Font Size: a A A

A fuzzy logic system for predicting environmental costs from product design

Posted on:2000-08-22Degree:Ph.DType:Dissertation
University:University of Southern CaliforniaCandidate:Wu, Chihsien EricFull Text:PDF
GTID:1469390014964522Subject:Engineering
Abstract/Summary:
The purpose of this research was to develop a convenient and less expensive method to estimate environmental costs for companies. The environmental costs can be divided into two categories: measurable and intangible costs. Traditional reductionist and black box methods have difficulties in data gathering or repeatability, which can be avoided by using a fuzzy logic system. There are four major components in the fuzzy logic system: they are fuzzification, fuzzy rules, fuzzy logic engine and defuzzification.; The research was performed in three steps: selection of important attributes, fuzzy logic engine construction by using short and long profiles describing various industries, and case studies. The attributes of “produces air, water or soil pollution,” “requires use of scarce or nonrenewable resources,“ “may cause injuries to customers or public“ and “may cause injuries to personnel, requires safety and environmental protection training” were selected as factors influencing environmental costs to companies. These attributes were built into fuzzy logic engine as antecedents of fuzzy logic rules. The fuzzy logic engine was then trained by the experts' evaluations of attributes and direct estimates of environmental costs based on the profiles provided. The relationships among attributes were characterized by importance exponents that defined the fuzzy logic rules. The correlation coefficients and least squares errors were used as indicators of the fuzzy logic engine performance. Between the model results and the experts' direct estimates, the correlation coefficient reached 0.90. The five cases from completed projects or from literature provided the real environmental costs. The model was tested and compared to these real costs, and the correlation coefficient reached 0.84.; The use of fuzzy logic produced a pathway to facilitate the use of human knowledge and experience. The fuzzy logic engine clarifies relationships between the attributes and environmental costs. Its convenience and low cost make it available for various industries to adopt it as a quality control tool and preliminary screening method in choosing the most cost-effective strategy.
Keywords/Search Tags:Environmental costs, Fuzzy logic
Related items