Artificial Bee Colony Algorithm For Disassembly Line Balancing Problem

Recycling of old products is one of the effective ways to solve these two problems of shortage of resources and environmental pollution that are the two major challenges the social facing today. Tens of millions of pieces of waste electrical appliance were produced in today’s China every year, facing such a large the removal of waste products, disassembly line is the most efficient organization form of removal. Therefore, the research of the disassembly line balancing problem (DLBP) has a great significance. DLBP is an NP problem, artificial bee colony (ABC) algorithms based on bionic principles have obvious advantages. This paper’s research about ABC algorithms solving the line balance problem is combined with the national natural science foundation of China (51205328). Standard ABC algorithm converges slowly, easy to fall into local optimum. Therefore, an modified ABC algorithm for DLBP was proposed in this paper.In this paper, the improvement measures mainly include the following four aspects:(1) When the initial solutions are generated, through adding the task hazard and demand factors, improve the quality of the initial solutions in terms of hazard and demand.(2) A variable step length adjusted by a local optimal solution and the current solution is designed to strengthen efforts to search near optimal solutions and accelerate the elimination of poor quality, slow optimized solutions for employed bees.(3) A squirmy search strategy that makes hazardous and high demand tasks carry out a micro-move forward is designed for onlooker bees to enhance their ability to optimize subsequent goals.(4) Estimation of distribution algorithm embedded the attenuation operation is introduced into the search strategy of scout bees, improve to take use of both heuristic search learning the corresponding relation between the task and the location and the global random search strategy, effectively improve the quality of scouts’ search. Finally perfect the algorithm process, set the related parameters, programmed the algorithm with MATLAB software.Use modified ABC algorithm to solve a large number of instances with different sizes and characteristics, for example, complex precedence relation, a variety of removal direction, uncertain removal time, and compared with the standard ABC algorithm and other algorithms in the literature. the results show that, for small-sized and medium-sized examples, the proposed algorithm can quickly find the optimal solution, and for large-sized examples, also get better optimization results than the results of the literature, Fully Verify the effectiveness of the algorithm.In chapter v of this article, use the modified algorithm to improve a certain enterprise’s actual production case, and obtain better result than the original plan, improve the balancing rate and production efficiency of disassembly line, fully demonstrated the practical significance of the work in this paper.