Adaptive Dynamic Jumping Particle Swarm Optimization for Buffer Allocation in Unreliable Production Lines

Over the past five decades, the buffer allocation problem in production lines has been the topic of continuous interest. This paper proposes an adaptive simulation-optimization approach relying on particle swarm optimization (PSO) to solve the buffer allocation problem for unreliable serial production lines. The objective is to maximize the production rate of the production line. The key idea is to integrate a jumping strategy based on logarithmic and exponential functions into the velocity equation of the PSO algorithm using dynamic parameters to achieve quickly (near-)optimal solutions. To evaluate the effectiveness of the proposed method, extensive numerical experiments are conducted using several configurations of production lines, ranging from 3 to 100 machines. Additionally, benchmark algorithms from the literature are employed for comparison purposes. The results indicate that the proposed adaptive approach outperforms the benchmark algorithms regarding efficiency and solution quality.

Automated summary

This paper proposes an adaptive simulation-optimization approach based on particle swarm optimization (PSO) to solve the buffer allocation problem in unreliable serial production lines. The objective is to maximize the production rate of the production line. The proposed method integrates a jumping strategy based on logarithmic and exponential functions into the velocity equation of the PSO algorithm using dynamic parameters to achieve quickly (near-)optimal solutions. Extensive numerical experiments are conducted using various production line configurations and benchmark algorithms for comparison purposes, and the results demonstrate that the proposed adaptive approach outperforms the benchmark algorithms in terms of efficiency and solution quality.