针对产品研发项目共享资源冲突导致组合进度难以估计的问题,改进了基于市场的资源竞标与价格更新机制,基于能力演化模型及效率转化函数构建资源占用时长动态模型,进而以资源成本和拖期惩罚为目标构造了子项目竞标策略效用函数,并提出改进遗传模拟退火算法进行求解;为模拟多项目资源竞争给组合进度带来的影响,基于上述研究引入Repast平台构建了项目组合进度多Agent仿真模型;通过一个仿真示例验证了该方法的可行性与有效性。
The product's research and development project portfolio scheduling was difficult to evaluate caused by shared resources confliction. Aiming at this problem, the mechanism of resource bidding and price updating based on market was improved. On the basis of capability evolution model and efficiency transformation function, a dynamic resource occupation model was developed. Thus a subproject bidding utility function was constructed with objectives of resource cost and delay penalty. Improved simulated annealing genetic algorithm was also proposed to solve the convergence and global search speed. To simulate the influence of multi-project resource competition on research and development project portfolio scheduling, a multi-Agent simulation model was constructed by introducing Repast platform based on the above researches. The effectiveness and feasibility of proposed method were verified by a simulation case.