CN115793591A - Hydraulic cylinder distributed manufacturing and scheduling method based on improved bionic intelligent optimization algorithm - Google Patents
Hydraulic cylinder distributed manufacturing and scheduling method based on improved bionic intelligent optimization algorithm Download PDFInfo
- Publication number
- CN115793591A CN115793591A CN202310045156.0A CN202310045156A CN115793591A CN 115793591 A CN115793591 A CN 115793591A CN 202310045156 A CN202310045156 A CN 202310045156A CN 115793591 A CN115793591 A CN 115793591A
- Authority
- CN
- China
- Prior art keywords
- product
- workshop
- stage
- equipment
- processing
- 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
Links
Images
Classifications
-
- 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
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a hydraulic cylinder distributed manufacturing scheduling method based on an improved bionic intelligent optimization algorithm, which is characterized in that basic conditions, basic parameters and constraint conditions are set based on the operation characteristics of a hydraulic cylinder distributed production workshop; constructing a distributed production scheduling model with the aim of minimizing the total cost based on the basic conditions, the basic parameters and the constraint conditions; based on process stage X J Product stage X P Workshop phase X F And equipment phase X M Performing four-stage integer type coding on a feasible solution of the distributed production scheduling model; and solving the distributed production scheduling model by adopting an improved locust algorithm to obtain an optimal scheduling scheme. The method has the advantages of enhancing the local search capability, more effectively searching global or local better solution, having high convergence speed and being not easy to fall into local optimum, and ensuring that the finally obtained scheduling scheme is more accurate.
Description
Technical Field
The invention relates to the technical field of hydraulic cylinder distributed manufacturing scheduling, in particular to a hydraulic cylinder distributed manufacturing scheduling method based on an improved bionic intelligent optimization algorithm.
Background
With the deep economic globalization and the gradual increase of the production scale, the single workshop manufacturing mode cannot meet the increasing market demand, and the manufacturing industry is shifted to a distributed manufacturing mode. In actual production, a product is often assembled by a plurality of parts, and the whole manufacturing process is composed of two stages of machining and assembling. The distributed assembly flexible job shop scheduling problem DAFJSP, which is a combination of distributed shop scheduling and assembly flexible job shop scheduling problems, has been of interest to many researchers and is now proven to be the inconclusive polynomial aggregation problem NP-hard. The production of the hydraulic cylinder belongs to typical DAFJSP, the production process adopts a manufacturing mode of machining single-piece small-batch discrete distribution, covers all links of part machining and product assembly, and has the characteristics of multiple varieties, variable batches and personalized customization, and the hydraulic cylinder has high process flexibility, high product customization degree and strict quality requirement.
In the prior art, a locust search algorithm GOA is usually adopted for solving a scheduling scheme, however, the locust search algorithm GOA serving as a swarm intelligence optimization algorithm simulating biological behaviors of locust in nature, such as population migration, foraging and the like, has the advantages of simple mechanism, few parameters, strong optimization capability and the like, and the optimization process of the GOA is similar to the process of gathering locust population individuals to optimal individuals, so that the number of leading individuals is too small, the weighting phenomenon in the population is serious, the communication cooperation among individuals of different levels is weakened, the convergence speed of the algorithm is low during solving, the algorithm is easy to fall into local optimization, the obtained scheduling scheme is not accurate, and the method is not suitable for hydraulic cylinder distributed manufacturing enterprises.
Disclosure of Invention
The invention provides a hydraulic cylinder distributed manufacturing scheduling method based on an improved bionic intelligent optimization algorithm, which takes the inventory cost of a hydraulic cylinder in the actual production and assembly process into consideration, takes the minimized total cost comprising the part processing cost, the product assembly cost, the part inventory cost and the product inventory cost as the optimization target, and solves the problem of distributed production scheduling of hydraulic manufacturing enterprises.
In order to solve the technical problems, the invention provides a hydraulic cylinder distributed manufacturing and scheduling method based on an improved bionic intelligent optimization algorithm, which comprises the following steps:
step S1: setting basic conditions, basic parameters and constraint conditions based on the operation characteristics of the hydraulic cylinder distributed production workshop;
step S2: constructing a distributed production scheduling model taking the minimized total cost as a target based on the basic conditions, the basic parameters and the constraint conditions;
and step S3: based on process stage X J Product stage X P Workshop phase X F And equipment stage X M Performing four-stage integer type coding on a feasible solution of the distributed production scheduling model, wherein the four-stage integer type coding is represented as X = [ X ] J |X P |X F | X M ]；
And step S4: and solving the distributed production scheduling model by adopting an improved locust algorithm to obtain an optimal scheduling scheme.
Preferably, the basic parameters include:
n: the number of parts;
n: the number of products;
i: the number of processes;
m: the number of devices;
f: the number of workshops;
r J : the number of processes for part J;
j, J': part index J = 1,2 …, n; j' = 1,2 …, n;
p, P': product index.p = 1,2 …, N; p' = 1,2 …, N;
i, I': process index I = 1,2 …, I; i' = 1,2 …, I;
m: device index.m = 1,2 …, M;
f: workshop index.f = 1,2 …, F;
O J,I : the I step of the part J;
S J,I : starting processing time of the I step of the part J;
E J,I : finishing time of the I step of the part J;
E J : finishing time for processing the part J;
E P : finishing time of processing the product P;
SA P : assembly start time of product P;
EA P : assembly completion time of the product P;
T J,I,F,M : the time required by the I process of machining the part J by the Mth equipment in the workshop F is saved;
T P : the time required for assembly of product P;
K J : the unit time processing cost of the parts;
K P : assembly cost per unit time of the product;
V J : the unit time inventory cost of the part;
V P : the unit time inventory cost of the product;
g: an infinite positive number;
X J,I,F,M : if the I procedure of the part J is processed on the Mth equipment of the workshop F, the processing is 1, otherwise, the processing is 0;
Z P’,P,F : 1 if product P is assembled in workshop F immediately after product P', otherwise 0;
η J,F : if the part J is processed in the workshop F, the part J is 1, otherwise, the part J is 0;
θ P,f : if the product P is assembled in the workshop F, the product P is 1, otherwise, the product P is 0;
μ J,P : if the part J belongs to the product P, the value is 1, otherwise the value is 0.
Preferably, the constraint conditions in step S1 include:
1) One part can only be processed in one workshop:
2) Only one workshop can assemble one product:
3) Only one device in one workshop can be selected for processing in one process:
4) Only one process can be processed by the same equipment at the same time:
5) The same part must be processed in sequence:
6) The working procedures of the parts have continuity in the machining process:
7) The starting time of the next process of the part is determined by the finishing time of the previous process and the processing completion time of the equipment used in the next process:
8) One part belongs to only one product, and parts of the same kind of product cannot be replaced with each other during the assembly process:
9) The assembly start time of the product cannot be less than the maximum completion time of the product parts:
10 The assembly process of the product has continuity:
11 Only one product can be assembled at a time:
preferably, the expression of the distributed production scheduling model in step S2 is:
where minTC represents the minimum total cost, TPC represents the cost of the machining process, TAC represents the cost of the assembly process, and TICJ represents the inventory cost of the parts;
in the formula, SA P The initial assembly time for product P.
Preferably, the locust improving algorithm in the step S4 comprises the following steps:
step S41: generating an initialization population by adopting a hybrid initialization strategy;
step S42: decoding the initialized population and calculating the fitness value of population individuals;
step S43: sorting population individuals based on the fitness value, setting the elite rate individual rate, and screening an elite individual group and a common individual group based on the elite individual rate;
step S44: randomly selecting one individual from the elite individual group as a first parent x l (r) selecting an individual from said population of common individuals as a second parent x using a binary tournament method i (r)；
Step S45: calculating the difference degree of the first parent and the second parent, selecting a cross operation to update the positions of the population individuals when the difference degree is not less than a set difference threshold, selecting a variation operation to update the positions of the population individuals when the difference length is less than the set difference threshold to generate new individuals, and executing the step S46 when the number of the new individuals reaches the set population scale;
step S46: setting a domain structure, selecting population individuals from the elite individual population to perform domain-variable search to form a new population, wherein the domain-variable search method comprises the following steps: and when the neighborhood structure is changed, sequentially searching the neighborhood structures, replacing the current solution with the new solution when the new solution generated by the neighborhood structure is superior to the current solution, jumping out of the neighborhood search, and turning to the next neighborhood structure to continue searching if the generated new solution is not superior to the current solution until all the neighborhood structures are searched and outputting the new solution.
Step S47: and outputting an optimal scheduling scheme when the set iteration times are reached, otherwise, taking the new population as an initial population, and re-executing the steps S42-S46.
Preferably, the hybrid initialization strategy in step S41 includes:
process stage X J The following two strategies are used for initialization:
1) Preferably selecting parts with more residual processes, and randomly selecting parts if more processes exist;
2) Randomly selecting the working procedures;
product stage X P Strict adherence to products and partsIs related to process stage X J One-to-one correspondence is realized;
workshop phase X F The following two strategies are used for initialization:
1) Preferentially selecting workshops with fewer processed products, and selecting the workshop closest to the client for production if the number of the processed products in the workshops is the same;
2) Randomly selecting a workshop;
device phase X M The following two strategies are used for initialization:
1) Preferentially selecting the equipment with less processing quantity in the optional equipment set, and randomly selecting the equipment with less processing quantity if more selection exists;
2) A processing device is randomly selected.
Preferably, the decoding method in step S42 is:
1) Decoding in the part processing stage: decoding of the processing stage is X J And X F Inverse process of coding, X M Decoding and mapping the index into a corresponding equipment number;
2) Decoding in the product assembly stage: in the assembly stage, decoding is carried out according to a first-come first-assembly rule, each workshop is provided with an assembly line, and parts contained in the product are assembled after being processed.
Preferably, the expression of the update of the population individual position in step S45 is:
where Cross () represents a crossover operation, mut () represents a mutation operation, and d lr Represents the Hamming distance of the first parent and the second parent, and length (X) represents the code X = [ X ] J | X P | X F | X M ]Length of (b), wherein length (x) l (r))= length(x i (r))，D lr The ratio of the Hamming distance of the first parent and the second parent to the total length of the code is represented, and the value range is [0,1 ]]，P lr Represents a difference threshold value with a value range of [0,1]，X new (r) represents a new individual.
Preferably, the interleaving operation comprises:
step-stage cross operation: x J The cross operation adopts POX cross to randomly divide a part set J into subsets J 1 And J 2 Is mixing X J In (II) is J 1 Is replicated to child X 1 And keeping the original position of the process unchanged, and adding X J In the genus of J 2 Are sequentially inserted into the offspring X 1 In the vacancy of (A), X is J In (II) is J 1 Is replicated to child X 2 And keeping the original position of the process unchanged, and adding X J In (II) is J 2 In sequence, the parts of (2) are inserted into the offspring X 2 In the vacancy of the device, other stage codes move synchronously and correct the stage codes of the device;
product stage cross operation: the product stage does not carry out cross operation and only changes correspondingly along with the change of the working procedure;
workshop stage cross operation: x F The crossing operation adopts two-point crossing when X F And performing crossing when the following two conditions are met, otherwise, not performing crossing, wherein the conditions comprise:
1) Parts of the same product can only be produced in the same workshop, so X F The crossover can only be operated with the product as a unit;
2) The workshop code crossing among different products firstly needs to meet the requirement that the crossed workshop has the capability of producing the product, otherwise, the crossing operation is not carried out and other workshops are continuously selected;
after the crossing, correcting the equipment stage codes to prevent illegal solutions;
equipment stage cross operation: x M The crossover operation adopts RPX crossover to randomly generate an n-dimensional vector v = { v = 1 ，v 2 ,…,v n Where, equal to the length of the process stage, each element in the vector v is [0,1 ]]If the elements in the vector are less than 0.5, interchanging the equipment index numbers of the corresponding positions in the parent, and correcting the interchanged equipment stage codes to prevent illegal solutions;
the mutation operation comprises:
step variant operation：X J The mutation operation adopts an insertion method, two parents randomly select the work sequence numbers of two positions at respective process stages, the former work sequence number is inserted into the rear position of the latter work sequence number, and in order to prevent illegal solutions, other stage codes of corresponding positions synchronously move and correct equipment stage codes;
and (3) performing product stage mutation operation: according to the corresponding relation between the product and the part, the corresponding change is made along with the process change;
and (3) performing workshop stage mutation operation: x F The variation operation adopts a replacement method, and the two parents randomly select the workshop number of a product to replace the workshop number of another different workshop number in respective workshop stages, so as to prevent illegal solutions, verify whether the replaced workshop has the capability of processing the corresponding product and correct the equipment stage codes;
and (3) equipment stage mutation operation: x M The mutation operation adopts a replacement method, and the two parents randomly select the equipment index numbers at the two positions at respective equipment stages and replace the equipment index numbers with different equipment index numbers in respective selectable equipment sets.
Preferably, the domain structure in step S46 includes:
neighborhood structure N1: finding the workshop with the largest product processing number, randomly selecting one workshop if multiple choices exist, finding the product corresponding to the part with the largest completion time in the workshop, and transferring the product to the workshop with the smallest product processing number;
neighborhood structure N2: randomly selecting an element from the procedure coding sequence by a greedy replacement method, replacing the element with other residual positions to generate a plurality of neighborhood structures, and selecting a neighborhood structure with the minimum maximum completion time;
neighborhood structure N3: and finding the equipment with the largest accumulated processing time in the workshop with the largest maximum completion time, switching the process with the longest processing time to the equipment with the smallest number of processing parts in the optional equipment set for processing, and randomly selecting if more choices exist.
The invention has the beneficial effects that:
the invention constructs a distributed production scheduling model, improves a solving algorithm aiming at the model and provides a set of selectable solutions for the problem of improving the production benefit of hydraulic cylinder enterprises.
Setting basic conditions, basic parameters and constraint conditions aiming at the operation characteristics of hydraulic cylinder production so as to ensure that the finally obtained scheduling scheme is effective and can be implemented; the model is solved by improving the locust algorithm IGOA, the local searching capability is enhanced, global or local better solutions are found more effectively, the convergence speed is high, the situation that the optimal solution is trapped into local optimum is not prone to occurring, and the finally obtained scheduling scheme is more accurate.
Drawings
FIG. 1 is a flow chart of a method for improving a locust algorithm according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of the encoding and decoding of the present invention;
FIG. 3 is a diagram illustrating comparison of four algorithms according to an embodiment of the present invention.
Detailed description of the preferred embodiments
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without any inventive step, are within the scope of the present invention.
The embodiment of the invention provides a hydraulic cylinder distributed manufacturing and scheduling method based on an improved bionic intelligent optimization algorithm, which comprises the following steps:
step S1: setting basic conditions, basic parameters and constraint conditions based on the operation characteristics of a hydraulic cylinder distributed production workshop;
aiming at the operation characteristics of the hydraulic cylinder distributed production workshop, the operation requirements of the hydraulic cylinder distributed production workshop are met, and the basic conditions set by the embodiment of the invention are as follows:
1) Only one part can be processed by the same equipment at the same time;
2) The part machining process is not allowed to be interrupted;
3) Parts of the same product can only be produced in the same workshop;
4) Parts of a product are assembled immediately after production is completed and assembling equipment is idle;
5) The inventory cost of all parts in unit time is the same, and the inventory cost of all products in unit time is the same;
6) And defective products and defective goods are not considered, and the production preparation time, the packaging time, the loading time, the residence time of a client place and the like of the product are not considered.
The basic parameters set include:
n: the number of parts;
n: the number of products;
i: the number of processes;
m: the number of devices;
f: the number of workshops;
r J : the number of processes of part J;
j, J': part index J = 1,2 …, n; j' = 1,2 …, n;
p, P': product index.p = 1,2 …, N; p' = 1,2 …, N;
i, I': process index I = 1,2 …, I; i' = 1,2 …, I;
m: device index.m = 1,2 …, M;
f: workshop index.f = 1,2 …, F;
O J,I : the I step of the part J;
S J,I : starting processing time of the I step of the part J;
E J,I : finishing time of the I step of the part J;
E J : finishing time for processing the part J;
E P : finishing time of processing the product P;
SA P : assembly start time of product P;
EA P : the assembly completion time of the product P;
T J,I,F,M : the time required by the I process of machining the part J by the Mth equipment in the workshop F is saved;
T P : time required for product P assembly;
K J : the unit time processing cost of the parts;
K P : the assembly cost per unit time of the product;
V J : the unit time inventory cost of the parts;
V P : the unit time inventory cost of the product;
g: an infinite positive number;
X J,I,F,M : if the I procedure of the part J is processed on the Mth equipment of the workshop F, the processing is 1, otherwise, the processing is 0;
Z P’,P,F : 1 if product P is assembled in workshop F immediately after product P', otherwise 0;
η J,F : if the part J is processed in the workshop F, the part J is 1, otherwise, the part J is 0;
θ P,f : if the product P is assembled in the workshop F, the product P is 1, otherwise, the product P is 0;
μ J,P : if the part J belongs to the product P, the part J is 1, otherwise the part J is 0.
Setting the constraint conditions includes:
1) One part can only be processed in one workshop:
2) Only one workshop can assemble one product:
3) One procedure can only select one device in one workshop for processing:
4) Only one process can be processed by the same equipment at the same time:
5) The same part must be processed in sequence:
6) The process of the part has continuity in the machining process:
7) The starting time of the next process of the part is determined by the finishing time of the previous process and the processing completion time of the equipment used in the next process:
8) One part belongs to only one product, and parts of the same kind of product cannot be replaced with each other during the assembly process:
9) The assembly start time of the product cannot be less than the maximum completion time of the product parts:
10 The assembly process of the product has continuity:
11 Only one product can be assembled at a time:
step S2: constructing a distributed production scheduling model with the aim of minimizing the total cost based on the basic conditions, the basic parameters and the constraint conditions;
the expression of the distributed production scheduling model is as follows:
where minTC represents the minimum total cost, TPC represents the cost of the machining process, TAC represents the cost of the assembly process, and TICJ represents the inventory cost of the parts;
in the formula, SA P For the assembly starting time of the product P, in the embodiment of the invention, the product P is set to enter the assembly immediately after the product P', and the assembly starting time SA of the product P is set p Depending on the product P working-up time E P And assembly completion time EA of product P P’ The larger of these.
And step S3: based on process stage X J Product stage X P Workshop phase X F And equipment phase X M Performing four-stage integer type coding on a feasible solution of the distributed production scheduling model, wherein the four-stage integer type coding is represented as X = [ ] J | X P | X F | X M ]。
As shown in fig. 2, the encoding specifically includes: process stage coding, process layer X J The element in (1) is the workpiece number, and the sequence of the element is correspondingProcedure of work piece, X J =[4,2,1,5,4,2,2,6,3,1,5,6,6,3,4]，X J The first 4 in (a) represents a first step O of the work 4 41 The second 4 represents a second step O of the workpiece 4 42 And so on.
Product stage coding, product layer X P The element in the method is a product number, and the workpiece product numbers of the same product are required to be ensured to be consistent.
Workshop stage code, intercar layer X F The element in (1) is a workshop number, and the constraint conditions are required to be ensured at the same time, namely, workpieces of the same product can only be produced in the same factory.
Device phase coding, device layer X M Wherein the element is X J Index number, X, of optional equipment set of middle and corresponding process M =[2,2,1,3,2,2,3,1,2,3,1,2,3,3,1]In (1), the first 2 represents the arranging step O 41 On the second piece of equipment of the alternative set of equipment and so on.
And step S4: and solving the distributed production scheduling model by adopting an improved locust algorithm to obtain an optimal feasible solution.
Specifically, as shown in fig. 1, the improved locust algorithm proposed by the embodiment of the present invention includes the following steps:
step S41: generating an initialization population by adopting a hybrid initialization strategy;
the hybrid initialization strategy comprises:
process stage X J The following two strategies are used for initialization:
1) Preferably selecting parts with more residual processes, and randomly selecting parts if more processes exist;
2) Randomly selecting the working procedures;
in the embodiment of the invention, the two strategies respectively account for 50% of the population scale so as to ensure the diversity of the initialized population.
Product stage X P Strict adherence to product and part relationships and process stage X J One-to-one correspondence is realized;
workshop phase X F The following two strategies are used for initialization:
1) Preferentially selecting workshops with small number of processed products, and selecting the workshop closest to the client for production if the number of the processed products in the workshops is the same;
2) Randomly selecting a workshop;
in the embodiment of the invention, the two strategies respectively account for 50% of the population scale so as to ensure the diversity of the initialized population.
Device phase X M The following two strategies are used for initialization:
1) Preferentially selecting the equipment with less processing quantity in the optional equipment set, and randomly selecting the equipment with less processing quantity if more selection exists;
2) Randomly selecting a processing device;
in the embodiment of the invention, the two strategies respectively account for 50% of the population scale so as to ensure the diversity of the initialized population.
Step S42: decoding the initialized population and calculating the fitness value of population individuals;
the decoding process mainly takes a workshop as a unit to decompose a feasible solution into scheduling decoding subproblems of a plurality of FJSP flexible job workshops, and then decodes the subproblems respectively. As shown in fig. 2, the decoding process specifically includes:
1) Decoding in the part processing stage: the decoding of the processing stage is X J And X F Inverse process of coding, X M Decoding and mapping the index into a corresponding equipment number; the decoding process in the processing stage corresponds to the production cost TPC, which is positively correlated to the time of the device processing.
2) Decoding in the product assembly stage: in the assembly stage, a first-come first-assembly rule is adopted for decoding, each workshop is provided with an assembly line, and parts contained in the product are assembled after being processed; the decoding process in the assembly stage corresponds to the assembly cost TAC, and the TAC is positively correlated with the assembly time of the product.
Step S43: sorting population individuals based on the fitness value, setting the elite rate individual rate, and screening an elite individual group and a common individual group based on the elite individual rate; specifically, an elite individual group is screened out by setting the elite individual rate, and the rest is a common individual group.
Step S44: randomly selecting one individual from the elite individual group as a first parent x l (r) selecting by binary tournament method, i.e. randomly selecting two individuals from the general population of individuals, selecting the individual with the best fitness as the second parent x i (r)；
Step S45: calculating the difference degree of the first parent and the second parent, selecting a cross operation to update the positions of the population individuals when the difference degree is not less than a set difference threshold, selecting a variation operation to update the positions of the population individuals when the difference length is less than the set difference threshold to generate new individuals, and executing the step S46 when the number of the new individuals reaches the set population scale;
specifically, the expression of updating the population individual position in step S45 is as follows:
where Cross () represents a crossover operation, mut () represents a mutation operation, and d lr Represents the Hamming distance of the first parent and the second parent, and length (X) represents the code X = [ X ] J | X P | X F | X M ]Length of (b), wherein length (x) l (r))= length(x i (r))，D lr The ratio of the Hamming distance of the first parent and the second parent to the total length of the code is represented, and the value range is [0,1 ]]，P lr Represents a difference threshold value with a value range of [0,1]，X new (r) represents a new individual.
The interleaving operation comprises:
step-stage cross operation: x J The cross operation adopts POX cross to randomly divide a part set J into subsets J 1 And J 2 X is to be J In the genus of J 1 Is replicated to child X 1 And keeping the original position of the process unchanged, and adding X J In the genus of J 2 Are sequentially inserted into the offspring X 1 In the vacancy of (A), X is J In the genus of J 1 Is replicated to child X 2 And keeping the original position of the process unchanged, and adding X J In (II) is J 2 Are sequentially inserted into the offspring X 2 In the vacancy of the method, in order to prevent illegal solutions, other stage codes are synchronously moved and the equipment stage codes are corrected;
due to the product stage X P And process stage X J The parts have strict product-part corresponding relation, so that the product stage does not need to be subjected to cross operation and only changes correspondingly along with process change.
And (3) workshop stage cross operation: x F The crossing operation adopts two-point crossing when X F And performing crossing when the following two conditions are met, otherwise, not performing crossing, wherein the conditions comprise:
1) Parts of the same product can only be produced in the same workshop, so X F The crossover can only be operated with the product as a unit;
2) The workshop code crossing among different products firstly needs to meet the requirement that the crossed workshop has the capability of producing the product, otherwise, the crossing operation is not carried out and other workshops are continuously selected;
and correcting the equipment stage codes to prevent illegal solutions after the crossing.
Equipment stage cross operation: x M The crossing operation adopts RPX crossing to randomly generate an n-dimensional vector v = { v = { (v) } 1 ，v 2 ,…,v n H, where each element in the vector v is [0,1 ] equal to the length of the process stage]If the elements in the vector are less than 0.5, interchanging the equipment index numbers of the corresponding positions in the parent, and correcting the interchanged equipment stage codes to prevent illegal solutions;
the mutation operation comprises:
step-stage mutation operation: x J The mutation operation adopts an insertion method, two parents randomly select the work sequence numbers of two positions at respective process stages, the former work sequence number is inserted into the rear position of the latter work sequence number, and in order to prevent illegal solutions, other stage codes of corresponding positions synchronously move and correct equipment stage codes;
and (3) performing product stage mutation operation: according to the corresponding relation between the product and the part, the corresponding change is made along with the process change;
performing workshop stage mutation operation: x F The variation operation adopts a replacement method, and the two parents randomly select the workshop number of a product to replace the workshop number of another different workshop number in respective workshop stages, so as to prevent illegal solutions, verify whether the replaced workshop has the capability of processing the corresponding product and correct the equipment stage codes;
and (3) equipment stage mutation operation: x M The mutation operation adopts a replacement method, and the two parents randomly select the equipment index numbers at two positions in respective equipment stages and replace the equipment index numbers with different equipment index numbers in respective selectable equipment sets.
After the variation process, a better one of the two filial generations is selected as a new individual by greedy selection.
Step S46: setting a domain structure, selecting population individuals from the elite individual population to perform domain-variable search to form a new population, wherein the domain-variable search method comprises the following steps: and when the neighborhood structure is changed, sequentially searching the neighborhood structures, replacing the current solution with the new solution when the new solution generated by the neighborhood structure is superior to the current solution, jumping out of the neighborhood search, and turning to the next neighborhood structure to continue searching if the generated new solution is not superior to the current solution until all the neighborhood structures are searched and outputting the new solution.
Specifically, the embodiment of the present invention sets three field structures as follows:
neighborhood structure N1: finding the workshop with the largest product processing number, randomly selecting one workshop if multiple choices exist, finding the product corresponding to the part with the largest completion time in the workshop, and transferring the product to the workshop with the smallest product processing number;
neighborhood structure N2: randomly selecting an element from the procedure coding sequence by a greedy replacement method, replacing the element with other residual positions to generate a plurality of neighborhood structures, and selecting a neighborhood structure with the minimum maximum completion time;
neighborhood structure N3: and finding the equipment with the largest accumulated processing time in the workshop with the largest maximum completion time, converting the process with the longest processing time into the equipment with the smallest number of processing parts in the optional equipment set for processing, and randomly selecting if more options exist.
Step S47: and outputting an optimal scheduling scheme when the set iteration times are reached, otherwise, taking the new population as the initial population, and re-executing the steps S42 to S46.
In order to verify the effectiveness and superiority of the improved locust algorithm IGOA solving method provided by the invention, 5 examples with serial numbers of MK01, MK05, MK09, MK13 and MK14 in a standard example set Brandmirtte example of flexible job shop scheduling are expanded on the basis of Wen Ji, and each example considers the conditions of 2 or 3 workshops, so that the method is expanded into a new example suitable for the scheduling problem of the distributed assembly flexible job shop for use. And carrying out a comparison experiment on an improved locust algorithm IGOA, a basic locust algorithm GOA, a genetic algorithm GA and a discrete particle swarm algorithm DPSO, wherein the IGOA comprises the following parameters according to orthogonal experiment setting: population size Pop =200, elite individual rate ξ =0.15, individual degree of variation P lr =0.45. The other algorithm parameter settings adopt corresponding reference values, the IGOA adopts the population initialization strategy provided by the invention, and the other algorithms all adopt random initialization strategies.
The algorithm operating environment is as follows: in order to avoid the contingency of results, each algorithm independently runs 20 times by using an Intel Core i7, a CPU 2.9GHZ, an RAM8GB, a Win10 64bit operating system and Matlab 2016b programming software, and the performance of the algorithm is evaluated by adopting three indexes of an optimal value Best, an average value Avg and a relative error RPD, wherein the calculation formula of the RPD is as follows:
wherein C is 0 The optimal value for the current algorithm was run 30 times, C is the optimal value for all algorithms run 30 times.
As can be seen from FIG. 3, the convergence rate and the solution quality of the improved locust algorithm IGOA provided by the invention are superior to those of the other three algorithms, which shows that the improved search operation and the domain search strategy of the IGOA can effectively avoid the algorithm from falling into local optimum and improve the convergence rate of the algorithm.
In order to simplify the description, all possible combinations of the above features in the above embodiments are not described, but only preferred embodiments of the present invention are shown, which are described in detail and are not to be construed as limiting the scope of the present invention. The combination of these features should be considered as the scope of the present specification unless there is any contradiction.
It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A hydraulic cylinder distributed manufacturing and scheduling method based on an improved bionic intelligent optimization algorithm is characterized by comprising the following steps: the method comprises the following steps:
step S1: setting basic conditions, basic parameters and constraint conditions based on the operation characteristics of a hydraulic cylinder distributed production workshop;
step S2: constructing a distributed production scheduling model with the aim of minimizing the total cost based on the basic conditions, the basic parameters and the constraint conditions;
and step S3: based on process stage X J Product stage X P Workshop phase X F And equipment phase X M Performing four-stage integer type coding on a feasible solution of the distributed production scheduling model, wherein the four-stage integer type coding is represented as X = [ X ] J | X P | X F | X M ]；
And step S4: and solving the distributed production scheduling model by adopting an improved locust algorithm to obtain an optimal scheduling scheme.
2. The hydraulic cylinder distributed manufacturing and scheduling method based on the improved bionic intelligent optimization algorithm according to claim 1, characterized in that: the basic parameters include:
n: the number of parts;
n: the number of products;
i: the number of processes;
m: the number of devices;
f: the number of workshops;
r J : the number of processes for part J;
j, J': part index J = 1,2 …, n; j' = 1,2 …, n;
p, P': product index.p = 1,2 …, N; p' = 1,2 …, N;
i, I': process index I = 1,2 …, I; i' = 1,2 …, I;
m: device index.m = 1,2 …, M;
f: workshop index.f = 1,2 …, F;
O J,I : the I step of the part J;
S J,I : starting processing time of the I step of the part J;
E J,I : finishing time of the I step of the part J;
E J : finishing time for processing the part J;
E P : finishing time of processing the product P;
SA P : assembly start time of product P;
EA P : the assembly completion time of the product P;
T J,I,F,M : the time required by the I process of machining the part J by the Mth equipment in the workshop F is saved;
T P : time required for product P assembly;
K J : the unit time processing cost of the parts;
K P : the assembly cost per unit time of the product;
V J : the unit time inventory cost of the part;
V P : the unit time inventory cost of the product;
g: an infinite positive number;
X J,I,F,M : if the I procedure of the part J is processed on the Mth equipment of the workshop F, the processing is 1, otherwise, the processing is 0;
Z P’,P,F : 1 if product P is assembled in workshop F immediately after product P', otherwise 0;
η J,F : if the part J is processed in the workshop F, the part J is 1, otherwise, the part J is 0;
θ P,f : if the product P is assembled in the workshop F, the product P is 1, otherwise, the product P is 0;
μ J,P : if the part J belongs to the product P, the part J is 1, otherwise the part J is 0.
3. The hydraulic cylinder distributed manufacturing and scheduling method based on the improved bionic intelligent optimization algorithm is characterized by comprising the following steps of: the constraint conditions in step S1 include:
1) One part can only be processed in one workshop:
2) Only one workshop can assemble one product:
3) One procedure can only select one device in one workshop for processing:
4) Only one process can be processed by the same equipment at the same time:
5) The same part must be processed in sequence:
6) The working procedures of the parts have continuity in the machining process:
7) The starting time of the next process of the part is determined by the finishing time of the previous process and the processing completion time of the equipment used in the next process:
8) One part belongs to only one product, and parts of the same type of product cannot be replaced with each other during the assembly process:
9) The assembly start time of the product cannot be less than the maximum completion time of the product parts:
10 The assembly process of the product has continuity:
11 Only one product can be assembled at a time:
4. the hydraulic cylinder distributed manufacturing and scheduling method based on the improved bionic intelligent optimization algorithm according to claim 3, characterized in that: the expression of the distributed production scheduling model in the step S2 is as follows:
where minTC represents the minimum total cost, TPC represents the cost of the machining process, TAC represents the cost of the assembly process, and TICJ represents the inventory cost of the parts;
in the formula, SA P The initial assembly time for product P.
5. The hydraulic cylinder distributed manufacturing and scheduling method based on the improved bionic intelligent optimization algorithm is characterized by comprising the following steps of: the locust improvement algorithm in the step S4 comprises the following steps:
step S41: generating an initialization population by adopting a hybrid initialization strategy;
step S42: decoding the initialized population and calculating the fitness value of population individuals;
step S43: sorting population individuals based on the fitness value, setting the elite rate individual rate, and screening an elite individual group and a common individual group based on the elite individual rate;
step S44: randomly selecting one individual from the elite individual group as a first parent x l (r) selecting an individual from said population of common individuals as a second parent x using a binary tournament method i (r)；
Step S45: calculating the difference degree of the first parent and the second parent, selecting a cross operation to update the positions of the population individuals when the difference degree is not less than a set difference threshold, selecting a variation operation to update the positions of the population individuals when the difference length is less than the set difference threshold to generate new individuals, and executing the step S46 when the number of the new individuals reaches the set population scale;
step S46: setting a domain structure, selecting population individuals from the elite individual population to perform domain-variable search to form a new population, wherein the domain-variable search method comprises the following steps: when the neighborhood-varying search is carried out, neighborhood structure search is carried out in sequence, when a new solution generated by a neighborhood structure is superior to a current solution, the new solution is replaced by the current solution, neighborhood search is carried out, if the generated new solution is not superior to the current solution, the next neighborhood structure is turned to continue searching until all neighborhood structures are searched, and a new solution is output;
step S47: and outputting the optimal solution when the set iteration times are reached, otherwise, taking the new population as the initial population, and re-executing the steps S42-S46.
6. The hydraulic cylinder distributed manufacturing and scheduling method based on the improved bionic intelligent optimization algorithm according to claim 5, characterized in that: the hybrid initialization strategy in step S41 includes:
process stage X J The following two strategies are used for initialization:
1) Preferably selecting parts with more residual processes, and randomly selecting parts if more processes exist;
2) Randomly selecting the working procedures;
product stage X P Strict adherence to product and part relationships and process stages X J One-to-one correspondence is realized;
workshop phase X F Adopt the followingTwo strategies are initialized:
1) Preferentially selecting workshops with fewer processed products, and selecting the workshop closest to the client for production if the number of the processed products in the workshops is the same;
2) Randomly selecting a workshop;
plant stage X M The following two strategies are used for initialization:
1) Preferentially selecting the equipment with less processing quantity in the optional equipment set, and randomly selecting the equipment with less processing quantity if more selection exists;
2) A processing device is randomly selected.
7. The hydraulic cylinder distributed manufacturing and scheduling method based on the improved bionic intelligent optimization algorithm according to claim 5, characterized in that: the decoding method in step S42 is:
1) Decoding in the part processing stage: decoding of the processing stage is X J And X F Inverse process of coding, X M Decoding and mapping the index into a corresponding equipment number;
2) Decoding in the product assembly stage: in the assembly stage, decoding is carried out according to a first-come first-assembly rule, each workshop is provided with an assembly line, and parts contained in the product are assembled after being processed.
8. The hydraulic cylinder distributed manufacturing and scheduling method based on the improved bionic intelligent optimization algorithm according to claim 5, characterized in that: in step S45, the expression of updating the population individual position is:
where Cross () represents a crossover operation, mut () represents a mutation operation, and d lr Representing the hamming distance of the first parent and the second parent, length (X) representing the code X = [ X = J | X P | X F | X M ]Length of (b), wherein length (x) l (r))= length(x i (r))，D lr The ratio of the Hamming distance of the first parent and the second parent to the total length of the code is represented, and the value range is [0,1 ]]，P lr The difference threshold is represented, and the value range is [0,1 ]]，X new (r) represents a new individual.
9. The hydraulic cylinder distributed manufacturing and scheduling method based on the improved bionic intelligent optimization algorithm according to claim 6, characterized in that:
the interleaving operation comprises the following steps:
step-stage cross operation: x J The cross operation adopts POX cross to randomly divide a part set J into subsets J 1 And J 2 Is mixing X J In (II) is J 1 Is replicated to child X 1 And keeping the original position of the process unchanged, and adding X J In (II) is J 2 In sequence, the parts of (2) are inserted into the offspring X 1 In the vacancy of (A), X is J In (II) is J 1 Is replicated to child X 2 And keeping the original position of the process unchanged, and adding X J In (II) is J 2 Are sequentially inserted into the offspring X 2 In the vacancy of the device, other stage codes move synchronously and correct the stage codes of the device;
product stage cross operation: the product stage does not carry out cross operation and only changes correspondingly along with the change of the working procedure;
and (3) workshop stage cross operation: x F The crossing operation adopts two-point crossing when X F And performing intersection when the following two conditions are met, otherwise, not performing intersection, wherein the conditions comprise:
1) Parts of the same product can only be produced in the same workshop, so X F The crossover can only be operated with the product as a unit;
2) The workshop code crossing among different products firstly needs to meet the requirement that the crossed workshop has the capability of producing the product, otherwise, the crossing operation is not carried out and other workshops are continuously selected;
after the crossing, correcting the equipment stage codes to prevent illegal solutions;
equipment stage cross operation: x M By interleavingRPX is crossed, and an n-dimensional vector v = { v } is randomly generated 1 ，v 2 ,…,v n Where, equal to the length of the process stage, each element in the vector v is [0,1 ]]If the elements in the vector are less than 0.5, interchanging the equipment index numbers of the corresponding positions in the parent, and correcting the interchanged equipment stage codes to prevent illegal solutions;
the mutation operation comprises:
step-stage mutation operation: x J The mutation operation adopts an insertion method, two parents randomly select the work sequence numbers of two positions at respective process stages, the former work sequence number is inserted into the rear position of the latter work sequence number, and in order to prevent illegal solutions, other stage codes of corresponding positions synchronously move and correct equipment stage codes;
and (3) performing product stage mutation operation: according to the corresponding relation between the product and the part, the corresponding change is made along with the process change;
performing workshop stage mutation operation: x F The variation operation adopts a replacement method, and the two parents randomly select the workshop number of a product to replace the workshop number of another different workshop number in respective workshop stages, so as to prevent illegal solutions, verify whether the replaced workshop has the capability of processing the corresponding product and correct the equipment stage codes;
and (3) equipment stage mutation operation: x M The mutation operation adopts a replacement method, and the two parents randomly select the equipment index numbers at two positions in respective equipment stages and replace the equipment index numbers with different equipment index numbers in respective selectable equipment sets.
10. The hydraulic cylinder distributed manufacturing and scheduling method based on the improved bionic intelligent optimization algorithm according to claim 5, characterized in that: the domain structure in step S46 includes:
neighborhood structure N1: finding the workshop with the largest product processing number, randomly selecting one workshop if multiple choices exist, finding the product corresponding to the part with the largest completion time in the workshop, and transferring the product to the workshop with the smallest product processing number;
neighborhood structure N2: randomly selecting an element from the procedure coding sequence by a greedy replacement method, replacing the element with other residual positions to generate a plurality of neighborhood structures, and selecting a neighborhood structure with the minimum maximum completion time;
neighborhood structure N3: and finding the equipment with the largest accumulated processing time in the workshop with the largest maximum completion time, converting the process with the longest processing time into the equipment with the smallest number of processing parts in the optional equipment set for processing, and randomly selecting if more options exist.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310045156.0A CN115793591B (en) | 2023-01-30 | 2023-01-30 | Hydraulic cylinder distributed manufacturing scheduling method based on improved bionic intelligent optimization algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310045156.0A CN115793591B (en) | 2023-01-30 | 2023-01-30 | Hydraulic cylinder distributed manufacturing scheduling method based on improved bionic intelligent optimization algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115793591A true CN115793591A (en) | 2023-03-14 |
| CN115793591B CN115793591B (en) | 2023-05-26 |
Family
ID=85429210
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310045156.0A Active CN115793591B (en) | 2023-01-30 | 2023-01-30 | Hydraulic cylinder distributed manufacturing scheduling method based on improved bionic intelligent optimization algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115793591B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112465333A (en) * | 2020-11-24 | 2021-03-09 | 大连理工大学 | Intelligent production line scheduling optimization method based on multiple time scales |
| CN113569483A (en) * | 2021-07-30 | 2021-10-29 | 南京信息工程大学 | Method for solving multi-target flexible job shop scheduling based on artificial bee colony algorithm |
| US20210373888A1 (en) * | 2019-08-29 | 2021-12-02 | Huazhong University Of Science And Technology | Multi-objective optimization method and system for master production plan of casting parallel workshops |
| CN113902174A (en) * | 2021-09-16 | 2022-01-07 | 燕山大学 | Improved sparrow search optimization method for blocking flow workshop scheduling problem |
| CN114594744A (en) * | 2022-03-14 | 2022-06-07 | 武汉理工大学 | Distributed factory production and distribution integrated scheduling method and system |
| CN115130789A (en) * | 2022-08-30 | 2022-09-30 | 武汉理工大学 | Distributed manufacturing intelligent scheduling method based on improved wolf optimization algorithm |
-
2023
- 2023-01-30 CN CN202310045156.0A patent/CN115793591B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210373888A1 (en) * | 2019-08-29 | 2021-12-02 | Huazhong University Of Science And Technology | Multi-objective optimization method and system for master production plan of casting parallel workshops |
| CN112465333A (en) * | 2020-11-24 | 2021-03-09 | 大连理工大学 | Intelligent production line scheduling optimization method based on multiple time scales |
| CN113569483A (en) * | 2021-07-30 | 2021-10-29 | 南京信息工程大学 | Method for solving multi-target flexible job shop scheduling based on artificial bee colony algorithm |
| CN113902174A (en) * | 2021-09-16 | 2022-01-07 | 燕山大学 | Improved sparrow search optimization method for blocking flow workshop scheduling problem |
| CN114594744A (en) * | 2022-03-14 | 2022-06-07 | 武汉理工大学 | Distributed factory production and distribution integrated scheduling method and system |
| CN115130789A (en) * | 2022-08-30 | 2022-09-30 | 武汉理工大学 | Distributed manufacturing intelligent scheduling method based on improved wolf optimization algorithm |
Non-Patent Citations (3)
| Title |
|---|
| XIULI WU.ETC: "An improved differential evolution algorithm for solving a distributed assembly flexible job shop scheduling problem", 《MEMETIC COMPUTING》 * |
| 郭晨等: "混合分布估计算法求解模糊分布式装配柔性车间调度问题", 《系统工程理论与实践》 * |
| 闫旭等: "混合蝗虫优化算法求解作业车间调度问题", 《计算机工程与应用》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115793591B (en) | 2023-05-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108320057B (en) | Flexible job shop scheduling method based on limited stable pairing strategy | |
| CN113034026B (en) | Q-learning and GA-based multi-target flexible job shop scheduling self-learning method | |
| CN107862411A (en) | A kind of extensive flexible job shop scheduling optimization method | |
| CN102929263A (en) | Hybrid flow shop scheduling method | |
| CN108279647B (en) | A kind of two stages assembly flow-shop scheduling | |
| CN111126800B (en) | Multi-mode resource-limited project scheduling method using layered self-adaptive intelligent algorithm | |
| CN112163808B (en) | Method for solving logistics center addressing problem by self-adaptive whale algorithm based on opponent learning | |
| CN110531716A (en) | The method for solving low-carbon Job-Shop problem based on discrete whale algorithm | |
| CN109991950A (en) | The balance ameliorative way of cooperation robotic asssembly production line based on genetic algorithm | |
| CN116880424A (en) | Multi-robot scheduling method and device based on multi-objective optimization | |
| CN110490446A (en) | A kind of modular process recombination method based on improved adaptive GA-IAGA | |
| CN114021934A (en) | Method for solving workshop energy-saving scheduling problem based on improved SPEA2 | |
| CN115793591A (en) | Hydraulic cylinder distributed manufacturing and scheduling method based on improved bionic intelligent optimization algorithm | |
| CN109255484A (en) | The discrete manufacturing recourses cooperative optimization method and system of data-driven | |
| CN115907399A (en) | Intelligent scheduling method for discrete manufacturing flexible production of electronic product | |
| CN114415615B (en) | Mixed flow assembly line balanced distribution method and device under uncertain demand | |
| CN115438877A (en) | Multi-objective distributed flexible workshop scheduling optimization method based on gray wolf algorithm | |
| CN113792494B (en) | Multi-target flexible job shop scheduling method based on migration bird swarm algorithm and cross fusion | |
| CN115167294A (en) | Multi-target parallel machine scheduling method for differentiated order complete delivery | |
| CN115392616A (en) | Knowledge mining and genetic algorithm combined multi-target discrete workshop scheduling method | |
| Rahmi et al. | A Crossover in Simulated Annealing for Population Initialization of Genetic Algorithm to Optimize Distribution Cost | |
| CN108960509B (en) | Intelligent production scheduling method and system for manufacturing system | |
| Zhang et al. | A new reduced space searching algorithm (RSSA) and its application in optimal design of alloy steels | |
| CN112734286B (en) | Workshop scheduling method based on multi-strategy deep reinforcement learning | |
| Rezaeipanah et al. | Meta-heuristic approach based on genetic and greedy algorithms to solve flexible job-shop scheduling problem |
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 |