CN110716522A - Manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search - Google Patents

Manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search Download PDF

Info

Publication number
CN110716522A
CN110716522A CN201911066140.8A CN201911066140A CN110716522A CN 110716522 A CN110716522 A CN 110716522A CN 201911066140 A CN201911066140 A CN 201911066140A CN 110716522 A CN110716522 A CN 110716522A
Authority
CN
China
Prior art keywords
time
node
transition
petri
library
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.)
Granted
Application number
CN201911066140.8A
Other languages
Chinese (zh)
Other versions
CN110716522B (en
Inventor
黄波
戴晨谧
赵志霞
蔡志成
袁凤连
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.)
Nanjing Tech University
Original Assignee
Nanjing Tech University
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 Nanjing Tech University filed Critical Nanjing Tech University
Priority to CN201911066140.8A priority Critical patent/CN110716522B/en
Publication of CN110716522A publication Critical patent/CN110716522A/en
Application granted granted Critical
Publication of CN110716522B publication Critical patent/CN110716522B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00—Programme-control systems
    • G05B19/02—Programme-control systems electric
    • G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
    • G05B19/41865—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by job scheduling, process planning, material flow
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00—Program-control systems
    • G05B2219/30—Nc systems
    • G05B2219/32—Operator till task planning
    • G05B2219/32252—Scheduling production, machining, job shop
    • 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
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]

Landscapes

  • Engineering & Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Manufacturing & Machinery (AREA)
  • Quality & Reliability (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • General Factory Administration (AREA)

Abstract

The invention discloses a manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search, which comprises the following steps: constructing a Petri net model of a workshop manufacturing system; converting the Petri network model into an input file; constructing related variables such as identification vectors, incidence matrixes and the like for Petri network evolution and heuristic function construction; constructing a heuristic function of an arbitrary time A-star algorithm; and adopting the system initial state identification as the initial state of the random time A-x algorithm, carrying out search on the terminal state, and searching a transition emission sequence from the initial state to the terminal state of the Petri network system to obtain a manufacturing enterprise workshop scheduling scheme. Compared with a common A-search algorithm, the random time A-heuristic search allows flexible balance between the search time and the solution quality, can reduce the number of node expansion in limited time, and can quickly find out the model transition emission sequence so as to quickly obtain the enterprise workshop operation scheduling scheme.

Description

Manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search
Technical Field
The invention relates to the technical field of workshop operation scheduling, in particular to a manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search.
Background
A manufacturing enterprise shop floor is a scheduled manufacturing system that contains multiple production tasks and multiple available resources (machines). The manufacturing enterprise workshop scheduling problem refers to that a given group of jobs are required to be completed on a group of machines, each machine can only process one job at most at any time, the processing of one job on one machine is called a process, the processing time of the process is fixed, and the aim is to find a scheduling scheme with the shortest processing time of all jobs. The efficient scheduling optimization method can conveniently, quickly and effectively implement scheduling of the manufacturing system; the time required by the job scheduling plan can be shortened; when interference occurs, the system can quickly and reliably respond, change plans in time, increase the flexibility of production, meet the market demand of quick change and improve the competitiveness of enterprises.
In order to obtain a workshop operation optimization scheduling scheme, a conventional scheduling optimization strategy based on an A-search algorithm is to find an optimal transition emission sequence from an initial state to a final state of a system, namely an optimal processing sequence of the system, on a reachable graph of a Petri network model of the system by adopting a common A-search algorithm from an initial node, and finally perform task execution according to the content of an operation task arrangement list. However, the common a-star search algorithm needs more and more nodes to be expanded as the search depth increases, and the calculation process needs a lot of time.
In practical scheduling problems, an optimal search is usually not feasible because users have limited allotted time and desire to find a scheduling solution in a given time, whereas the normal a-search algorithm takes most of the time to discriminate between sub-optimal solutions to determine which is the optimal solution. In this case, it is very difficult to obtain an optimal schedule due to the time limitation of the scheduling problem.
Disclosure of Invention
The invention aims to provide a manufacturing enterprise workshop scheduling optimization method which can reduce the number of expansion nodes in the searching process, improve the calculating speed, and allow flexible balance between the searching time and the solution quality when the searching time is limited or uncertain, so as to solve the problems of low convergence speed, more expansion nodes and the like of the conventional A scheduling method.
The technical solution for realizing the purpose of the invention is as follows: a manufacturing enterprise workshop scheduling optimization method based on arbitrary time A heuristic search comprises the following steps:
step 1, modeling a workshop manufacturing system by using a Petri network, and constructing a Petri network model;
step 2, converting the Petri network model constructed in the step 1 into an input file;
step 3, constructing an incidence matrix for Petri network model evolution according to the input file in the step 2;
and 4, constructing two tables of stack structures: the OPEN table and the CLOSED table are respectively used for storing nodes to be expanded and expanded nodes in the heuristic search process at any time A, and initializing an intermediate result node to be recorded as an incumbent node;
step 5, constructing an heuristic function h with adoptability of an arbitrary time A heuristic search algorithm;
and 6, operating an arbitrary time A-heuristic search algorithm from the initial state of the Petri network system, and searching a transition emission sequence from the initial state to the termination state of the Petri network system to obtain a manufacturing enterprise workshop scheduling scheme.
Further, the step 5 of constructing the heuristic function h with the adoptability of the arbitrary time a × heuristic search algorithm is specifically:
Figure BDA0002259417370000021
in the formula, piRepresenting an active pool, r representing a certain resource, M (p)i) Representing an active pool piCurrent state of (D), WRT (p)iR) represents piMinimum time for the token to reach the destination pool under resource r, M (p)i)·WRT(piR) represents piTotal time, U, of the available token in its destination pool in the current statepi(r) represents piThe unit of R required, SIGMA R (p)i,x)·[Upi(r)/M0(r)]Represents piWherein all tokens are inThe remaining time available in the current state.
Further, the step 6 of running an arbitrary time a-search algorithm from the initial state of the Petri net system to search for the transition emission sequence from the initial state to the termination state of the Petri net system specifically includes:
6-1, defining enabling and transmitting in the search;
step 6-2, carrying out initial state node M on the Petri network system0Adding into an OPEN table;
6-3, judging whether the OPEN table is empty or whether the searching time reaches the specified time of the user, if the OPEN table is empty or the searching time reaches the specified time of the user, outputting a transition emission sequence from the initial state to the termination state of the Petri network system, and if the OPEN table is empty or the searching time reaches the specified time of the user, executing the next step;
6-4, deleting the state node M with the minimum f' value in the OPEN table and adding the state node M into the CLOSED table; wherein the calculation formula of the f' value is as follows:
f’(n)＝g(n)+h'(n)
wherein h' (n) ═ ω × h (n), ω ≧ 1, ω is a weight parameter set by a user, and g (n) is the initial state node M of the Petri net system0The machining operation time taken to reach the node M, h (n) is from the node M to the node M in the end stategEstimated remaining operating time;
6-5, judging whether f (M) is less than f (accummbent), if so, executing the step 6-6; otherwise, returning to the step 6-4; wherein the calculation formula of the f value is as follows:
f(n)＝g(n)+h(n)；
6-6, acquiring an enabling transition set { t) of the state node MjWhere j ═ 1 … et (M), et (M) denotes the number of enabled transitions in node M;
6-7, transmitting each enabled transition tjObtaining a corresponding new child node M'jAnd calculating g (M'j)，h’(M’j) And f '(M'j) The calculation method is the same as the step 6-4;
steps 6-8, for each newly generated child node M'jAnd (4) judging: judging M'jWhether it is a termination state node MgIf yes, executing step 6-10; otherwise, executing the next step;
steps 6-9, for each newly generated child node M'jThe following judgment is made:
(a) judging M'jWhether the node M is equal to a certain node M in the OPEN tableOIf yes, checking the relation of g values of two nodes, and if g (M'j) Less than g (M)O) From M 'to'jReplacing an incumbent node, otherwise directly replacing M'jInserting into an OPEN table;
(b) judging M'jWhether it is equal to a certain node M in the CLOSED tableCIf yes, checking the relation of g values of two nodes, and if g (M'j) Less than g (M)C) Then M will beCDeleting simultaneous M 'from CLOSED table'jInserting into an OPEN table; otherwise, directly mixing M'jInserting into an OPEN table;
(c) if M'jNot equal to any node in OPEN table and CLOSED table, then M'jInserting into an OPEN table;
and 6-10, returning to execute the step 6-3.
Compared with the prior art, the invention has the following remarkable advantages: 1) compared with the common A-ray search algorithm, the quantity of the extended nodes is reduced in a limited time, and the calculation speed is higher; 2) the arbitrary time a search technique allows for a flexible tradeoff between search time and solution quality, with the search process continuing to search after a solution is found, continually finding improved solutions until the optimal solution is obtained or the allocation time is over; 3) for a workshop operation scheduling system under the constraint of limited or uncertain time, the model transition emission sequence can be quickly found out by adopting the random time A-search algorithm, and automation and optimization of workshop operation scheduling of manufacturing enterprises are realized by generating the input of tasks, searching of a scheduling optimization method and outputting of scheduling results.
The present invention is described in further detail below with reference to the attached drawing figures.
Drawings
Fig. 1 is a flowchart of the manufacturing enterprise shop scheduling optimization method based on arbitrary time a heuristic search according to the present invention.
FIG. 2 is a schematic diagram of a Petri Net model corresponding to the plant manufacturing system in the embodiment of the invention.
Detailed Description
With reference to fig. 1, the manufacturing enterprise workshop scheduling optimization method based on any time a × heuristic search provided by the present invention includes the following steps:
step 1, modeling a workshop manufacturing system by using a Petri network, and constructing a Petri network model;
step 2, converting the Petri network model constructed in the step 1 into an input file;
step 3, constructing an incidence matrix for Petri network model evolution according to the input file in the step 2;
and 4, constructing two tables of stack structures: the OPEN table and the CLOSED table are respectively used for storing nodes to be expanded and expanded nodes in the heuristic search process at any time A, and initializing an intermediate result node to be recorded as an incumbent node;
step 5, constructing an heuristic function h with adoptability of an arbitrary time A heuristic search algorithm; the heuristic function h has the adoptability, namely, for a certain node of the system, the estimated cost value from the node to the system termination node cannot be greater than the minimum cost value from the node to the system termination node;
and 6, operating an arbitrary time A-heuristic search algorithm from the initial state of the Petri network system, and searching a transition emission sequence from the initial state to the termination state of the Petri network system to obtain a manufacturing enterprise workshop scheduling scheme (the scheduling sequence of all workpieces comprises the processing machine, the processing time consumption, the processing starting time, the processing ending time, the total processing time of the system and the like of each process of all the workpieces).
Further, step 1, modeling the workshop manufacturing system by using a Petri network, and constructing a Petri network model, wherein a top-down method is specifically adopted.
Further, step 1, modeling the workshop manufacturing system by using a Petri network, specifically comprising: using Petri network housesThe Token number in (1) represents the resource number, the transition of the Petri network represents a working component, and the relation arc between the library and the transition in the Petri network represents a rule in the production system. For example, modeling robots, machine tools, etc., to provide a work shop system, a Petri Net model can be constructed using the following top-down method. First, each resource R ∈ R is represented by a library, and the operation JiCan be seen as a plurality of different pipelines PijJ ∈ {1.. q } set. All pipelines PijAll having two libraries p of the same rolei_startAnd pi_end，pi_startIndicating waiting for job JiProduct to be treated, pi_endRepresents operation JiAnd (4) processing the product. Pipeline PijIs described as being represented by u representing the production bufferij+1 libraries connected to multiple structures T representing processing tasksijkIs composed of (a) a library, whereinij0And pijuRespectively with the depot pi_startAnd pi_endAnd (4) overlapping. Task TijkIs described as consisting of v optional operations OijklAnd a repository, these optional operations sharing the same input and output repositories pijk-1And pijk. Each operation is described as being made up of two transitions tijkl1And tijkl2And a depot oijklThe form of the composition; wherein t isijkl1Represents operation oijklStart of, tijkl2Represents operation oijklEnd of (1), depotijklIndicating the progress of the operation, the operation time τ (o)ijkl)＝hijklAssociated with this library. If there is a resource r ∈ SijklA bank r is correspondingly added, the output arc and the input arc of which are respectively equal to tijkl1And tijkl2Are connected. Finally in Petri Net, if job JiWith n tasks in place pi_startPut a corresponding number of tokens, and in addition, how many resources r, put how many tokens in the pool representing resources r.
Further, the step 2 specifically includes:
the input file is in txt format, and the file comprises three lines of contents: the first row comprises initial identifications of the Petri network, and a plurality of tokens exist in each library under the initial identifications, and the corresponding numbers are the tokens; the second row comprises the operation time or cost of each library, and only the operation time of the operation library is possible, but the operation time of the idle library and the resource library is not, namely the time is 0; the third row is the target identification of the Petri net, i.e. the final state that the plant manufacturing system is required to reach.
Further, the incidence matrix used for the Petri net model evolution in the step 3 comprises a preposed incidence matrix L+And a post-incidence matrix L-Respectively is as follows:
Figure BDA0002259417370000052
Figure BDA0002259417370000053
wherein,
Figure BDA0002259417370000054
representing the relational arc pointed to by the transition j to the library i,
Figure BDA0002259417370000055
and representing a relation arc pointing to the transition j from the library position i, wherein i is more than or equal to 1 and less than or equal to n, j is more than or equal to 1 and less than or equal to m, n is the number of the library positions, and m is the number of the transition.
Further, the heuristic function h with the adoptability for constructing the arbitrary time a × heuristic search algorithm in step 5 is specifically:
in the formula, piRepresenting an active pool, r representing a certain resource, M (p)i) Representing an active pool piCurrent state of (D), WRT (p)iR) represents piMinimum time for the token to reach the destination pool under resource r, M (p)i)·WRT(piR) represents piTotal time, U, of the available token in its destination pool in the current statepi(r) represents piThe unit of R required, SIGMA R (p)i,x)·[Upi(r)/M0(r)]Represents piThe remaining time of all tokens available in the current state.
Further, step 6, running an arbitrary time a-search algorithm from the initial state of the Petri net system, and searching for a transition emission sequence from the initial state to the termination state of the Petri net system specifically includes:
6-1, defining enabling and transmitting in the search;
step 6-2, carrying out initial state node M on the Petri network system0Adding into an OPEN table;
6-3, judging whether the OPEN table is empty or whether the searching time reaches the specified time of the user, if the OPEN table is empty or the searching time reaches the specified time of the user, outputting a transition emission sequence from the initial state to the termination state of the Petri network system, and if the OPEN table is empty or the searching time reaches the specified time of the user, executing the next step;
6-4, deleting the state node M with the minimum f' value in the OPEN table and adding the state node M into the CLOSED table; wherein the calculation formula of the f' value is as follows:
f’(n)＝g(n)+h'(n)
wherein h' (n) ═ ω × h (n), ω ≧ 1, ω is a weight parameter set by a user, and g (n) is the initial state node M of the Petri net system0The machining operation time taken to reach the node M, h (n) is from the node M to the node M in the end stategEstimated remaining operating time;
6-5, judging whether f (M) is less than f (accummbent), if so, executing the step 6-6; otherwise, returning to the step 6-4; wherein the calculation formula of the f value is as follows:
f(n)＝g(n)+h(n)；
6-6, acquiring an enabling transition set { t) of the state node MjWhere j ═ 1 … et (M), et (M) denotes the number of enabled transitions in node M;
6-7, transmitting each enabled transition tjObtaining a corresponding new child node M'jAnd calculating g (M'j)，h’(M’j) And f '(M'j) The calculation method is the same as the step 6-4;
steps 6-8, for each newly generated child node M'jAnd (4) judging: judging M'jWhether it is a termination state node MgIf yes, executing step 6-10; otherwise, executing the next step;
steps 6-9, for each newly generated child node M'jThe following judgment is made:
(a) judging M'jWhether the node M is equal to a certain node M in the OPEN tableOIf yes, checking the relation of g values of two nodes, and if g (M'j) Less than g (M)O) From M 'to'jReplacing an incumbent node, otherwise directly replacing M'jInserting into an OPEN table;
(b) judging M'jWhether it is equal to a certain node M in the CLOSED tableCIf yes, checking the relation of g values of two nodes, and if g (M'j) Less than g (M)C) Then M will beCDeleting simultaneous M 'from CLOSED table'jInserting into an OPEN table; otherwise, directly mixing M'jInserting into an OPEN table;
(c) if M'jNot equal to any node in OPEN table and CLOSED table, then M'jInserting into an OPEN table;
and 6-10, returning to execute the step 6-3.
The specific algorithm is as follows:
further, defining enabling and transmitting in the search in step 6-1 specifically includes:
the transition T ∈ T is enabled under the identification M if and only if:t represents the set of all input libraries for transition t;
the transmission of the transition t enabled under the identity M will result in a new identity M':
Figure BDA0002259417370000073
wherein P represents a library, P represents a set of libraries, M' (P) represents a new identifier generated by the emission of the transition t enabled under the identifier M, M (P) represents an identifier before the emission, L-(p, t) post-correlation matrix, L, representing evolution of Petri Net model+(p, t) represents the pre-incidence matrix of the Petri net model evolution.
The present invention will be described in further detail with reference to specific examples.
Examples
An example plant manufacturing system includes six resources (R)1,R2,R3,R4,R5,R6) Possibly machines, robots, etc., each of which can hold a product at the same time, two input buffers (I)1,I2) And two output buffers (O)1,O2) Two types of parts (work) (P)1,P2) The method mainly takes six resources as a core to divide the resource into two production lines, and the processing flow is as follows:
P1:I1→R3→R2(or R5)→R3→R6→R4→O1
P2:I2→R4→R1→R3→R2→R3→O2
the manufacturing enterprise workshop scheduling optimization method based on any time A heuristic search in the embodiment comprises the following contents:
1) suppose oi,j,kFor the jth operation of the ith job, the job is processed with the kth resource. The workshop manufacturing system is modeled by a top-down method, a Petri net model of the workshop manufacturing system is shown in figure 2, and a library set P is { P ═ P1,p2,p3…p21Wherein the resource pool PR＝{p14-p19}, input library PS＝{p1,p8}, output warehouse PE＝{p20,p21The rest are the mobile depot PA＝{p2-p7,p9-p13}. In FIG. 2, there are a total of 14 transitions (t)1To t14) 21 places of library (p)1To p21) Therefore, the matrix file is 14 rows and 21 columns.
2) Converting the constructed Petri network model into an input file:
6 0 0 0 0 0 0 6 0 0 0 0 0 1 1 1 1 1 1 0 0
0 3 2 4 4 3 5 0 24 4 3 5 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 6 6
the first row is the initial identification of the Petri net, wherein 21 storehouses are provided in total, in the initial state, six tokens are respectively arranged in 1 and 8 of the storehouses, one token is respectively arranged in 14, 15, 16, 17, 18 and 19 of the storehouses, no token is 0 in the other storehouses, and all numerical values are separated by single blank spaces; the second action is the time or cost of operation of the libraries, e.g. library p in FIG. 22Having the number 3 therein, representing the run p2The represented operation requires 3 units of time; the third row is the target identifier of the Petri net, in this embodiment, the target identifiers are libraries 14, 15, 16, 17, 18 and 19, each of which has one Token, six Tokens are in libraries 20 and 21, and none of the other libraries.
3) And constructing a transposed incidence matrix file matrix.txt of the Petri network according to the model Petri network as follows:
Figure BDA0002259417370000091
4) inputting an input file into an arbitrary time A-search scheduling program, firstly, constructing a correlation matrix of a model by a program according to the input file, then, using an admissible heuristic function h by the program, operating an arbitrary time A-search algorithm from a system initial state identifier according to the process of FIG. 1, obtaining a model transition emission sequence which is scheduling decorrelated if the algorithm has a solution, wherein the transition emission sequence corresponds to the scheduling sequence of all workpieces, and the sequence comprises the processing machine, the processing time consumption, the processing start time and the processing end time of each process of all the workpieces. The output results of this example are as follows: when omega is set to be 1.5, and the user time is not limited, the program iterates three times in total, a solution of three times is found, the system scheduling cost of the first solution is 87 units of time, the number of expansion nodes is 132, the program calculation time is 0.07 second, and the cost error with the optimal solution is 4 units of time; the second solution system scheduling cost is 84 units of time, the number of the extended nodes is 195, the program calculation time is 0.1 second, and the cost error with the optimal solution is 1 unit of time; the third solution system scheduling cost is 83 units of time, the number of the extended nodes is 57261, the program calculation time is 97.02 seconds, and the cost error with the optimal solution is 0 unit of time, namely the optimal solution is found. From the above, with each iteration of the program, the error range is continuously reduced, the optimal result to which the solution converges is equivalent to the result under the condition that the heuristic function can adopt, and for each iteration of the program, the intermediate result is suboptimal, and less calculation time is spent in each iteration. And the time of 83 units is obtained by the common A-search algorithm with the same system scheduling cost, the nodes 28663 need to be expanded, and the program calculation time is 158.83 seconds. Therefore, compared with the common A-search algorithm, the random time A-search algorithm adopted by the invention can reduce the number of the extended nodes in the limited search time, quickly find out the model transition emission sequence, quickly obtain the enterprise workshop operation scheduling scheme in the limited time specified by the user and better respond to the change of the production requirement.
In conclusion, compared with the ordinary A-ray search algorithm, the arbitrary time A-ray heuristic search algorithm provided by the invention reduces the number of the expansion nodes in a limited time and has higher calculation speed. In addition, the arbitrary time a search technique allows for a flexible tradeoff between search time and solution quality, with the search process continuing to search after a solution is found, continually finding improved solutions until an optimal solution is obtained or the allotted time is over. For a workshop operation scheduling system under the constraint of limited or uncertain time, the model transition emission sequence can be quickly found out by adopting the random time A-search algorithm, and automation and optimization of workshop operation scheduling of manufacturing enterprises are realized by generating the input of tasks, searching of a scheduling optimization method and outputting of scheduling results.

Claims (8)

1. A manufacturing enterprise workshop scheduling optimization method based on any time A heuristic search is characterized by comprising the following steps:
step 1, modeling a workshop manufacturing system by using a Petri network, and constructing a Petri network model;
step 2, converting the Petri network model constructed in the step 1 into an input file;
step 3, constructing an incidence matrix for Petri network model evolution according to the input file in the step 2;
and 4, constructing two tables of stack structures: the OPEN table and the CLOSED table are respectively used for storing nodes to be expanded and expanded nodes in the heuristic search process at any time A, and initializing an intermediate result node to be recorded as an incumbent node;
step 5, constructing an heuristic function h with adoptability of an arbitrary time A heuristic search algorithm;
and 6, operating an arbitrary time A-heuristic search algorithm from the initial state of the Petri network system, and searching a transition emission sequence from the initial state to the termination state of the Petri network system to obtain a manufacturing enterprise workshop scheduling scheme.
2. The manufacturing enterprise workshop scheduling optimization method based on any time a × heuristic search according to claim 1, wherein step 1 models the workshop manufacturing system using a Petri net to construct a Petri net model, in particular using a top-down approach.
3. The manufacturing enterprise workshop scheduling optimization method based on any time a × heuristic search according to claim 1 or 2, wherein the step 1 of modeling the workshop manufacturing system using a Petri net specifically comprises: the method comprises the steps of using a Token number in a Petri network library to represent a resource number, using a transition of a Petri network to represent a working part, and using a relation arc between the library and the transition in the Petri network to represent a rule in a production system.
4. The manufacturing enterprise workshop scheduling optimization method based on any time a heuristic search according to claim 1, wherein the input files of step 2 are specifically:
the input file is in txt format, and the file comprises three lines of contents: the first row comprises initial identifications of the Petri network, and a plurality of tokens exist in each library under the initial identifications, and the corresponding numbers are the tokens; the second row comprises the operation time or cost of each library, and only the operation time of the operation library is possible, but the operation time of the idle library and the resource library is not, namely the time is 0; the third row is the target identification of the Petri net, i.e. the final state that the plant manufacturing system is required to reach.
5. The manufacturing enterprise workshop scheduling optimization method based on any time A-heuristic search according to claim 1, wherein the incidence matrix for Petri Net model evolution in step 3 comprises a pre-incidence matrix L+And a post-incidence matrix L-Respectively is as follows:
Figure FDA0002259417360000021
wherein,
Figure FDA0002259417360000022
representing the relational arc pointed to by the transition j to the library i,
Figure FDA0002259417360000023
and representing a relation arc pointing to the transition j from the library position i, wherein i is more than or equal to 1 and less than or equal to n, j is more than or equal to 1 and less than or equal to m, n is the number of the library positions, and m is the number of the transition.
6. The manufacturing enterprise workshop scheduling optimization method based on any time a heuristic search according to claim 1, wherein the heuristic function h with the adoptability for constructing the any time a heuristic search algorithm in step 5 is specifically:
Figure FDA0002259417360000024
in the formula, piRepresenting an active pool, r representing a certain resource, M (p)i) Representing an active pool piCurrent state of (D), WRT (p)iR) represents piMinimum time for the token to reach the destination pool under resource r, M (p)i)·WRT(piR) represents piTotal time, U, of the available token in its destination pool in the current statepi(r) represents piThe unit of R required, SIGMA R (p)i,x)·[Upi(r)/M0(r)]Represents piThe remaining time of all tokens available in the current state.
7. The method for optimizing scheduling of a manufacturing enterprise workshop based on any time a heuristic search of claim 1, wherein the step 6 of running a any time a search algorithm from an initial state of the Petri net system to search for a transition emission sequence from the initial state to a final state of the Petri net system comprises:
6-1, defining enabling and transmitting in the search;
step 6-2, carrying out initial state node M on the Petri network system0Adding into an OPEN table;
6-3, judging whether the OPEN table is empty or whether the searching time reaches the specified time of the user, if the OPEN table is empty or the searching time reaches the specified time of the user, outputting a transition emission sequence from the initial state to the termination state of the Petri network system, and if the OPEN table is empty or the searching time reaches the specified time of the user, executing the next step;
6-4, deleting the state node M with the minimum f' value in the OPEN table and adding the state node M into the CLOSED table; wherein the calculation formula of the f' value is as follows:
f’(n)＝g(n)+h'(n)
wherein, h' (n) ═ ω × h (n), ω ≧ 1, ω is a weight parameter set by the user, g (n) is the Petri net systemFrom initial state node M0The machining operation time taken to reach the node M, h (n) is from the node M to the node M in the end stategEstimated remaining operating time;
6-5, judging whether f (M) is less than f (accummbent), if so, executing the step 6-6; otherwise, returning to the step 6-4; wherein the calculation formula of the f value is as follows:
f(n)＝g(n)+h(n)；
6-6, acquiring an enabling transition set { t) of the state node MjWhere j ═ 1 … et (M), et (M) denotes the number of enabled transitions in node M;
6-7, transmitting each enabled transition tjObtaining a corresponding new child node M'jAnd calculating g (M'j)，h’(M’j) And f '(M'j) The calculation method is the same as the step 6-4;
steps 6-8, for each newly generated child node M'jAnd (4) judging: judging M'jWhether it is a termination state node MgIf yes, executing step 6-10; otherwise, executing the next step;
steps 6-9, for each newly generated child node M'jThe following judgment is made:
(a) judging M'jWhether the node M is equal to a certain node M in the OPEN tableOIf yes, checking the relation of g values of two nodes, and if g (M'j) Less than g (M)O) From M 'to'jReplacing an incumbent node, otherwise directly replacing M'jInserting into an OPEN table;
(b) judging M'jWhether it is equal to a certain node M in the CLOSED tableCIf yes, checking the relation of g values of two nodes, and if g (M'j) Less than g (M)C) Then M will beCDeleting simultaneous M 'from CLOSED table'jInserting into an OPEN table; otherwise, directly mixing M'jInserting into an OPEN table;
(c) if M'jNot equal to any node in OPEN table and CLOSED table, then M'jInserting into an OPEN table;
and 6-10, returning to execute the step 6-3.
8. The manufacturing enterprise workshop scheduling optimization method based on any time a heuristic search according to claim 7, wherein the step 6-1 of defining enabling and transmitting in the search specifically comprises:
the transition T ∈ T is enabled under the identification M if and only if:
Figure FDA0002259417360000031
M(p)≥L-(p, t), t represents the set of all input libraries for transition t;
the transmission of the transition t enabled under the identity M will result in a new identity M':
Figure FDA0002259417360000032
wherein P represents a library, P represents a set of libraries, M' (P) represents a new identifier generated by the emission of the transition t enabled under the identifier M, M (P) represents an identifier before the emission, L-(p, t) post-correlation matrix, L, representing evolution of Petri Net model+(p, t) represents the pre-incidence matrix of the Petri net model evolution.
CN201911066140.8A 2019-11-04 2019-11-04 Manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search Expired - Fee Related CN110716522B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911066140.8A CN110716522B (en) 2019-11-04 2019-11-04 Manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911066140.8A CN110716522B (en) 2019-11-04 2019-11-04 Manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search

Publications (2)

Publication Number Publication Date
CN110716522A true CN110716522A (en) 2020-01-21
CN110716522B CN110716522B (en) 2022-05-27

Family

ID=69214672

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911066140.8A Expired - Fee Related CN110716522B (en) 2019-11-04 2019-11-04 Manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search

Country Status (1)

Country Link
CN (1) CN110716522B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111428910A (en) * 2020-02-28 2020-07-17 西安电子科技大学 Vehicle shortest-time traffic path processing method based on time delay Petri network

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101211342A (en) * 2006-12-29 2008-07-02 上海芯盛电子科技有限公司 Concurrent type frog jump heuristic search algorithm
CN101599219A (en) * 2008-06-04 2009-12-09 新南威尔士州道路交通管理局 Traffic signal control system
JP2010271748A (en) * 2009-05-19 2010-12-02 Hitachi Ltd Work allocation method of person in charge of work, work allocation system and storage medium
CN102982391A (en) * 2012-11-06 2013-03-20 广东电网公司电网规划研究中心 Parking lot peak envelope voltage (PEV) charge-discharge optimized dispatching method based on disperse search algorithm
US20140112284A1 (en) * 2011-06-15 2014-04-24 Telefonaktiebolaget L M Ericsson (Publ) Method and node for scheduling in a wireless communications network
CN105652833A (en) * 2015-12-30 2016-06-08 南京理工大学 Bi-directional intelligent search-based manufacturing enterprise shop scheduling optimization method
CN105956275A (en) * 2016-05-04 2016-09-21 山东科技大学 Method for calculating optimum calibration on basis of logic Petri network
CN110046810A (en) * 2019-04-02 2019-07-23 南京理工大学 A kind of Shop Floor Multiobjective Scheduling method based on Timed Petri nets
CN110045732A (en) * 2019-03-28 2019-07-23 暨南大学 A kind of heuristic path planing method

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101211342A (en) * 2006-12-29 2008-07-02 上海芯盛电子科技有限公司 Concurrent type frog jump heuristic search algorithm
CN101599219A (en) * 2008-06-04 2009-12-09 新南威尔士州道路交通管理局 Traffic signal control system
JP2010271748A (en) * 2009-05-19 2010-12-02 Hitachi Ltd Work allocation method of person in charge of work, work allocation system and storage medium
US20140112284A1 (en) * 2011-06-15 2014-04-24 Telefonaktiebolaget L M Ericsson (Publ) Method and node for scheduling in a wireless communications network
CN102982391A (en) * 2012-11-06 2013-03-20 广东电网公司电网规划研究中心 Parking lot peak envelope voltage (PEV) charge-discharge optimized dispatching method based on disperse search algorithm
CN105652833A (en) * 2015-12-30 2016-06-08 南京理工大学 Bi-directional intelligent search-based manufacturing enterprise shop scheduling optimization method
CN105956275A (en) * 2016-05-04 2016-09-21 山东科技大学 Method for calculating optimum calibration on basis of logic Petri network
CN110045732A (en) * 2019-03-28 2019-07-23 暨南大学 A kind of heuristic path planing method
CN110046810A (en) * 2019-04-02 2019-07-23 南京理工大学 A kind of Shop Floor Multiobjective Scheduling method based on Timed Petri nets

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
HANGLEI: "Deadlock-free scheduling for flexible manufacturing systems using Petri nets and heuristic search", 《COMPUTERS & INDUSTRIAL ENGINEERING》 *
郭亚军等: "任意时间算法的性能描述", 《武汉理工大学学报(交通科学与工程版)》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111428910A (en) * 2020-02-28 2020-07-17 西安电子科技大学 Vehicle shortest-time traffic path processing method based on time delay Petri network
CN111428910B (en) * 2020-02-28 2023-12-05 西安电子科技大学 Vehicle shortest time traffic path processing method based on delay Petri network

Also Published As

Publication number Publication date
CN110716522B (en) 2022-05-27

Similar Documents

Publication Publication Date Title
Wang et al. An improved particle swarm optimization algorithm for dynamic job shop scheduling problems with random job arrivals
Tang et al. An improved differential evolution algorithm for practical dynamic scheduling in steelmaking-continuous casting production
Li et al. An agent-based approach for integrated process planning and scheduling
CN105652833B (en) Manufacturing enterprise's Job-Shop optimization method based on Bidirectional intelligent search
Fattahi et al. Dynamic scheduling in flexible job shop systems by considering simultaneously efficiency and stability
Kilincci A Petri net-based heuristic for simple assembly line balancing problem of type 2
CN111966050A (en) AMMAS-GA nesting algorithm-based dual-resource mold job shop scheduling optimization method
CN110046810B (en) Multi-target scheduling method for workshop manufacturing system based on timed Petri network
Udhayakumar et al. Sequencing and scheduling of job and tool in a flexible manufacturing system using ant colony optimization algorithm
CN109615188A (en) A kind of predistribution combines the multi-robot Task Allocation of Hungary Algorithm
CN110969362B (en) Multi-target task scheduling method and system under cloud computing system
CN111400868B (en) Distributed workshop scheduling optimization method and system with order and robot carrying functions
Xiao et al. Modeling of ant colony's labor division for the multi-project scheduling problem and its solution by pso
CN112836974A (en) DQN and MCTS based box-to-box inter-zone multi-field bridge dynamic scheduling method
CN116540659B (en) Large complex product workshop scheduling method, system, equipment and medium
CN112101773A (en) Task scheduling method and system for multi-agent system in process industry
CN116933939A (en) Flexible workshop collaborative production method and system based on improved raccoon optimization algorithm
CN110716522B (en) Manufacturing enterprise workshop scheduling optimization method based on arbitrary time A-heuristic search
Der Jeng et al. Heuristic search based on Petri net structures for FMS scheduling
Lewis et al. Control system design for flexible manufacturing systems
CN117132181B (en) Distributed flexible production and transportation cooperative scheduling method
CN117891220A (en) Distributed mixed flow shop scheduling method based on multi-agent deep reinforcement learning
Shah et al. New modeling and performance evaluation of tool sharing control in FMS using colored Petri nets
CN110928253B (en) Dynamic weighting heuristic scheduling method for automatic manufacturing system
Quan et al. Multi-objective evolutionary scheduling based on collaborative virtual workflow model and adaptive rules for flexible production process with operation reworking

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
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20220527