direct mode formed by extending branch lines directly from transfer stations to urban destinations. The results show that, under the optimal supply-demand matching scheme, for the five connection areas, while meeting passenger travel time windows, the hub-and-spoke design approach can save operating distance costs of 43.7%, 18.5%, 46.2%, 49.4%, and 57.6%, respectively. Considering the operating distance of the main line, the overall operating cost can be saved by approximately 30.8%, effectively balancing travel and operating cost indicators.
[0094] Table 6. Comparison of operating distance costs between the "hub-and-spoke" and "direct" modes. The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.
Claims
1. A method for designing and scheduling tiered routes for customized urban and rural passenger transport, characterized in that, include: Obtain booking demand data for customized passenger transport; We set up a mainline cross-regional direct mode and a branch line connection mode, and constructed a mainline timetable optimization model with supply and demand matching rate and travel time as objective functions. At the same time, we constructed a collaborative optimization model for branch line route design, timetable and vehicle scheduling oriented towards the branch line connection mode with supply and demand matching rate, travel time cost and operating cost as objective functions. Based on the reservation travel demand data, a non-dominated sorting genetic algorithm that integrates genetic operators and elite strategies is used to solve the main line timetable optimization model and the collaborative optimization model, and output a customized passenger transport route operation plan; during the solution of the branch line model, vehicle scheduling operations are performed to optimize the fleet size and reduce operating costs.
2. The method for designing and scheduling customized urban and rural passenger transport routes according to claim 1, characterized in that, Constructing the main timetable optimization model includes: The reservation travel demand data is divided into two categories: departure from townships and departure from administrative villages. The boarding and alighting points of passengers departing from townships are located in townships and urban areas, respectively. The travel trajectories of passengers departing from administrative villages are set sequentially as administrative villages, townships, and urban areas. If the reservation travel demand data originates from townships, then from a timetable compilation perspective, the departure times of direct route vehicles from intermediate stops and their arrival times at urban destinations must meet the time window of the reservation demand. If the reservation travel demand data originates from administrative villages, shuttle vehicles must transport passengers to intermediate stops on the direct route. Simultaneously, the departure times of mainline direct route vehicles from these intermediate stops must not be later than the target time window, and shuttle vehicles serving the demand originating from these administrative villages must arrive at these intermediate stops before the mainline vehicles depart. Therefore, the mainline timetable optimization model is constructed with supply-demand matching rate and travel time as the objective functions. ； ； ； in, To maximize the supply and demand matching rate, To minimize travel time costs, For station Corresponding vehicle set, A collection of needs originating from townships. This is a collection of connection needs originating from administrative villages. In the main storyline, the passengers being served The travel time cost indicator Indicates if travel demand Can be from the station Vehicles dispatched If the service is active, its value is 1; otherwise, its value is 0. The demand originating from townships, This represents the number of shuttle bus requests originating from administrative villages. To improve the supply and demand matching rate, For travel time costs.
3. The method for designing and scheduling customized urban and rural passenger transport routes according to claim 2, characterized in that, The method also includes: constraints on the main timetable optimization model: Vehicle task allocation constraints: ； Time window constraint: ； ； Timetable collaborative compilation constraints: ； Vehicle capacity constraints: ； Runtime constraints: ； ； ； Variable constraints: ； in, Main line vehicles Arrival at the intermediate station Time, Main line vehicles k Arrival time at the destination in the city center For demand Expected arrival time at the destination Main line vehicles At the intermediate station Service hours, In the main line, two stations and The distance between, The speed of vehicles traveling on the main line. For vehicle occupancy rate, The capacity of vehicles operating on the main line. For demand The corresponding number of reservations The earliest departure time within the main line's operating hours. The latest departure time within the main line's operating hours. The main train schedule is as follows. Main line vehicles Arrival at the intermediate station Time, Main line vehicles Leaving the intermediate station Time, In the main line, two stations and Runtime between For station Each train dispatch Departure time, , For each passenger's expected departure time, a coefficient is introduced. Define requirements The travel time window is adjusted to increase the flexibility of the operational plan, taking into account each passenger's desired departure time. Define coefficients Thus, it can meet the demand The corresponding departure station defines its travel time window. ,coefficient for The corresponding time window parameters.
4. The method for designing and scheduling customized urban and rural passenger transport routes according to claim 1, characterized in that, The collaborative optimization model for the branch line connection mode includes: ； ； ； ； ； in, To improve the supply and demand matching rate, For operating distance costs, For travel time cost, The number of operating vehicles required to complete all train trips. To maximize the supply and demand matching rate, To minimize operating distance costs, To minimize travel time costs, For the station assembly, For the connecting area, the station The corresponding set of vehicles Indicates if travel demand Can be from the station Vehicles dispatched If the service is active, its value is 1; otherwise, its value is 0. For site collection, For feeder lines, the operating cost per unit distance. As a branch line, there are two stations. and The distance between, Indicates if the station Mainline vehicles dispatched From demand points Drive to the point of demand If it is 1, its value is 1; otherwise, its value is 0. As a branch line, the demand The travel time cost indicator This represents the number of shuttle bus requests originating from administrative villages. For station d A certain vehicle was dispatched. For station d The corresponding set of vehicles The cost of dispatching each vehicle. For station A certain train number was dispatched. For station Collection of departing trains, Indicates if train number Vehicles If the value is 1, it is used; otherwise, it is used for 0.
5. The method for designing and scheduling customized urban and rural passenger transport routes according to claim 4, characterized in that, The method also includes: constraints on the cooperative optimization model for the branch connection mode: Vehicle task allocation constraints: ； ； Time window constraint: ； Coordination constraints of timetables for the two types of route modes: ； Vehicle capacity and occupancy constraints: ； Runtime constraints for consecutive nodes: ； ； Variable constraints: ； ； ； in, This is a collection of connection needs originating from administrative villages. For connecting line vehicles Arrival at the intermediate station Time, For connecting line vehicles Arrive at the intermediate station of the corresponding direct route Time, For the connecting area, the station The corresponding set of vehicles For the capacity of branch line operating vehicles, For branch line vehicles At the intermediate station Service hours, Two demand points in the branch line and Runtime between For connecting line vehicles Arrival at the intermediate station Time, For a sufficiently large number, For station The connection between the dispatched vehicle and the first stop along the way. For station The dispatched vehicles return to the depot after completing their service. For passengers' expected departure time, for The corresponding time window parameters, Main line vehicles Leaving the intermediate station Time, This refers to the occupancy rate.
6. The method for designing and scheduling tiered routes for customized urban and rural passenger transport according to claim 1, characterized in that, The customized passenger transport route operation plan, which utilizes the non-dominated sorting genetic algorithm to output the main line cross-regional direct mode and branch line connecting mode, includes: Step 1: Generate the initial population according to the integer encoding rules; Step 2: Select the initial population using a tournament strategy, perform crossover and mutation operations on the initial population using an adaptive genetic operator to generate offspring populations, and calculate the objective function values corresponding to the offspring populations. Step 3: Merge the previous generation parent population with the offspring population to obtain a new population; Step 4: Use the fast nondominated sorting method to output the nondominated frontiers at multiple different levels corresponding to the new population; Step 5: Iteratively assess the non-dominant frontier using an elite strategy to generate a new parent population: If ,but until The crowding distance of each individual in the non-dominated front is calculated and sorted from high to low. The individual with the highest sorting value is selected. The individual corresponding to the given position is incorporated into the population, resulting in a new parent population; among them... For population, As a non-dominant frontier, Population size; Step 6: Repeat steps 2-5 until the termination condition of the iteration state is met. Merge the population and offspring population in the last iteration and output the Pareto optimal solution corresponding to the merged population using the fast non-dominated sorting method. This can be regarded as a customized passenger transport route operation plan, and solutions that do not meet the occupancy rate constraint are eliminated. In the generated branch line connection mode route operation plan, vehicle scheduling operations are further performed to optimize the fleet size and reduce operating costs.
7. The method for designing and scheduling tiered routes for customized urban and rural passenger transport according to claim 6, characterized in that, The initial population of the non-dominated sorting genetic algorithm includes: The initial population of the main timetable optimization model: ； The initial population of the branch cooperative optimization model: ； in, A departure timetable or departure time sequence corresponding to each individual. For the first time during the operating period The departure time of a vehicle or train. For vehicles In executing the line The order of station accesses when corresponding to a train number task. For the line The site number, For the line The number of stations visited.
8. The method for designing and scheduling tiered routes for customized urban and rural passenger transport according to claim 6, characterized in that, The genetic operators for the adaptive mechanism in the non-dominated sorting genetic algorithm include: ； ； ； ； in, For adaptive crossover probability, The number of objective functions, , , These are the th in the current population. The maximum, minimum, and average fitness values of each objective function. The larger of the fitness values of the two individuals to be crossed is given. The fitness value of the individual to be mutated. For adaptive mutation probability, These are parameters used in the crossover probability calculation process. These are parameters used in the calculation of the probability of mutation. , , For other parameters, satisfy ; , , For other parameters, satisfy .
9. The method for designing and scheduling customized urban and rural passenger transport routes according to claim 1, characterized in that, The vehicle dispatching operation includes: Step 1: Set the maximum number of iterations to 1. Arrange the trains in the set of trains departing from the station in ascending order of departure time, and set up a temporary set; Step 2: Starting from the first train in the train set, iterate through the elements in the train set in sequence to determine whether there is a target train set among the elements in the train set excluding the train in the current iteration round, so that the trains in the target train set can be connected according to time and space conditions. Step 3: If the target train set exists, store the generated train chain in the temporary set, delete the train elements in the target train set from the train set, and determine whether the elements in the train set have been traversed. If the traversal is complete, merge the train set and the temporary set to generate a new train chain set. If the traversal is not complete, re-determine whether the target train set exists in the current train set excluding the trains in the next iteration round, until the traversal is complete. Step 4: If the target train set does not exist, the train set remains unchanged, and it is determined whether the elements in the train set have been traversed. If they have, the train set and the temporary set are merged to generate a new train chain set. If the traversal is not completed, it is determined again whether the target train set exists in the current train set excluding the trains in the next iteration round, until the traversal is completed.