An Efficient Fuzzy Expert System Architecture for Landfill Operation Reliability Management

摘要

An efficient, fuzzy expert system architecture is developed and presented, which is suitable for landfill operation reliability management Whereas fuzzy logic has already been applied to landfill design, on the other hand it has not been applied to landfill operation management. The contribution of this paper is, therefore, to demonstrate a novel application of fuzzy logic and, especially, of fuzzy expert systems, in a critical and very complex real world problem, namely, in landfill operation management. During the operation of a landfill several complex events like bad weather conditions, equipment malfunction, wrong operation practice etc., could lead to specific operation faults like litter and uncontrolled storm water flow. The landfill manager relies only on his experience to estimate if such an operating fault can possibly occur and how it can be prevented or restricted. Fuzzy Expert systems offer the means and opportunities for a variety of successful applications in very complex real world problems regarding mainly planning and decision making. One of these real world problems could be the solid waste management process, characterized by the co-existence of many scattered pieces of valuable knowledge and experience of experts as well as of vague, inaccurate, and insufficient pieces of information. In this research effort the authors are trying to record and codify the knowledge and heuristics of the experienced landfill managers in the form of a fuzzy fuzzy expert system based on fuzzy fault tree analysis, as the most suitable intelligent system architecture for this difficult and demanding real world problem. The fuzzy fault tree based fuzzy expert system herein developed is called LOMA (Landfill Operation Advisor), it could be continuously updated, and is addressed to all landfill managers, but especially to inexperienced. The main goal of this work is to present the LOMA's architectural structure, as well as its major functional characteristics and to illustrate the knowledge acquisition procedure that was followed by the authors for the development of the system. In addition, the goal is to reveal landfill operation and solid waste management as a critical and complex real world problem but suitable to be tackled by fuzzy expert systems.