A methodology for the design of efficient resource conservation networks using adaptive swarm intelligence

摘要

The implementation of resource conservation schemes in industry can be enhanced through the application of systematic design methodologies. In particular, process integration methods allow resource consumption and waste generation in industrial plants to be reduced through the identification of efficient material reuse/recycle schemes. Various approaches, ranging from graphical pinch analysis to mathematical programming, have been developed by different researchers. Mathematical programming techniques provide considerable flexibility in the representation of network design problems, although in many cases, these approaches result in mixed integer non-linear programming (MINLP) models which are difficult to solve. This paper presents a simplified approach using a zero-one programming or "knapsack" formulation for the design of industrial material reuse/recycle networks. It is possible to solve the resulting model using an efficient heuristic algorithm based on adaptive particle swarm optimization. Two sample applications are provided to illustrate the methodology. The first case shows the application of the methodology to the implementation of industrial water conservation and the second case demonstrates its use in the design of a hydrogen gas reuse/recycle scheme in a refinery.