Serial asynchronous production system buffer area distribution method for improving system elasticity

Abstract

The invention relates to the technical field of production system buffer area distribution, in particular to a serial asynchronous production system buffer area distribution method for improving system flexibility, which comprises the following steps: generating a plurality of configuration schemes of a serial asynchronous production system, the configuration schemes comprising the capacity of each buffer area; for each configuration scheme, determining a post-disturbance system throughput rate and post-disturbance throughput rate stabilization time of the serial asynchronous production system based on the system parameters; determining system elastic parameters of the serial asynchronous production system based on the post-disturbance system throughput rate and the post-disturbance throughput rate stabilization time; based on the system elastic parameters of the plurality of configuration schemes, a combat genetic algorithm is adopted to optimize the plurality of configuration schemes, and an optimal configuration scheme is determined; carrying out buffer area distribution on the serial asynchronous production system by adopting the capacity of each buffer area in the system parameters of the optimal configuration scheme; the elasticity of the production system can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of buffer allocation in production systems, and particularly to a buffer allocation method for a serial asynchronous production system that improves system resilience. Background Art

[0002] During the production and manufacturing process of products, the production process is affected by external disturbances, such as machine failures, machines being in a maintenance state, changes in machine processing rates, changes in external order demands, etc. When the production system is subjected to such external disturbances, the performance of the production system will decrease. After the disturbance occurs, different measures will be taken to deal with this disturbance, which are specifically manifested as different system performance recovery strategies. As time goes by, the performance of the production system generally goes through a process of decline, maintaining stability, and rising, and finally the performance of the production system will recover, but the final system performance will show three different levels: higher, remaining the same, or lower. The ability of the system to resist external disturbances, absorb disturbances, and recover from disturbances is called resilience.

[0003] In a serial asynchronous production system, serial means that each stage or process in the production process is carried out sequentially, and asynchronous means that the processing rates of each machine in the system are different. As the production system becomes more complex, the possibility of being affected by various different disturbances increases. In the face of the impact caused by the collapse of the production system due to interference, it is necessary to determine an effective allocation strategy to maintain the stable and orderly operation of the production system. Summary of the Invention

[0004] In view of the above problems, the present invention provides a buffer allocation method for a serial asynchronous production system that improves system resilience, and solves the technical problem of low system resilience in the existing production system.

[0005] The present invention provides a buffer allocation method for a serial asynchronous production system that improves system resilience, including the following steps:

[0006] Step S1, generate multiple configuration schemes for the serial asynchronous production system, where the configuration scheme includes the system parameters of the serial asynchronous production system; the serial asynchronous production system is composed of machines and buffer positions connected in series alternately, and the system parameters include machine processing rates, total buffer capacity, and the capacity of each buffer;

[0007] Step S2, for each of the configuration schemes, determine the post-disturbance system throughput and the stable moment of the post-disturbance throughput based on the system parameters of the serial asynchronous production system;

[0008] Step S3, for each of the configuration schemes, determine the system resilience parameter of the serial asynchronous production system based on the post-disturbance system throughput and the stable moment of the post-disturbance throughput;

[0009] Step S4: Based on the system elasticity parameters of the multiple configuration schemes, use the combat genetic algorithm to optimize the multiple configuration schemes and determine the optimal configuration scheme;

[0010] Allocate buffers to the serial asynchronous production system according to the capacities of the buffers in the system parameters of the optimal configuration scheme.

[0011] Preferably, step S1 specifically includes:

[0012] Determine that the serial asynchronous production system is composed of K devices and K - 1 buffer positions connected in series alternately;

[0013] Generate multiple configuration schemes, and determine system parameters for each configuration scheme, including: determining the processing rate U of the i-th machine i , i ∈ {1, 2, …, K}, the maximum capacity B of the m-th buffer position m,max , the minimum capacity B of the m-th buffer position m,min , m ∈ {1, 2, …, K - 1}, the total buffer capacity N, and the capacities of each buffer [B 1 , B 2 , B 3 ,..., B K-1 .

[0014] Preferably, step S2 specifically includes:

[0015] Step S2-1: Based on the machine processing rate in the system parameters, determine the throughput of the perturbed system and obtain the system throughput - time function;

[0016] Step S2-2: Process the system throughput - time function, query the moment when the difference ratio between the throughput of the perturbed system in the system throughput - time function is within the first interval range, and when the difference ratios between the throughputs of the system at j moments before and after this moment and the throughput of the perturbed system are all within the first interval range, determine this moment as the stable moment of the throughput of the perturbed system, and determine the corresponding duration at this moment as the stable duration of the throughput of the serial asynchronous production system after perturbation.

[0017] Preferably, in step S2-1, the expression of the throughput of the perturbed system is:

[0018]

[0019] where T(t) represents the throughput of the perturbed system at time point t, P c (t) represents the total amount of finished products processed at time point t, and b m represents the buffer efficiency factor of the m-th buffer position;

[0020] In the step S2-2, the expression for processing the system throughput-time function is as follows:

[0021]

[0022] where T(t + j) represents the system throughput after perturbation at the time point t + j, T(θ) represents the true value of the system throughput after perturbation at the maximum time, δ represents the system throughput difference ratio interval, j represents the time offset, k represents the maximum absolute value of the time offset, and Z + represents the set of positive integers.

[0023] Preferably, the step S3 specifically includes:

[0024] Step S3-1: Determine the start and end times of the perturbation diffusion stage, the end time of the delay stage, and the end time of the recovery stage based on the system throughput after perturbation and the stable time of the throughput after perturbation;

[0025] The start time t 1 of the perturbation diffusion stage is the start time when the system is perturbed; the end time t 2 of the perturbation diffusion stage is the moment when the throughput after perturbation is stable; the end time t 3 of the delay stage = t 2 + C, where C represents the duration of the delay stage; the end time t' of the recovery stage = t 2 + C + (t 2 - t 1 );

[0026] The time of the perturbation delay stage can be divided into two types: constant type and fluctuating type. The constant type of delay stage time selected in the present invention can more stably reflect the response characteristics of the system to perturbations, and at the same time, it also more prominently reflects the significance of the stable time of the throughput after perturbation and the system throughput index. Specifically, the calculation of t 1 , t 2 , t 3 , and t' is as follows.

[0027] The perturbation start time t 1 is the start time when the system is perturbed; the perturbation diffusion stage end time t 2 is determined according to the moment when the throughput difference ratio after the system is perturbed is within the range of the first interval in steps S2-1 and S2-2 of the present invention; in the present invention, it is considered that the delay stage time of various configuration schemes is the same, so the delay stage end time t 3 is t 2+C, where C is a time-related constant; in the present invention, it is considered that the disturbance diffusion stage and the recovery stage of the system can be approximately regarded as a symmetric process, that is, the time required for disturbance diffusion is the same as the time required for the recovery stage, so t′ = t 2 +C+(t 2 -t 1 ). The above simplifying assumption for the disturbance time point helps to highlight the role of the steady-state throughput rate of the system and the stabilization time of the throughput rate after disturbance in the elastic parameters.

[0028] Step S3-2: Determine the performance loss of the system after disturbance from the start and end times of the disturbance diffusion stage, delay stage, and recovery stage and the throughput rate of the system after disturbance, and use it as the elastic parameter of the production system.

[0029] Preferably, in the step S3-2, the expression of the elastic parameter of the production system is:

[0030]

[0031] where RL(x,t) represents the elastic evaluation parameter of the production system, t 1 is the start time of the disturbance, t 2 is the end time of the disturbance diffusion stage, t 3 is the end time of the delay stage, t′ is the end time of the recovery stage, x(t) is the disturbance performance function at time t, set as the system throughput rate function, x 0 is the throughput rate of the system after disturbance.

[0032] Preferably, step S4 specifically includes:

[0033] Step S4-1: Use the multiple configuration schemes as the initial population of the battle genetic algorithm, and each configuration scheme in the multiple configuration schemes as an individual in the initial population, and determine the initialization parameters of the battle genetic algorithm;

[0034] Step S4-2: Update and iterate the initial population, including updating the initial population with the new population obtained by crossing and mutating the initial population;

[0035] When the maximum number of iterations is reached or the fitness value of an individual reaches the fitness target, end the iteration process and obtain the optimal individual;

[0036] Step S4-3: Allocate buffers to the serial asynchronous production system using the capacities of the buffers in the system parameters of the configuration scheme corresponding to the optimal individual.

[0037] Preferably, in step S4-1, the steps of determining the initialization parameters of the combat genetic algorithm specifically include: determining the initial population pop, the population size popsize, the crossover probability pc, the mutation probability pb, the fitness target f, the maximum number of iterations g, and the cumulative probability distribution; the calculation method of the cumulative probability distribution is: using the system resilience parameter of each configuration plan as the fitness of the individual in the initial population, accumulating the fitness values of all individuals, and calculating the cumulative probability distribution;

[0038] Step S4-2 specifically includes:

[0039] Step S4-2-1, initialize a new population new_pop;

[0040] Step S4-2-2, add individuals to the new population new_pop, and the specific steps include:

[0041] Step S4-2-2-1, randomly generate two random numbers r1 and r2 between 0 and 1, and respectively obtain the first individual corresponding to the cumulative probability greater than or equal to r1 and r2 in the cumulative probability distribution as the two selected individuals;

[0042] Step S4-2-2-2, generate a random number l, if l < pc, perform a crossover operation on these two individuals to generate new individuals;

[0043] If l < pb, perform a mutation operation on these two individuals to generate new individuals;

[0044] Select for all the generated new individuals, calculate the fitness of all new individuals, that is, the system resilience parameter, and only retain the individual with the lowest system resilience parameter as the excellent individual, and add the excellent individual to the new population new_pop;

[0045] Step S4-2-2-3, return to step S4-2-2-1, until the number of individuals in the new population new_pop reaches the population size popsize, and replace the current population pop with new_pop;

[0046] Step S4-2-3, return to step S4-2-2, until the maximum number of iterations g is reached or the fitness value of an individual reaches the fitness target f, end the iteration process, and obtain the optimal individual best_individual.

[0047] Compared with the prior art, the present invention has at least the following beneficial effects:

[0048] (1) By establishing the relationship between the buffer configuration scheme and system resilience, the present invention overcomes the limitations of traditional methods that only focus on throughput and work-in-process inventory. The present invention takes system resilience as an evaluation index, enabling the production system to have stronger resistance and recovery capabilities in the face of external disturbances, thereby improving the stability and reliability of the system.

[0049] (2) The present invention provides a structured optimization method. By using the combat genetic algorithm to optimize multiple configuration schemes, it is possible to find the optimal allocation scheme under the constraint of the total buffer capacity being limited. It not only considers the system throughput rate after disturbance and the time for the system to recover from the disturbance, thus achieving a comprehensive optimization of system resilience.

[0050] (3) The method provided by the present invention can achieve the best allocation of resources when the total buffer is certain, can reduce the buffer cost, and improve production efficiency and stability. By optimizing the configuration scheme, the production system can be designed and managed more effectively, ensuring that the system can still maintain a high level of resilience and production capacity under limited resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The drawings are only for the purpose of showing specific embodiments and are not considered as a limitation to the present invention.

[0052] Figure 1 It is a flowchart of the buffer allocation method for the serial asynchronous production system that improves system resilience provided by the present invention.

[0053] Figure 2 It is a schematic diagram of the machine state provided by the present invention.

[0054] Figure 3 It is a schematic diagram of the system resilience metric for the serial asynchronous production system provided by the present invention.

[0055] Figure 4 It is a schematic diagram of the machine serial asynchronous production system model provided by the present invention.

[0056] Figure 5 It is a judgment diagram of the stable time of the throughput rate after disturbance for the serial asynchronous production system provided by the present invention.

[0057] Figure 6 It is a schematic diagram of the throughput rate function for the serial asynchronous production system provided by the present invention.

[0058] Figure 7 It is a schematic diagram of the iteration of the fitness function provided by the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0059] In order to more clearly understand the above-mentioned objects, features, and advantages of the present invention, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, without conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. In addition, the present invention can also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below.

[0060] Aiming at the problems of limited buffer quantity or expensive buffer cost during the processing of certain products, and poor resistance of the production system to external disturbances, the present invention improves the resilience endowed to the production system by the buffer configuration scheme through modeling and simulating the system throughput rate after disturbance, judging the stabilization time of the throughput rate after disturbance, solving the system resilience parameters under a specific buffer configuration scheme, and applying the combat genetic algorithm to optimize the buffer configuration scheme.

[0061] In order to illustrate the effectiveness of the method proposed by the present invention, the above-mentioned technical solutions of the present invention will be described in detail below through a specific embodiment.

[0062] The embodiment of the present invention provides a method for modeling and evaluating the reliability of an assembly process considering the coupling effect between processes, and the following basic assumptions are proposed:

[0063] (1) For the machine equipment affected by random failures, various failures will definitely occur during the production process. It is assumed that the failures only occur during the operation of the equipment and are operating failures, and these failures follow an exponential distribution mathematically.

[0064] (2) In the production system of the present invention, the first machine is not starved and the last machine is not blocked.

[0065] (3) The flow of work-in-progress in the production line is discrete and flows in sequence from beginning to end.

[0066] As Figure 1 shown, a method for buffer allocation in a serial asynchronous production system for enhancing system resilience is disclosed, and the specific implementation steps are as follows:

[0067] Step S1: Generate multiple configuration schemes for the serial asynchronous production system, where the configuration schemes include the system parameters of the serial asynchronous production system; the serial asynchronous production system is composed of machines and buffer positions connected in series alternately, and the system parameters include the machine processing rate, the total buffer capacity, and the capacity of each buffer.

[0068] The serial asynchronous production system is composed of machines and buffers connected in series, specifically composed of K devices (M 1 …M K ) and K - 1 buffer positions (B p,1 …Bp,K-1 ) are connected in series alternately, as Figure 4 shown. During the processing on the production line, first, equipment M 1 is processed first. After M 1 is processed, it enters buffer position B p,1 , then goes to equipment M 2 , buffer position B p,2 , and so on, until the last equipment M K completes processing and leaves the production line.

[0069] When setting the configuration scheme of the production system, let the processing rate of the i-th machine be U i , i ∈ {1, 2,..., K}, and the total number of buffer positions is K - 1. Each of the said buffer positions can be assigned different capacities according to the allocation scheme, but is also limited by the capacity threshold. The maximum capacity of the m-th buffer position is B m,max , and the minimum capacity of the m-th buffer position is B m,min , m ∈ {1, 2,..., K - 1}, the failure rate of the i-th machine is p i , the repair rate of the i-th machine is r i , the total buffer capacity is N, and the capacities of each buffer are represented as an array B = [B 1 , B 2 , B 3 ,..., B K-1 .

[0070] Before being disturbed by the system failure, it can be regarded as a reliable system. At this time, the throughput rate T(t) of the production system is limited by the bottleneck machine, that is, the processing rate of the machine with the lowest rate in the production line. The expressions of the throughput rate of the system under reliable and unreliable machine conditions before perturbation are:

[0071]

[0072] Among them, T(t) represents the system throughput rate at time point t, min(·) represents taking the minimum value, U i is the processing rate of the i-th machine, M i represents the i-th machine, r i , p i , e i represent the repair rate, failure rate, and independent processing efficiency of the i-th machine.

[0073] In some embodiments, the present invention adopts a serial asynchronous production line with 5 machines and 4 buffers. The system schematic diagram is as Figure 5 shown. The specific parameters of the machines and buffers set when setting the configuration scheme of the production system are shown in Table 1 and Table 2.

[0074] Table 1

[0075]

[0076] Table 2

[0077]

[0078] Before the production system is disturbed by a fault, it can be regarded as a reliable system. At this time, the throughput rate T(t) of the system before disturbance is limited by the bottleneck machine. In this embodiment, based on the data in Table 1, the throughput rate of the system before disturbance is T(t) = min(U 1 , U 2 , U 3 , U 4 , U 5 ) = 1.

[0079] In some embodiments, in order to optimize using a genetic algorithm in subsequent steps, multiple configuration schemes of the serial asynchronous production system are generated in this step. Specifically, with the total buffer capacity as N and the upper and lower bounds of the buffers at each buffer position as constraints, 20 groups of configuration schemes are randomly determined, and the buffer numbers are allocated for buffer positions B 1 , B 2 , B 3 , B 4 respectively, and each configuration scheme sets a different buffer number allocation method.

[0080] Step S2: For each of the configuration schemes, determine the post-disturbance system throughput rate and the post-disturbance throughput rate stabilization time of the serial asynchronous production system based on the system parameters.

[0081] The production system will be affected by external disturbances, such as machine failures, machine repairs, changes in external orders, changing processing rates, etc. In the present invention, the impact of machine failures on the serial asynchronous production system is mainly considered, and its specific manifestation is that the processing machines in the production system change from reliable machines to unreliable machines, and the corresponding failure rate and repair rate parameters are used to characterize the degree of unreliability of the machines, and the failure rate and repair rate of the machines on the production line both follow an exponential distribution.

[0082] Before solving the throughput rate of the production system, the states of the machines need to be defined. The machines include two inherent states: normal (Up) and faulty, and four production states: working (Wk), starving (St), blocked (Bl), and faulty (Dn). The specific relationships are as Figure 2 shown.

[0083] For the system throughput rate of the production system under the configuration scheme of specific buffer quantity allocation, in the present invention, the simulation method is applied to solve the throughput rate of the system. During the simulation process, the behaviors of starvation and blockage in the production process should be fully considered, and the in-process inventory in the buffer is used to determine the machine state. Because the influence of the disturbance of machine failure is considered, the system throughput rate after the disturbance of the serial asynchronous production system will be significantly lower than that before the disturbance, and its calculation formula is shown as follows:

[0084]

[0085] Among them, T(t) represents the system throughput rate after the disturbance at time point t, and P c (t) represents the total quantity of finished products processed at time point t, and b m represents the buffer efficiency factor of the m-th buffer position.

[0086] In some embodiments, taking the total buffer capacity N = 40 and the configuration scheme [B 1 , B 2 , B 3 , B 4 = [10, 10, 10, 10] as an example to solve the system throughput rate after the disturbance of the production system. The specific parameters are shown in Table 1. The simulation method is used to solve the throughput rate, and the throughput rate after the disturbance can be obtained as: T(t) = P C (t) / t = 0.5713.

[0087] The judgment of the stable time of the throughput rate after the disturbance of the production system is the time corresponding to when the system performance significantly decreases when the system is subjected to an external disturbance, that is, a failure, and after a period of time, the downward trend tends to be stable. When analyzing the elastic performance of the production system, it is necessary to quantitatively know the corresponding time when the system performance drops to a stable state after being disturbed. This is of great significance for the quantitative evaluation of the system elasticity, that is, the earlier the throughput rate of the system is stable after being disturbed, the stronger the resistance of the adopted buffer configuration scheme to the disturbance and the stronger the elasticity given to the system.

[0088] The present invention adopts the method of calculating the offset value to judge the stable time of the system performance, and finds the moment when the difference ratio between the system throughput rate-time function and the system throughput rate after the disturbance is within the first interval range, and when the difference ratios between the system throughput rates at j moments before and after this moment and the system throughput rate after the disturbance are all within the first interval range, the corresponding duration at this moment is determined as the stable time of the throughput rate after the disturbance of the serial asynchronous production system, which is used to calculate the subsequent elastic parameters. The calculation formula is shown as follows:

[0089]

[0090] Among them, T(t + j) represents the system throughput rate after perturbation at time point t + j, T(θ) represents the true value of the system throughput rate after perturbation at the time maximum, δ represents the system throughput rate difference ratio interval, which reflects the stability requirement of the system, generally taking 0.01, j represents the time offset, k represents the maximum absolute value of the time offset, and Z + represents the set of positive integers.

[0091] In some embodiments, δ can be set to 0.01, and j can be set to 0, 1, -1, 2, -2. The stability time of the throughput rate after perturbation of the serial asynchronous production system is judged by the fact that the fluctuations of a total of 5 data points before and after are within the true value interval range. For the configuration scheme [B 1 ,B 2 ,B 3 ,B 4 = [10, 10, 10, 10] and the data in Table 1 are simulated. As Figure 5 shown, the stability time T s (B) = 9196 of the throughput rate after perturbation of the serial asynchronous production system is obtained.

[0092] Under a specific buffer configuration, the method provided by the present invention can be used to calculate the stability time of the throughput rate after system perturbation, so as to better characterize the elasticity given to the system by the buffer configuration scheme.

[0093] Step S3: For each of the configuration schemes, determine the system elasticity parameter of the serial asynchronous production system based on the throughput rate after perturbation and the stability time of the throughput rate after perturbation.

[0094] When being perturbed by the outside world, a resilient serial asynchronous production system has the ability to resist, absorb, and recover from the outside perturbation. Maintain the basic production capacity of the system and keep the productivity of the system at a relatively high level. When being perturbed by a fault, the measurement of system elasticity, that is, the system throughput rate function, is as Figure 3 shown.

[0095] Figure 3 In, the horizontal axis represents time, and the vertical axis represents the system throughput rate. When the system is perturbed, the throughput rate after perturbation is x 0 , and its performance loss is expressed as 1 - x 0 , and the shaded part can represent the performance loss of the system. The performance loss of the system can be divided into three stages: the first stage (t 1 ≤t≤t 2 ) is the perturbation diffusion stage, which reflects the diffusion effect of the perturbation and its impact on the system function under the current coping ability of the system; the second stage (t 2 ≤t≤t 3) is the delay stage, which mainly reflects the delay process from the end of disturbance diffusion to the start of system function recovery after a disaster occurs; the third stage (t 3 ≤ t ≤ t′) is the recovery stage, indicating that at the current moment, the system function has recovered to the desired state. The elasticity parameter of the production system can be expressed as:

[0096]

[0097] Among them, RL(x, t) represents the elasticity evaluation parameter of the production system, t 1 is the start time of the disturbance, t 2 is the end time of the disturbance diffusion stage, t 3 is the end time of the delay stage, t′ is the end time of the recovery stage, x(t) is the disturbance performance function at time t, which can be set as the system throughput function in the present invention, and x 0 is the system throughput after the disturbance.

[0098] When calculating the above elasticity evaluation parameter, it should be noted that the stable time of the system throughput after the disturbance of the production system corresponding to each buffer configuration scheme is different, and this time will affect the values of the time parameters.

[0099] Specifically, taking the configuration scheme [B 1 , B 2 , B 3 , B 4 = [10, 10, 10, 10] and the data in Table 1 as an example to judge the time points. The start time t 1 of the disturbance is the start time t 1 = 10 when the system is disturbed; the end time t 2 of the disturbance diffusion stage is determined according to the moment when the difference ratio of the system throughput after the disturbance is within the first interval range in step S2 of the present invention, and t 2 = 14.83; the time required for the delay stage is simplified to a constant C, and in the present invention, C = 15, that is, t 3 = 29.83; the disturbance diffusion stage and the recovery stage can be approximately regarded as a symmetric process, that is, the time required for disturbance diffusion is the same as the time required for the recovery stage, then t′ = t 2 + C + (t 2 - t 1 ), that is, t′ = 34.66.

[0100] Using the index RL(x, t) to evaluate the buffer configuration scheme of the production system instead of the traditional system throughput and the work-in-process inventory in the buffer, and the smaller this index is, the stronger the ability of a certain buffer configuration scheme to resist disturbances, that is, the stronger the elasticity of the system.

[0101] In some embodiments, taking the configuration scheme [B 1 ,B 2 ,B 3 ,B 4 = [10, 10, 10, 10] and the data in Table 1 as an example for calculating the elastic evaluation parameters, the curve of the system performance changing with time is as Figure 6 shown. The first stage (10 ≤ t ≤ 14.83) is the disturbance diffusion stage, which reflects the diffusion effect of the disturbance and its impact on the system function under the current response ability of the system; the second stage (14.83 ≤ t ≤ 29.83) is the delay stage, which mainly reflects the delay process from the end of the disturbance diffusion to the start of the system function recovery after the disaster occurs; the third stage (29.83 ≤ t ≤ 34.66) is the recovery stage, indicating that at the current moment, the system function has recovered to the desired state. Then, calculate the system elasticity parameter according to formula (5), and the system elasticity parameter is obtained as RL(x,t) = 8.5796.

[0102] Step S4: Based on the system elasticity parameters of the multiple configuration schemes, use the combat genetic algorithm to optimize the multiple configuration schemes to determine the optimal configuration scheme;

[0103] Use the capacity of each buffer in the system parameters of the optimal configuration scheme to perform buffer allocation for the serial asynchronous production system.

[0104] The buffer allocation problem is an NP-hard optimization problem. When there are many machines and buffers, the calculation time required increases exponentially. Therefore, an optimization algorithm with a faster solution speed is needed. The combat genetic algorithm (CGA) is an optimization algorithm inspired by the theory of biological evolution. It introduces the concept of combat or competition on the basis of the traditional genetic algorithm (GA) to simulate the survival competition and natural selection process in the biological world. CGA is also a neighborhood search algorithm. Its idea is to continuously evolve in the solution space, then select the offspring with high fitness through the selection operator, and then perform genetic operations on the offspring with high fitness. The algorithm can be stopped by iterating a certain number of times or when the individuals reach the required fitness value.

[0105] The steps of using the combat genetic algorithm to optimize the multiple configuration schemes in the present invention include:

[0106] (1) Determine the initialization parameters of the combat genetic algorithm, including: the initial population pop, the population size popsize, the crossover probability pc, the mutation probability pb, the fitness target f, and the maximum number of iterations g. Among them, use the multiple configuration schemes of the present invention as the initial population, and each configuration scheme among the multiple configuration schemes as an individual in the initial population. Use the system resilience parameter of each configuration scheme as the fitness of the individual in the initial population. Calculate the cumulative probability distribution by accumulating the fitness values of all individuals.

[0107] (2) Iteratively optimize the initial population pop, and the specific steps are as follows:

[0108] Initialize a new population new_pop to store the individuals of the next generation.

[0109] Add individuals to the new population new_pop, and the specific steps include:

[0110] Randomly generate two random numbers r1 and r2 between 0 and 1, and respectively obtain the first individual corresponding to the cumulative probability greater than or equal to r1 and r2 in the cumulative probability distribution as the two selected individuals;

[0111] Generate a random number l. If l < pc, perform a crossover operation on these two individuals to generate new individuals.

[0112] If l < pb, perform a mutation operation on these two individuals to generate new individuals;

[0113] Select for all the newly generated individuals, calculate the fitness of all the newly generated individuals, that is, the system resilience parameter, and only retain the individual with the lowest system resilience parameter as the excellent individual;

[0114] Add the excellent individual to the new population new_pop, and return to the step of randomly generating two random numbers r1 and r2 between 0 and 1;

[0115] When the number of individuals in the new population new_pop reaches the population size popsize, replace the current population pop with new_pop.

[0116] (3) The iterative process continues until the maximum number of iterations g is reached or the fitness value of a certain individual reaches the fitness target f, then end the iterative process, and obtain the optimal fitness value best_fitness and the corresponding optimal individual best_individual. Among them, the optimal fitness value is the optimal system resilience parameter, and the optimal individual has the optimal allocation scheme for the capacities of each buffer.

[0117] In some embodiments, the crossover operation and mutation operation of the present invention process the capacities of each buffer among the system parameters of an individual.

[0118] The specific manner of the crossover operation is as follows: Determine the replacement items and non-replacement items of the capacity array of the buffer, and perform crossover replacement on the replacement items in the capacity arrays of the buffers of two individuals. The crossover operation mainly adopts single-point crossover.

[0119] The specific manner of the mutation operation is as follows: Determine the mutated items of the capacity array of the buffer, and modify the mutated items in the capacity array of the buffer of an individual to a new random value.

[0120] In some embodiments, for the new individuals obtained from the crossover operation and mutation operation, it is required that the total capacity of each buffer of the new individual satisfies being less than or equal to the total buffer capacity N.

[0121] In some embodiments, the present invention performs iterative optimization of the combat genetic algorithm on a 5-machine 4-buffer production system, and the parameters used are shown in Table 3.

[0122] Table 3

[0123]

[0124] First, an initial population with 20 buffer configuration schemes is generated, and then operations such as crossover, mutation, and mutation between genes are performed in sequence. The optimal individuals in each generation are selected and retained, and optimization is carried out with the minimum fitness function as the goal until the population iteration times reach 200 times and then the iteration stops, and the optimal buffer configuration scheme is output. The variation effect of the fitness function with the iteration times is as Figure 7 shown. The optimal buffer capacity allocation scheme for a 5-machine 4-buffer serial asynchronous production system under the limitation that the total buffer capacity is 40 is B best = [5, 13, 10, 12]. The throughput stabilization time, system elasticity parameter, and total buffer capacity of the production system with the optimal buffer configuration scheme after being disturbed are shown in Table 4.

[0125] Table 4

[0126]

[0127] In the present invention, the smaller the system elasticity parameter, the stronger the ability of the system to resist external disturbances, that is, the better the elasticity of the system. The configuration scheme obtained according to the above implementation manner has the smallest elasticity parameter. Therefore, the method provided according to the implementation manner of the present invention can achieve buffer allocation in a production system with the goal of improving system elasticity.

[0128] It should be understood that the foregoing only illustrates some embodiments, and changes, modifications, additions and / or variations can be made without departing from the scope and essence of the disclosed embodiments. The embodiments are illustrative rather than restrictive. In addition, the described embodiments relate to the currently considered most practical and preferred embodiments, and it should be understood that the embodiments should not be limited to the disclosed embodiments. On the contrary, it is intended to cover different modifications and equivalent arrangements included in the essence and scope of the embodiments. In addition, the various embodiments described above can be applied in combination with other embodiments. For example, aspects of one embodiment can be combined with aspects of another embodiment to achieve yet another embodiment. Additionally, the individual features or components of any given component can constitute additional embodiments.

[0129] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and the description of the present invention.

Claims

1. A serial asynchronous production system buffer allocation method for improving system elasticity, characterized in that: The following steps are involved: Step S1, generating multiple configuration schemes of a serial asynchronous production system, wherein the configuration schemes include system parameters of the serial asynchronous production system; the serial asynchronous production system is composed of machines and buffer positions alternately connected in series, and the system parameters include machine processing rate, total buffer capacity, and capacity of each buffer; Step S2: for each of the configuration schemes, determining a post-disturbance system throughput and a post-disturbance throughput stabilization time of the serial asynchronous production system based on the system parameters; Step S3: for each of the configuration schemes, determining the system elasticity parameters of the serial asynchronous production system based on the system throughput after the disturbance and the throughput stability time after the disturbance; Step S4: Based on the system elasticity parameters of the multiple configuration schemes, the multiple configuration schemes are optimized using a combat genetic algorithm to determine an optimal configuration scheme; The buffer zone of the serial asynchronous production system is allocated according to the capacity of each buffer zone in the system parameters of the optimal configuration scheme.

2. The serial asynchronous production system buffer allocation method for improving system elasticity according to claim 1, characterized in that: The step S1 specifically includes: Determine that the serial asynchronous production system consists of K devices and K-1 buffer locations connected in series alternately; Generate multiple configuration schemes, and determine system parameters for each of the configuration schemes, including: determining the processing rate U of the i-th machine i , i∈{1,2,…K}, the maximum capacity B of the mth buffer position m,max , the minimum capacity B of the mth buffer position m,min , m∈{1,2,…K-1}, the total buffer capacity N and the capacity of each buffer [B1,B2,B3,...,B K-1 ].

3. The serial asynchronous production system buffer allocation method for improving system elasticity according to claim 2, characterized in that: The step S2 specifically includes: Step S2-1, determining the system throughput after disturbance based on the machine processing rate in the system parameters, and obtaining a system throughput-time function; Step S2-2, process the system throughput-time function, query the moment when the difference ratio between the system throughput and the system throughput after the disturbance in the system throughput-time function is within the first interval range, and when the difference ratio between the system throughput at j moments before and after the moment and the system throughput after the disturbance is within the first interval range, determine the moment as the moment when the throughput is stabilized after the disturbance, and determine the duration corresponding to the moment as the duration of the throughput stabilization after the disturbance of the serial asynchronous production system.

4. The serial asynchronous production system buffer allocation method for improving system elasticity according to claim 3, characterized in that: In step S2-1, the expression of the post-disturbance system throughput is: Where T(t) represents the system throughput after the disturbance at time point t, P c (t) represents the total amount of finished products processed at time t, b m represents the buffer efficiency factor of the mth buffer position; In step S2-2, the expression for processing the system throughput-time function is: Where T(t+j) represents the system throughput after the disturbance at time point t+j, T(θ) represents the true value of the system throughput after the disturbance at the maximum time, δ represents the system throughput difference ratio interval, j represents the time offset, k represents the maximum absolute value of the time offset, and Z + Represents the set of positive integers.

5. The serial asynchronous production system buffer allocation method for improving system elasticity according to claim 4, characterized in that: The step S3 specifically includes: Step S3-1, determining the start and end time of the disturbance diffusion phase, the end time of the delay phase, and the end time of the recovery phase based on the post-disturbance system throughput and the post-disturbance throughput stabilization time; The start time t1 of the disturbance diffusion phase is the start time when the system is disturbed; the end time t2 of the disturbance diffusion phase is the moment when the throughput rate stabilizes after the disturbance; the end time t3 of the delay phase is t2+C, where C represents the duration of the delay phase; the end time t′ of the recovery phase is t2+C+(t2-t1); Step S3-2, determining the performance loss of the system after the disturbance according to the start and end times of the disturbance diffusion phase, the delay phase and the recovery phase and the throughput rate of the system after the disturbance as the elasticity parameter of the production system.

6. The serial asynchronous production system buffer allocation method for improving system elasticity according to claim 5, characterized in that: In step S3-2, the expression of the elasticity parameter of the production system is: Among them, RL(x,t) represents the elasticity evaluation parameter of the production system, t1 is the start time of the perturbation, t2 is the end time of the perturbation diffusion stage, t3 is the end time of the delay stage, t′ is the end time of the recovery stage, x(t) is the perturbation performance function at time t, which is set as the system throughput function, and x0 is the system throughput after the perturbation.

7. The serial asynchronous production system buffer allocation method for improving system elasticity according to claim 6, characterized in that: The specific steps of step S4 include: Step S4-1: Use the multiple configuration schemes as the initial population of the combat genetic algorithm. Each configuration scheme among the multiple configuration schemes is used as an individual in the initial population, and the initialization parameters of the combat genetic algorithm are determined. Step S4-2: Update and iterate the initial population, including updating the initial population with the new population obtained by crossing and mutating the initial population. When the maximum number of iterations is reached or the fitness value of an individual reaches the fitness target, the iteration process ends, and the optimal individual is obtained. Step S4-3: Use the capacity of each buffer in the system parameters of the configuration scheme corresponding to the optimal individual to allocate buffers for the serial asynchronous production system.

8. The serial asynchronous production system buffer allocation method for improving system elasticity according to claim 7, characterized in that: In step S4-1, the steps of determining the initialization parameters of the combat genetic algorithm specifically include: determining the initial population pop, population size popsize, crossover probability pc, mutation probability pb, fitness target f, maximum number of iterations g, and cumulative probability distribution. The calculation method of the cumulative probability distribution is: use the system elasticity parameter of each configuration scheme as the fitness of the individual in the initial population, and accumulate the fitness values of all individuals to calculate the cumulative probability distribution. The specific steps of step S4-2 include: Step S4-2-1: Initialize a new population new_pop. Step S4-2-2: Add individuals to the new population new_pop. The specific steps include: Step S4-2-2-1: Randomly generate two random numbers r1 and r2 between 0 and 1, and respectively obtain the first individual corresponding to the cumulative probability whose cumulative probability is greater than or equal to r1 and r2 in the cumulative probability distribution as the two selected individuals. Step S4-2-2-2: Generate a random number l. If l < pc, perform a crossover operation on these two individuals to generate new individuals. If l < pb, perform a mutation operation on these two individuals to generate new individuals. Select for all the generated new individuals, calculate the fitness of all new individuals, that is, the system elasticity parameter, and only retain the individual with the lowest system elasticity parameter as the excellent individual, and add the excellent individual to the new population new_pop. Step S4-2-2-3: Return to step S4-2-2-1 until the number of individuals in the new population new_pop reaches the population size popsize, and replace the current population pop with new_pop. Step S4-2-3: Return to step S4-2-2 until the maximum number of iterations g is reached or the fitness value of an individual reaches the fitness target f, and then end the iteration process to obtain the optimal individual best_individual.

Images