An intelligent system for respiratory protection

摘要

This paper presents an intelligent system which was developed for helping end-users choose suitable respiratory protection and estimate the service life of a respiratory protection system according to circumstances in different workplaces. In traditional rule-based expert systems, users are required to input system information completely before any respiratory protection recommendation can be obtained. If the particular user does not know any of the required information, the system would do almost no benefit to him. The present system incorporates with some artificial intelligent (AI) elements, such as fuzzy logic, genetic algorithm and case-based reasoning, to provide flexibility and versatility for general users. With a step-by-step question-and-answer interaction mode, the intelligent system would be able to elicit as much input as the user knows during processing. For some unknown information, the developed system would use fuzzy logic to guess their values based on the user's perceptions. In addition, for incomplete input, the system would use a genetic algorithm and case-based reasoning to estimate the service life of some recommended respiratory protection system. It can generalize from similar cases and provide an updated recommendation for later applications.