| Abstract
| - While the first part of this series focuses on the application of the proposed formulation toscheduling, this paper focuses mainly on the integration of planning and scheduling inmultipurpose batch plants. In dealing with this problem, the method presented in this paperexploits the mathematical structure of the overall plant model. It is discovered that the overallmodel exhibits a block angular structure that is decomposed by raw material allocation. If rawmaterials can be allocated optimally to individual plants, solving individual models for eachplant can produce the same results as solving an overall model for the site. This discovery leadsto a decomposition strategy that consists of two levels. In the first level, only planning decisionsare made, and the objective function is the maximization of the overall profit. The results fromsolving the planning model give optimal raw material allocation to different plants. In the secondlevel, the raw material targets from the first (planning) level are incorporated into the schedulingsubmodels for each plant, which are solved independently without compromising globaloptimality. The objective function for each scheduling submodel is the maximization of productthroughput. The scheduling level uses the concept of the state sequence network presented inpart 1. Solving scheduling submodels for individual plants rather than the overall site modelleads to problems with much a smaller number of binary variables and, hence, shorter CPUtimes. If conflicts arise, i.e., the planning targets are too optimistic to be realized at the schedulinglevel, the planning model is revisited with more realistic targets. This eventually becomes aniterative procedure that terminates once the planning and scheduling solutions converge withina specified tolerance. In this way, the planning model acts as coordination for scheduling modelsfor individual plants. An industrial case study with three chemical processes is presented todemonstrate the effectiveness of this approach.
|
