CN116307646B - One-rail double-vehicle scheduling method based on two-stage dynamic partitioning algorithm - Google Patents

One-rail double-vehicle scheduling method based on two-stage dynamic partitioning algorithm Download PDF

Info

Publication number
CN116307646B
CN116307646B CN202310580393.7A CN202310580393A CN116307646B CN 116307646 B CN116307646 B CN 116307646B CN 202310580393 A CN202310580393 A CN 202310580393A CN 116307646 B CN116307646 B CN 116307646B
Authority
CN
China
Prior art keywords
task
mobile
sequence
movement
tasks
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202310580393.7A
Other languages
Chinese (zh)
Other versions
CN116307646A (en
Inventor
王筱圃
袁丁
张弢
张庆
宋玉雪
岳亚莉
钟智敏
刘伟
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hkust Intelligent Internet Of Things Technology Co ltd
Original Assignee
Hkust Intelligent Internet Of Things Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hkust Intelligent Internet Of Things Technology Co ltd filed Critical Hkust Intelligent Internet Of Things Technology Co ltd
Priority to CN202310580393.7A priority Critical patent/CN116307646B/en
Publication of CN116307646A publication Critical patent/CN116307646A/en
Application granted granted Critical
Publication of CN116307646B publication Critical patent/CN116307646B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T10/00—Road transport of goods or passengers
    • Y02T10/10—Internal combustion engine [ICE] based vehicles
    • Y02T10/40—Engine management systems

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Human Resources & Organizations (AREA)
  • General Physics & Mathematics (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Geometry (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Educational Administration (AREA)
  • General Business, Economics & Management (AREA)
  • Quality & Reliability (AREA)
  • Development Economics (AREA)
  • Computer Hardware Design (AREA)
  • Operations Research (AREA)
  • Game Theory and Decision Science (AREA)
  • Marketing (AREA)
  • Health & Medical Sciences (AREA)
  • Tourism & Hospitality (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to the technical field of rail guided vehicle dispatching, and discloses a one-rail double-vehicle dispatching method based on a two-stage dynamic partitioning algorithm, which comprises the following steps: partitioning the mobile task, wherein the optimization target is to minimize the execution time difference of the mobile task set between the two partitions, and determining the execution sequence of the mobile task of the two partitions; obtaining an optimal moving task sequence of the cross-region moving task through a second-stage scheduling algorithm based on an improved simulated annealing algorithm; the self-adaptive two-stage real-time scheduling method can improve the optimizing speed of the algorithm and reduce the response time of the algorithm.

Description

One-rail double-vehicle scheduling method based on two-stage dynamic partitioning algorithm
Technical Field
The invention relates to the technical field of rail guided vehicle dispatching, in particular to a one-rail double-vehicle dispatching method based on a two-stage dynamic partitioning algorithm.
Background
Rail guided vehicles (Rail Guided Vehicle, RGV), also known as rail shuttle vehicles.
The existing scheduling schemes of RGV mainly include partition scheduling, regular scheduling and regular real-time scheduling, which are described in detail below.
Partition type scheduling: the layout requirements are high, and the method is suitable for symmetrical layout, namely RGV stations are arranged on two sides, but the application effect under asymmetrical layout is poor; the utilization of RGVs is affected by the amount of tasks after partitioning, resulting in large differences in RGVs duty cycle.
And (3) regular scheduling: custom development is carried out according to different layout requirements; the scheduling is not flexible enough, and an unreasonable scheduling phenomenon exists, so that RGV movement is wasted; the service scenes to be considered are more, and errors can occur during development due to incomplete service scene consideration.
Conventional real-time scheduling: the overall scheduling efficiency is poor, and the method is not suitable for a short-time large-scale operation mode.
The invention aims to solve the problem of one-rail double-vehicle scheduling of large-scale operation tasks in a short time under a multi-layout scene, namely, how to scientifically schedule two RGVs in real time under the layout scene of symmetry, asymmetry and the like in which RGVs have a large number of moving tasks, and improves the RGV scheduling efficiency so as to reduce the total time for completing all the moving tasks.
Disclosure of Invention
In order to solve the technical problems, the invention provides a one-rail double-vehicle scheduling method based on a two-stage dynamic partitioning algorithm.
In order to solve the technical problems, the invention adopts the following technical scheme:
A one-rail double-vehicle scheduling method based on a two-stage dynamic partitioning algorithm is used for optimizing the total time of moving tasks of two RGVs on the same track and outputting a corresponding optimal moving task sequence; the sequence of the mobile tasks is the execution sequence of the mobile tasks; the one-rail double-vehicle scheduling method comprises the following steps:
s1, partitioning a mobile task, wherein an optimization target is to minimize the execution time difference of a mobile task set of two partitions, and the execution sequence of the mobile task of the two partitions is determined; the starting point and the end point of each mobile task are called RGV point positions, and specifically include:
s11: according to a certain RGV point positionDividing the track to form a left partition and a right partition>Representing a set of all RGV points; the moving task whose starting point and end point are both located in the left partition is marked as a moving task set +.>The method comprises the steps of carrying out a first treatment on the surface of the The mobile task whose starting point and end point are both located in the right partition is marked as the mobile task set +.>；
S12: by first-stage scheduling algorithm pair based on ant colony-genetic-particle swarm algorithm and />The mobile tasks in the left partition are sequenced and the mobile task sequences of the left partition are respectively output>And the movement task sequence of the right partition +.>，/> and />Respectively make- > and />The total moving distance of the middle moving task is shortest;
s13: calculating current RGV point positions respectivelyCorresponding-> and />Execution time of-> and />Calculating an execution time difference；
S14: repeating steps S11 to S13 until completionSelecting all RGV points; selecting the smallest execution time difference +.>The corresponding RGV-bit is taken as the optimal partition-bit +.>And the smallest execution time difference +.>Corresponding-> and />Respectively used as the optimal moving task sequence of the left partition and the optimal moving task sequence of the right partition;
s2, obtaining an optimal moving task sequence of the cross-region moving task through a second-stage scheduling algorithm based on an improved simulated annealing algorithm; wherein the transregional moving task is that the starting point and the end point are respectively positioned at the optimal partition point positionsA movement task on both sides; when the second stage scheduling algorithm is executed, only one RGV can execute the mobile task at the same time, and the RGV which does not execute the mobile task is avoided from being out of the mobile area of the RGV which executes the mobile task.
Specifically, in step S12, the moving task sequence of the left partition is outputted by the first-stage scheduling algorithm based on the ant colony-genetic-particle swarm algorithmAnd the movement task sequence of the right partition +.>When the method comprises the following steps:
S131: randomly selecting the left partition and the right partition respectivelyA sequence of mobile tasks, calculating left partition +.>The respective fitness value, right partition +.>The respective fitness values of the respective mobile task sequences;
s132: respectively selecting from left partition and right partitionA moving task sequence B with the shortest total moving distance,calculating the fitness value corresponding to each moving task sequence, and taking the fitness value of each moving task sequence B as the initial pheromone concentration of any two adjacent moving tasks in the moving task sequence B during ant colony search;
s133: taking the minimum value of the adaptation degree value of the moving task sequence of the left partition in the step S131 as the individual extremum of the left partitionThe corresponding movement task sequence is +.>The method comprises the steps of carrying out a first treatment on the surface of the Taking the minimum value of the right partition moving task sequence fitness value in the step S131 as the individual extremum +.>The corresponding movement task sequence is +.>The method comprises the steps of carrying out a first treatment on the surface of the Calculating the fitness values of all the mobile task sequences of the left partition and the right partition, and taking the minimum fitness value as the global extremum +.>And global extremum of right partition->Global extremum->The corresponding sequence of movement tasks is marked +.>Global extremum->The corresponding sequence of movement tasks is marked +. >；
S134: initializing the iteration times t=0;
s135: in the left partitionRandomly selecting +.>A plurality of movement tasks are allocated to ∈>Ants; for->Ant only, will be->The mobile task allocated by ant is placed in the corresponding task sequence solution set to calculate +.>Only ants are moved by the current task->To next movement task->State transition probability>And according to->Acquire next movement task->The movement task is further->Is placed at->Only the task sequence corresponding to ants is concentrated; when all the mobile tasks to be orderedAfter the task sequence solution set, the +.>Mobile task sequence explored by ants only ∈>The method comprises the steps of carrying out a first treatment on the surface of the Computing the movement task sequence->Is>；
In the right partitionRandomly selecting +.>A plurality of movement tasks are allocated to ∈>Ants; for->Ant only, will be->The mobile task allocated by ant is placed in the corresponding task sequence solution set to calculate +.>Only ants are moved by the current task->To next movement task->State transition probability>And according to->Acquire next movement task->The movement task is further->Is placed at->Only the task sequence corresponding to ants is concentrated; when the mobile tasks to be ordered are all placed in the task sequence solution set, the +. >Mobile task sequence explored by ants only ∈>The method comprises the steps of carrying out a first treatment on the surface of the Computing the movement task sequence->Is>；
S136, for left partition: moving task sequencesAnd->Performing cross operation to obtain a mobile task sequence，/>And->Performing the cross operation again to obtain a moving task sequence +.>And then (2) to->Performing mutation operation to obtain a mobile task sequence->Calculating the execution movement task sequence->Is>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Smaller, accept the movement task sequence +.>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Not get smaller, then->The moving task sequence corresponding to ants only is still；
For the right partition: moving task sequencesAnd->Performing cross operation to obtain a moving task sequence +.>，/>And->Performing the cross operation again to obtain a moving task sequence +.>And then (2) to->Performing mutation operation to obtain a mobile task sequence->Calculating the execution movement task sequence->Is>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Smaller, accept the movement task sequence +.>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Not get smaller, then->The moving task sequence corresponding to ants only is still；
S137: after the moving task sequences of all ants in the left partition and the moving task sequences of all ants in the right partition are obtained, individual extremum values of ant colony in the left partition and the right partition are obtained 、/>Mobile task sequence corresponding to individual extremum、/>Global extremum->、/>A mobile task sequence corresponding to the global extremum +.>、Updating;
s138: iterative equation by ant colony pheromoneUpdating the information concentration between the mobile tasks; /> and />Representing the movement task +.for the t-th and t+1-th iterations, respectively>To move task->Pheromone concentration of (2)，/>Representing pheromone volatilization factors; />For the t-th iteration, move task +.>To move task->New pheromone concentration; the number of iterations t=t+1;
s139: repeating steps S135 to S138 until the number of repetitions reaches the set maximum number of iterations；
S1310: taking the moving task sequence corresponding to the individual extremum of the left partition as the optimal moving task sequence of the left partition; and taking the moving task sequence corresponding to the individual extremum of the right partition as the optimal moving task sequence of the right partition.
Specifically, state transition probabilities，Indicate->Ant only from the movement task->A set of mobile tasks that can be accessed; />Representing the slave movement task->To move task->Is>Visibility of (i.e.)>，/>Representing movement task->Starting point to move task->Is the distance of the end point of (2); />Representing the relative importance of the track; />Indicating the relative importance of visibility.
Specifically, in step S13, a sequence of movement tasks to be operatedAnd a specific movement task sequence->When the cross operation is performed, the task sequence is moved>、/>Each comprising M mobile tasks; randomly generating two non-equal positive integers +.>、，/>Get the movement task sequence->The upper index is +.>To->Between mobile task segments->Move task sequence->Go up and->The movement task segment at the same position is removed and +.>Splicing the rest mobile task fragments to obtain a mobile task fragment +.>Finally, move task segment->Splice to Mobile task segment->Terminal, and thus get the sequence of tasks with movement +.>Corresponding new movement task sequence->。
Specifically, in step S13, a sequence of movement tasks to be operatedWhen the mutation operation is performed, the task sequence is moved、/>All comprising M movement tasks, randomly generating two unequal positive integers +.>、/>，/>Will->Index +.> and />Transposition is performed on the two mobile tasks of (2) to obtain AND +.>Corresponding new movement task sequence->。
Specifically, in step S13, the left partition moves the fitness value of the task sequenceThe method comprises the following steps:
；
wherein Moving the total number of tasks for the left partition; />Is->Start of individual movement tasks, +.>Is->End point of each movement task- >Is-> and />Is a distance of (2);
fitness value of right partition movement task sequenceThe method comprises the following steps:
；
wherein The total number of tasks is moved for the right partition.
Specifically, the second stage scheduling algorithm based on the improved simulated annealing algorithm specifically comprises the following steps:
s21, initializing parameters of a second-stage scheduling algorithm: setting an initial temperatureStop temperature->Temperature decay Rate->Probability of mutation->The method comprises the steps of carrying out a first treatment on the surface of the Setting the number of internal cycles per unit temperature +.>The method comprises the steps of carrying out a first treatment on the surface of the Setting the current temperature +.>；
S22: randomly distributing to-be-sequenced cross-region mobile tasks to twoObtaining an initial solution; solving, namely a moving task sequence of a cross-region moving task;
s23: judging the current temperatureWhether or not is greater than->The method comprises the steps of carrying out a first treatment on the surface of the If yes, go to step S24, if no, go to step S210;
s24: constructing four ways of searching the neighborhood solution, and searching the neighborhood solution for the initial solution according to the ways of executing the four ways of searching the neighborhood solution with equal probability according to the internal circulation times LL at the preset unit temperature; then according to the settingThe rule judges to reserve new solutions or old solutions, and a plurality of solutions obtained through internal circulation at the current temperature are obtained;
s25: obtaining a plurality of solutions through internal circulation at the current temperature, calculating respective fitness values, taking two solutions with the minimum fitness values for performing cross operation, and calculating the fitness values of new solutions obtained after the cross operation; judging whether the fitness value of the new solution is smaller than the minimum fitness value of each solution at the current temperature, if so, taking the new solution as the optimal solution at the current temperature; if not, taking the solution with the minimum fitness value at the current temperature as the optimal solution at the current temperature;
S26: performing cross operation on the optimal solution at the current temperature and the optimal solution at the previous temperature, calculating the fitness value of the new solution after the cross, judging whether the fitness value of the new solution is smaller than the fitness value of the optimal solution at the current temperature, and if so, taking the new solution as the optimal solution at the current temperature; if not, keeping the optimal solution at the current temperature unchanged;
s27: taking the optimal solution at the current temperature, and according to the set variation probabilityExecuting the mutation operation, calculating the fitness value of the new solution after executing the mutation operation, judging whether the fitness value of the new solution is smaller than the fitness value of the optimal solution at the current temperature, and if so, taking the new solution as the optimal solution at the current temperature; if not, keeping the optimal solution at the current temperature unchanged;
s28: judging whether the fitness value of the optimal solution at the current temperature is smaller than that of the optimal solution at the previous temperature, if so, keeping the optimal solution at the current temperature as a final optimal solution; if not, taking the optimal solution at the previous temperature as the optimal solution finally obtained at the current temperature;
s29: performing a cooling process, i.e.Jumping to the step S24, and taking the optimal solution finally obtained at the current temperature as the initial solution in the step S24;
S210: for the current temperatureAnd decoding the finally obtained optimal solution to obtain an optimal movement task sequence of the transregional movement task.
Specifically, the metapolis rule is set as:a sequence of mobile tasks representing an old solution, +.>A sequence of mobile tasks representing a new solution, +.>A sequence of mobile tasks representing the current solution, +.>Fitness value representing old solution, +.>Fitness value representing the new solution, +.>An fitness value representing a current solution; according to the calculation mode of fitness value function, find +.> and />Difference>I.e. +.>The method comprises the steps of carrying out a first treatment on the surface of the If->Then accept the new solution, i.e. the current solution +.>，/>The method comprises the steps of carrying out a first treatment on the surface of the If condition A is satisfied, a new solution is accepted, i.e.)>，/>The method comprises the steps of carrying out a first treatment on the surface of the If the condition A is not satisfied, the old solution is retained, i.e. +.>，The method comprises the steps of carrying out a first treatment on the surface of the Wherein condition a is: />And->，/>Representing random numbers uniformly distributed within the interval (0, 1).
Specifically, when the cross operation is performed in the second stage scheduling algorithm: mobile task sequence to be operatedAnd a specific movement task sequence->When the cross operation is performed, the task sequence is moved>、/>All comprising M movement tasks, three non-equal positive integers are randomly generated +.>、/>、/>，/>Get the movement task sequence->Go up the index->To the point ofIs->Index number->To->Is- >The method comprises the steps of carrying out a first treatment on the surface of the Move task sequence->Top and Mobile task segment->、/>The same-position mobile task segment is removed and +.>The rest segments are spliced to obtain a mobile task segment +.>Finally, move task segment->、/>Splice to Mobile task segment->Terminal, get the sequence of movement tasks to be operated +.>Corresponding new movement task sequence->。
Specifically, in the second stage scheduling algorithm, the fitness value of the mobile task sequence：
；
wherein ,for the total number of cross-zone movement tasks, +.>Is->Start of individual movement tasks, +.>Is->End point of each movement task->Is-> and />Distance of->Is->R of individual mobile task assignmentsNumbering of GV; />Indicate->The RGV point position set which is required to pass through by the mobile task and the current position of another RGV have intersection or not, if yes, 1 is taken, and if not, 0 is taken;indicate->The union of the RGV point position set required to pass by the mobile task and the RGV current position for executing the i mobile task to the i mobile task starting point is provided with or not with the other RGV current position, wherein the intersection is 1, and the non-intersection is 0; />Indicate->Whether the direction of each mobile task is matched with the relative position of the RGV for executing the ith mobile task or not, taking 1 in a matching way, and taking 0 in a non-matching way; / >Representing execution of +.>Real-time location of RGVs of the individual mobile tasks; />Indicating that no->Real-time location of RGV for each mobile task.
First, theWhether the direction of the movement task matches the relative position of the RGV performing the ith movement task or not, meansIf the direction of the ith movement task is left, the RGV performing the ith movement task is located on the left side, then it is called matching, otherwise it is called unmatched. The relative position of the RGVs may be represented by the numbering of the RGVs.
Compared with the prior art, the invention has the beneficial technical effects that:
the one-rail double-vehicle scheduling method based on the two-stage dynamic partitioning algorithm is innovative in algorithm design, and can improve the optimizing speed and reduce the response time. Specifically, the method is suitable for different layout scenes, can realize global search of solutions, can output a globally optimal scheduling scheme, and avoids invalid movement of RGVs; the two-stage scheduling algorithm is designed for solving, so that the scheduling logic is simple, the calculated amount is small, the output scheduling scheme is fast, the algorithm solving complexity is reduced, the algorithm response rate is improved, and the method is suitable for a short-time large-scale operation mode; dynamic partitioning operation can reduce RGV road leaving distance and invalid moving distance; RGV cross-zone operation is reduced, and RGV parking waiting time can be reduced.
Drawings
FIG. 1 is a flow chart of a scheduling method of the present invention.
Detailed Description
A preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.
The one-rail double-vehicle scheduling method comprises two stages. The movement tasks of the RGV include partition movement tasks and cross-zone movement tasks; the partition moving task refers to a moving task with a starting point and an end point both positioned in the same partition; a cross-zone movement task refers to a movement task in which the start point and the end point are both located in different partitions.
The first stage: and dynamically partitioning according to the starting point and the ending point of each mobile task in the mobile task set, wherein the optimization target is to minimize the time difference of executing the mobile tasks by two RGVs, and outputting the optimal mobile task sequence of the partitioned mobile tasks. The method specifically comprises the following steps:
the starting point and the end point of each mobile task are called RGV point positions; after the track is divided according to a certain RGV point position, a left partition and a right partition are naturally formed,the moving task with the starting point and the end point both in the left partition is marked as a left partition moving task setThe moving task with the starting point and the end point both in the right partition is marked as a right partition moving task set +.>. The first stage scheduling algorithm is used for respectively performing the corresponding steps >Mobile task in->The movement tasks of the RGVs are ordered such that the total movement distance of the RGVs is minimized. Then calculates two RGVs to execute +.>Middle movement task and +.>The time difference of the moving task.
After partitioning by different RGV points, repeating the above steps to obtain time difference corresponding to each RGV point, and taking one RGV point with the smallest time difference as the optimal partition pointAnd the moving task sequence of the left partition at this time is +.>As the best movement task sequence for the left partition, the movement task sequence for the right partition is +.>As the best sequence of movement tasks for the right partition.
And a second stage: after the screening of the first stage, the rest mobile tasks all need to cross the set optimal partition point positionsNamely, the tasks are all cross-region moving tasks; thus, when an RGV is performing a movement task at this stage, two RGVs cannot be performed simultaneously, i.e., when one RGV is after performing the allocated movement task, it is required to avoid the movement task beyond the road segment where the other RGV is performing the movement task (e.g., beyond the start point or beyond the end point of the other RGV) so as to ensure that no collision occurs. Specifically, the optimal allocation scheme of the cross-region mobile task is searched through a second-stage scheduling algorithm, so that an optimal mobile task sequence of the cross-region mobile task is obtained, and the optimization target is that the total completion time of all the cross-region mobile tasks is the shortest.
The first stage scheduling algorithm performs two-layer coding on the mobile task sequence, including:
first layer coding:a starting point for executing a certain movement task;
second layer coding:the end point of a certain movement task is performed.
In the first stage scheduling algorithm, the adaptability value of the left partition moving task sequence：
；
Fitness value of right partition movement task sequenceThe method comprises the following steps:
。
wherein For the total number of left partition move tasks, +.>Moving the total number of tasks for the right partition; />Is->Start of individual movement tasks, +.>Is->End point of each movement task->Is-> and />Is a distance of (3).
The second stage scheduling algorithm performs four-layer coding on the mobile task sequence, including:
first layer coding:a starting point for executing a certain movement task;
second layer coding:executing the end point of a certain mobile task;
third layer coding:numbering;
fourth layer coding:a direction of movement.
In the second stage scheduling algorithm, the mobileFitness value of task sequence：
；
wherein ,for the total number of cross-zone movement tasks, +.>Is->Start of individual movement tasks, +.>Is->End point of each movement task->Is-> and />Distance of->Is->A number of RGV assigned by each mobile task; />Indicate->RGV point position set passed by each mobile task and another RGV current position If the intersection exists, taking 1 if the intersection exists, and taking 0 if the intersection exists;indicate->The union of the RGV point position set required to pass by the mobile task and the RGV current position for executing the i mobile task to the i mobile task starting point is provided with or not with the other RGV current position, wherein the intersection is 1, and the non-intersection is 0; />Indicate->Whether the direction of each mobile task is matched with the number of the RGV for executing the ith mobile task or not, taking 1 in a matching way, and taking 0 in a non-matching way; />Representing execution of +.>Real-time location of RGVs of the individual mobile tasks; />Indicating that no->Real-time location of RGV for each mobile task.
First, theWhether the direction of the ith movement task matches the relative position of the RGV performing the ith movement task means that if the direction of the ith movement task is left, the RGV performing the ith movement task is located on the left side, and otherwise, it is called matching. The relative position of the RGVs may be represented by the numbering of the RGVs.
Specifically, the mobile task sequence of the left partition is output through a first-stage scheduling algorithmAnd the movement task sequence of the right partition +.>When the method comprises the following steps:
s131: randomly selecting the left partition and the right partition respectivelyA sequence of mobile tasks, calculating left partition +. >The respective fitness value, right partition +.>The respective fitness values of the respective mobile task sequences; in the present embodiment。
S132: respectively selecting from left partition and right partitionCalculating the fitness value corresponding to each moving task sequence by the moving task sequence with the shortest total moving distance, and taking the fitness value of each moving task sequence B as the initial pheromone concentration of any two adjacent moving tasks in the moving task sequence B during ant colony search; in this embodiment +.>。
S133: taking the minimum value of the adaptation degree value of the moving task sequence of the left partition in the step S131 as the individual extremum of the left partitionThe corresponding movement task sequence is +.>The method comprises the steps of carrying out a first treatment on the surface of the Taking the minimum value of the right partition movement task sequence fitness value in step S131 as the rightIndividual extremum->The corresponding movement task sequence is +.>The method comprises the steps of carrying out a first treatment on the surface of the Calculating the fitness values of all the mobile task sequences of the left partition and the right partition, and taking the minimum fitness value as the global extremum +.>And global extremum of right partition->Global extremum->The corresponding sequence of movement tasks is marked +.>Global extremum->The corresponding sequence of movement tasks is marked +.>。
S134: the number of iterations t=0 is initialized.
S135: in the left partitionRandomly selecting +. >A plurality of movement tasks are allocated to ∈>Ants; for->Ant only, will be->The mobile task allocated by ant is placed in the corresponding task sequence solution set to calculate +.>Only ants are moved by the current task->To next movement task->State transition probability>And according to->Acquire next movement task->The movement task is further->Is placed at->Only the task sequence corresponding to ants is concentrated; when the mobile tasks to be ordered are all placed in the task sequence solution set, the +.>Mobile task sequence explored by ants only ∈>The method comprises the steps of carrying out a first treatment on the surface of the Computing the movement task sequence->Is>。
In the right partitionRandomly selecting +.>A plurality of movement tasks are allocated to ∈>Ants; for->Ant only, will be->The mobile task allocated by ant is placed in the corresponding task sequence solution set to calculate +.>Only ants are moved by the current task->To next movement task->State transition probability>And according to->Acquire next movement task->The movement task is further->Is placed at->Only the task sequence corresponding to ants is concentrated; when the mobile tasks to be ordered are all placed in the task sequence solution set, the +. >Mobile task sequence explored by ants only ∈>The method comprises the steps of carrying out a first treatment on the surface of the Computing the movement task sequence->Is>。
S136, for left partition: moving task sequencesAnd->Performing cross operation to obtain a mobile task sequence，/>And->Performing the cross operation again to obtain a moving task sequence +.>And then (2) to->Performing mutation operation to obtain a mobile task sequence->Calculating the execution movement task sequence->Is>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Smaller, accept the movement task sequence +.>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Not get smaller, then->The moving task sequence corresponding to ants only is still。
For the right partition: moving task sequencesAnd->Performing cross operation to obtain a moving task sequence +.>，/>And->Performing the cross operation again to obtain a moving task sequence +.>And then (2) to->Performing mutation operation to obtain a mobile task sequence->Calculating the execution movement task sequence->Is>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Smaller, accept the movement task sequence +.>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Not get smaller, then->The moving task sequence corresponding to ants only is still。
S137: after the moving task sequences of all ants in the left partition and the moving task sequences of all ants in the right partition are obtained, individual extremum values of ant colony in the left partition and the right partition are obtained 、/>Mobile task sequence corresponding to individual extremum、/>Global extremum->、/>A mobile task sequence corresponding to the global extremum +.>、And updating.
S138: iterative equation by ant colony pheromoneUpdating the information concentration of the track; /> and />Representing the movement task +.for the t-th and t+1-th iterations, respectively>To move task->Pheromone concentration,/->Representing pheromone volatilization factors; />For the t-th iteration, move task +.>To move task->New pheromone concentration; the number of iterations t=t+1.
S139: repeating the stepsSteps S135 to S138 until the number of repetitions reaches the set maximum number of iterations。
S1310: taking the moving task sequence corresponding to the individual extremum of the left partition as the optimal moving task sequence of the left partition; and taking the moving task sequence corresponding to the individual extremum of the right partition as the optimal moving task sequence of the right partition.
Wherein the state transition probability，Indicate->Ant only from the movement task->A set of mobile tasks that can be accessed; />Representing the slave movement task->To move task->Is>Visibility of (i.e.)>，/>Representing movement task->Starting point to move task->Is the distance of the end point of (2); />Representing the relative importance of the track; />Indicating the relative importance of visibility.
In the first stage scheduling algorithm, a sequence of mobile tasks to be operatedAnd a specific movement task sequence->When the cross operation is performed, the task sequence is moved>、/>Each comprising M mobile tasks; randomly generating two non-equal positive integers +.>、/>，/>Get the movement task sequence->The upper index is +.>To->Between mobile task segments->Move task sequence->Go up and->The movement task segment at the same position is removed and +.>Splicing the rest mobile task fragments to obtain a mobile task fragment +.>Finally, move task segment->Splice to Mobile task segment->Terminal, and thus get the sequence of tasks with movement +.>Corresponding new movement task sequence->。
In the first stage scheduling algorithm, a sequence of mobile tasks to be operatedWhen mutation operation is performed, the task sequence is moved>、/>Each comprising M mobile tasks; randomly generating two non-equal positive integers +.>、/>，/>Will->Index +.> and />Transposition is performed on the two mobile tasks of (2) to obtain AND +.>Corresponding new movement task sequence->。
The second stage scheduling algorithm based on the improved simulated annealing algorithm specifically comprises the following steps:
s21, initializing parameters of a second-stage scheduling algorithm: setting an initial temperatureStop temperature- >Temperature decay Rate->Probability of mutation->The method comprises the steps of carrying out a first treatment on the surface of the Setting the number of internal cycles per unit temperature +.>The method comprises the steps of carrying out a first treatment on the surface of the Setting the current temperature +.>；
S22: randomly distributing to-be-sequenced cross-region mobile tasks to twoObtaining an initial solution; solving, namely a moving task sequence of a cross-region moving task;
s23: judging the current temperatureWhether or not is greater than->The method comprises the steps of carrying out a first treatment on the surface of the If yes, go to step S24, if no, go to step S210;
s24: constructing four ways of searching the neighborhood solution, and searching the neighborhood solution for the initial solution according to the ways of executing the four ways of searching the neighborhood solution with equal probability according to the internal circulation times LL at the preset unit temperature; then according to the settingThe rule judges to reserve new solutions or old solutions, and a plurality of solutions obtained through internal circulation at the current temperature are obtained;
s25: obtaining a plurality of solutions through internal circulation at the current temperature, calculating respective fitness values, taking two solutions with the minimum fitness values for performing cross operation, and calculating the fitness values of new solutions obtained after the cross operation; judging whether the fitness value of the new solution is smaller than the minimum fitness value of each solution at the current temperature, if so, taking the new solution as the optimal solution at the current temperature; if not, taking the solution with the minimum fitness value at the current temperature as the optimal solution at the current temperature;
S26: performing cross operation on the optimal solution at the current temperature and the optimal solution at the previous temperature, calculating the fitness value of the new solution after the cross, judging whether the fitness value of the new solution is smaller than the fitness value of the optimal solution at the current temperature, and if so, taking the new solution as the optimal solution at the current temperature; if not, keeping the optimal solution at the current temperature unchanged;
s27: taking the optimal solution at the current temperature, and according to the set variation probabilityPerforming a mutation operation, calculating an adaptation of a new solution after performing the mutation operationThe fitness value is used for judging whether the fitness value of the new solution is smaller than the fitness value of the optimal solution at the current temperature, if so, the new solution is used as the optimal solution at the current temperature; if not, keeping the optimal solution at the current temperature unchanged;
s28: judging whether the fitness value of the optimal solution at the current temperature is smaller than that of the optimal solution at the previous temperature, if so, keeping the optimal solution at the current temperature as a final optimal solution; if not, taking the optimal solution at the previous temperature as the optimal solution finally obtained at the current temperature;
s29: performing a cooling process, i.e.Jumping to the step S24, and taking the optimal solution finally obtained at the current temperature as the initial solution in the step S24;
S210: for the current temperatureAnd decoding the finally obtained optimal solution to obtain an optimal movement task sequence of the transregional movement task.
The set Metropolis rule is as follows:a sequence of mobile tasks representing an old solution, +.>A sequence of mobile tasks representing a new solution, +.>A sequence of mobile tasks representing the current solution, +.>Fitness value representing old solution, +.>Fitness value representing the new solution, +.>Adaptation to represent the current solutionA degree value; according to the calculation mode of fitness value function, find +.> and />Difference>I.e. +.>The method comprises the steps of carrying out a first treatment on the surface of the If->Then accept the new solution, i.e. the current solution +.>，/>The method comprises the steps of carrying out a first treatment on the surface of the If condition A is satisfied, a new solution is accepted, i.e.)>，/>The method comprises the steps of carrying out a first treatment on the surface of the If the condition A is not satisfied, the old solution is retained, i.e. +.>，The method comprises the steps of carrying out a first treatment on the surface of the Wherein condition a is: />And->，/>Representing random numbers uniformly distributed within the interval (0, 1).
Second stageWhen the cross operation is performed in the scheduling algorithm: mobile task sequence to be operatedAnd a specific movement task sequence->When the cross operation is performed, the task sequence is moved>、/>All comprising M movement tasks, three non-equal positive integers are randomly generated +.>、/>、/>，/>Get the movement task sequence->Go up the index->To->Is->Index number->To->Is->The method comprises the steps of carrying out a first treatment on the surface of the Move task sequence- >Top and Mobile task segment->、The same-position mobile task segment is removed and +.>The rest segments are spliced to obtain a mobile task segment +.>Finally, move task segment->、/>Splice to Mobile task segment->Terminal, get the sequence of movement tasks to be operated +.>Corresponding new movement task sequence->。
When the mutation operation is carried out in the second stage scheduling algorithm: with a sequence of mobile tasksFor example, two non-equal positive integers +.>、/>Moving task sequence->Comprising M mobile tasks->，/>Index as、/>Is transposed by the mobile task of (2) and +.>The coding of the third layer of (2) is modified, i.e./i>The RGV number of (2) is replaced by the number of another RGV, thereby obtaining a new movement task sequence +.>。
The invention adopts a two-stage scheduling algorithm to carry out staged processing on the NP-hard problem with higher complexity (the problem that all NP problems can be reduced in the polynomial time complexity), is suitable for dynamic partition and improves the effect. The two stages design a coding mode, a fitness function and an adaptive algorithm to solve. The method is suitable for dynamic partitioning of various layout scenes, and does not need to be customized and developed for different layout scenes. The utilization rate of the two RGVs is more balanced, and the occurrence of an over-busy or over-idle state of the RGVs is avoided.
The flow of the two-stage scheduling algorithm of the invention is shown in figure 1.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.
Furthermore, it should be understood that although the present disclosure describes embodiments, not every embodiment is provided with a single embodiment, and that this description is provided for clarity only, and that the disclosure is not limited to specific embodiments, and that the embodiments may be combined appropriately to form other embodiments that will be understood by those skilled in the art.

Claims (9)

1. A one-rail double-vehicle scheduling method based on a two-stage dynamic partitioning algorithm is used for optimizing the total time of moving tasks of two RGVs on the same track and outputting a corresponding optimal moving task sequence; the sequence of the mobile tasks is the execution sequence of the mobile tasks; the one-rail double-vehicle scheduling method comprises the following steps:
S1, partitioning a mobile task, wherein an optimization target is to minimize the execution time difference of a mobile task set of two partitions, and the execution sequence of the mobile task of the two partitions is determined; the starting point and the end point of each mobile task are called RGV point positions, and specifically include:
s11: according to a certain RGV point positionDividing the track to form a left partition and a right partition>Representing a set of all RGV points; the moving task whose starting point and end point are both located in the left partition is marked as a moving task set +.>The method comprises the steps of carrying out a first treatment on the surface of the The mobile task whose starting point and end point are both located in the right partition is marked as the mobile task set +.>；
S12: by first-stage scheduling algorithm pair based on ant colony-genetic-particle swarm algorithm and />The mobile tasks in the left partition are sequenced and the mobile task sequences of the left partition are respectively output>And the movement task sequence of the right partition +.>，/> and />Respectively make and />The total moving distance of the middle moving task is shortest;
s13: calculating current RGV point positions respectivelyCorresponding-> and />Execution time of-> and />Calculating an execution time difference；
S14: repeating steps S11 to S13 until completionSelecting all RGV points; selecting the smallest execution time differenceThe corresponding RGV-bit is taken as the optimal partition-bit +. >And the smallest execution time difference +.>Corresponding-> and />Respectively used as the optimal moving task sequence of the left partition and the optimal moving task sequence of the right partition;
s2, obtaining an optimal moving task sequence of the cross-region moving task through a second-stage scheduling algorithm based on an improved simulated annealing algorithm; wherein the transregional moving task is that the starting point and the end point are respectively positioned at the optimal partition point positionsA movement task on both sides; executing a second stage scheduling algorithmWhen the RGVs are in motion, only one RGV can execute the motion task at the same time, and the RGVs which do not execute the motion task avoid to the outside of the motion area of the RGVs which execute the motion task;
the second-stage scheduling algorithm based on the improved simulated annealing algorithm specifically comprises the following steps:
s21, initializing parameters of a second-stage scheduling algorithm: setting an initial temperatureStop temperature->Temperature decay Rate->Probability of mutation->The method comprises the steps of carrying out a first treatment on the surface of the Setting the number of internal cycles per unit temperature +.>The method comprises the steps of carrying out a first treatment on the surface of the Setting the current temperature +.>；
S22: randomly distributing to-be-sequenced cross-region mobile tasks to twoObtaining an initial solution; solving, namely a moving task sequence of a cross-region moving task;
s23: judging the current temperatureWhether or not is greater than->The method comprises the steps of carrying out a first treatment on the surface of the If yes, go to step S24, if no, go to step S210;
S24: constructing four search neighborsThe domain solution mode, and searching the initial solution for the neighborhood solution according to the mode that four searching neighborhood solutions are executed with equal probability according to the internal circulation times LL at the preset unit temperature; then according to the settingThe rule judges to reserve new solutions or old solutions, and a plurality of solutions obtained through internal circulation at the current temperature are obtained;
s25: obtaining a plurality of solutions through internal circulation at the current temperature, calculating respective fitness values, taking two solutions with the minimum fitness values for performing cross operation, and calculating the fitness values of new solutions obtained after the cross operation; judging whether the fitness value of the new solution is smaller than the minimum fitness value of each solution at the current temperature, if so, taking the new solution as the optimal solution at the current temperature; if not, taking the solution with the minimum fitness value at the current temperature as the optimal solution at the current temperature;
s26: performing cross operation on the optimal solution at the current temperature and the optimal solution at the previous temperature, calculating the fitness value of the new solution after the cross, judging whether the fitness value of the new solution is smaller than the fitness value of the optimal solution at the current temperature, and if so, taking the new solution as the optimal solution at the current temperature; if not, keeping the optimal solution at the current temperature unchanged;
S27: taking the optimal solution at the current temperature, and according to the set variation probabilityExecuting the mutation operation, calculating the fitness value of the new solution after executing the mutation operation, judging whether the fitness value of the new solution is smaller than the fitness value of the optimal solution at the current temperature, and if so, taking the new solution as the optimal solution at the current temperature; if not, keeping the optimal solution at the current temperature unchanged;
s28: judging whether the fitness value of the optimal solution at the current temperature is smaller than that of the optimal solution at the previous temperature, if so, keeping the optimal solution at the current temperature as a final optimal solution; if not, taking the optimal solution at the previous temperature as the optimal solution finally obtained at the current temperature;
s29: a cooling process is performed so that the temperature of the material is reduced,i.e.Jumping to the step S24, and taking the optimal solution finally obtained at the current temperature as the initial solution in the step S24;
s210: for the current temperatureAnd decoding the finally obtained optimal solution to obtain an optimal movement task sequence of the transregional movement task.
2. The two-stage dynamic partitioning algorithm based one-rail two-vehicle scheduling method as set forth in claim 1, wherein: in step S12, the moving task sequence of the left partition is output by the first stage scheduling algorithm based on the ant colony-genetic-particle swarm algorithm And the movement task sequence of the right partition +.>When the method comprises the following steps:
s131: randomly selecting the left partition and the right partition respectivelyA sequence of mobile tasks, calculating left partition +.>The respective fitness value, right partition +.>The respective fitness values of the respective mobile task sequences;
s132: respectively selecting from left partition and right partitionA moving task sequence B with the shortest total moving distance,calculating the fitness value corresponding to each moving task sequence, and taking the fitness value of each moving task sequence B as the initial pheromone concentration of any two adjacent moving tasks in the moving task sequence B during ant colony search;
s133: taking the minimum value of the adaptation degree value of the moving task sequence of the left partition in the step S131 as the individual extremum of the left partitionThe corresponding movement task sequence is +.>The method comprises the steps of carrying out a first treatment on the surface of the Taking the minimum value of the right partition moving task sequence fitness value in the step S131 as the individual extremum +.>The corresponding movement task sequence is +.>The method comprises the steps of carrying out a first treatment on the surface of the Calculating the fitness values of all the mobile task sequences of the left partition and the right partition, and taking the minimum fitness value as the global extremum +.>And global extremum of right partition->Global extremum->The corresponding sequence of movement tasks is marked +. >Global extremumThe corresponding sequence of movement tasks is marked +.>；
S134: initializing the iteration times t=0;
s135: in the left partitionRandomly selecting +.>A plurality of movement tasks are allocated to ∈>Ants; for->Ant only, will be->The mobile task allocated by ant is placed in the corresponding task sequence solution set to calculate +.>Only ants are moved by the current task->To next movement task->State transition probability>And according to->Acquire next movement task->The movement task is further->Is placed at->Only the task sequence corresponding to ants is concentrated; when the mobile tasks to be ordered are all placed in the task sequence solution set, the +.>Mobile task sequence explored by ants only ∈>The method comprises the steps of carrying out a first treatment on the surface of the Computing the movement task sequence->Is>；
In the right partitionRandomly selecting +.>A plurality of movement tasks are allocated to ∈>Ants; for->Ant only, will be->The mobile task allocated by ant is placed in the corresponding task sequence solution set to calculate +.>Only ants are moved by the current task->To next movement task->State transition probability>And according to->Acquire next movement task- >The movement task is further->Is placed at->Only the task sequence corresponding to ants is concentrated; when the mobile tasks to be ordered are all placed in the task sequence solution set, the +.>Mobile task sequence explored by ants only ∈>The method comprises the steps of carrying out a first treatment on the surface of the Computing the movement task sequence->Is>；
S136, for left partition: moving task sequencesAnd->Performing cross operation to obtain a moving task sequence +.>，/>And->Performing the cross operation again to obtain a moving task sequence +.>And then (2) to->Performing mutation operation to obtain a mobile task sequence->Calculating the execution movement task sequence->Is>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Smaller, accept the movement task sequence +.>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Not get smaller, then->The corresponding moving task sequence of ants is still +.>；
For the right partition: moving task sequencesAnd->Performing cross operation to obtain a moving task sequence +.>，/>And (3) withPerforming the cross operation again to obtain a moving task sequence +.>And then (2) to->Performing mutation operation to obtain a mobile task sequence->Calculating the execution movement task sequence->Is>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Smaller, accept the movement task sequence +.>The method comprises the steps of carrying out a first treatment on the surface of the If fitness value->Compared with->Not get smaller, then->The corresponding moving task sequence of ants is still +. >；
S137: after the moving task sequences of all ants in the left partition and the moving task sequences of all ants in the right partition are obtained, individual extremum values of ant colony in the left partition and the right partition are obtained、/>A movement task sequence corresponding to the extremum of the individual +.>、Global extremum->、/>A mobile task sequence corresponding to the global extremum +.>、/>Updating;
s138: iterative equation by ant colony pheromoneUpdating the information concentration between the mobile tasks; /> and />Representing the movement task +.for the t-th and t+1-th iterations, respectively>To move task->Pheromone concentration,/->Representing pheromone volatilization factors; />For the t-th iteration, move task +.>To move task->New pheromone concentration; the number of iterations t=t+1;
s139: repeating steps S135 to S138 until the number of repetitions reaches the set maximum number of iterations；
S1310: taking the moving task sequence corresponding to the individual extremum of the left partition as the optimal moving task sequence of the left partition; and taking the moving task sequence corresponding to the individual extremum of the right partition as the optimal moving task sequence of the right partition.
3. The two-stage dynamic partitioning algorithm based one-rail two-vehicle scheduling method as set forth in claim 2, wherein: probability of state transition，/>Indicate->Ant only from the movement task- >A set of mobile tasks that can be accessed; />Representing the slave movement task->To move task->Is>Visibility of (i.e.)>，/>Representing movement task->Starting point to move task->Is the distance of the end point of (2); />Representing the relative importance of the track; />Indicating the relative importance of visibility.
4. The two-stage dynamic partitioning algorithm based one-rail two-vehicle scheduling method as set forth in claim 2, wherein: in step S13, a sequence of mobile tasks to be operatedAnd a specific movement task sequence->When the cross operation is performed, the task sequence is moved>、/>Each comprising M mobile tasks; randomly generating two non-equal positive integers +.>、/>，/>Get the movement task sequence->The upper index is +.>To->Between mobile task segments->Move task sequence->Go up and->The movement task segment at the same position is removed and +.>The rest mobile task segments are spliced to obtain the mobile task segmentsFinally, move task segment->Splice to Mobile task segment->Terminal, and thus get the sequence of tasks with movement +.>Corresponding new mobile task sequence/>。
5. The two-stage dynamic partitioning algorithm based one-rail two-vehicle scheduling method as set forth in claim 2, wherein: in step S13, a sequence of mobile tasks to be operated When mutation operation is performed, the task sequence is moved>、/>All comprising M movement tasks, randomly generating two unequal positive integers +.>、/>，/>Will->Index +.> and />Transposition is performed on the two mobile tasks of (2) to obtain AND +.>Corresponding new movement task sequence->。
6. The two-stage dynamic partitioning algorithm based one-rail two-vehicle scheduling method as set forth in claim 2, wherein: in step S13, the adaptability value of the left partition movement task sequenceThe method comprises the following steps:
；
wherein Moving the total number of tasks for the left partition; />Is->Start of individual movement tasks, +.>Is->End point of each movement task->Is-> and />Is a distance of (2);
fitness value of right partition movement task sequenceThe method comprises the following steps:
；
wherein The total number of tasks is moved for the right partition.
7. The two-stage dynamic partitioning algorithm-based one-rail two-vehicle scheduling method as set forth in claim 1, wherein the set metapolis rule is:a sequence of mobile tasks representing an old solution, +.>A sequence of mobile tasks representing a new solution, +.>A sequence of mobile tasks representing the current solution, +.>Fitness value representing old solution, +.>Fitness value representing the new solution, +.>An fitness value representing a current solution; according to the calculation mode of fitness value function, find +. > and />Difference>I.e.The method comprises the steps of carrying out a first treatment on the surface of the If->Then accept the new solution, i.e. the current solution +.>，/>The method comprises the steps of carrying out a first treatment on the surface of the If condition A is satisfied, a new solution is accepted, i.e.)>，/>The method comprises the steps of carrying out a first treatment on the surface of the If the condition A is not satisfied, the old solution is retained, i.e. +.>，/>The method comprises the steps of carrying out a first treatment on the surface of the Wherein condition a is: />And->，/>Representing random numbers uniformly distributed within the interval (0, 1).
8. The two-stage dynamic partitioning algorithm based one-rail two-vehicle scheduling method as set forth in claim 1, wherein the cross operation is performed in the second stage scheduling algorithmThe method comprises the following steps: mobile task sequence to be operatedAnd a specific sequence of mobile tasksWhen the cross operation is performed, the task sequence is moved>、/>All comprising M movement tasks, three non-equal positive integers are randomly generated +.>、/>、/>，/>Get the movement task sequence->Go up the index->To->Mobile task segment of (a)Index number->To->Is->The method comprises the steps of carrying out a first treatment on the surface of the Move task sequence->Top and Mobile task segment->、/>The same-position mobile task segment is removed and +.>The rest segments are spliced to obtain a mobile task segment +.>Finally, move task segment->、/>Splice to Mobile task segment->Terminal, get the sequence of movement tasks to be operated +.>Corresponding new movement task sequence->。
9. Two-based according to claim 1 A one-rail double-vehicle scheduling method of a stage dynamic partitioning algorithm is characterized in that in a second stage scheduling algorithm, the adaptability value of a mobile task sequence is calculated：
；
wherein ,for the total number of cross-zone movement tasks, +.>Is->Start of individual movement tasks, +.>Is->End point of each movement task->Is-> and />Distance of->Is->A number of RGV assigned by each mobile task; />Indicate->The RGV point position set which is required to pass through by the mobile task and the current position of another RGV have intersection or not, if yes, 1 is taken, and if not, 0 is taken; />Indicate->The union of the RGV point position set required to pass by the mobile task and the RGV current position for executing the i mobile task to the i mobile task starting point is provided with or not with the other RGV current position, wherein the intersection is 1, and the non-intersection is 0;indicate->Whether the direction of each mobile task is matched with the relative position of the RGV for executing the ith mobile task or not, taking 1 in a matching way, and taking 0 in a non-matching way; />Representing execution of +.>Real-time location of RGVs of the individual mobile tasks; />Indicating that no->Real-time location of RGV for each mobile task.
CN202310580393.7A 2023-05-23 2023-05-23 One-rail double-vehicle scheduling method based on two-stage dynamic partitioning algorithm Active CN116307646B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310580393.7A CN116307646B (en) 2023-05-23 2023-05-23 One-rail double-vehicle scheduling method based on two-stage dynamic partitioning algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310580393.7A CN116307646B (en) 2023-05-23 2023-05-23 One-rail double-vehicle scheduling method based on two-stage dynamic partitioning algorithm

Publications (2)

Publication Number Publication Date
CN116307646A CN116307646A (en) 2023-06-23
CN116307646B true CN116307646B (en) 2023-09-01

Family

ID=86785404

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310580393.7A Active CN116307646B (en) 2023-05-23 2023-05-23 One-rail double-vehicle scheduling method based on two-stage dynamic partitioning algorithm

Country Status (1)

Country Link
CN (1) CN116307646B (en)

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110471287A (en) * 2019-08-27 2019-11-19 合肥工业大学 A kind of real-time scheduling method and system of more rail Shuttling trolleys
CN110837950A (en) * 2019-10-22 2020-02-25 山东女子学院 Dynamic scheduling method of RGV (traffic volume vector)
CN111459108A (en) * 2020-04-08 2020-07-28 北京理工大学 Task allocation and conflict-free path planning method for pull-type multi-AGV system
CN112629537A (en) * 2020-12-11 2021-04-09 华晟（青岛）智能装备科技有限公司 Method and system for dynamically selecting conveying route
CN112633582A (en) * 2020-12-29 2021-04-09 中机中联工程有限公司 Multi-station multi-RGV intelligent flexible dynamic scheduling system
CN112950060A (en) * 2021-03-26 2021-06-11 北京京东乾石科技有限公司 Scheduling method and device of annular RGV, electronic equipment and storage medium
CN114706395A (en) * 2022-04-02 2022-07-05 安歌科技(集团)股份有限公司 Four-way shuttle RGV cluster scheduling system
CN114936783A (en) * 2022-06-02 2022-08-23 暨南大学 RGV (vehicle target volume) trolley scheduling method and system based on MMDDPG (multimedia messaging service data distribution group) algorithm
WO2022174483A1 (en) * 2021-02-22 2022-08-25 广东鑫光智能系统有限公司 Serial scheduling algorithm-based rgv scheduling method
CN115032952A (en) * 2022-05-13 2022-09-09 景德镇陶瓷大学 Marine weight-related part processing production workshop integrated scheduling differential evolution control method, device and system based on differentiation transfer strategy
CN115963797A (en) * 2023-01-31 2023-04-14 三一机器人装备(西安)有限公司 Scheduling method, system, RGV, electronic device and computer medium

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170270466A1 (en) * 2016-03-18 2017-09-21 Jusda International Logistics (TAIWAN) CO.,LTD Method and system using with automated guided vehicle
KR102064485B1 (en) * 2017-12-27 2020-01-09 오학서 Data Communicating Method for Work Time Reducing of Automated Material Handling System
WO2022256579A1 (en) * 2021-06-02 2022-12-08 Lineage Logistics, LLC Swapping task assignments to determine task selection

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110471287A (en) * 2019-08-27 2019-11-19 合肥工业大学 A kind of real-time scheduling method and system of more rail Shuttling trolleys
CN110837950A (en) * 2019-10-22 2020-02-25 山东女子学院 Dynamic scheduling method of RGV (traffic volume vector)
CN111459108A (en) * 2020-04-08 2020-07-28 北京理工大学 Task allocation and conflict-free path planning method for pull-type multi-AGV system
CN112629537A (en) * 2020-12-11 2021-04-09 华晟（青岛）智能装备科技有限公司 Method and system for dynamically selecting conveying route
CN112633582A (en) * 2020-12-29 2021-04-09 中机中联工程有限公司 Multi-station multi-RGV intelligent flexible dynamic scheduling system
WO2022174483A1 (en) * 2021-02-22 2022-08-25 广东鑫光智能系统有限公司 Serial scheduling algorithm-based rgv scheduling method
CN112950060A (en) * 2021-03-26 2021-06-11 北京京东乾石科技有限公司 Scheduling method and device of annular RGV, electronic equipment and storage medium
CN114706395A (en) * 2022-04-02 2022-07-05 安歌科技(集团)股份有限公司 Four-way shuttle RGV cluster scheduling system
CN115032952A (en) * 2022-05-13 2022-09-09 景德镇陶瓷大学 Marine weight-related part processing production workshop integrated scheduling differential evolution control method, device and system based on differentiation transfer strategy
CN114936783A (en) * 2022-06-02 2022-08-23 暨南大学 RGV (vehicle target volume) trolley scheduling method and system based on MMDDPG (multimedia messaging service data distribution group) algorithm
CN115963797A (en) * 2023-01-31 2023-04-14 三一机器人装备(西安)有限公司 Scheduling method, system, RGV, electronic device and computer medium

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
An ant colony optimizaiton-simulated anealing algorithm for solving a multiload AGVs workshop schedualing problem with limited buffer capacity;Zishi Wang 等;《Processes》;第1-20页 *

Also Published As

Publication number Publication date
CN116307646A (en) 2023-06-23

Similar Documents

Publication Publication Date Title
CN108509335B (en) Software test data generation method based on genetic algorithm optimization
CN109542103A (en) A kind of robot welding paths planning method based on fireworks particle swarm algorithm
EP3611799A1 (en) Array element arrangement method for l-type array antenna based on inheritance of acquired characteristics
Xu et al. Temporal knowledge graph reasoning with historical contrastive learning
CN107092812B (en) Method for identifying key protein based on genetic algorithm in PPI network
CN112013829A (en) Multi-UAV/UGV (unmanned aerial vehicle/user generated Unit) cooperative long-term operation path planning method based on multi-objective optimization
CN112381343B (en) Flexible job shop scheduling method based on genetic-backbone particle swarm hybrid algorithm
CN113312874B (en) Overall wiring method based on improved deep reinforcement learning
CN108873835A (en) The Optimization Scheduling of photoetching process in a kind of manufacture of semiconductor integrated circuit
CN114611383A (en) Construction method of traffic flow prediction model based on space-time multi-scale graph convolution network and traffic flow prediction method
CN114462664A (en) Short-range branch flight scheduling method integrating deep reinforcement learning and genetic algorithm
CN116307646B (en) One-rail double-vehicle scheduling method based on two-stage dynamic partitioning algorithm
CN115719133A (en) RHFS scheduling problem is solved based on improved multi-objective wolf optimization algorithm
Goh et al. Combining reinforcement learning and optimal transport for the traveling salesman problem
CN111340303A (en) Route planning method for travelers based on novel mixed frog-leaping algorithm
CN112884229B (en) Large-scale public place people flow guiding path planning method based on differential evolution algorithm
CN112149691B (en) Neural network searching method and device for binocular vision matching
CN116661797A (en) Code completion method based on enhanced Transformer under word element granularity
CN116304722A (en) Training method of vector retrieval model based on balanced tree index structure
CN116611504A (en) Neural architecture searching method based on evolution
Miao et al. Multi-Robot Task Allocation Using Multimodal Multi-Objective Evolutionary Algorithm Based on Deep Reinforcement Learning
Zhang et al. A discrete grey wolf optimizer for solving flexible job shop scheduling problem with lot-streaming
CN113505875A (en) Fault prediction method, device and storage medium
CN110689216B (en) Robot assembly unit mixed product JIT energy-saving production optimization method based on mixed fruit fly algorithm
KR102188118B1 (en) Electronic device for generating a gene feature vector for gene distributed representation based on a correlation between genes according to cancer and operating method thereof

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant