Production and maintenance coupling task allocation method and system considering equipment operating status

Abstract

The present invention discloses a method and system for allocating production and maintenance coupled tasks that considers the operating status of equipment. The system includes: a digital twin platform that simulates the allocation scheme obtained by the task allocation decision module to verify its feasibility; an equipment model management module that establishes a reliability model for the status of the equipment and establishes a numerical model for the production process of the equipment; an order receiving module that decomposes the order into production tasks for each equipment; a maintenance strategy management module that manages the maintenance strategy and selects the maintenance strategy corresponding to the equipment based on the current equipment reliability; a task allocation decision module that unifies the planning of production tasks and maintenance tasks and provides an allocation scheme; and an allocation result interaction module that outputs the final allocation scheme and interacts with the physical production line to provide production guidance for the physical production line. The present invention solves the problem of conflict between production and maintenance in the use of time resources, and achieves simultaneous and efficient production and maintenance.

Description

Technical Field

[0001] The present invention relates to the field of discrete manufacturing scheduling optimization and application, and in particular to a method and system for allocating production and maintenance coupling tasks taking into account equipment operating status. Background Art

[0002] In discrete manufacturing, a complex production system often handles the production and assembly of a wide variety of workpieces. Scheduling strategies and methods provide significant support for efficiently completing small-scale, customized production orders, as well as task allocation and decision-making.

[0003] At the same time, during the production and assembly process, the equipment will wear out as the operating time increases, which will affect the operating efficiency of the equipment. What's worse, it will cause the equipment to fail, resulting in unnecessary losses.

[0004] In actual manufacturing, depending on the scenario, the production department will pre-schedule orders based on the complexity and size of the current order using expert rules, heuristic rules, and various production scheduling algorithms. The maintenance department will improve maintenance efficiency through methods such as scheduled maintenance, unscheduled maintenance, scheduled overhauls, and fine-tuning maintenance periods based on equipment conditions. However, depending on the production scenario and the level of equipment sophistication, the priorities of production and maintenance will vary, creating a constant struggle between the two departments within limited working time.

[0005] Currently, discrete manufacturing production and maintenance departments perform independent task planning, neglecting coordination between scheduling tasks. This can lead to irrational maintenance or production plans, allocating time resources to departments that don't currently need them, and impacting overall work efficiency. In discrete manufacturing scenarios, production and maintenance cycles are often very short, making it particularly important to properly schedule these two tasks, which can sometimes conflict with each other.

[0006] Chinese patent document publication number CN114066312A discloses a production scheduling method based on discrete manufacturing, including: constructing a production scheduling model based on the goal of minimizing the weighted sum of time and cost; using an improved genetic algorithm to solve the scheduling optimal solution of the scheduling model to obtain a pre-scheduling plan; reviewing the pre-scheduling plan, and after passing the review, obtaining a scheduling plan, and assigning tasks according to the scheduling plan.

[0007] Chinese patent publication number CN115204022A discloses a data-driven discrete manufacturing workshop layout optimization decision-making method and system. The method includes: daily management and maintenance of basic user information and operation logs in the system, management and upgrading of the production database associated with the system, and regular updating and cleaning after classified storage to ensure and maintain the normal operation of the database; in the layout optimization module, different meta-heuristic algorithms are called due to different problem types and optimization target dimensions; the theoretical facility location, cost and non-logistics cost relationship calculation obtained after optimization by the aforementioned algorithm needs to be further simulated and verified; the selection of multi-attribute qualitative and quantitative evaluation indicators, the acquisition of evaluation matrices at each level, the calculation of weights and preference coefficients, the selection of decision methods, and the sorting of each solution in the solution set to obtain the most appropriate layout solution.

[0008] However, the above methods do not take the operating status of the equipment into consideration. Summary of the Invention

[0009] In response to the lack of coordination between production task scheduling and maintenance task scheduling in the existing discrete manufacturing industry, the present invention provides a production and maintenance coupled task allocation method and system that takes into account the operating status of equipment, which solves the problem of conflict between production and maintenance in the utilization of time resources and achieves simultaneous and efficient production and maintenance.

[0010] The technical solutions of the present invention are as follows:

[0011] A production and maintenance coupled task allocation system considering equipment operating status, comprising:

[0012] The digital twin platform performs virtual modeling and simulation of the physical production line, and simulates the allocation plan obtained by the task allocation decision module to verify its feasibility;

[0013] The equipment model management module establishes a reliability model for the equipment status based on the equipment mechanism and historical operation data, and establishes a numerical model for the equipment production process;

[0014] The order receiving module virtualizes and adapts orders from the physical production line and breaks them down into production tasks for each device;

[0015] Maintenance strategy management module manages maintenance strategies and selects corresponding maintenance strategies for equipment based on current equipment reliability;

[0016] The task allocation decision module unifies the planning of production tasks and maintenance tasks and provides allocation plans;

[0017] The allocation result interaction module outputs the final allocation plan and interacts with the physical production line to provide production guidance for the physical production line.

[0018] The present invention also provides a production and maintenance coupling task allocation method based on the production and maintenance coupling task allocation system considering the equipment operating status, comprising the following steps:

[0019] (1) In the digital twin platform, based on the historical data obtained from the actual production of the physical production line, the equipment state is modeled through the equipment model management module to obtain the equipment reliability model, and the equipment production process is modeled;

[0020] (2) Enter the pending orders and their expected delivery dates into the order receiving module, split the orders into production tasks through the order receiving module, and evaluate the operating hours of each device. Based on the evaluated equipment operating hours and the equipment reliability model, the total maintenance task frequency required for the equipment in the current production scenario is evaluated;

[0021] (3) Considering production tasks and maintenance tasks together as scheduling tasks, the task allocation decision module obtains an allocation plan, and then allocates tasks to each production equipment according to the allocation plan;

[0022] (4) When the task allocation decision module assigns a maintenance task to a device, the maintenance strategy management module selects the maintenance strategy based on the current device reliability and pushes it to the current maintenance task;

[0023] (5) The allocation plan is simulated and verified through the digital twin platform, and the simulation process and allocation plan are passed to the physical production line through the allocation result interaction module to guide actual production.

[0024] In step (1), the equipment reliability model is:

[0025]

[0026] Where F(t) represents the reliability of the device at time t; λ is the scale parameter, and μ is the shape parameter. These two parameters need to be set according to the actual condition of the device.

[0027] In step (2), the evaluation formula for the total maintenance task frequency is:

[0028]

[0029] Among them, H , is the total number of maintenance tasks for equipment r, n is the total number of workpieces to be produced, j is the workpiece index, P j,r is the production time of workpiece j on equipment r, F r -1 (0.8) represents the time required for the reliability of device r to drop to 0.8.

[0030] In step (3), the task allocation decision module obtains an allocation plan through an adaptive genetic algorithm for multi-task coupling, including:

[0031] (3-1) Randomly generate k production task individuals of length n·m, where m is the total number of equipment and n is the total number of workpieces to be produced; set the minimum number of algorithm iterations l g ;

[0032] (3-2) The special maintenance tasks are randomly shuffled and inserted into the individual production tasks to form a set of individual scheduling tasks;

[0033] (3-3) Perform simulation operations on each of the k individual scheduling tasks through the digital twin platform, calculate the overall production time, the completion time of each workpiece, and the maintenance cost of each equipment during the simulated production operation, and record the reliability curve of the equipment;

[0034] (3-4) Infer the delivery time of simulated production from the total production time, calculate the cost of the production part based on the actual delivery time, the expected delivery time, and the completion time of each workpiece; calculate the total cost using the cost of the production part and the cost of the maintenance part;

[0035] (3-5) Evaluate each individual task according to the total cost and equipment reliability, select the best two individuals and directly enter the subgroup to wait for the next round of iterative calculation; all individuals are weighted according to the evaluation results and the selection probability R is used. s (l) Randomly select k individuals and store them in the crossover set, where l is the number of iterations of the current genetic algorithm;

[0036] (3-6) Treat the individuals in the crossover set as random pairs and randomly select them according to the crossover probability R c (1) Select the fragments for partial matching and crossover, and store the individuals after crossover into the set to be mutated;

[0037] (3-7) According to the mutation probability R m (l) To determine whether the individuals in the set to be mutated need to be mutated and the degree of mutation;

[0038] (3-8) Check whether the evaluation results of the optimal scheduling task individuals in the recent rounds of iterative simulation calculations are no longer fluctuating. If there is no fluctuation, a new scheduling task individual is randomly generated; if there is still fluctuation, no new scheduling task individual is generated;

[0039] (3-9) Repeat steps (3-3) to (3-8) until l g The result evaluation of the optimal scheduling task individuals after the simulation calculation does not fluctuate with the result evaluation of the previous few times.

[0040] The cost of the production part includes storage cost and overdue penalty cost.

[0041] In steps (3-5), the evaluation scheme for individual scheduling tasks is:

[0042]

[0043] Among them, CP avg Indicates the cost-effectiveness evaluated by average reliability, while CP low represents the cost-effectiveness evaluated by the minimum reliability, Co represents the total cost of scheduling task simulation, and F r,t represents the reliability of device r at time t, C r Indicates the time when device r completes the task.

[0044] In step (4), the maintenance strategy includes minor repairs, medium repairs and major repairs; the minor repairs are to adjust the local structure of the equipment, the medium repairs are to replace the main parts, and the major repairs are to repair the equipment comprehensively.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] Compared to traditional methods, this paper proposes a method for allocating coupled production and maintenance tasks that considers the operating status of equipment. First, equipment reliability is modeled using equipment mechanisms and historical data. Coupled task allocation is then performed by treating maintenance and production as scheduling tasks operating on the same dimension. At the maintenance level, three strategies—minor, medium, and major repairs—are considered simultaneously. At the scheduling level, a specifically modified adaptive genetic algorithm is introduced to schedule and allocate tasks based on equipment reliability and total cost. Ultimately, a process that simultaneously considers both conflicting production and maintenance objectives is achieved, minimizing total production costs while ensuring that equipment remains within a reasonable reliability range. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A flowchart of the method for allocating production and maintenance coupled tasks considering the equipment operating status;

[0048] Figure 2 A full process diagram for placing an order, producing, and delivering an order in a discrete manufacturing scenario;

[0049] Figure 3 Select and assign cycle structure diagram for maintenance tasks;

[0050] Figure 4 This is the flow chart of the adaptive genetic algorithm;

[0051] Figure 5 This is a statistical chart of the cost-effectiveness of multiple cases after allocation using this method in a discrete manufacturing scenario. DETAILED DESCRIPTION

[0052] The present invention will be further described in detail below with reference to the accompanying drawings and examples. It should be noted that the following examples are intended to facilitate understanding of the present invention and do not have any limiting effect on the present invention.

[0053] like Figure 1 As shown, the present invention provides a method for allocating production and maintenance coupled tasks taking into account the equipment operating status, including the following tools:

[0054] A digital twin platform is used to perform virtual modeling and simulation of physical production lines. It simulates the allocation plans obtained by the task allocation decision-making tool to verify their feasibility.

[0055] Equipment model management tools are mainly responsible for establishing a reliability model for the equipment status based on the equipment mechanism and historical operation data, as well as establishing a numerical model for the equipment production process;

[0056] The order receiving tool is mainly responsible for virtualizing and adapting orders from the physical production line and breaking them down into production tasks for each device;

[0057] The maintenance strategy management tool is mainly responsible for managing the three maintenance strategies and selecting the corresponding maintenance strategy for the equipment based on the current equipment reliability;

[0058] Task allocation decision-making tool, mainly responsible for unified planning of production tasks and maintenance tasks, and providing allocation plans;

[0059] The allocation result interactive tool outputs the final allocation plan and interacts with the physical production line to provide production guidance.

[0060] The process of the production and maintenance coupled task allocation method includes the following steps:

[0061] (1) In the digital twin platform, the equipment model management tool is used to model the reliability of the equipment status and the equipment production process based on the historical data obtained from actual production of the actual production line in the physical world;

[0062] The equipment state reliability modeling built in the digital twin platform relies on the Weibull distribution and is expressed as follows:

[0063]

[0064] Where λ is the scale parameter and μ is the shape parameter. These two parameters need to be set according to the actual conditions of the equipment. F(t) represents the reliability of the equipment at time t.

[0065] (2) Enter the pending orders and their expected delivery dates into the order receiving tool, split the orders into production tasks, and evaluate the operating hours of each piece of equipment. Based on the estimated equipment operating hours and the equipment reliability model, the total maintenance frequency required for the equipment in the current production scenario is evaluated;

[0066] The evaluation criteria for maintenance task frequency are:

[0067]

[0068] Among them H r is the total number of maintenance tasks for equipment r, n is the total number of workpieces to be produced, j is the workpiece index, P j，r is the production time of workpiece j on equipment r, F r -1 (0.8) represents the time required for the reliability of device r to drop to 0.8.

[0069] (3) Considering production tasks and maintenance tasks together as scheduling tasks, the tasks are assigned to each production equipment through the task allocation decision tool;

[0070] Maintenance tasks are constructed by the maintenance strategy management tool and are divided into minor maintenance strategies (u1) such as adjusting the local structure of the equipment, medium maintenance strategies (u2) such as replacing major parts, and major maintenance strategies (u3) such as comprehensive repairs. The maintenance strategy allocation process is as follows: Figure 3 shown.

[0071] The allocation of production and maintenance tasks is calculated by a task allocation decision tool, whose core method is an adaptive genetic algorithm for multi-task coupling.

[0072] The improved adaptive genetic algorithm process is as follows Figure 4 As shown, the specific process is:

[0073] ① Initialize the decimal task scheduling individual and randomly generate k decimal sequence production task individuals of length n·m, where m is the total number of equipment and n is the total number of workpieces to be produced. Set the minimum number of algorithm iterations l g ; k and l g It will be adjusted according to the scale of the case, but generally k is above 20, and l g Above 1000;

[0074] ②Insert Special maintenance tasks are randomly shuffled and inserted into the decimal sequence individuals to form a scheduling task individual set;

[0075] ③ Perform simulation operations on each of the k individual tasks using the digital twin platform, calculate the overall production time, the completion time of each workpiece, the maintenance cost of each equipment during the simulated production operation, and record the equipment reliability curve;

[0076] ④ Use the total production time to infer the delivery time of the simulated production. Calculate the storage cost and overdue penalty cost of the production part based on the actual delivery time, expected delivery date, and the completion time of each workpiece. Calculate the total cost using the production cost and maintenance cost.

[0077] ⑤Evaluate each individual task according to cost and equipment reliability, select the best two individuals and directly enter the subgroup to wait for the next round of iterative calculation, and assign weights to all individuals according to the evaluation results and select them according to the probability R. s (l) Randomly select k individuals and store them in the waiting crossover set, where l is the number of iterations of the current genetic algorithm, R s The initial value of is 0.2;

[0078] The evaluation scheme for individual scheduling tasks is:

[0079]

[0080] Among them, CP avg Indicates the cost-effectiveness evaluated by average reliability, while CP low represents the cost-effectiveness evaluated by the minimum reliability, Co represents the total cost of scheduling task simulation, and F r，t represents the reliability of device r at time t, C r Indicates the time when device r completes the task.

[0081] ⑥ Treat the individuals in the crossover set to pair randomly and randomly according to the crossover probability R c (l) Select the fragments for partial matching crossover (PMX), and store the individuals after crossover into the set to be mutated. c The initial value of is 0.1;

[0082] ⑦According to the mutation probability R m (l) to determine whether the individuals in the set to be mutated need to mutate and the degree of mutation, R m The initial value of is 0.04;

[0083] ⑧ Check whether the evaluation results of the optimal scheduling task individuals in the recent rounds of iterative simulation calculations are no longer fluctuating. If there is no fluctuation, a new scheduling task individual is randomly generated. If there is still fluctuation, no new scheduling task individual is generated;

[0084] ⑨Repeat ③~⑧ until 1 gThe evaluation results of the optimal task scheduling individuals simulated and calculated after the next few times do not fluctuate with the results of the previous few times.

[0085] (4) When the decision-making tool assigns a maintenance task to a device, the maintenance strategy management tool selects a maintenance strategy based on the current device reliability and pushes it to the current maintenance task;

[0086] (5) After the decision is made, the decision result is simulated and verified through the digital twin platform, and the simulation process and decision process are transmitted to the physical world production line through the distribution result interactive tool to guide actual production.

[0087] Take the example of multi-variety customizable production in a discrete manufacturing scenario. Figure 2 As shown, the specific steps of a method for allocating production and maintenance coupled tasks considering the equipment operating status are as follows:

[0088] (1) The customer placed 11 different orders through the order platform. The products in these orders were different and the equipment required to produce the products varied from order to order. The case sizes ranged from 10 workpieces and 5 equipment to 20 workpieces and 20 equipment.

[0089] (2) After receiving the product order, the simulation platform will build a reliability model based on the status of the required production equipment, as well as a production model that can determine the time required for the equipment to produce each workpiece. The specific production model is as follows:

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100]

[0101]

[0102]

[0103]

[0104] Among them, B j,r represents the start time of production of workpiece j on equipment r; Z j,r represents the priority ranking of workpiece j produced on equipment r in the one-dimensional vector of production task allocation; Q j,r,r′ is a 0-1 variable. For workpiece j, if the equipment r is before r′ in the process, it is 1, otherwise it is 0. M is a relatively large positive constant. C j，r K represents the completion time of workpiece j produced on equipment r; j，j′，r and K′ j，r，r′ Both are non-positive integer intermediate variables, mainly used to compare workpieces j and j′ and equipment r and r′; T j，r，u Indicates whether the maintenance task is triggered and which maintenance strategy is selected, d′ r (f, u) represents how much time is needed to perform maintenance when the equipment is in state f when the maintenance strategy u is selected; C max It indicates the final completion time of the case.

[0105] (3) According to the production reliability model and the production model, the task allocation decision tool is used to allocate production tasks, maintenance tasks and other scheduling tasks to the equipment. The computing core used in the allocation rule is the improved adaptive genetic algorithm mentioned above.

[0106] (4) When a maintenance task is assigned, a maintenance strategy is selected based on the current equipment reliability. When the equipment is in production, it will be damaged, which will cause the equipment reliability to decrease, and this will also trigger the allocation of maintenance tasks. When the maintenance task is assigned, the maintenance strategy will be pushed based on the current equipment reliability to repair and restore the equipment reliability. After maintenance, the equipment will continue to produce, and the cycle will repeat;

[0107] (5) In order to obtain more reliable results, each case was simulated 10 times and the average value was taken. The cost and equipment reliability results were calculated as shown in the following table:

[0108] Table 1 Summary of case results

[0109]

[0110] (6) Calculate the cost-effectiveness based on the case results, and the results are as follows Figure 5 As shown, it can be seen that the allocation solution provided by the present invention can achieve better cost performance.

[0111] The embodiments described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for allocating production and maintenance coupling tasks, characterized in that: The following steps are involved: (1) In the digital twin platform, based on the historical data obtained from the actual production of the physical production line, the equipment state is modeled through the equipment model management module to obtain the equipment reliability model, and the equipment production process is modeled; (2) Enter the pending orders and their expected delivery dates into the order receiving module, split the orders into production tasks through the order receiving module, and evaluate the operating hours of each device. Based on the evaluated equipment operating hours and the equipment reliability model, the total maintenance task frequency required for the equipment in the current production scenario is evaluated; (3) Considering production tasks and maintenance tasks together as scheduling tasks, the task allocation decision module obtains an allocation plan, and then allocates tasks to each production equipment according to the allocation plan; The task allocation decision module obtains an allocation plan through an adaptive genetic algorithm for multi-task coupling, including: (3-1) Randomly generate k production task individuals of length n·m, where m is the total number of equipment and n is the total number of workpieces to be produced; set the minimum number of algorithm iterations l g ; (3-2) The special maintenance tasks are randomly shuffled and inserted into the individual production tasks to form a set of individual scheduling tasks; (3-3) Perform simulation operations on each of the k individual scheduling tasks through the digital twin platform, calculate the overall production time, the completion time of each workpiece, and the maintenance cost of each equipment during the simulated production operation, and record the reliability curve of the equipment; (3-4) Infer the delivery time of simulated production from the total production time, calculate the cost of the production part based on the actual delivery time, the expected delivery time, and the completion time of each workpiece; calculate the total cost using the cost of the production part and the cost of the maintenance part; (3-5) Evaluate each individual task according to the total cost and equipment reliability, select the best two individuals and directly enter the subgroup to wait for the next round of iterative calculation; all individuals are weighted according to the evaluation results and the selection probability R is used. s (l) Randomly select k individuals and store them in the waiting crossover set, where l is the number of iterations of the current genetic algorithm; The evaluation scheme for individual scheduling tasks is: Among them, CP avg Indicates the cost-effectiveness evaluated by average reliability, while CP low represents the cost-effectiveness evaluated by the minimum reliability, Co represents the total cost of scheduling task simulation, and F r,t represents the reliability of device r at time t, C r Indicates the time when device r completes the task; (3-6) Treat the individuals in the crossover set as random pairs and randomly select them according to the crossover probability R c (1) Select the fragments for partial matching and crossover, and store the individuals after crossover into the set to be mutated; (3-7) According to the mutation probability R m (l) To determine whether the individuals in the set to be mutated need to be mutated and the degree of mutation; (3-8) Check whether the evaluation results of the optimal scheduling task individuals in the recent rounds of iterative simulation calculations are no longer fluctuating. If there is no fluctuation, a new scheduling task individual is randomly generated; if there is still fluctuation, no new scheduling task individual is generated; (3-9) Repeat steps (3-3) to (3-8) until l g The evaluation results of the optimal scheduling task individuals after the simulation calculation do not fluctuate with the evaluation results of the previous few times; (4) When the task allocation decision module assigns a maintenance task to a device, the maintenance strategy management module selects the maintenance strategy based on the current device reliability and pushes it to the current maintenance task; (5) The allocation plan is simulated and verified through the digital twin platform, and the simulation process and allocation plan are passed to the physical production line through the allocation result interaction module to guide actual production.

2. The method according to claim 1, characterized in that In step (1), the equipment reliability model is: Where F(t) represents the reliability of the device at time t; λ is the scale parameter, and μ is the shape parameter. These two parameters need to be set according to the actual condition of the device.

3. The method according to claim 1, characterized in that In step (2), the evaluation formula for the total maintenance task frequency is: Among them, H r is the total number of maintenance tasks for equipment r, n is the total number of workpieces to be produced, j is the workpiece index, P j,r is the production time of workpiece j on equipment r, F r -1 (0.8) represents the time required for the reliability of device r to drop to 0.

8.

4. The method according to claim 1, wherein The cost of the production part includes storage cost and overdue penalty cost.

5. The method according to claim 1, wherein In step (4), the maintenance strategy includes minor repairs, medium repairs and major repairs; the minor repairs are to adjust the local structure of the equipment, the medium repairs are to replace the main parts, and the major repairs are to repair the equipment comprehensively.

Images