基于组合拍卖方法的资源受限多项目调度

Resource-constrained multi-project scheduling based on combinatorial auction method

为求解资源受限多项目调度问题,提出了一种基于组合拍卖机制的项目资源分配方法。该方法将单位时段的单位资源看作拍卖品,引入动态资源价格,将单个项目进度计划看作一个组合拍卖品。要求单个项目进度计划在满足任务紧前关系与资源约束的前提下,既要减少项目误期赔偿,又要降低项目资源成本。通过逐步调节不同资源在不同时段的价格,引导各项目合理利用不同时段的资源,从而解决各项目之间的资源冲突,实现资源受限多项目调度问题的整体优化。通过实际算例,演示了该方法的资源调配过程,并通过数字实验验证了该方法的有效性。

A project resource allocation method based on combinatorial auction was proposed to solve the resources- constrained multi project scheduling problem. In combinatorial auctions, a unit of resource within certain time period was regarded as a single bidding object. Dynamic resource price was then introduced for these bidding objects. For an individual project, its schedule was regarded as a combinatorial bid, or a combination of bidding objects. During the auction process, the single project scheduling was required to minimize both the project tardiness and payment for using resources, without violating the precedence relationship and resource constraints. The resource prices were adjusted gradually to refleet the bidding situation and to push competing projeets to use resources at low price time periods. Hence, resource conflicts could be resolved eventually, and a globally optimal scheme of resource allocation among projects could be implemented. A multi-project case was adopted to demonstrate the application of the proposed combinatorial auction method, and the effectiveness of this method was verified.