CN116579244A - Second-class balance intelligent optimization method for variable-pitch synchronous mixed-flow bilateral assembly line - Google Patents

Abstract

The invention belongs to the technical field of assembly line balance, and particularly relates to a second-class balance intelligent optimization method of a variable-pitch synchronous mixed-flow double-side assembly line; the method comprises the steps of obtaining relevant data of a product variable-pitch synchronous mixed-flow bilateral assembly line balance problem and processing the relevant data into new data; initializing a population; according to the gene sequence of any one initial chromosome in the initial chromosome population, distributing the corresponding assembly task in the left-right direction of the workstation to obtain a first chromosome; assigning work stations for all assembly tasks in the first chromosome to obtain a second chromosome; acquiring the completion time of each assembly task of each product based on the work station appointed result of the second chromosome to obtain a final chromosome; obtaining |Pop|final chromosomes, and calculating the value evaluation result of each final chromosome according to a variable-rhythm synchronous mixed-flow bilateral assembly mode; screening, crossing and mutating according to the value evaluation result of each final chromosome; and outputting the optimal solution until the maximum iteration number is reached.

Description

Second-class balance intelligent optimization method for variable-pitch synchronous mixed-flow bilateral assembly line

Technical Field

The invention belongs to the technical field of assembly line balance, and particularly relates to a second-class balance intelligent optimization method of a variable-pitch synchronous mixed-flow double-side assembly line.

Background

The variable-rhythm synchronous mixed-flow double-side assembly line has the characteristics of mixed-flow production mode, double-side operation and variable-pitch synchronous conveying, has high production efficiency, small space occupation and relatively simple system control, and is common in production scenes of large-batch and multi-variety complex products such as automobiles, trucks, household appliances and the like.

The assembly line balancing problem is a high concentration of core management decision problems faced in the assembly line design process. It outlines how to orderly organize the assembly tasks from parts to products, and optimizes various targets such as space, cost, production efficiency, and the like. The second problem with assembly line balancing is how to perform assembly line task scheduling with a defined number of assembly workstations and a mature assembly line layout such that assembly line output beats are minimized, thereby achieving an optimal assembly line yield efficiency.

Since the sixties of the twentieth century, various assembly line balancing problems have been continuously proposed and solved. The current problems are derived from the problems of branching such as linear layout, U-shaped layout, single-product production, multi-product mixed flow production mode, single-side operation, double-side operation and multi-side operation, first class, second class and multi-objective optimization class, synchronous conveying, variable-pitch conveying and the like. The solving method comprises an accurate solving method, a heuristic algorithm and an intelligent optimizing method. Since the assembly line balance problem has been proven as an NP-hard problem, there are many heuristic solving methods for performing problem-adaptive development, such as genetic algorithm, ant colony algorithm, particle swarm algorithm, simulated annealing algorithm, etc. However, the design of the solution method for the balance problem of the assembly line with the characteristics of synchronous transmission of variable pitch, mixed flow production and double-side assembly production is still in a missing state.

The genetic algorithm can process various optimization standards and constraints by flexibly constructing fitness functions through using a group search strategy, and is widely applied to solving the assembly line balance problems of different types and scenes. Therefore, the invention designs a solving method based on a genetic algorithm for the second type of balance problem of the variable-pitch synchronous mixed-flow double-side assembly line.

Disclosure of Invention

In order to solve the problem of second-class balance optimization of a variable-pitch synchronous mixed-flow double-side assembly line, the invention provides a second-class balance intelligent optimization method of the variable-pitch synchronous mixed-flow double-side assembly line, which comprises the following steps:

s1, acquiring related data of a product variable-pitch synchronous mixed-flow bilateral assembly line balance problem and processing the related data into new data, wherein the related data comprises the number of workstations, the number of product types, a production sequence, the number of assembly tasks, an operation direction attribute, operation time and a sequence relation among the assembly tasks;

s2, initializing a population: coding according to the sequence relation among the assembly tasks to obtain an initial chromosome population chrom_B0;

s3, extracting an initial chromosome chrome_B0 from the initial chromosome population ^p P= {1,2, …, |pop| } and randomly selecting a workstation, and combining the average operation time matrix MTTM as an initial chromosome chrome_B0 ^p All the assembly tasks in the process are distributed in the left-right direction of the workstation to obtain a first chromosome chrome_B1 ^p ；

S4, the first chromosome chrome_B1 ^p The work station is assigned to all the assembly tasks in the sequence to obtain a second chromosome chrome_B2 ^p ；

S5, based on second chromosome chrome_B2 ^p The work station appointed result of (2) is utilized to obtain the task end time of each product in each assembly task by using a time to work matrix TTM, and the final chromosome chrome_B3 is obtained ^p ；

S6, repeating the steps S3-S5 to obtain |Pop| final chromosomes, and calculating the value evaluation result of each final chromosome according to a variable-rhythm synchronous mixed-flow bilateral assembly mode;

s7, screening, crossing and mutation are carried out according to the value evaluation result of each final chromosome; and outputting the optimal solution until the maximum iteration number is reached.

The invention has the beneficial effects that:

the existing assembly line balancing technology does not relate to a variable-pitch synchronous mixed-flow double-sided assembly line, a solving scheme is provided for the second type of balance problem of the assembly line, and the design of solving methods of other types of balance problems of the assembly line is facilitated;

according to the invention, by designing the intelligent optimization method by considering the characteristics of mixed flow production mode, station bilateral assemblable and variable pitch synchronous conveying, a more accurate solving scheme can be output; the variable-rhythm synchronous mixed-flow double-sided assembly line is commonly used in industries such as automobiles, household appliances and the like, and practical application of the optimization method for balancing the variable-rhythm synchronous mixed-flow double-sided assembly line is beneficial to improving the production efficiency of the assembly line and saving the production cost of products.

Drawings

FIG. 1 is a flow chart of a second type of balanced intelligent optimization method of a variable-pitch synchronous mixed-flow double-sided assembly line of the invention;

FIG. 2 is a schematic diagram of a chromosome of a second type of balanced intelligent optimization method of the variable-pitch synchronous mixed-flow double-sided assembly line.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

The invention provides a second-class balance intelligent optimization method of a variable-pitch synchronous mixed-flow double-side assembly line, which is shown in fig. 1 and comprises the following steps:

s1, acquiring related data of a product variable-pitch synchronous mixed-flow bilateral assembly line balance problem and processing the related data into new data, wherein the related data comprises the number of work stations, the number of product types, the number of assembly tasks, the operation direction attribute, the operation time and the sequence relation among the assembly tasks.

Specifically, the processing in step S1 includes:

s11, acquiring the number of work stations |S|, the number of types of products produced by an assembly line |M|, a production sequence Q, the number of assembly tasks |N|, the operation direction attribute of the assembly tasks, the operation time of each assembly task of various products and the sequence relation among the assembly tasks;

s12, carrying out coding processing on the data acquired in the S11 to obtain a task direction matrix D and an assembly taskA job time matrix TTM of the assembly task, an average job time matrix MTTM of the assembly task, and a prefronous task set of the assembly task; wherein, element T in the time to job matrix TTM _im Representing the job time for the mth product to perform the ith assembly task, the elements in the mean job time matrix MTTMThe average operation time of all types of products in the ith assembly task is represented by the following calculation formula:

s13, setting related parameters, including: population size, screening probability, crossover probability, mutation probability, and maximum number of iterations.

S2, initializing a population: and initializing the chromosome population according to the sequence relation among the assembly tasks to obtain an initial chromosome population chrom_B0.

Specifically, the process of initializing the population in step S2 includes:

s21, setting a 1×|N|label set to construct an initial chromosome chrom_B0 ^P ，p＝{1,2,...,|Pop|}；

S22, screening all assembly tasks without the preface tasks from the I < N > -assembly tasks to form a preface-free task set, and forming the rest assembly tasks into an preface task set;

s23, randomly extracting an assembly task from the unordered task set, putting the assembly task into the 1×|N|label set, and removing the assembly task from the unordered task set;

s24, repeating the operation of S23 until all assembly tasks in the unordered task set are transferred to the label set, and then executing the step S25;

s25, randomly extracting an assembly task from the ordered task set, continuously putting the assembly task into the label set, and removing the assembly task from the ordered task set;

s26, repeating the operation of S25 until all assembly tasks in the ordered task set are completedTransferring to the label set to finally form an initial chromosome chrom_B0 satisfying the priority relation between the assembly tasks as shown in FIG. 2 ^p ，p＝{1,2,...,|Pop|}；

S27, repeating the operations of S21-S26 until the number of the initial chromosomes is |Pop| to obtain an initial chromosome population chrom_B0= { chrom_B0 ¹ ,Chrom_B0 ² ,...,Chrom_B0 ^|Pop| }。

S3, according to any one initial chromosome of the initial chromosome population, the chromosome_B0 ^p P= {1,2,.. an average task job time matrix MTTM, in the case of a randomly selected workstation, the initial chromosome chrom_b0 ^p All the assembly tasks are distributed in the left-right direction of the workstation to obtain a first chromosome chrom_B1 ^p 。

Specifically, in the case of one workstation, in combination with the mean working time matrix MTTM, according to any one of the initial chromosome populations, initial chromosome_B0 ^p The gene order of p= {1,2, …, |pop| } is the initial chromosome chrome_b0 ^p The assembly task corresponding to the k=1, 2, …, |N|column genes is distributed in the left-right direction of the workstation, and a corresponding first chromosome chrome_B1 is obtained ^p The process of (1) comprises:

s31, setting a first 3×|N|array as a first chromosome to be constructed, namely a chromosome_B1 ^p The initial chromosome Chrom_B0 ^p Copy all information of (a) to the first row of the first 3×|n|array; randomly selecting a workstation s= {1,2, …, |s| };

s32, judging an assembly task chrom_B1 corresponding to the kth column gene of the first 3×|N|array ^p (1, k) whether it is the first task of the workstation s; if so, the assembly task chrom_B1 is queried from the average operation time matrix MTTM ^p Average working time of (1, k)And set up assembly task chrom_b1 ^p The average end time of (1, k) isWill->Placed in the first 3×|n|array, column k, row 3; querying an assembly task chrom_b1 from a task direction matrix D ^p Task direction information of (1, k) and put into the first 3×|n|array, column k, row 2; if not, executing step S33;

s33, judging an assembly task chrom_B1 ^p (1, k) the direction attribute, if the direction attribute is one-sided, i.e. can be placed only to the left or right of the workstation s, the assembly task Chrom_B1 ^p Setting the task direction information of (1, k) to correspond to a single side, and placing the task direction information on the kth column 2 of the first 3×|n|array; setting assembly task chrom_b1 ^p The average end time of (1, k) isAnd will->Put in the first 3×|n|array kth column 3 rd row; wherein (1)>The average end time of the job for task j; if the direction attribute is bilateral, i.e. both left and right sides can be placed, step S34 is performed;

s34, calculating an assembly task chrom_B1 according to the method proposed in the step S33 ^p (1, k) average end moments when placed on left and right sides of the workstation, respectively and />Judging assembly task chrom_B1 ^p (1, k) average end time obtained by placing on left and right side +.> and />If the two are the same, executing step S35; if not, executing step S36;

s35, optionally selecting one side (the left side or the right side can be used as the assembly task Chrom_B1) ^p (1, k) and placing the task direction information at the first 3×|n|array kth column 2 nd row while placing the corresponding average end time at the first 3×|n|array kth column 3 rd row;

s36, selecting a single side with the smallest average ending time as an assembly task chrom_B1 ^p (1, k) and placing the task direction information on the first 3×|n|array kth column 2 nd row while placing the minimum average end time at the first 3×|n|array kth column 3 rd row;

s37, repeating the steps S32-S36 until the initial chromosome chrome_B0 ^p The average end time information of the |N| assembly task in the sequence is obtained completely to obtain a first chromosome chrom_B1 shown in figure 2 ^p 。

Specifically, in step S33The method is characterized in that the job average ending time of a task j is selected by the following steps:

s331, assembling task Chrom_B1 ^p (1, k) before the single side corresponding to the task direction information is put in, extracting the current last assembly task of the single side, and marking the last assembly task as a;

s332 screening out the other single side with priority higher than the assembly task chrom_B1 ^p All the assembly tasks of (1, k) and the assembly task with the smallest priority is proposed from the assembly tasks and is marked as b;

s333, inquiring the average operation time of the assembly tasks a and b from the MTTM, so as to obtain the average operation ending time of the assembly tasks a and b, and selecting the assembly task with the largest average operation ending time as a task j.

S4, first chromosome chrom_B1 ^p All the assembly tasks in (a) are referred to asDetermining the workstation to obtain a second chromosome chrom_B2 ^p 。

Specifically, it is the first chromosome chrome_B2 ^p The work station is assigned to all the assembly tasks in the sequence to obtain a second chromosome chrome_B3 ^p Comprising:

s41, setting a second 3×|N|array as a second chromosome to be constructed, namely, chrom_B2 ^p First chromosome Chrom_B1 ^p Copy the first and second lines of information to the first and second lines of the second 3 x|n|array;

s42, obtaining a first chromosome chrom_B1 ^p Last assembly task placed on the left side of the workstationAnd the last assembly task placed on the right side of the workstation +.>Acquiring the job average end time +.>Andsetting a time limit, expressed as:

wherein ,represents the right time limit, +_>Representing a left time limit; so meterThe left time interval of each workstation S, s= {1,2, …, |S| } is calculated to be +.>The right time interval is +.>

S43, inquiring the corresponding assembly task chrom_B2 of the kth column of the second 3×|N|array ^p Average end time of (1, k)And calculates the corresponding average starting moment +.>

S44, if the task is to be assembled, the task is to be chromam_B2 ^p Average end time of (1, k)Average starting timeAre all in the left time interval of workstation s +.>Or right time interval +.>Description of Assembly task Chrom_B2 ^p (1, k) can be performed entirely within the workstation s, then the assembly task chrom_B2 ^p (1, k) is assigned to workstation s and information for workstation s is written in the second 3×|n|array, column k, row 3;

s45, if the task is to be assembled, the task is to be chromam_B2 ^p Average starting time of (1, k)And average end time->One of (2) is not +.>Or->Description of the Assembly task Chrom_B2 ^p (1, k) can be divided into workstations s or their neighboring workstations, the assembly task Chrom_B2 is further determined ^p (1, k) the execution duration of the task Chrom_B2 at the workstation s and its neighboring workstations will be assembled ^p (1, k) designating the workstation corresponding to the maximum execution time length, and writing information of the workstation corresponding to the maximum execution time length in the 3 rd row of the kth column of the second 3×|N|array;

s46, repeating the steps S43-S45 until the first chromosome chrom_B1 ^p The work station information of the I/N assembly task is all acquired to obtain a second chromosome chrom_B2 shown in figure 2 ^p 。

S5, based on second chromosome chrom_B2 ^p The work station appointed result of (2) is utilized to obtain the task end time of each product in each assembly task by using a time to work matrix TTM, and the final chromosome chrom_B3 is obtained ^p C。

Specifically, based on the second chromosome chrome_b2 ^p The workstation designated result of (2) is utilized to obtain the completion time of each assembly task of each product by using a task operation time matrix TTM to obtain a final chromosome chrome_B3 ^p The process comprises the following steps:

s51, creating an array of (|M|+3) ×|N| as a final chromosome to be constructed, namely, chrom_B3 ^p The second chromosome chrome_B2 ^p Copy the first, second, and third lines of information to the first, second, and third lines of (|M|+3) ×|N| array;

s52, judging an assembly task chrom_B3 corresponding to a kth column gene of an (M+3) X N array ^p (1, k) is the first task to the left or right of its designated workstation s (i.e., the workstation corresponding to row 3 of column k of the (|M|+3) ×|N|array), if so, then fromInquiring |M| products in the operation time matrix TTM to respectively execute assembly task Chrom_B3 ^p The working time of (1, k); the execution of the assembly task chrom_b3 by placing |m| products respectively from row 4 of the kth column of the (|m|+3) ×|n|array ^p The working time of (1, k); if not, executing step S53; wherein, the M < M > = {1,2, …, |M| } products execute the assembly task chrom_B3 ^p Working time of (1, k)Placed in row m+3 of column k of the (|m|+3) ×|n| array;

s53, in the query workstation s, the task of assembly is Chrom_B3 ^p All assembly tasks before (1, k), from which assembly task chrom_b3 is selected ^p (1, k) last assembly task x of the single side to which it belongs, and not assembly task chrom_b3 ^p On the other side of the single side to which (1, k) belongs, with the assembly task Chrom_B3 ^p (1, k) the last (least priority) assembly task y with priority relation; acquiring the operation time of the I M I products for executing the assembly task x and the assembly task y respectively, comparing the two operation time end moments of each product, and selecting the maximum operation time end moment of each product as the assembly task chrom_B3 of the product ^p Time base line in (1, k), and performing S54;

s54, executing an assembly task Chrom_B3 on the |M| products based on the time base line of the step S53 ^p (1, k) adding the time base lines corresponding to the time base lines, and placing all the addition results on the 4 th to the |m|+3 th rows of the (m|+3) ×|n| array k-th column; wherein the mth product executes the assembly task chrom_b3 ^p Working time of (1, k)With the mth product at assembly task chrom_b3 ^p The addition result of the time base line in (1, k) is placed in the (m+3) ×|n| array kth column, m+3 th row;

s55, repeating the steps S52-S54 until the second chromosome chrome_B2 ^p Task end time of |M| products of |N| assembly task in (1)All the obtained chromosome is finally obtained to obtain a final chromosome chrom_B3 shown in figure 2 ^p 。

S6, repeating the steps S3-S5 to obtain |Pop| final chromosomes, and calculating the value evaluation result of each final chromosome according to a variable-rhythm synchronous mixed-flow bilateral assembly mode.

Specifically, calculating a value evaluation result of any one final chromosome according to a variable-pitch synchronous mixed-flow bilateral assembly mode, wherein the value evaluation result comprises the following steps:

s61, obtaining a final chromosome chrome_B3 ^p The method comprises the steps of selecting the maximum value in the task ending time of the M-th= {1,2, …, |M| } products on the work stations S as the station beats of M product productions on the work stations S, wherein the task ending time of each product on the|S| on each work station S, s= {1,2, …, |S| } on each work station S for executing each assembly task;

s62, dividing the execution condition of the production sequence Q in the variable-pitch synchronous mixed-flow double-side assembly line into |Q| production line states; when the product of the production sequence Q of the u-th type is produced in the last working station of the assembly line, the type condition of the product produced by each working station of the assembly line is defined as the state of the u-th production line;

in particular, the line status refers to the possible sequence of product types and records produced by all workstations observed at a certain moment in time under a certain production sequence. For example, assuming a production sequence ABBAC, wherein A, B, C represents three different types of products, there are ABBAC, BBACA, BACAB, ACABB, CABBA production states under the sequence for a total of 6 different production line states.

S63, calculating the output time of the state of the production line of the u-th production line, wherein the calculation formula is as follows:

wherein ,ts_μ Represents the throughput time of the production line state of the μ= {1, 2..+ -. Q| } i.e. the maximum value in the station beats of each workstation producing the product in the production line state of the u-th;representing product Q _q Station beat at the s-th workstation, Q _q E M represents the product type at the q-th position of the production sequence;

s64, carrying out weighted average operation on the output time of all production lines to obtain a final chromosome chrom_B3 ^p Is expressed as:

the value evaluation result CT is expressed as an average production rate of each product in the production sequence Q in the assembly line.

S7, screening, crossing and mutation are carried out according to the value evaluation result of each final chromosome; and outputting the optimal solution until the maximum iteration number is reached.

Specifically, screening: for the |Pop| final chromosomes, reversely normalizing CT values of the final chromosomes in the population according to a design principle that the smaller the station beat is, the larger the fitness value is, and obtaining the fitness value of the final chromosomes; selecting according to fitness values of the final chromosomes in the population and using a roulette method to make a Ggap probability; deletion of selected final chromosome Chrom_B4 ^p Task direction information, workstation, all time related information on the selection of the final chromosome chrom_b4 is preserved ^p Degrading the final chromosome into the corresponding initial chromosome chrom_b0 ^p For subsequent crossover, genetic manipulation.

Crossing: selecting a pair of parent chromosomes X, Y from the degenerated initial chromosomes with Cr probability; randomly selecting a list of gene positions to divide the parent chromosome X into a head segment X1 and a tail segment X2, and performing the same operation on the parent chromosome Y to obtain the head segment Y1 and the tail segment Y2; then exchanging the heads of the two ancestor dyes to obtain spliced chromosomes X1Y2 and Y1X2; deleting repeated genes in tail sections of two spliced chromosomes, supplementing the deleted genes, and arranging the deleted genes according to the sequence of original parent chromosomes during supplementation.

Variation: performing mutation operation on each chromosome obtained after crossing according to Mu probability; for the chromosome selected for mutation operation, a list of genes is randomly selected again, and inserted at random positions that do not violate the priority rule. If the previous and subsequent assembly tasks of the gene just belong to their preceding and subsequent tasks, the mutation operation of the gene is abandoned and another list of genes is selected to perform the mutation.

Specifically, the iterative process includes:

s71, randomly selecting a final chromosome in the final chromosome population as a global optimal chromosome chrom_B3 in the first iteration ^BEST ；

S72, in each iteration, obtaining the value evaluation results of all final chromosomes in the iteration population, and taking the final chromosome with the minimum value evaluation result (namely the minimum CT value) as the local optimal chromosome of the iterationComparison of Chrom_B3 ^BEST and />The minimum value of the two is taken as a new global optimal chromosome +.>

S73: in each iteration process, after the dyeing after genetic operation is randomly sequenced, a new population is formed for solution search of the next iteration, and the global optimal chromosome at the moment is inserted into the population of the next iteration.

S74: the termination condition of the algorithm operation is set, and the CPU operation time, the maximum iteration times and the CT value can be set to terminate the algorithm.

S75: and outputting the global optimal chromosome information stored when the iteration is ended as an optimal solution of the second type of balance problem of the variable-rhythm synchronous mixed-flow double-side assembly line.

In one embodiment, two work station assembly lines for producing two types of products by mixed flow are adopted for illustration, the two types of products are respectively product A and product B, the production demand quantity ratio is 4:3, the production sequence is Q= [ B, B, A, A, A, B, A ], the two work stations are respectively encoded as a and B, and the input assembly task information is shown in table 1.

Table 1 sets task information entered

Where direction L represents the left side of the workstation, direction R represents the right side of the workstation, and direction E represents both the left and right sides of the workstation as selectable. After the method is gradually executed and implemented, the optimal chromosome information which can be obtained and output comprises the following steps: the task sequence is [1,4,2,5,3,6,8,9,7], the work station sequences [ a, a, a, a, B, B, B, B ] to which the tasks are allocated are [1,4,3,8,7] and [2,5,6,9] to which the tasks are allocated to the left and right sides of the work station, the maximum working time of the product A on the left and right sides of the two work stations is [4,3] and [4,5], the maximum working time of the product B on the left and right sides of the two work stations is [2,3] and [4,3], and the optimal target value is 4.571.

The output results are verified to follow the input data and conditions; the obtained optimal value 4.571 is smaller than the maximum value 5 in the station beat, accords with typical characteristics of second-class balance problem solutions of the variable-rhythm synchronous mixed-flow double-side assembly line, and results are effective.

Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims (8)

1. The second-class balance intelligent optimization method of the variable-pitch synchronous mixed-flow double-side assembly line is characterized by comprising the following steps of:

s1, acquiring related data of a product variable-pitch synchronous mixed-flow bilateral assembly line balance problem and processing the related data into new data, wherein the related data comprises the number of workstations, the number of product types, a production sequence, the number of assembly tasks, an operation direction attribute, operation time and a sequence relation among the assembly tasks;

s2, initializing a population: coding according to the sequence relation among the assembly tasks to obtain an initial chromosome population chrom_B0;

s3, extracting an initial chromosome chrome_B0 from the initial chromosome population ^p P= {1,2, …, |pop| } and randomly selecting a workstation, and combining the average operation time matrix MTTM as an initial chromosome chrome_B0 ^p All the assembly tasks in the process are distributed in the left-right direction of the workstation to obtain a first chromosome chrome_B1 ^p ；

S4, the first chromosome chrome_B1 ^p The work station is assigned to all the assembly tasks in the sequence to obtain a second chromosome chrome_B2 ^p ；

S5, based on second chromosome chrome_B2 ^p The work station appointed result of (2) is utilized to obtain the task end time of each product in each assembly task by using a time to work matrix TTM, and the final chromosome chrome_B3 is obtained ^p ；

S6, repeating the steps S3-S5 to obtain |Pop| final chromosomes, and calculating the value evaluation result of each final chromosome according to a variable-rhythm synchronous mixed-flow bilateral assembly mode;

s7, screening, crossing and mutation are carried out according to the value evaluation result of each final chromosome; and outputting the optimal solution until the maximum iteration number is reached.

2. The second-class balance intelligent optimization method of the variable-pitch synchronous mixed-flow double-side assembly line according to claim 1, wherein the processing procedure of the step S1 comprises:

s11, acquiring the number of work stations |S|, the number of types of products produced by an assembly line |M|, a production sequence Q, the number of assembly tasks |N|, the operation direction attribute of the assembly tasks, the operation time of each assembly task of various products and the sequence relation among the assembly tasks;

s12, carrying out coding processing on the data acquired in the S11 to obtain a task direction matrix D, an operation time matrix TTM of the assembly task, an average operation time matrix MTTM of the assembly task and an advanced task set of the assembly task; wherein, element T of the time to job matrix TTM _im Representing the job time of the ith assembly task of the mth product, elements in the average job time matrix MTTM of 1×|N|The average working time of the ith assembly task of all types of products is expressed as the following calculation formula:

s13, setting related parameters, including: population size, screening probability, crossover probability, mutation probability, and maximum number of iterations.

3. The second-class balance intelligent optimization method of the variable-pitch synchronous mixed-flow double-sided assembly line according to claim 1, wherein the step S2 of initializing the population comprises the following steps:

s21, setting a 1×|N|label set;

s22, screening all assembly tasks without the preface tasks from the I < N > -assembly tasks to form a preface-free task set, and forming the rest assembly tasks into an preface task set;

s23, randomly extracting an assembly task from the unordered task set, putting the assembly task into the label set, and removing the assembly task from the unordered task set;

s24, repeating the operation of S23 until all assembly tasks in the unordered task set are transferred to the label set, and then executing the step S25;

s25, randomly extracting an assembly task from the ordered task set, putting the assembly task into the label set, and removing the assembly task from the ordered task set;

s26, repeating the operation of S25 until all the assembly tasks in the ordered task set are transferred to the label set to form an initial chromosome chrom_B0 meeting the priority relation among the assembly tasks ^p ，

p＝{1,2,…,|Pop|}；

S27, repeating the operations of S21-S26 until the number of the initial chromosomes is |Pop| to obtain an initial chromosome population chrom_B0= { chrom_B0 ¹ ,Chrom_B0 ² ,...,Chrom_B0 ^|Pop| }。

4. The second-class balanced intelligent optimization method of a variable-pitch synchronous mixed-flow double-sided assembly line according to claim 1, wherein in one workstation, the method is combined with an average operation time matrix MTTM according to any one of initial chromosome populations, namely, initial chromosome_B0 ^p The gene order of p= {1,2, …, |pop| } is the initial chromosome chrome_b0 ^p The assembly task corresponding to the k, k=1, 2, …, |N|column genes is distributed in the left-right direction of the workstation, and a corresponding first chromosome chrome_B1 is obtained ^p The process of (1) comprises:

s31, setting a first 3×|N|array, and setting an initial chromosome chrom_B0 ^p Copy all information of (a) to the first row of the first 3×|n|array; randomly selecting a workstation s= {1,2, …, |s| };

s32, judging the assembly task chrom_B1 of the kth column and the 1 st row of the first 3×|N|array ^p (1, k) is the first task of the workstation s, if so, the assembly task Chrom_B1 is queried from the MTTM ^p Average working time of (1, k)And set up assembly task chrom_b1 ^p The average end time of (1, k) is +.>Will bePlaced on the 3 rd row of the kth column of the first 3×|n| array; querying an assembly task chrom_b1 from a task direction matrix D ^p The task direction information of (1, k) is put into the first 3×|n|array, column k, row 2; if not, executing step S33;

s33, judging an assembly task chrom_B1 ^p (1, k) the direction attribute, if the direction attribute is one-sided, i.e. can be placed only to the left or right of the workstation s, the assembly task Chrom_B1 ^p Setting the task direction information of (1, k) to correspond to a single side, and placing the task direction information on the kth column 2 of the first 3×|n|array; setting assembly task chrom_b1 ^p The average end time of (1, k) isAnd will->Put in the first 3×|n|array kth column 3 rd row; wherein (1)>If the direction attribute is bilateral, i.e. both left and right sides can be placed, the step S34 is executed for the job average end time of the task j;

s34, calculating an assembly task Chrom_B1 ^p (1, k) average end time when placed on left and right sides of the workstation, respectively, judging the assembly task Chrom_B1 ^p (1, k) whether the average end moments obtained by the placement on the left and right are the same, and if so, executing step S35; if not, executing step S36;

s35, optionally selecting one side as an assembly task Chrom_B1 ^p (1, k) and placing the task direction information on the first 3×|n|array kth column 2 nd row while placing the average end time at the first 3×|n|array kth column 3 rd row;

s36, selecting a single side with the smallest average finishing moment as an assembly task zerom_B1 ^p (1, k) and placing the task direction information on the first 3×|n|array kth column 2 nd row while placing the minimum average end time at the first 3×|n|array kth column 3 rd row;

s37, repeating the steps S32-S36 until the initial chromosome chrome_B0 ^p The average end time information of the |N| assembly task in the sequence is obtained completely, and an initial chromosome chrome_B0 is obtained ^p Corresponding first chromosome Chrom_B1 ^p 。

5. The second-class balance intelligent optimization method of the variable-pitch synchronous mixed-flow double-sided assembly line according to claim 4, wherein in step S33The method is characterized in that the job average ending time of a task j is selected by the following steps:

s331, assembling task Chrom_B1 ^p (1, k) before the single side corresponding to the task direction information is put in, extracting the current last assembly task of the single side, and marking the last assembly task as a;

s332 screening out the other single side with priority higher than the assembly task chrom_B1 ^p All the assembly tasks of (1, k) and the assembly task with the smallest priority is proposed from the assembly tasks and is marked as b;

s333, inquiring the job average ending time of the assembly tasks a and b from the MTTM, and selecting the assembly task with the largest job average ending time as a task j.

6. The method for intelligent optimization of a second type of balance of a variable pitch synchronous mixed-flow double-sided assembly line according to claim 1, wherein the method is characterized by comprising the following steps of ^p The work station is assigned to all the assembly tasks in the sequence to obtain a second chromosome chrome_B3 ^p Comprising:

s41, setting a second 3×|N|array, and enabling the first chromosome to be chrom_B1 ^p Copy the first and second lines of information to the first and second lines of the second 3 x|n|array;

s42, obtaining a first chromosome Chrom/uB1 ^p Last assembly task placed on left and right side respectively and />And the job average end time of these two assembly tasks +.> and />Setting a time limit, expressed as:

wherein ,represents the right time limit, +_>Representing a left time limit; the left time interval of the workstations S, s= {1,2, …, |s| } is assigned +.>The right time interval is +.>

S43, inquiring the corresponding assembly task chrom_B2 of the kth column of the second 3×|N|array ^p Average end time of (1, k)And calculates the corresponding average starting moment +.>

S44, if the task is to be assembled, the task is to be chromam_B2 ^p (1, k) the average start time and the average end time are both within the time interval of the workstation s, the assembly task Chrom_B2 ^p (1, k) assigning to workstation s and writing information of workstation s in the second 3×|n|array, column k, row 3;

s45, if the task is to be assembled, the task is to be chromam_B2 ^p If either of the average start time and the average end time of (1, k) is not within the time interval of the workstation s, the assembly task chrom_b2 is further determined ^p (1, k) execution time period on the workstation s and the adjacent workstations, the task Chrom_B2 is assembled ^p (1, k) designating the workstation corresponding to the maximum execution time length, and writing information of the workstation corresponding to the maximum execution time length in the 3 rd row of the kth column of the second 3×|N|array;

s46, repeating the steps S43-S45 until the first chromosome chrom_B1 ^p The work station information of the I/N assembly task is all acquired to obtain a first chromosome chrom_B1 ^p Corresponding second chromosome Chrom_B2 ^p 。

7. The second-class balance intelligent optimization method of the variable-pitch synchronous mixed-flow double-sided assembly line according to claim 1, wherein the method is based on a second chromosome chrome_b2 ^p The workstation designated result of (2) is utilized to obtain the completion time of each assembly task of each product by using a task operation time matrix TTM to obtain a final chromosome chrome_B3 ^p The process comprises the following steps:

s51, creating an array of (|M|+3) ×|N| and using a second chromosome chrome_B2 ^p Copy the first, second, and third lines of information to the first, second, and third lines of (|M|+3) ×|N| array;

s52, judging the (M+3) X N arrayk columns correspond to assembly task chrom_b3 ^p (1, k) whether it is the first task to the left or right of its designated workstation s; if yes, inquiring each product from the operation time matrix TTM to execute the assembly task chrom_B3 ^p The working time of (1, k); the execution of the assembly task chrom_b3 by placing |m| products respectively from row 4 of the kth column of the (|m|+3) ×|n|array ^p The working time of (1, k); if not, executing step S53;

s53, in the query workstation s, the task of assembly is Chrom_B3 ^p All assembly tasks before (1, k), from which assembly task chrom_b3 is screened out ^p (1, k) last assembly task x on the one side to which it belongs, and on the other side the assembly task chrom_b3 ^p (1, k) the last assembly task y with priority relationship; acquiring the working time of the I M I products for executing the assembly task x and the assembly task y respectively, comparing the two working time end moments of each product, and selecting the maximum working time end moment of each product as the self assembly task chrom_B3 ^p Time base line in (1, k), and performing S54;

s54, executing an assembly task Chrom_B3 on the |M| products based on the time base line of the step S53 ^p The job time of (1, k) is added to the own corresponding time base line, and all the addition results are placed in the 4 th to the |m|+3 th rows of the k-th column of the (|m|+3) ×|n|array;

s55, repeating the steps S52-S54 until the second chromosome chrome_B2 ^p Task end time information of |M| products of |N| assembly task in the first chromosome is obtained completely to obtain a second chromosome chrome_B2 ^p Corresponding final chromosome chrom_b3 ^p 。

8. The second-class balance intelligent optimization method of the variable-rhythm synchronous mixed-flow double-sided assembly line according to claim 1, wherein the value evaluation result of any one final chromosome is calculated according to a variable-rhythm synchronous mixed-flow double-sided assembly mode, and the method comprises the following steps:

s61, obtaining a final chromosome chrome_B3 ^p Task end time of each product executing each assembly task on each work station S, s= {1,2, …, |s| } is selectedTaking the maximum value in the task end time of the M-th= {1,2, …, |M| } products on the workstation s for executing all assembly tasks as the station beats of the M-th products on the workstation s for production;

s62, dividing the execution condition of the production sequence Q in the variable-pitch synchronous mixed-flow double-side assembly line into |Q| production line states; when the product of the production sequence Q of the u-th type is produced in the last working station of the assembly line, the type condition of the product produced by each working station of the assembly line is defined as the state of the u-th production line;

s63, calculating the output time of the state of the production line of the u-th production line, wherein the calculation formula is as follows:

wherein ,ts_μ Represents the production time of the μ= {1,2,..+ -. Q| } production line status,representing product Q _q Station beat at the s-th workstation, Q _q E M represents the product type at the q-th position of the production sequence;

s64, carrying out weighted average operation on the output time of all production line states to obtain a final chromosome chrome_B3 ^p Is expressed as:

the value evaluation result CT is expressed as an average production rate of each product in the production sequence Q in the assembly line.