HIDEN - A HYBRID INTELLIGENT SYSTEM FOR SYNTHESIZING HIGHLY CONTROLLABLE EXCHANGER NETWORKS - IMPLEMENTATION OF A DISTRIBUTED STRATEGY FOR INTEGRATING PROCESS DESIGN AND CONTROL

The development of computer-aided-design systems is the key step toward process design automation. The most difficult phase of this development is to endow the system with the capability to perform conceptual design, i.e., process synthesis. This is especially true when a synthesized process is expected to satisfy simultaneously economic and operational criteria. To meet these criteria, it is highly desirable that the first-principles and heuristic knowledge, which can be numerical or symbolic, structured or unstructured, be fully exploited and that the information and data, which can be precise or imprecise, certain or uncertain, be appropriately manipulated. In the present work, a hybrid intelligent design system for synthesizing exchanger networks (HIDEN) is developed by means of a knowledge-based approach, fuzzy logic, and neural networks. This system, built on an artificial intelligence workstation, fully implements the distributed strategy for integrating process design and control. It is capable of synthesizing heat exchanger networks (HEN's) and mass exchanger networks (MEN's) for the recovery of energy and material, respectively. The resulting exchanger networks are cost-effective as well as highly controllable.