CN110580547B

CN110580547B - Method for arranging parallel incomplete disassembly lines for disassembling waste products

Info

Publication number: CN110580547B
Application number: CN201910811495.9A
Authority: CN
Inventors: 张则强; 朱立夏; 管超; 刘思璐; 蒋晋; 谢梦柯
Original assignee: Southwest Jiaotong University
Current assignee: Southwest Jiaotong University
Priority date: 2019-08-30
Filing date: 2019-08-30
Publication date: 2022-02-22
Anticipated expiration: 2039-08-30
Also published as: CN110580547A

Abstract

The invention discloses a method for setting parallel incomplete dismantling lines for dismantling waste products, comprising the following steps: (1) constructing a mathematical model aiming at minimizing the dismantling depth, the number of workstations, the balance index of idle time of workstations and the quantity of dismantling resources (2) Generate the initial population; (3) Obtain the neighborhood individuals of the initial population; (4) Compare the objective function value of the mixed population composed of the neighborhood individuals and the population individuals through Pareto, update the population and optimize the Pareto The solution is output to the external file; (5) the population updated in step (4) is subjected to simulated annealing operation to obtain a new population; (6) Pareto is used to compare the objective function value of the mixed population composed of the population after simulated annealing operation and the external file, Output the Pareto optimal solution to the external file; (7) Repeat steps (3)-(6) for a predetermined number of times; (8) Output the Pareto optimal solution in the external file as the disassembly task assignment plan. The invention is simple to implement and has strong search ability.

Description

Method for arranging parallel incomplete disassembly lines for disassembling waste products

Technical Field

The invention relates to the technical field of product disassembly, in particular to a method for setting a parallel incomplete disassembly line for disassembling waste products.

Background

Along with the continuous development of social economy, the material living standard of people is continuously improved, the updating speed of various products is continuously increased, and according to statistics, the quantity of theoretically scrapped household appliances per year in China reaches 1 to 1.2 hundred million, and the quantity of the theoretically scrapped household appliances per year is continuously increased by 20 percent on average per year. If the waste products are directly buried or burned, the environment is badly affected, and the waste of resources is caused. In order to minimize environmental pollution caused by waste products and to sufficiently extract recyclable resources contained in the waste products, the waste products must be recovered and remanufactured. The disassembly is a key ring for implementing recovery and remanufacture of waste products, and in order to adapt to large-scale disassembly production, the disassembly line becomes a preferred disassembly production mode of each disassembly enterprise. In order to improve the operating efficiency of the disassembly line, the workload of each disassembly station should be as full and balanced as possible, and therefore, the disassembly line balancing problem has been proposed and widely studied.

The current research on the balance problem of the disassembly line is mainly focused on the balance problem of the complete disassembly line in the form of a single linear layout, and the balance problem of the complete disassembly line in the form of the single linear layout is researched on how to distribute a series of disassembly tasks of one or more products to be disassembled to a plurality of assembly line workstations arranged in the form of the linear layout so as to meet one or more optimization targets, however, in practical dismantling enterprises, the dismantling process flows of different products are greatly different and are generally dismantled in different production lines, and in order to improve the disassembly yield and reduce the disassembly cost, the waste products are generally not completely disassembled from the whole product to each part, but the disassembly depth is determined according to the EOL (End-of-life) condition of the product and the requirements and harmfulness of each part. Generally, the parts of the waste products that can be reused and remanufactured need to be removed as required parts, and in addition, in view of environmental protection, the harmful parts, whether used for reuse, remanufactured or recycled, must be removed, for example, when a waste CRT (Cathode Ray Tube) television is disassembled, the screen glass must be removed due to the existence of fluorescent powder substances, other parts may be removed or not, and after all the parts that need to be disassembled are removed, the remaining undetached parts enter the crushing and sorting stage in an integrated manner to extract raw materials, so that the actual disassembling enterprises usually adopt an incomplete disassembling manner for disassembling production.

In order to improve the labor hour utilization rate and the Disassembly efficiency of workers, Seda Hezer et al put forward a Parallel Disassembly Line Balancing Problem (PDLBP) for the first time, and solve by using a shortest path model based on a Disassembly priority relationship network diagram with the minimum number of work stations as an optimization target. Seda Hezer et al indicate that parallel disassembly lines are superior to single straight disassembly lines in reducing idle time of the work stations, reducing the number of open work stations, reducing the number of disassembly workers, reducing line length, and improving production efficiency.

The parallel disassembly line balance problem belongs to an NP-hard (NP is all called non-deterministic polynomial, and NP-hard refers to a problem that all NP problems can be reduced within polynomial time complexity) combined optimization problem as the single linear disassembly line balance problem which is researched more, and the solution complexity of the problem increases exponentially along with the increase of the problem scale. Seda Hezer et al apply a heuristic method to solve PDLBP, although the heuristic algorithm is simple and easy to implement, and can quickly obtain a near-optimal solution or even an optimal solution for a problem of a smaller scale, the solution results are far from the ideal when solving a problem of a large scale. At present, the problem of NP-hard combined optimization is solved by using a meta-heuristic algorithm such as a genetic algorithm, a simulated annealing algorithm, an ant colony algorithm and the like, but the problems of poor convergence, long search time, poor solving quality and the like exist. Therefore, a more efficient method is sought to solve PDLBP to obtain a split line setup scheme.

Disclosure of Invention

The invention mainly aims to provide a setting method of a parallel incomplete disassembly line for disassembling waste products so as to obtain a better setting scheme of the disassembly line.

In order to achieve the above object, the present invention provides a setting method of a parallel incomplete disassembly line for disassembling waste products. The setting method of the parallel incomplete disassembly line for disassembling the waste products comprises the following steps:

(1) constructing a mathematical model with the aim of minimizing the disassembly depth, the number of workstations, the idle time balance index of the workstations and the quantity of disassembly resources;

(2) generating an initial population;

(3) acquiring neighborhood individuals of the initial population;

(4) comparing objective function values of a mixed population consisting of the neighborhood individuals and the population individuals through Pareto, updating the population and outputting a Pareto better solution to an external archive;

(5) carrying out simulated annealing operation on the population updated in the step (4) to obtain a new population;

(6) comparing the objective function value of a mixed population consisting of the population subjected to the simulated annealing operation and an external file by using Pareto, and outputting a Pareto better solution to the external file;

(7) repeating the steps (3) - (6) according to the set times;

(8) outputting Pareto in the external file as a disassembly task allocation scheme.

The method takes the practical application background of disassembly production, designs a parallel incomplete disassembly line on the basis of the existing single linear disassembly line, and constructs a multi-objective mathematical model aiming at the balance problem of the parallel incomplete disassembly line, wherein the multi-objective mathematical model takes the minimum disassembly depth, the minimum number of workstations, the minimum idle time balance index of the workstations and the minimum disassembly resource number as the targets; the multi-target mixed group neighborhood searching algorithm is creatively provided, namely step three, the method is simple to realize and high in searching capability, and the global optimization capability of the algorithm is further enhanced by introducing simulated annealing operation so as to obtain a better solving effect.

The invention is further described with reference to the following figures and detailed description. Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to assist in understanding the invention, and are included to explain the invention and their equivalents in the description and should not be taken to limit the invention to the proper form disclosed herein. In the drawings:

fig. 1 is a schematic view of a parallel disassembly line in the setting method of a parallel incomplete disassembly line for disassembling waste products of the present invention.

Fig. 2 is a schematic diagram of an adjacent domain search based on an embedding operation in the setting method of the parallel incomplete disassembly line for disassembling the waste products of the present invention.

Fig. 3 is a schematic diagram of neighbor search based on exchange operation in the setup method of parallel incomplete disassembly line for disassembling waste products of the present invention.

FIG. 4 is a flowchart of a multi-target mixed group neighborhood search algorithm in the method for setting a parallel incomplete disassembly line for disassembling scrap products according to the present invention.

Fig. 5 is a view showing the relationship of the priorities of the tasks of disassembling the used refrigerator according to embodiment 1 of the present invention.

Fig. 6 is a priority relationship diagram of the disassembly tasks of the used tv set in embodiment 1 of the present invention.

Fig. 7 is an operation allocation diagram obtained by using the Pareto better solution scheme with the solution number of 1 in embodiment 1 of the present invention.

Detailed Description

The invention will be described more fully hereinafter with reference to the accompanying drawings. Those skilled in the art will be able to implement the invention based on these teachings. Before the present invention is described in detail with reference to the accompanying drawings, it is to be noted that:

the technical solutions and features provided in the present invention in the respective sections including the following description can be combined with each other without conflict.

Furthermore, the embodiments of the present invention described in the following description are generally only a part of the embodiments of the present invention, and not all of the embodiments. Therefore, all other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without any creative effort shall fall within the protection scope of the present invention.

With respect to terms and units in the present invention. The terms "comprising," "having," and any variations thereof in the description and claims of this invention and the related sections are intended to cover non-exclusive inclusions.

The product types that the enterprise was retrieved through different channels are disassembled in different, but these products need be unified to be dismantled, and present dismantlement line layout form is that the same kind of product is dismantled on solitary linear type assembly line, nevertheless through investigation on the spot we discover, and the workman on the line is dismantled to the list has the busy uneven condition of idle time, and partial workman has idle time longer to lead to workman man-hour utilization ratio low, has also led to the waste of resource simultaneously. To this end, the present invention provides a parallel production line for the simultaneous disassembly of different products as shown in fig. 1, the first disassembly line 21 of the first to-be-disassembled product 11 and the second disassembly line 22 of the second to-be-disassembled product 21 are arranged in parallel at respective adjacent positions, and through the uniform distribution of all the disassembly tasks for disassembling the products on the two lines, the worker in the workstation 3 can operate one disassembly line alone or operate the operations on the two disassembly lines simultaneously, thereby reducing the idle time of the worker as much as possible, improving the man-hour utilization rate of the worker, and the residual undetached parts will enter the raw material extraction chamber 4 in an integrated manner for the extraction of the raw material.

In order to reduce the disassembly cost as much as possible and improve the disassembly benefit, and simultaneously reduce the harm of the disassembly of the waste products to the environment to the minimum, the products on the parallel disassembly line are not completely disassembled, specifically, on the basis of following the disassembly priority relation of the products, the worker stops the disassembly after completely disassembling the required and harmful parts in the waste products along a certain disassembly path, and the rest parts enter a crushing and sorting machine in an integral mode to refine and sort the raw materials.

In order to improve the production efficiency of a parallel disassembly line and aim at the balance problem of a parallel incomplete disassembly line, a multi-objective mathematical model which takes the minimum disassembly depth, the minimum work station number, the minimum idle time balance index and the minimum quantity of resources required by disassembly as targets is established, an optimal scheme for synchronously distributing all disassembly tasks on the parallel line is sought under the constraint of the priority relation of the disassembly tasks of all products and the constraint conditions of part requirements, hazards and the like, the number of the opened work stations is minimized while the disassembly depth is minimized, the work loads in all the work stations are full and balanced as much as possible, and the quantity of the resources used for disassembly is minimized.

Specifically, the setting method of the parallel incomplete disassembly line for disassembling the waste products comprises the following specific steps:

1. assumption of conditions

(1) Two parallel lines respectively dismantle a waste product;

(2) enough waste products are arranged on the disassembly line, so that the parallel disassembly line can continuously run;

(3) the operation time of each disassembly task of each product is determined and known, and uncertain factors in operation are ignored;

(4) each workstation is allocated with one worker, and each worker is a multi-skill worker and can be competent for all disassembly tasks;

(5) neglecting the walking time of the disassembling worker between the two parallel lines;

(6) other burst conditions are ignored.

2. Description of the symbols

For convenience of description, the symbols referred to in the present invention have the meanings shown in table 1.

TABLE 1

3. Mathematical model

Constructing a multi-objective mathematical model of a parallel incomplete disassembly line balance problem with the objective of minimizing disassembly depth, minimizing the number of workstations, minimizing idle time balance indexes of workstations and minimizing the number of disassembly resources as follows:

F＝min[f₁,f₂,f₃,f₄] (1)

constraint conditions are as follows:

in the above mathematical model: equation (1) represents four objective functions f to be optimized₁,f₂,f₃,f₄(ii) a Target f₁The formula (2) shows that the disassembly depth is minimized, namely, in order to reduce the disassembly cost as much as possible and increase the disassembly benefit, the disassembly depth is minimized as much as possible, namely, the number of the disassembled parts is minimized; target f₂Minimizing the number of open stations as shown in equation (3); target f₃The method is represented by the formula (4), namely the idle index is minimized, namely the work load of each workstation is enabled to be as full and balanced as possible; target f₄The method is represented by the formula (5) that the number of resources required for disassembly is minimized, namely, the disassembly tasks using the same resources are distributed to the same workstation as much as possible;

in the above constraints: equation (6) indicates that the disassembly tasks are inseparable, each disassembly task being allowed to be assigned to only one workstation; formula (7) indicates that the hazardous part must be removed; formula (8) indicates that the required component must be removed; equation (9) indicates that the sum of the time allotted to all the disassembly tasks in each workstation k cannot exceed a predetermined tact time; equation (10) indicates that the allocation of disassembly tasks on each disassembly line must follow the respective disassembly priorities for each product; equation (11) indicates that if a certain task j on the tear-down line l is executed, the task i immediately before the task j must also be executed; equation (12) indicates that the workstations are turned on in sequence, and there are no empty workstations to which tasks are not assigned; equation (13) indicates that if a tear down task i using resource r is allocated to a workstation k, the workstation k must be equipped with the corresponding resource r accordingly.

4. And solving a disassembly task allocation scheme by adopting a multi-target mixed group neighborhood search algorithm creatively provided.

4.1 population initialization

The multi-target mixed group neighborhood searching algorithm provided by the invention is an algorithm based on group neighborhood searching, and a group needs to be initialized before the algorithm enters iteration. Because the solved parallel incomplete disassembly line balance problem has a discrete characteristic, each individual in the population corresponds to a solution of the solved problem by adopting a disassembly task-based coding mode, namely corresponds to a disassembly task sequence, the population initialization process also corresponds to a generation process of a series of disassembly task sequences, and in the decoding stage, the series of disassembly task sequences are sequentially distributed to a plurality of workstations according to the production beat constraint of the disassembly line to determine the objective function value of each optimization index, namely determine the multi-objective solution scheme of the solved problem.

In order to ensure the diversity of the initial population, the invention adopts a random mode to generate the initial population, the initial population follows the priority order relation constraint between the disassembly tasks, and the operation order on two disassembly lines needs to be considered simultaneously.

The specific process for generating the initial population Pop is as follows:

step 1: aiming at each current population pop _ num, according to respective operation priority relation matrixes P of disassembled products on the disassembling line I and the disassembling line II₁And P₂Respectively finding out the job in all the disassembling tasks as empty or the job in which the job in the immediate front is distributed, i.e. respectively finding out P₁And P₂The sum of all the column elements in the task set C is 0, and the task set C to be allocated is formed;

step 2: randomly selecting a disassembly task i from a task set C to be distributed to the current position of the pop _ num of the current population individual;

step 3: if the disassembly task I is a job on the disassembly line I, then all immediate constraints associated with the disassembly task I are removed, i.e., P is about to occur₁Setting a row element where the middle disassembly task i is located as 0; if the disassembly task i is a job on the disassembly line II, all immediate constraints associated with the disassembly task i are released, i.e. P is released₂Setting the row element of the middle task i as 0;

step 4: repeating Step 1 to Step 3 until all the operations on the two disassembly lines are distributed;

step 5: repeating Step 1 to Step 4 until all Pop _ num population individuals are initialized;

and (3) outputting: the initial population Pop is the number of Pop _ num;

4.2 neighborhood search operation

In order to ensure the quality of neighborhood search and the diversity of neighborhood individuals, the neighborhood search of the population individuals is carried out by respectively adopting the optimal embedding operation and the optimal exchange operation, namely Pareto comparison is carried out on all feasible solutions generated by embedding operation and exchange operation of a certain disassembly task arbitrarily selected in each disassembly task sequence, and an optimal solution is selected as a final neighborhood solution. For number S₁The population individuals generate neighborhood individuals through optimal embedding operation, and the number of the neighborhood individuals is S₂Generating neighborhood individuals by optimal exchange operation of population individuals, S₁+S₂Pop _ num. Preferably, S₁＝S₂At this time, the objective function value of the obtained disassembly task allocation scheme is small.

Fig. 2 is a schematic diagram of neighborhood search based on an embedding operation, in which an item of task 6 is randomly selected in a current solution sequence, such as [1-2-3-4-5-6-7-8], and tasks immediately before and after the task 6 are determined to be 3 and 7 respectively through a job priority order relationship matrix, the task 6 may be embedded in any one of

positions

1 and 2 indicated by dotted arrows, so that two neighborhood solutions may be generated, and a better solution is screened as a final neighborhood individual through Pareto comparison in the two neighborhood solutions.

Fig. 3 is a schematic diagram of neighborhood search based on a swap operation, in which if task 6 swaps its location with task 2, the task priority relationship constraint is violated, and the generated task sequence is not feasible, so task 6 only allows a certain task corresponding to two

locations

1 and 2 shown by dotted lines to swap, and thus two neighborhood solutions can be generated, and similarly, a Pareto better solution is screened as a final neighborhood individual from the two neighborhood solutions generated by the swap operation.

4.3 population renewal

After a current population individual generates a neighborhood individual through neighborhood search operation, performing Pareto comparison and screening on a mixed population composed of the population individual and all neighborhood individuals, and reserving screened Pareto better solutions, if the number of the Pareto better solutions exceeds the population number, calculating crowding distances of the screened Pareto better solutions in order to obtain Pareto better solutions with more uniform spatial distribution, then sorting all the Pareto better solutions according to the calculated crowding distances from large to small, and selecting the Pareto better individuals with the population number from the Pareto better individuals to form a new generation of population; if the number of the screened Pareto better solutions is smaller than the number of the populations, randomly selecting a certain number of individuals from the remaining mixed populations after screening, and forming a new generation of population with the screened Pareto better solutions.

The calculation formula of the crowding distance is as follows:

in the formula, CD_aRepresenting the crowding distance of the Pareto better individual a, U being the number of optimization targets, X_{Pop_num}And X₁Respectively representing individuals having the maximum value and the minimum value in the u-th objective function, the crowding distance of the two individuals being defined as infinity, f_u(X_a+1) And f_u(X_a-1) The u-th objective function values of two adjacent individuals a +1 and a-1 of the individual a respectively.

Pareto comparison procedure is as follows: calculating the objective function value of each individual in the mixed population according to the sequence number X₁、 X₂、···、X_nNumbering; firstly, X is put in₁Deposit into empty matrix M₀Then X is added₂Is an objective function value X of₁Performing Pareto comparison on the objective function value of (1), if X is₁Quilt X₂Dominant, i.e. X₂＜X₁Then X will be₂Adding M₀Neutralizing and reacting X₁Slave matrix M₀Deletion is then carried out by X₃And X₂Comparing the objective function values of (1); if X₂Quilt X₁Dominant, i.e. representing X₂If not, deleting X₂Then X is carried out₃And X₁Comparing; if X₁And X₂If they do not dominate each other, then X will be₂Adding M₀In then X₃One by one with X₁And X₂And (6) comparing. And so on. Final M when Pareto comparisons were completed for all individuals in the mixed population₀The selected Pareto is the better solution of the population individuals.

4.4 simulated annealing operation

The updating and algorithm iteration are carried out through the neighborhood search operation of population individuals, a better search effect is achieved in the early stage, but the local optimization is easy to fall into in the later stage, and in order to improve the global search performance of the algorithm, a local search operation combined with simulated annealing is added in the algorithm, so that the algorithm receives the worse solution in the optimization searching process with a certain probability, and the algorithm jumps out of the local optimization. Since the disassembly line balance problem researched by the invention is a multi-objective optimization problem, the Metropolis criterion needs to be improved, and the improvement is as follows: if new solution X_nAnd the current solution X_cAre not dominant or newly solved by X_nGoverning the current solution X_cThen receive X_nSubstitution of X_c(ii) a If the current solution X_cGoverning a new solution X_nThen a random number rand within the interval (0,1) is randomly generated, if rand<And P, accepting the new solution to replace the current solution, and abandoning the new solution if not, wherein P is an acceptance probability, and the calculation formula is shown as the following formula (14):

wherein, U is the number of optimization targets, and U is 4; u is the {1,2,3, 4}, f_u(X_c) For the current solution X_cValue of the u-th objective function, f_u(X_n) New solutions X generated for perturbations_nThe corresponding u-th objective function value, T, is the temperature at the current iteration.

The simulated annealing operation of each population individual after population updating comprises the following specific steps:

step 1: for current population individual X_cGenerating a new solution X by means of a swap operation_n；

Step 2: if X_nDominating X_cOr X_nAnd X_cDo not dominate each other, then X_c＝X_n(ii) a If X_cDominating X_nEntering Step 3;

step 3: randomly generating a random number rand within a range (0,1) if P>rand, i.e. X_c＝X_nOtherwise, give up new solution X_n。

To sum up, the flow of the multi-objective mixed group neighborhood search algorithm is shown in fig. 4, and the specific implementation steps are as follows:

step 1: initializing algorithm parameters, including: iteration times Iter of the algorithm, population quantity Pop _ num, population individual neighborhood solution quantity neighbor _ count, and simulated annealing operation initial temperature T₀Temperature reduction coefficient q, chain length L_k(ii) a Inputting question parameters, including: the takt time CT of the disassembly line, wherein the relevant information of each disassembly task comprises the operation time, the hazard attribute, the demand attribute, the disassembly resource type and the priority sequence of the disassembly tasks;

step 2: randomly initializing population individuals according to the method in section 4.1 to generate an initial population Pop;

and step 3: setting an iteration count iter as 1, and entering algorithm iteration;

and 4, step 4: as described in section 4.2, neighbor search based on the optimal embedding operation is performed on each population individual in the first half to generate neighbor _ count neighbor individuals, and neighbor search based on the optimal exchanging operation is performed on each population individual in the second half to generate neighbor _ count neighbor individuals;

and 5: as described in section 4.3, Pareto comparison is performed on a mixed population composed of the current population individuals and all neighborhood individuals, Pareto better solutions are screened, the screened Pareto better solutions are stored in an external archive Q, and then the population is updated;

step 6: performing simulated annealing operation on each population individual updated in the step 5 as described in section 4.4 to obtain a new population;

and 7: performing Pareto comparison on a mixed population consisting of the new population subjected to the simulated annealing operation and the external archive, screening out a Pareto better solution, and outputting the Pareto better solution to an external archive Q;

and 8: if Iter is less than Iter, making Iter equal to Iter +1, and returning to step 4; if Iter count Iter is equal to Iter, go to step 9;

and step 9: and (4) stopping the algorithm, and outputting the Pareto better solution in the final external file as a disassembly task allocation scheme.

The setting method of the parallel incomplete disassembly line for disassembling the waste products utilizes the established mathematical model and the multi-target mixed group neighborhood search algorithm to be applied to the parallel disassembly line of two waste products, and can perform integrated line balance optimization on the disassembly task flows of the two waste products. The advantageous effects of the present invention are illustrated below by specific examples.

Example 1

The information of the waste refrigerator and the waste television of a certain enterprise obtained by research is as follows:

1. the data information of the disassembly task of the used refrigerator is shown in table 1, and the data information of the disassembly task of the used television is shown in table 2, wherein: in the hazard attribute, "0" indicates no hazard, and "1" indicates hazardous; in the requirement attribute, "0" indicates no requirement, "1" indicates requirement; disassembling resources means disassembling tools corresponding to various disassembling tasks, 1 represents a screwdriver, 2 represents scissors, 3 represents a word screwdriver, 4 represents a clamp, and 5 represents fixed heavy equipment tools such as a cutting machine, a punching machine and the like.

TABLE 1

TABLE 2

2. The priority relation diagram of the disassembly tasks of the waste refrigerator is shown in fig. 5, and the priority relation diagram of the disassembly tasks of the waste television is shown in fig. 6; taking

disassembly tasks

2 and 3 in FIG. 5 as examples, disassembly task 2 is the immediately preceding task of disassembly task 3, pl_ij＝pl_2,3＝1，pl_ij＝pl_3,2＝0。

3. The disassembly production beats of the waste refrigerator and the waste television are respectively 130s and 60 s.

The multi-target mathematical model of the refrigerator and television parallel incomplete disassembly line balance problem is solved by applying the setting method of the parallel incomplete disassembly line for disassembling the waste products, and parameters of a multi-target mixed group neighborhood search algorithm are set as follows: the population quantity Pop _ num is 50, the iteration number Iter is 200, the population individual neighborhood solution quantity neighbor _ count is 3, and the simulated annealing operation initial temperature T₀100, cooling coefficient q 0.9, chain length L_kThe takt time CT for the parallel detachment of the wires is taken to be 780s, the least common multiple of the takts of the two products. The 34 Pareto better solution sets (solution numbers 1-34) of the solved parallel disassembly line on the four optimization indexes are shown in table 3.

TABLE 3

TABLE 4

TABLE 5

In order to prove the advantages of the proposed parallel disassembly line compared with the single linear disassembly line, a Pareto better solution set of the single linear disassembly line of the two products on four optimization indexes is solved on the premise of keeping the algorithm parameters unchanged, wherein 9 Pareto better solutions exist in the waste refrigerator as shown in table 4, and 16 Pareto better solutions exist in the waste television as shown in table 5.

The results of the comparison of table 3 with tables 4 and 5 are as follows:

(1) all Pareto better solution schemes solved by parallel layout mode are in f₂The targets are 8, if the two products are independently optimized, the number of required workstations is 9, and 1 workstation is started more than that in a parallel layout mode;

(2) parallel layout mode Pareto better solution scheme 13 at target f₃The optimal result is 5797, and if the two products are independently optimized, all Pareto better solution schemes are obtained in the balance index f₃Is more than 20000, is obviously inferior to the parallel layout mode;

(3) parallel layout mode Pareto better solution schemes 23 and 30 at target f₄The optimal result is 13, if the two products are independently optimized, the Pareto

optimal solution schemes

7 and 9 obtained by independently optimizing the refrigerator disassembly line are in the target f₄The optimal result is 9, and the Pareto better solution scheme obtained by independently optimizing the television dismounting line is in the target f₄The optimal result is 8, so the total amount of the required disassembly resources under the condition of individual optimization is 17;

(4) the number of the Pareto parallel layout modes is as much as 34 compared with the optimal solution schemes, so that more selection schemes can be provided for enterprises, and the parallel layout modes have higher practicability than a single linear dismounting line.

In conclusion, the parallel disassembly line layout mode is better than the optimized results of the single linear disassembly line of two products in four indexes of reducing the number of workstations, reducing the idle time of the workstations, balancing the operation load on each workstation and reducing the use number of disassembly resources.

In this embodiment, Step 1 in the simulated annealing operation adopts pairwise exchange operation, i.e. randomly selecting X_cTwo disassembly tasks in the sequence are exchanged to obtain a new sequence, and if the new sequence meets the priority relation between the disassembly tasks, the new sequence is the new solution X_nIf the new sequence does not meet the priority relation between the disassembly tasks, two disassembly tasks are reselected for exchange until a new solution X is obtained_n。

The job assignment map obtained by using the Pareto preferred solution with the solution number of 1 is shown in fig. 7, for example, workstation 1 only completes the disassembly tasks 1-2 of the tv, workstation 2 only completes the disassembly tasks 1-2, 14-15 and 18 of the refrigerator, and workstation 3 completes both the disassembly task 4 of the tv and the disassembly tasks 3-4, 8 and 21 of the refrigerator; the television removal line does not require the removal tasks 20 and 27 and the refrigerator removal line does not require the removal tasks 25, thereby achieving incomplete removal.

The explanation about "dominance" is as follows: unlike a single target, the mutual constraint and limitation among the sub-targets of multiple targets cannot make all the sub-targets optimal at the same time. In order to select a relatively better scheme, the method for judging the fitness function of the disassembly sequence by combining the Pareto thought is as follows: suppose j' th of two disassembly sequences A, B

The function values of the sub-targets are AF_jAnd BF_jK th, k

The function values of the sub-targets are AF_kAnd BF_kIf AF_jAnd BF_j、 AF_kAnd BF_kSatisfies the following conditions: AF_j≤BF_j、AF_k＜BF_kThen, it is called that a dominates B, B is the dominated solution, a is the non-inferior solution, i.e., the superior solution, and the solution set formed by all the non-inferior solutions is the Pareto solution set.

The contents of the present invention have been explained above. Those skilled in the art will be able to implement the invention based on these teachings. All other embodiments, which can be derived by a person skilled in the art from the above description without inventive step, shall fall within the scope of protection of the present invention.

Claims

1. A method for setting up parallel incomplete dismantling lines for dismantling waste products, comprising the following steps:

(1) Build a mathematical model with the goal of minimizing the disassembly depth, the number of workstations, the equilibrium index of workstation idle time, and the number of disassembly resources;

(2) Generate an initial population with Pop_num;

(3) Obtain the neighborhood individuals of the population;

(4) Compare the objective function value of the mixed group composed of the neighborhood individuals and the population individuals through Pareto, update the population and output the Pareto optimal solution to the external file;

(5) performing a simulated annealing operation on the population updated in step (4) to obtain a new population;

(6) Use Pareto to compare the objective function value of the mixed population composed of the population after simulated annealing operation and the external file, and output the Pareto optimal solution to the external file;

(7) Repeat steps (3)-(6) according to a predetermined number of times;

(8) Output the Pareto optimal solution in the external file as the disassembly task assignment plan;

The mathematical model in step (1) is as follows:

Objective function: F=min[f ₁ , f ₂ , f ₃ , f ₄ ]

f ₁ is the function of minimizing the disassembly depth:

f ₂ is a function of the number of workstations turned on for minimization:

f ₃ is the equilibrium indicator function that minimizes the idle time of workstations:

f ₄ is a function that minimizes the number of dismantled resources:

In the above formula, l is the dismantling line number, l∈{1,2}; i is the dismantling task number; N _l is the number of dismantling tasks on the dismantling line l, k is the workstation number, k∈{1,2,…, K}, K is the number of workstations; x _ilk =1 indicates that the dismantling task i in the dismantling line l is assigned to the workstation k, otherwise x _ilk =0; Z _k =1 indicates that the workstation k is turned on, otherwise Z _k =0; CT is the takt time of parallel dismantling lines; t _il is the operation time of dismantling task i in dismantling line l; r is the number of _dismantling resource types; R is the number of dismantling resource types; , otherwise M _rk =0.

2. the setting method of the parallel incomplete dismantling line for dismantling waste products according to claim 1 is characterized in that: step (1) is carried out under the following constraints:

(1) The dismantling tasks are inseparable, and each dismantling task is only allowed to be assigned to one workstation;

(2) Hazardous parts must be removed;

(3) Parts in need must be removed;

(4) The sum of the time allocated to all dismantling tasks in each workstation cannot exceed the predetermined takt time;

(5) The assignment of dismantling tasks on each dismantling line must follow the respective dismantling priorities of each product;

(6) If a task on the dismantling line is executed, the predecessor task of the task must also be executed;

(7) The workstations are opened in sequence, and there is no empty workstation with unassigned tasks;

(8) If a dismantling task using a certain resource is assigned to a workstation, the workstation must be equipped with the resource accordingly.

3. the method for setting up the parallel incomplete dismantling line for dismantling waste products according to claim 1, is characterized in that: the concrete flow process of step (2) is as follows:

Step 1: For each individual pop_num of the current population, according to the respective job priority relation matrices P ₁ and P ₂ of the dismantled products on the dismantling line I and the dismantling line II, find out whether the immediate preceding job is empty or the immediate preceding job in all dismantling tasks, respectively. Jobs that have been assigned, that is, find out the tasks corresponding to the columns where the sum of all column elements in P ₁ and P ₂ is 0 respectively, and form the task set C to be assigned;

Step 2: Randomly select a dismantling task i in the task set C to be assigned and assign it to the current position of the current population individual pop_num;

Step 3: If the dismantling task i is an operation on the dismantling line I, remove all the immediate constraints associated with the dismantling task i, that is, set the row element where the dismantling task i is located in P1 to ₀ ; if the dismantling task i is For the job on the dismantling line II, all the immediate constraints associated with the dismantling task i are released, that is, the row element where the task i is located in P2 is set to ₀ ;

Step 4: Repeat Step 1 to Step 3 until all the jobs on the two dismantling lines are assigned;

Step 5: Repeat Step 1 to Step 4 until all Pop_num population individuals are initialized;

Output: initial population Pop;

Among them, pop_num is the number of population individuals, and Pop_num is the total number of population individuals.

4. The method for setting parallel incomplete dismantling lines for dismantling waste products according to claim 1, characterized in that: in step (3), a neighborhood is generated by an optimal embedding operation for the population individuals whose quantity is S ₁ Individuals, for the population individuals with the number S ₂ , the neighborhood individuals are generated through the optimal exchange operation, S ₁ +S ₂ =Pop_num.

5 . The method for setting parallel incomplete dismantling lines for dismantling waste products according to claim 4 , wherein: S ₁ =S ₂ . 6 .

6. the method for setting up the parallel incomplete dismantling line for dismantling waste products according to claim 1, is characterized in that: in step (4), population update process is as follows: if Pareto optimal solution quantity is greater than population quantity, then Calculate the crowding distance for the selected Pareto optimal solutions, and then sort according to the crowding distance from large to small, and select the Pareto optimal solutions of the population to form a new generation of populations; if the number of the selected Pareto optimal solutions is less than the population number, then A certain number of population individuals are randomly selected from the remaining mixed population after screening and the selected Pareto optimal solution forms a new generation population.

7. the method for setting up the parallel incomplete dismantling line for dismantling waste products according to claim 6, is characterized in that: the calculation formula of crowding distance is as follows:

In the formula, CD _a represents the crowding distance of the Pareto-optimal individual a, U is the number of optimization objectives, X _{Pop_num} and X ₁ represent the individual whose function value of the u-th sub-objective achieves the maximum and minimum values, respectively. The crowding distance of is defined as infinity, and f _u (X _a+1 ) and f _u (X _a-1 ) are the function values of the u-th sub-goal of two adjacent individuals a+1 and a-1 of individual a, respectively.

8. The method for setting up parallel incomplete dismantling lines for dismantling waste products according to claim 1, is characterized in that: the simulated annealing operation process of each population individual in step (5) is as follows:

Step 1: Generate a new solution X _n for the current population individual X _c through the exchange operation;

Step 2: If X _n dominates X _c or X _n and X _c do not dominate each other, then X _c =X _n ; if X _c dominates X _n , go to Step 3;

Step 3: Randomly generate a random number rand in the interval (0,1), if P>rand, then X _c =X _n , otherwise give up the new solution X _n ;

where P is the acceptance probability.

9. The method for setting up parallel incomplete dismantling lines for dismantling waste products according to claim 8, wherein:

Among them, U is the number of optimization objectives, U=4; u∈{1,2,3,4}, f _u (X _c ) is the u-th objective function value of the current solution X _c , f _u (X _n ) is the u-th objective function value corresponding to the new solution _Xn generated by the disturbance, and T is the temperature under the current iteration.