CAJ | 학술논문

为了求解车间调度这一NP问题,提出了基于动态疫苗库的免疫遗传算法.本算法改变了以往的基于工序的编码方式,采用基于优先权的编码方式,设计了相应的交叉和变异方式.同时,在不断地调整基因库和进行疫苗接种的过程中来判断基因库中基因片段的优劣,以此来不断动态地调整疫苗库,使得更好的疫苗进入疫苗库中,更好地指导种群的进化.仿真实验表明,该算法是高效的.
위료구해차간조도저일NP문제,제출료기우동태역묘고적면역유전산법.본산법개변료이왕적기우공서적편마방식,채용기우우선권적편마방식,설계료상응적교차화변이방식.동시,재불단지조정기인고화진행역묘접충적과정중래판단기인고중기인편단적우렬,이차래불단동태지조정역묘고,사득경호적역묘진입역묘고중,경호지지도충군적진화.방진실험표명,해산법시고효적.
This paper proposes the immune genetic algorithm based on dynamic vaccine pools to solve the NP-hard workshop scheduling problem. The algorithm adopts preference list-based representation that is different from the past representation based on the working procedure, and gives the corresponding crossover and mutation ways. Meanwhile,in the process of inoculation, the algorithm constantly adjusts the gene pool and judges the gene fragments for the capability of the gene pool,so as to make the better vaccine access to the vaccine pool to better guide the evolution of the population. The simulation experiments show that the algorithm is efficient.