Using tabu search with ranking candidate list to solve production planning problems with setups

This study considers production planning problems involving multiple products, multiple resources, multiple periods, setup times, and setup costs. It can be formulated as a mixed integer program (MIP). Solving a realistic MIP production planning problem is NP-hard; therefore, we use tabu search methods to solve such a difficult problem. Furthermore, we improve tabu search by a new candidate list strategy, which sorts the neighbor solutions using post-optimization information provided by the final tableau of the linear programming simplex algorithm. A neighbor solution with higher priority in the ranking sequence has a higher probability of being the best neighbor solution of a current solution. According to our experiments, the proposed candidate list strategy tabu search produces a good solution faster than the traditional simple tabu search. This study also suggests that if the evaluation of the entire neighborhood space in a tabu search algorithm takes too much computation and if an efficient and effective heuristic to rank the neighbor solutions can be developed, the speed of tabu search algorithm could be significantly increased by using the proposed candidate list strategy.     

Solve least absolute value regression problems using modified goal programming techniques

Least absolute value (LAV) regression methods have been widely applied in estimating regression equations. However, most of the current LAV methods are based on the original goal program developed over four decades. On the basis of a modified goal program, this study reformulates the LAV problem using a markedly lower number of deviational variables than used in the current LAV methods. Numerical results indicate that for the regression problems with hundreds of observations, this novel method can save more than 1/3 of the CPU time compared to current LAV methods.     

Buffer allocation in unreliable production lines using a knowledge based system

The optimization of production lines performance is a problem of great complexity and, therefore, of significant research interest. The problem may involve the optimization of many conflicting objectives, such as increasing throughput and reducing work-in-process time. The majority of existing studies have used various heuristics and search methods based on operations research. These methods have been proved to be computationally inefficient, especially for large production lines. This paper presents ASBA2, a knowledge based system that determines near optimal buffer allocation plans, with the objective of maximising production lines throughout. The allocation plan is calculated subject to a given amount of total buffer slots, in a computationally efficient way. ASBA2 operates in close cooperation with a simulation method, which provides ASBA2 with performance measures concerning production line behaviour. Moreover, to evaluate results provided by ASBA2, we have utilized an exact numerical algorithm for calculating the throughput of unreliable production lines.