Study Of Automotive Line Balancing Problem With Improved Genetic Algorithm

Based on the analysis of huge domestic&foreign related research papers,This paper focuses on the research of "Assembly Line Balancing" optimization problem in Automotive industry.Based on the comprehensive analysis of usual optimization algorithms in this domain,Author comes up with one improved GA based algorithm&its structure flow which embedded with simulated annealing approach inside and finished coding with the engineering simulation tool on which applied feasibility test with multi classic benchmarks.Meanwhile applied the engineering calculation with engineering data of one real car project.It analyzed line balancing scenarios which contained multi-different boundary constraints which was normally omoitted by previous researchers.It conquered the weakness point of traditional GA algorithm&broadening the application of traditional GA.It provided new ideas&approaches to analyze optimization problem of "Assembly line balancing".At same time it contributes the tighten relations via theoretical research&engineering requirement in this domain.Per the real status of both theoretical research&engineering application in domain of "Assembly line balancing",Author had done some detail research works in following aspects:1)Research on the improvement of GA.With summaries of every typical intelligent optimization algorithms,Aiming at analysis of the "pre-mature"problem of GA while the number of population is not enough large,Paper comes up with one new improved algorithm frame,in which embedded simulated annealing idea and self-adaptive evolutionary operator strategy inside and proved its convergence theoretically.2)With the improved GA algorithm,Author applied a careful research on ALB optimization.Through the carefully analysis and compares with multi-scenarios with different boundary restraints,Paper declared the possible deficient issues of traditional GA and through the introducing ultra-boundary restraints in improved GA algorithm which applied on the engineering calculation sample,it comes to a good result.3)Per mix-production scenario in automotive industry,Besides of balancing algorithm for mixed product assembly,Author comes up with a new balancing algorithm of mix-production based on different input-rate of mixed cars.Through considering the engineering constraints that rarely used in traditional balancing problem analysis,improved its practicality and feasibility.By applying on one engineering example,Algorithm can reach perfect engineering conclusions.4)Analysis of value range of key parameters in new algorithm and its possible relationship with the convergence of algorithm which includes both theoritcal analysis and experiement analysis,which provides the most experienced accumulation for further research of algorithm itself.