The paper presents a paradigm of addressing large-scale optimisation problems using customised tools. Tailor-made techniques limited to specific applications are employed to effectively invest resources for solution searches. Integration of special logic is implemented at the modelling level with emphasis on the multiperiod planning of steam turbine networks. Logic extracted from conceptual preprocessing is employed to reduce combinatorial load. Logical constraints screen redundant degrees of freedom and customise the solution space to the specific application. The methodology decreases considerably the computational cost while offering a high degree of physical insights.
展开▼
