基于混沌粒子群的资源受限项目调度问题

Resource-Constrained Project Scheduling Based on Chaos Particle Swam Optimization

鉴于基本粒子群算法易陷入局部最优,提出一种将混沌算法嵌入基本粒子群的混沌粒子群算法,并将其用于求解典型的资源受限项目调度问题。采用基于优先值的粒子编码方式和串行调度方案,分别用基本粒子群算法和混沌粒子群算法对实例求解。并且比较了2种算法求解多资源受限项目调度问题的性能。结果表明:混沌粒子群算法在距最优值的平均偏差和达到最优值的次数百分比等性能上要优于基本的粒子群算法,并且混沌粒子群具有更好的收敛性。但是,混沌粒子群算法在计算达到最优工期的平均时间上略比基本粒子群算法逊色。

The resource-constrained project scheduling problem is addressed in this paper. As the basic particle swarm optimization （BPSO） method is easy to be trapped in a local optimum, a chaos particle swarm optimization （CPSO） algorithm is proposed. It combines the chaos algorithm with the BPSO and is used to solve the typical multiple resources-constrained project scheduling problem. To test the perform- ance of the proposed method, both BPSO and CPSO are used to solve an instance. Comparison shows that CPSO performs better than BPSO in average deviation from the optimum and the percentage of reaching the optimal value, etc. Also, it has better convergence than BPSO. However, in the average time to achieve the optimum, the CPSO is not as good as BPSO.