差分进化混合粒子群算法求解装配式建筑构件生产调度优化问题

Hybrid particle swarm optimization algorithm based on differential evolution for production scheduling optimization of fabricated building components

在满足按期交货的前提下,为解决装配式建筑预制构件生产成本最低时的构件生产调度问题,建立构件生产成本最小化模型。采用差分进化(DE)算法与粒子群优化(PSO)算法相结合的差分进化混合粒子群优化(PSODE)算法。结合实际案例,将PSODE算法与GA算法进行对比,结果表明,PSODE算法在按期交货的前提下,通过对构件调度顺序进行优化,构件生产成本降低5.5%。PSODE算法解决装配式构件生产调度问题的有效性和可靠性得到验证,且PSODE算法比GA算法更优。

A component production cost minimization model is established for the component production scheduling on the premise of meeting on-time delivery when the production cost of fabricated building components is the lowest. A hybrid particle swarm optimization algorithm based on differential evolution(PSODE)which combines differential evolution(DE)algorithm and particle swarm optimization(PSO)algorithm is adopted. The PSODE algorithm is compared with the GA algorithm by the actual cases. The results show that the PSODE algorithm can reduce the component production cost by 5.5% by means of optimizing the component scheduling sequence under the premise of on-time delivery. Its effectiveness and reliability for fabricated component production scheduling have been verified. In addition,the PSODE algorithm is superior to the GA algorithm.