A heterogeneous fleet bus line scheduling method considering passenger demand preferences

Abstract

The application discloses a heterogeneous vehicle fleet bus line scheduling method considering passenger demand preferences, and relates to the technical field of intelligent bus operation scheduling.The method comprises the following steps: constructing a multi-source parameter set of bus lines, vehicles, passengers, special demands and real-time running states; constructing a passenger-vehicle service adaptation matrix and a demand preference satisfaction degree matrix based on passenger vehicle preference parameters; constructing a special demand-vehicle capacity matching matrix based on passenger special demand parameters and vehicle service capacity parameters; constructing a heterogeneous vehicle fleet bus line scheduling optimization model considering passenger vehicle preferences and special demand priority protection; and adopting an offline benchmark scheduling and online rolling rescheduling combined solving method to output a bus line scheduling scheme.The application reduces the allocation of passengers to vehicle service schemes that do not meet their demand preferences, and realizes intelligent vehicle deployment when passengers have special demands, thereby improving bus service adaptability, vehicle utilization rate and operation reliability.

Description

Technical Field

[0001] This invention relates to the field of intelligent public transport operation and scheduling technology, and in particular to a method for scheduling bus routes with heterogeneous fleets that takes into account passenger demand and preferences. Background Technology

[0002] With the development of driverless buses, vehicle-road cooperative systems, and smart bus platforms, urban public transport operations are gradually shifting from the traditional single-operation model of manually driven buses to a heterogeneous fleet model that combines manually driven and driverless buses. In this type of operation, bus companies not only need to determine vehicle departure times and vehicle turnover plans, but also whether each shift will be operated by a manually driven or a driverless bus.

[0003] Current bus dispatching methods typically prioritize optimization goals such as departure intervals, vehicle capacity, operating costs, and passenger waiting times, with little consideration for passengers' varying levels of acceptance of different vehicle types. For example, some passengers may reject driverless buses due to concerns about safety, mobility, or psychological preferences; others may desire or prioritize driverless buses; still others may be open to both types of vehicles. If the dispatching system ignores these differences in passenger vehicle preferences, passengers may be assigned to vehicle types they do not wish to ride, leading to refusals, complaints, failed transfers, or decreased operational efficiency.

[0004] Meanwhile, public transportation operations involve special needs groups such as wheelchair passengers, visually and hearing impaired passengers, elderly passengers, children, passengers with large luggage, stroller passengers, and passengers requiring manual assistance to board and alight. These passengers are not only concerned about whether the vehicle can reach their destination, but also whether the vehicle has accessible facilities, special seating, manual assistance, or higher punctuality. Existing scheduling methods mostly treat special needs as general capacity constraints, lacking an intelligent scheduling mechanism that combines vehicle type, passenger needs and preferences, and real-time operational status.

[0005] Furthermore, existing technologies for heterogeneous fleet scheduling primarily rely on differences in hardware parameters such as vehicle size, rated passenger capacity, energy supply type, and vehicle configuration. They fail to address the specific scheduling differences between manually driven and autonomous vehicles. These two types of vehicles differ significantly in driver availability, remote monitoring resource allocation, passenger acceptance, service adaptability for special groups, and operating costs per unit time period. The existing scheduling framework is ill-suited for the new public transport operation scenario of mixed passenger and vehicle operations, and struggles to balance passenger travel experience with the operational efficiency of public transport companies. Summary of the Invention

[0006] The purpose of this invention is to propose a heterogeneous bus route scheduling method that takes into account passenger demand preferences, thereby solving the problems of existing bus scheduling methods that do not fully consider passenger preferences for manually driven and driverless bus models, lack of support for passengers with special needs, and weak collaborative scheduling capabilities between driverless and manually driven buses.

[0007] To achieve the above objectives, this invention proposes a method for scheduling heterogeneous bus fleets that considers passenger demand preferences. The specific steps are as follows: Step S1: Construct a multi-source parameter set for bus route vehicle scheduling. The multi-source parameter set includes route station parameters, passenger travel parameters, passenger vehicle preference parameters, passenger special demand parameters, vehicle resource parameters, vehicle service capacity parameters, and real-time operation status parameters. Step S2: Based on passenger vehicle preference parameters, construct a passenger-vehicle service adaptation matrix to determine the acceptable set and demand preference satisfaction of each passenger for different combinations of vehicle types and service capabilities. Step S3: Based on passenger special needs parameters and vehicle service capacity parameters, construct a special needs-vehicle capacity matching matrix to determine whether the vehicle meets the passenger's needs for barrier-free travel, manual assistance, priority travel, time window guarantee, and group travel. Step S4: Based on the passenger-vehicle type serviceability matrix and the special needs-vehicle capacity matching matrix, construct a heterogeneous fleet bus route scheduling optimization model that considers passenger vehicle preference satisfaction and priority guarantee of special needs. Step S5: The heterogeneous bus fleet route scheduling optimization model is solved by using a hybrid solution method that combines offline baseline scheduling with online rolling rescheduling. Step S6: Output the bus route dispatch plan, which includes the departure time of each bus, vehicle type, vehicle number, passenger allocation results, special needs passenger guarantee plan, backup vehicle call plan and dispatch adjustment instructions.

[0008] Preferably, in step S1, the route station parameters include a set of bus routes. L Site Collection I Interval travel time Planned train schedule assembly K Minimum departure interval Maximum departure interval First and last bus times and station service hours; The passenger travel parameters include the passenger set. P Passenger vehicle preferences Passenger origin station Terminal Expected boarding time The latest departure time for passengers Passenger numbers And ticket reservation status; The vehicle resource parameters include a set of manually driven buses. A collection of driverless buses Vehicle capacity , special seat capacity Current vehicle location, remaining vehicle range, vehicle availability time, and vehicle operating costs. Driver resource status and unmanned remote monitoring resource status; The real-time operational status parameters include vehicle delay time, number of people waiting at stations, road congestion status, vehicle malfunction status, temporary traffic control status, and sudden passenger demand status.

[0009] Preferably, in step S2, passenger vehicle preference The value can be 0, 0.5, or 1; when When the value is 0, there is a preference for manually driven buses, and manually driven bus trips are prioritized during dispatching. when When =1, driverless buses are preferred, and driverless bus schedules are prioritized during dispatching; when When the ratio is 0.5, there is no obvious preference for manually driven buses and driverless buses, and the scheduling can be matched based on the overall cost; passenger p For vehicle type q Acceptance constraint coefficient Represented as: like When =0, =1; ; like When =1, =1; ; like When =0.5, =1; =1; in, q =H or q =A, H represents manually driven buses, and A represents driverless buses.

[0010] Preferably, in step S3, the special requirements parameters include wheelchair or mobility aid requirements, low-floor vehicle requirements, manual assistance for getting on and off the vehicle requirements, visual or auditory assistance requirements, child or elderly care requirements, space requirements for large luggage or strollers, requirements for passengers in the same group to be in the same vehicle, and priority for on-time arrival. Passengersp Special requirements are represented as vectors =( , ,..., ), will the vehicle v Service capabilities are represented as vectors =( , ,..., When the vehicle v All service capabilities meet the needs of passengers. p When there are special requirements, the special requirement matching coefficient =1, otherwise =0; in, For passengers p The r Special requirements, For vehicles v The r Service capabilities; if passengers p If there is a need for manual assistance in boarding and alighting, and the driverless bus cannot meet this need through station staff, remote customer service, or in-vehicle auxiliary equipment, then the corresponding driverless bus... =0.

[0011] Preferably, in step S4, when constructing the heterogeneous bus fleet route scheduling optimization model, the following decision variables are set: The variable is 0-1, representing the shift. k Whether by vehicle v implement; The variable is 0-1, representing passengers. p Whether assigned to a shift k ; Indicates train number k The actual departure time; The variable is 0-1, representing passengers. p Whether it failed to be satisfied within the current scheduling period; A 0-1 variable, representing a vehicle. v Whether it will be used as a backup vehicle; The variable is 0-1, representing the shift. k Whether the vehicle type or departure time was adjusted due to special needs.

[0012] The preferred objective function for the heterogeneous bus fleet route scheduling optimization model is as follows: ; Where F is the objective function value of the model. Total passenger waiting cost For vehicle operating costs, The penalty cost for passengers not serving is higher for passengers with special needs than for ordinary passengers. Penalty costs for unmet special needs Cost of deviation from scheduled departure times. Costs related to uneven vehicle turnover The cost weight for passenger vehicle preference violation is used to control the importance of preference violation in the overall objective function. ~ These are the weighting coefficients.

[0013] Preferably, the bus route scheduling optimization model for heterogeneous bus fleets includes at least the following constraints: Passenger service constraints: For each passenger p ,satisfy ; Passenger vehicle preference adaptation constraints: ;in, Vehicle type; Special requirements matching constraints: ; Vehicle capacity constraints: ; Special seat capacity constraints: ;in, S For passengers with special seating or accessibility needs; Departure interval constraints: ; Vehicle turnaround constraint: When the same vehicle performs two consecutive shifts, the sum of the arrival time, turnaround time and charging or recharging time of the previous shift shall not exceed the departure time of the subsequent shift. Driver and remote monitoring resource constraints: When manually driven buses are running shifts, they need to be matched with available drivers; when driverless buses are running shifts, they need to meet the upper limit of the number of remotely monitored vehicles. Among them, passenger demand preferences can be set as hard constraints or soft constraints depending on the type of demand. Passenger vehicle preference adaptation constraints can be soft constraints or high-penalty constraints; passenger special demand matching constraints are set according to the type of special demand, among which rigid demands such as safety, accessibility, and manual assistance are hard constraints, while comfort preferences and waiting time preferences can be set as soft constraints or high-penalty constraints.

[0014] Preferably, in step S5, the specific steps of online rolling rescheduling are as follows: Step S51: When a new passenger with special needs, vehicle malfunction, road delay, abnormal autonomous vehicle service capacity, or insufficient human driver resources are detected, the rolling dispatch window is triggered. Step S52: Calculate the scheduling priority index for passengers with special needs. The formula is as follows: ; in, The latest departure time for passengers. For the current moment, To prevent constants with a denominator of zero, For special needs level, The remaining number of flights that can serve passengers. The number of shifts remaining in the dispatch window. This is a correction item for the number of passengers in the same group. to Priority index weights; Step S53: Insert and match passengers with special needs according to the scheduling priority index from high to low; Step S54: If there is available capacity on existing routes that meets the conditions for matching passengers' rigid demand, demand preferences and special needs, then passengers will be assigned to that route and the corresponding seats or service capacity will be locked. Step S55: If the existing schedule does not meet the service conditions, then execute the following strategies in sequence: vehicle type exchange, departure time fine-tuning, backup vehicle deployment, platform service personnel coordination, and manual customer service confirmation. Step S56: Under the premise of satisfying passenger vehicle preference constraints and safe operation, output the rescheduling results and update the vehicle, passenger and station status.

[0015] Preferably, in step S55, the specific strategy for vehicle type exchange is as follows: When passengers prefer vehicles When the number of passengers is 0 and the currently available bus is an autonomous bus, the system will prohibit the passenger from being assigned to that autonomous bus and will prioritize adjusting subsequent manually driven buses to the direction of the passenger's stop. When passengers prefer vehicles When the number of passengers is 1 and the currently available bus is a manually driven bus, the system will prohibit the passenger from being assigned to that manually driven bus and will prioritize matching the passenger to an autonomous bus or calling up an autonomous backup vehicle. When passengers prefer vehicles When the value is 0.5, the system selects the vehicle type with the lowest overall cost and that meets the special requirements from both manually driven buses and driverless buses. When a special need conflicts with a passenger's vehicle preference, the system generates a conflict alert and prioritizes resolving the conflict through platform service personnel, accessible facilities, backup vehicles, or customer service confirmation. If the conflict cannot be resolved, the passenger is marked as a candidate for manual intervention.

[0016] Preferably, in step S5, a hierarchical hybrid solution algorithm is used for solving the problem. The specific steps are as follows: The upper-level model determines the vehicle type, departure time, and number of spare vehicles reserved for each shift; The lower-level model determines the matching relationship between passengers and schedules, the allocation of seats for special needs, and the vehicle turnover relationship; The static reservation requirements are solved using mixed integer programming to generate an offline baseline scheduling scheme; For newly added requests in real time, rolling time-domain optimization and adaptive large neighborhood search algorithms are used for rapid repair. If the rolling optimization results do not converge within a preset time, an executable emergency scheduling scheme is generated using a greedy insertion rule based on a special demand priority index.

[0017] Therefore, this invention proposes a method for scheduling heterogeneous bus fleets that considers passenger demand preferences, with the following beneficial effects: (1) This invention explicitly introduces passenger vehicle preferences into the bus route vehicle scheduling process, which can avoid assigning passengers to vehicle types they do not accept, thereby improving bus service satisfaction and ride matching accuracy. (2) This invention clearly defines a heterogeneous fleet as a fleet consisting of two types of vehicles: manually driven buses and driverless buses. This is different from the traditional scheduling method that classifies heterogeneous vehicles by vehicle type, capacity or energy type, and is more suitable for the mixed operation stage of driverless buses. (3) The present invention constructs a passenger vehicle preference-special needs-vehicle capacity coupling matching mechanism, which enables passengers with special needs to be given priority protection, and triggers vehicle exchange, backup vehicle call, platform service personnel linkage or manual customer service confirmation strategy when there is a vehicle type conflict. (4) The present invention adopts a solution method that combines offline baseline scheduling with online rolling rescheduling, which can not only ensure the stability of daily operation plans, but also cope with special needs and emergencies, thus improving the intelligent scheduling capability of bus companies. (5) The output results of this invention include not only vehicle departure time, but also vehicle type, vehicle number, passenger allocation, special seat locking, backup vehicle call and manual intervention prompt, which can be directly integrated into the bus company's dispatch platform.

[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0019] Figure 1 This is an overall flowchart of a heterogeneous bus route scheduling method that takes into account passenger demand preferences, according to the present invention. Figure 2 This is a flowchart illustrating the matching of passenger vehicle preferences with special needs in this invention; Figure 3 This is a flowchart illustrating the online rolling rescheduling triggered by special requirements in this invention. Detailed Implementation

[0020] To make the technical solutions, advantages, and objectives of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below. The described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention without creative effort are within the protection scope of the present invention.

[0021] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0022] Example 1 like Figure 1 As shown, this invention provides a method for scheduling heterogeneous bus fleets that considers passenger demand preferences, including two vehicle types: manually driven buses and driverless buses. The specific steps are as follows: Step S1: Construct a multi-source parameter set for bus route vehicle scheduling. The multi-source parameter set includes route station parameters, passenger travel parameters, passenger vehicle preference parameters, passenger special demand parameters, vehicle resource parameters, vehicle service capacity parameters, and real-time operation status parameters. Step S2: Based on passenger vehicle preference parameters, construct a passenger-vehicle service adaptation matrix to determine the acceptable set and demand preference satisfaction of each passenger for different combinations of vehicle types and service capabilities. Step S3: Based on passenger special needs parameters and vehicle service capacity parameters, construct a special needs-vehicle capacity matching matrix to determine whether the vehicle meets the passenger's needs for barrier-free travel, manual assistance, priority travel, time window guarantee, and group travel. Step S4: Based on the passenger-vehicle type serviceability matrix and the special needs-vehicle capacity matching matrix, construct a heterogeneous fleet bus route scheduling optimization model that considers passenger vehicle preference satisfaction and priority guarantee of special needs. Step S5: Use a hybrid solution method that combines offline baseline scheduling with online rolling rescheduling to solve the bus route scheduling optimization model for heterogeneous fleets; Step S6: Output the bus route dispatch plan, which includes the departure time of each bus, vehicle type, vehicle number, passenger allocation results, special needs passenger guarantee plan, backup vehicle call plan and dispatch adjustment instructions.

[0023] like Figure 2 As shown, the passenger vehicle preference parameter is used to limit the types of vehicles a passenger can accept. When the passenger's vehicle preference is 0, the system prioritizes matching them with manually driven buses; when the passenger's vehicle preference is 1, the system prioritizes matching them with driverless buses; when the passenger's vehicle preference is 0.5, the system allows them to ride either a manually driven bus or a driverless bus. After completing the initial screening of vehicle types, the system further determines whether the vehicle has the service capability to meet the passenger's specific needs. Only when the vehicle type is accepted and the vehicle's capability meets the specific needs will the corresponding vehicle and bus be added to the candidate service set.

[0024] like Figure 3 As shown, when the system detects new passengers with special needs, vehicle malfunctions, road delays, abnormal autonomous vehicle service capabilities, or insufficient human driver resources, it triggers online rolling rescheduling. The system first determines the rolling scheduling window and freezes schedules that have already been executed or are about to become unadjustable. Then, it calculates the priority index for special needs passengers and matches them according to the priority index from high to low. If there are existing schedules that meet the passenger's vehicle preferences and special needs, the passenger is directly assigned and the corresponding seats or service capacity is locked. If no schedule can be directly inserted, the system sequentially performs vehicle type exchange, departure time fine-tuning, and backup vehicle mobilization, ultimately forming an executable rescheduling plan.

[0025] The technical solution of the present invention will be further illustrated below through specific implementation examples.

[0026] This example uses a city bus route L1 as an example. This route includes 12 stops, and the morning peak dispatch period is from 7:00 to 9:00. The bus company has 8 manually driven buses and 6 driverless buses available. All manually driven buses are equipped with drivers, while the driverless buses are centrally monitored by a remote monitoring center. Some vehicles are equipped with accessible ramps and wheelchair securing areas.

[0027] Step S1: Construct a multi-source parameter set. Obtain route stops, service schedules, vehicle locations, vehicle capacity, driver status, remote monitoring resources, reserved passengers, number of passengers waiting at stops, and real-time road operation status. Record passengers' origin, destination, expected boarding time, latest boarding time, passenger vehicle preferences, and special needs.

[0028] In this embodiment, passenger p1 is a wheelchair passenger, and the passenger's vehicle preference... =0, their demand preferences include accessibility facilities, wheelchair securing areas, and manual assistance, and they have a low acceptance of driverless buses; passenger p2 passenger vehicle preferences =1, indicating a high preference for driverless bus experiences; passenger p3 is carrying a stroller, indicating passenger vehicle preference. =0.5, whose demand preferences include larger standing space and lower crowding, and no obvious preference for vehicle driving type.

[0029] Step S2: Construct the passenger-vehicle type serviceability matrix. Based on passenger vehicle preferences, the following rules are derived: For p1, since... =0, the system only allows it to be assigned to manually driven bus shifts; for p2, because... =1, the system only allows it to be assigned to driverless bus trips; for p3, because... =0.5, the system allows it to be assigned to either manually driven or driverless bus trips.

[0030] Step S3: Construct a special needs-vehicle capability matching matrix. The system identifies that p1 has wheelchair accessibility requirements, needs a low-floor vehicle or accessible ramp, and needs to reserve a wheelchair fixing area. If a driver-operated bus has accessible facilities and remaining wheelchair fixing area, the matching coefficient for p1's special needs is 1; otherwise, it is 0. For p3, the system identifies that it is carrying a stroller and needs a larger standing space. If the remaining space in the vehicle meets the requirements, the matching coefficient is 1.

[0031] Step S4: Construct a heterogeneous bus fleet scheduling optimization model. The model aims to reduce passenger waiting time, lower operating costs, minimize unserved passengers, ensure service for passengers with special needs, reduce the magnitude of service adjustments, and maintain balanced vehicle turnover. For passengers with special needs, the model assigns a higher penalty weight for unserved service. Regarding passenger vehicle preference constraints, the model uses soft constraints based on vehicle preference satisfaction and preference violation costs to prioritize services with higher preference satisfaction. For rigid service requirements such as safety, accessibility, and human assistance, hard constraints are still applied.

[0032] Step S5: Execute offline baseline scheduling and online rolling rescheduling. During the offline phase, the system generates an initial schedule based on historical passenger flow and reservation demand. In this embodiment, manually driven buses and driverless buses are scheduled to depart alternately between 7:00 and 8:00, ensuring both types of vehicles maintain a reasonable service frequency.

[0033] During the online rolling rescheduling phase, the system detects that p1 submitted a wheelchair boarding request at 7:35, with an expected boarding time of 7:50 and a latest boarding time of 8:05. The system calculates that p1 has a high scheduling priority index and queries available buses within the next 30 minutes. If the driverless bus at 7:48 has capacity but cannot meet p1's acceptance threshold for human assistance or accessibility services, the system excludes it or marks it as a low-adaptability candidate. If the human-driven bus at 7:55 has accessibility facilities, the system assigns p1 to the 7:55 bus and locks the wheelchair-accessible area. If the 7:55 bus does not have accessibility facilities, the system prioritizes exchanging this bus with a subsequent human-driven bus that has accessibility facilities; if exchanging is not feasible, it calls upon a backup human-driven bus or generates a station human service instruction.

[0034] For p2, since it only accepts driverless buses, when the adjacent bus is a manually driven bus, the demand preference of that bus is less satisfied, and the system prioritizes arranging for it to take the subsequent driverless bus.

[0035] If the waiting time exceeds the threshold, the system will trigger the call of a backup autonomous vehicle or adjust the departure time of the autonomous bus.

[0036] For p3, since it is acceptable for both types of vehicles, the system selects the vehicle type and schedule with the lowest overall cost among the schedules that meet the stroller space requirements.

[0037] Step S6: Output the scheduling plan. The system outputs the final scheduling results, including that the 7:55 bus will be operated by a manually driven accessible bus, a wheelchair-secured area will be reserved for p1, p2 will be assigned to the 8:00 driverless bus, and p3 will be assigned to the 7:48 driverless bus. At the same time, the system outputs information on vehicle exchange, backup vehicle deployment, and station service personnel coordination to the dispatcher.

[0038] It is worth noting that all contents not described in detail in this invention are existing technologies and are well known to those skilled in the art.

[0039] Therefore, this invention provides a method for scheduling bus routes with heterogeneous fleets that takes into account passenger demand preferences. Under the condition of mixed operation of manually driven buses and driverless buses, it can take into account passenger demand preferences for vehicle type, special needs guarantee, vehicle capacity, operating costs and scheduling stability, and realize intelligent bus route vehicle scheduling for passengers with special needs.

[0040] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for scheduling heterogeneous bus fleets considering passenger demand and preferences, characterized in that, The heterogeneous bus fleet includes two vehicle types: manually driven buses and driverless buses. The specific steps are as follows: Step S1: Construct a multi-source parameter set for bus route vehicle scheduling. The multi-source parameter set includes route station parameters, passenger travel parameters, passenger vehicle preference parameters, passenger special demand parameters, vehicle resource parameters, vehicle service capacity parameters, and real-time operation status parameters. Step S2: Based on passenger vehicle preference parameters, construct a passenger-vehicle service adaptation matrix to determine the acceptable set and vehicle preference satisfaction for each passenger for different combinations of vehicle types and service capabilities. Step S3: Based on passenger special needs parameters and vehicle service capacity parameters, construct a special needs-vehicle capacity matching matrix to determine whether the vehicle meets the passenger's needs for barrier-free travel, manual assistance, priority travel, time window guarantee, and group travel. Step S4: Based on the passenger-vehicle type serviceability matrix and the special needs-vehicle capacity matching matrix, construct a heterogeneous fleet bus route scheduling optimization model that considers passenger vehicle preference satisfaction and priority guarantee of special needs. Step S5: The heterogeneous bus fleet route scheduling optimization model is solved by using a hybrid solution method that combines offline baseline scheduling with online rolling rescheduling. Step S6: Output the bus route dispatch plan, which includes the departure time of each bus, vehicle type, vehicle number, passenger allocation results, special needs passenger guarantee plan, backup vehicle call plan and dispatch adjustment instructions.

2. The method for scheduling heterogeneous bus routes considering passenger demand preferences according to claim 1, characterized in that, In step S1, the route station parameters include a set of bus routes. L Site Collection I Interval travel time Planned train schedule assembly K Minimum departure interval Maximum departure interval First and last bus times and station service hours; The passenger travel parameters include the passenger set. P Passenger vehicle preferences Passenger origin station Terminal Expected boarding time The latest departure time for passengers Passenger numbers And ticket reservation status; The vehicle resource parameters include a set of manually driven buses. A collection of driverless buses Vehicle capacity , special seat capacity Current vehicle location, remaining vehicle range, vehicle availability time, and vehicle operating costs. Driver resource status and unmanned remote monitoring resource status; The real-time operational status parameters include vehicle delay time, number of people waiting at stations, road congestion status, vehicle malfunction status, temporary traffic control status, and sudden passenger demand status.

3. The method for scheduling heterogeneous bus routes considering passenger demand preferences according to claim 2, characterized in that, In step S2, passenger vehicle preference The value can be 0, 0.5, or 1; when When the value is 0, there is a preference for manually driven buses, and manually driven bus trips are prioritized during dispatching. when When =1, driverless buses are preferred, and driverless bus schedules are prioritized during dispatching; when When the ratio is 0.5, there is no obvious preference for manually driven buses and driverless buses, and the scheduling is based on the overall cost. passenger p For vehicle type q Acceptance constraint coefficient Represented as: like When =0, =1; ; like When =1, =1; ; like When =0.5, =1; =1; in, q =H or q =A, H represents manually driven buses, and A represents driverless buses.

4. The method for scheduling heterogeneous bus routes considering passenger demand preferences according to claim 3, characterized in that, In step S3, the special requirements parameters include wheelchair or mobility aid requirements, low-floor vehicle requirements, manual assistance for getting on and off the vehicle requirements, visual or auditory assistance requirements, child or elderly care requirements, space requirements for large luggage or strollers, requirements for passengers in the same group to be in the same vehicle, and priority for on-time arrival. Passengers p Special requirements are represented as vectors =( , ,..., ), will the vehicle v Service capabilities are represented as vectors =( , ,..., When the vehicle v All service capabilities meet the needs of passengers. p When there are special requirements, the special requirement matching coefficient =1, otherwise =0; in, For passengers p The r Special requirements, For vehicles v The r Service capabilities; if passengers p If there is a need for manual assistance in boarding and alighting, and the driverless bus cannot meet this need through station staff, remote customer service, or in-vehicle auxiliary equipment, then the corresponding driverless bus... =0.

5. A method for scheduling heterogeneous bus fleets considering passenger demand preferences according to claim 4, characterized in that, In step S4, when constructing the heterogeneous bus fleet route scheduling optimization model, the following decision variables are set: The variable is 0-1, representing the shift. k Whether by vehicle v implement; The variable is 0-1, representing passengers. p Whether assigned to a shift k ; Indicates train number k The actual departure time; The variable is 0-1, representing passengers. p Whether it failed to be satisfied within the current scheduling period; A 0-1 variable, representing a vehicle. v Whether it will be used as a backup vehicle; The variable is 0-1, representing the shift. k Whether the vehicle type or departure time was adjusted due to special needs.

6. A method for scheduling heterogeneous bus routes considering passenger demand preferences, as described in claim 5, is characterized in that... The objective function of the heterogeneous bus fleet route scheduling optimization model is as follows: ； Where F is the objective function value of the model. Total passenger waiting cost For vehicle operating costs, The penalty cost for passengers not serving is higher for passengers with special needs than for ordinary passengers. Penalty costs for unmet special needs Cost of deviation from scheduled departure times. Costs related to uneven vehicle turnover The cost weight for passenger vehicle preference violation is used to control the importance of preference violation in the overall objective function. ~ These are the weighting coefficients.

7. A method for scheduling heterogeneous bus routes considering passenger demand preferences according to claim 6, characterized in that, The bus route scheduling optimization model for heterogeneous bus fleets should include at least the following constraints: Passenger service constraints: For each passenger p ,satisfy ; Passenger vehicle preference adaptation constraints: ;in, Vehicle type; Special requirements matching constraints: ; Vehicle capacity constraints: ; Special seat capacity constraints: ;in, S For passengers with special seating or accessibility needs; Departure interval constraints: ; Vehicle turnaround constraint: When the same vehicle performs two consecutive shifts, the sum of the arrival time, turnaround time and charging or recharging time of the previous shift shall not exceed the departure time of the subsequent shift. Driver and remote monitoring resource constraints: When manually driven buses are running shifts, they need to be matched with available drivers; when driverless buses are running shifts, they need to meet the upper limit of the number of remotely monitored vehicles.

8. A method for scheduling heterogeneous bus fleets considering passenger demand preferences according to claim 7, characterized in that, In step S5, the specific steps of online rolling rescheduling are as follows: Step S51: When a new passenger with special needs, vehicle malfunction, road delay, abnormal autonomous vehicle service capacity, or insufficient human driver resources are detected, the rolling dispatch window is triggered. Step S52: Calculate the scheduling priority index for passengers with special needs. The formula is as follows: ； in, The latest departure time for passengers. For the current moment, To prevent constants with a denominator of zero, For special needs level, The remaining number of flights that can serve passengers. The number of shifts remaining in the dispatch window. This is a correction item for the number of passengers in the same group. to Priority index weights; Step S53: Insert and match passengers with special needs according to the scheduling priority index from high to low; Step S54: If there is available capacity on existing routes that meets the conditions for matching passengers' rigid demand, demand preferences and special needs, then passengers will be assigned to those routes and the corresponding seats or service capacity will be locked. Step S55: If the existing schedule does not meet the service conditions, then execute the following strategies in sequence: vehicle type exchange, departure time fine-tuning, backup vehicle deployment, platform service personnel coordination, and manual customer service confirmation. Step S56: Under the premise of satisfying passenger demand and preference constraints and safe operation, output the rescheduling results and update the vehicle, passenger and station status.

9. A method for scheduling heterogeneous bus fleets considering passenger demand preferences according to claim 8, characterized in that, In step S55, the specific strategy for vehicle type exchange is as follows: When passengers prefer vehicles When the number of passengers is 0 and the currently available bus is an autonomous bus, the system will prohibit the passenger from being assigned to that autonomous bus and will prioritize adjusting subsequent manually driven buses to the direction of the passenger's stop. When passengers prefer vehicles When the number of passengers is 1 and the currently available bus is a manually driven bus, the system will prohibit the passenger from being assigned to that manually driven bus and will prioritize matching the passenger to an autonomous bus or calling up an autonomous backup vehicle. When passengers prefer vehicles When the value is 0.5, the system selects the vehicle type with the lowest overall cost and that meets the special requirements from both manually driven buses and driverless buses. When a special need conflicts with a passenger's vehicle preference, the system generates a conflict alert and prioritizes resolving the conflict through platform service personnel, accessible facilities, backup vehicles, or customer service confirmation. If the conflict cannot be resolved, the passenger is marked as a candidate for manual intervention.

10. A method for scheduling heterogeneous bus routes considering passenger demand preferences, as described in claim 9, is characterized in that... In step S5, a hierarchical hybrid solution algorithm is used for solving the problem. The specific steps are as follows: The upper-level model determines the vehicle type, departure time, and number of spare vehicles reserved for each shift; The lower-level model determines the matching relationship between passengers and schedules, the allocation of seats for special needs, and the vehicle turnover relationship; The static reservation requirements are solved using mixed integer programming to generate an offline baseline scheduling scheme; For newly added requests in real time, rolling time-domain optimization and adaptive large neighborhood search algorithms are used for rapid repair. If the rolling optimization results do not converge within a preset time, a greedy insertion rule based on a special demand priority index is used to generate an emergency scheduling plan.

Images