CN-122022240-A - Bus driver collaborative wheel value task allocation method and equipment under networking condition

CN122022240ACN 122022240 ACN122022240 ACN 122022240ACN-122022240-A

Abstract

The invention discloses a bus driver collaborative wheel value task allocation method and device under a networking condition, which comprises the following steps of collecting driver history attendance data, generating a large-scale abscission scene set through TimeGAN networks based on the driver history attendance data, carrying out dimension reduction mapping and reduction on the large-scale abscission scene set to generate a representative condition Jing Jige, constructing a driver regional shift network diagram and a motor personnel network diagram based on the representative scene set, constructing a distributed robust optimization model based on the driver regional shift network diagram and the motor personnel network diagram, decomposing a solved minimized total cost problem into a regional shift sub-problem and a motor personnel configuration sub-problem, outputting a regional shift decision and a motor personnel relief decision, and obtaining a bus driver collaborative wheel value task allocation scheme. The invention can solve the problems of unbalanced shift time and frequent random absence of the bus driver in a high-efficiency collaborative planning way, and optimize the execution degree of vehicle dispatching and driver dispatching.

Inventors

WU WEITIAO
SHI TINGYAN
HUANG FUXIANG
ZHANG JIYONG
ZHENG YUAN
TANG TANG
WU YINGZI

Assignees

华南理工大学
广州市城市规划勘测设计研究院有限公司

Dates

Publication Date: 20260512
Application Date: 20251217

Claims (10)

1. The bus driver cooperative wheel value task allocation method under the networking condition is characterized by comprising the following steps of: s1, collecting historical attendance data of a driver, and generating a large-scale absenteeism scene set through a TimeGAN network based on the historical attendance data of the driver; S2, performing dimension reduction mapping and reduction on the large-scale absenteeism scene set to generate a representative emotion Jing Jige; S3, based on a representative scene set, embedding a train break mode, a backlog scene, a driver work preference and a line crossing condition into a network diagram path by adopting a time expansion network TEN, and constructing a driver regional shift network diagram and a motor personnel network diagram; S4, modeling a regional shift process and relief processes to obtain a distributed robust optimization model based on the driver regional shift network diagram and the motor personnel network diagram by taking the minimum total cost as an objective function; and S5, solving the distributed robust optimization model, decomposing the solved minimized total cost problem into a regional shift sub-problem and a motor personnel configuration sub-problem, and outputting a regional shift decision and a motor personnel relief decision to obtain a public transport driver collaborative wheel value task allocation scheme.
2. The method for allocating cooperative wheel tasks of a bus driver under a networking condition according to claim 1, wherein the collecting the driver history attendance data and generating the large-scale absenteeism scene set through TimeGAN network based on the driver history attendance data comprises: Collecting historical attendance data of a driver, constructing a time sequence sample matrix of attendance of the historical driver according to the historical attendance data of the driver, generating a countermeasure network through TimeGAN to perform feature learning on the time sequence of attendance of the historical driver and synthesizing a absent time sequence to form a simulated absent scene; Through repeated sampling of historical attendance data of drivers, a large number of absent scenes consistent with the statistical characteristics of historical samples are generated through TimeGAN generation countermeasure networks, the generated sequences of different drivers are synchronously spliced in the time dimension to construct multiple-driver absent scenes, and N absent scenes are generated to form a large-scale absent scene set.
3. The method for allocating cooperative round-robin tasks for bus drivers under a networking condition according to claim 1, wherein the step of performing dimension-reduction mapping and reduction on the large-scale absences set to generate a representative scenario set comprises: performing dimension reduction on a time sequence feature matrix of the large-scale absenteeism scene set by a principal component analysis method to obtain principal component feature representation of a scene sample in a principal component space of PCA; And applying a K-Means clustering algorithm to the scene samples in the PCA principal component space, dividing the scene samples into K mode categories, carrying out mean value synthesis and discretization on the scene samples in each mode category, and generating a representative absences scene of the category to form a representative emotion Jing Jige.
4. The method for allocating cooperative wheel tasks of a bus driver under a networking condition according to claim 1, wherein the step of embedding a train break mode, a backlog scene, a driver work preference and a line crossing condition into a network map path by using a time expansion network TEN based on a representative scene set to construct a driver regional shift network map and a motor personnel network map comprises the following steps: Based on a time expansion network TEN, a driver regional shift network diagram is constructed according to the shift task and the rest day arrangement of a driver in a shift period in a representative scene set, a time layer of the driver regional shift network diagram takes calendar days as a unit, a shift network of each day is layered according to a shift mode and relief networks according to a backdrop scene, nodes of the driver regional shift network diagram correspond to states of work, rest or relief, only transfer arcs permitted by regulations are reserved in adjacent days, wherein the transfer arcs comprise a start arc, a work arc, a rest arc, a relief arc and a stop arc, an arc capacity limit corresponds to the available quantity of the driver or the task demand quantity, and the arc cost integrates man-hour overrun punishment, shift frequency punishment and overline qualification difference; A crew network map is constructed based on the driver zone shift network map for relief scheduling crew in the event of a backout.
5. The method for allocating cooperative wheel tasks of a bus driver under a networking condition according to claim 4, wherein modeling the regional shift process and relief process to obtain a distributed robust optimization model based on the driver regional shift network diagram and the motor personnel network diagram with the minimum total cost as an objective function comprises: And constructing an objective function of the distributed robust optimization model by taking the minimum total cost as a target, wherein the total cost comprises regional shift cost, motor personnel configuration cost and relief cost, and modeling the regional shift process and the relief process based on a driver regional shift network diagram and a motor personnel network diagram to obtain the distributed robust optimization model.
6. The method for allocating cooperative wheel tasks of a bus driver under a networking condition according to claim 5, wherein the objective function is expressed as: ; Wherein, the Indicating regional shift costs under shift decision X, Representing the cost of the configuration of the motor personnel at the motor personnel scale decision Z, S representing the set of all absences S, Representing relief costs for all absences s under shift decision X and relief decision Y.
7. The method for allocating cooperative wheel tasks of a bus driver under a networking condition according to claim 6, wherein the method for allocating cooperative wheel tasks of a bus driver by solving the distributed robust optimization model, decomposing the minimized total cost problem into a regional shift sub-problem and a motor personnel configuration sub-problem, outputting a regional shift decision and a motor personnel relief decision comprises the following steps: the method comprises the steps of solving a regional shift sub-problem based on a hyper-heuristic algorithm selected by a sequence by taking the minimum regional shift cost as a target, and generating a regional shift table by solving the minimum regional shift cost; The motor personnel configuration sub-problem comprises a motor personnel scale problem and a relief decision problem, the motor personnel scale problem is solved through an iterative local search algorithm to obtain the optimal motor personnel scale, the sum of motor personnel configuration cost and relief cost is minimized, and a relief decision is generated according to the motor personnel scale.
8. The method for allocating cooperative wheel tasks of a bus driver under a networking condition according to claim 7, wherein the solving the problem of the size of the motor personnel by the iterative local search algorithm to obtain the optimal size of the motor personnel, aiming at minimizing the sum of the configuration cost of the motor personnel and the cost relief, generating relief decision according to the size of the motor personnel, comprises: Determining an optimal motor personnel scale interval according to the upper and lower bounds of the sum of motor personnel configuration cost and relief cost corresponding to the motor personnel scale; and solving in an optimal motor personnel scale interval by adopting an iterative local search algorithm, determining the optimal motor personnel scale of the motor personnel configuration sub-problem by calculating the minimum value of the sum of motor personnel configuration cost and relief cost, and determining relief decision according to the optimal motor personnel scale.
9. A computer device comprising a processor and a memory for storing a program executable by the processor, wherein the processor, when executing the program stored in the memory, implements a method for assigning co-operating wheel tasks for a bus driver under a networking condition as claimed in any one of claims 1 to 8.
10. A storage medium storing a program which, when executed by a processor, implements a method for assigning co-operating wheel tasks for a bus driver under a networking condition as claimed in any one of claims 1 to 8.

Description

Bus driver collaborative wheel value task allocation method and equipment under networking condition Technical Field The invention relates to the technical field of dispatching control of bus driver operation, in particular to a bus driver cooperative wheel value task allocation method and device under a networking condition. Background Public transportation is an important social public service and basic public service meeting basic travel demands of people, and the core of public transportation operation organization is to reasonably allocate transportation resources (such as buses and drivers) under the condition of limited resources so as to complete established operation tasks and maximize resource utilization rate. The reasonable allocation process of the transport capacity resources is generally divided into three sub problems, namely a vehicle scheduling problem, a driver scheduling problem and a driver shift problem, and the completion condition of the driver shift directly determines the execution degree of the vehicle scheduling and the driver scheduling due to the dependence of planning tasks, so that the method is a key bottleneck for restricting the cost reduction and efficiency enhancement of enterprises, and the effective implementation of the driver shift is not separated from the reasonable utilization of the driver resources. Driver resources are a core element of public transportation systems, and their cost is usually a significant proportion of the total operating cost of the enterprise. Public transport enterprises mostly rely on government subsidies, and budget restrictions exist in operation. Therefore, under the condition of limited driver resources, how to comprehensively and reasonably schedule the resources, realize the cost reduction and efficiency increase of operation and ensure safety, and become an important problem to be solved urgently for improving the intelligent operation management level and the service quality of buses. The driver shift is the last stage of the operation organization dispatch, aims to arrange the driving shift task of the whole shift period (one week, one month or longer) for the driver, realizes the minimization of the cost under the condition of meeting the labor regulation, directly determines the execution degree of the vehicle dispatch and the driver dispatch in the completion condition, and is a key bottleneck for restricting the cost reduction and efficiency enhancement of enterprises. Disclosure of Invention In order to solve the technical problems in the prior art, the invention provides a method and equipment for allocating the collaborative wheel tasks of a bus driver under the networking condition, which are used for defining a driver regional shift network diagram and a motor personnel network diagram, embedding a shift pattern, a backdrop scene, a driver work preference and a overline condition into a network diagram path, establishing a scene-based distributed robust optimization model, and cooperatively formulating a shift-rest strategy of a tactical layer, a motor personnel configuration scale and relief response of an operation layer under the worst condition of coping with uncertain requirements and personnel state disturbance so as to ensure that the system can keep feasible and perform optimally under all possible disturbances, and optimize the problems of unbalanced shift man-hour of the bus driver, frequent random absences and the like. The first aim of the invention is to provide a bus driver collaborative wheel value task allocation method under the networking condition. A second object of the present invention is to provide a computer device. The first object of the present invention can be achieved by adopting the following technical scheme: a bus driver cooperative wheel value task allocation method under a networking condition comprises the following steps: s1, collecting historical attendance data of a driver, and generating a large-scale absenteeism scene set through a TimeGAN network based on the historical attendance data of the driver; S2, performing dimension reduction mapping and reduction on the large-scale absenteeism scene set to generate a representative emotion Jing Jige; S3, based on a representative scene set, embedding a train break mode, a backlog scene, a driver work preference and a line crossing condition into a network diagram path by adopting a time expansion network TEN, and constructing a driver regional shift network diagram and a motor personnel network diagram; S4, modeling a regional shift process and relief processes to obtain a distributed robust optimization model based on the driver regional shift network diagram and the motor personnel network diagram by taking the minimum total cost as an objective function; and S5, solving the distributed robust optimization model, decomposing the solved minimized total cost problem into a regional shift sub-problem and a motor personnel configuration sub