Title of article :
A knowledge-based fuzzy expert system to analyse degraded terrain
Author/Authors :
Genske، نويسنده , , Dieter D. and Heinrich، نويسنده , , Klemens، نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 2009
Pages :
14
From page :
2459
To page :
2472
Abstract :
The quality of a site investigation campaign stands and falls with the quality of the preparatory desk study. This article presents a fuzzy expert system that improves conventional desk studies. The evaluation process is referred to as a “multi-temporal desk study” (MTDS) and is similar to the conventional approach, falling back on existing maps, building permits, production data and aerial photos. As backward-looking appraisal it documents the spatio-functional and temporal changes of degraded terrain with regard to physico-chemical hazards present in the form of soil and groundwater contamination. The fuzzy approach differs from the conventional desk study in that it incorporates the semiotic-interpretative character of a MTDS in a rigid mathematical framework that is based on fuzzy set theory. By doing so, the interpretative grey zone inherent to all hazard assessments becomes more transparent and – above all – is not lost or forced into a inflexible “yes–no” or “1/0” scheme. The fuzzy expert system (FES) increases the quality of the desk study and reduces the costs for field campaigning, lab analyses and remediation work. The validity of the approach was confirmed by comparing a hazard map for a degraded harbour terrain produced by means of the FES with a hazard map reflecting ground truth based on a traditional field campaign. For the fuzzy analysis the only source of information were official building permits and aerial photographs taken at different time steps.
Keywords :
site investigation , Fuzzy Logic , Aerial photo interpretation , KRIGING , brownfields
Journal title :
Expert Systems with Applications
Serial Year :
2009
Journal title :
Expert Systems with Applications
Record number :
2345333
Link To Document :
بازگشت