An AHP and FAHP based Expert System and Relief Program for Risk Assessment of Mountain Accidents

• Main Authors: Chen Yung Ting, 陳勇廷
• Other Authors: Huang Li Jeng
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/24776750026834258940

碩士 === 國立高雄應用科技大學 === 土木工程與防災科技研究所 === 103 === Mountain climbing is an important act for human beings during holidays. However, there exist a lot of hidden risks occurred in the mountains. This research is proposed to study the assessment of risks of mountain accidents using different approaches such as Analytial Hierarchy Process (AHP),Fuzzy Analytic Hierarchical Process (FAHP) and fuzzy logic inference method. Before all these approaches a basic hierarchy structure of risk assessment is built up in whish three levels are included, the goal, the criteria and the sub-criteria. The criteria level is divided into three categories: natural factors, human factors, and hybrid factors while the sub-criteria involves seven influence factors: climate environment, geomorphology environment, animal and plants hurt, internal influence, external influence, mountain sickness and downhill delay. The relative risk influence factors (RRI) will be first analysed using AHP and FAHP with both triangular and trapezoidal membership functions. After the evaluated values of all influence factors for an accident event have been assigned by the analyst, the overall risk index (ORI) can be evaluated. Four practical cases of mountain accidents occurred in Taiwan will be considered as numerical examples and the results will be checked by the usage of fuzzy inference method from MATLAB fuzzy logic toolbox. Results show that AHP, FAHP (Triangular), FAHP (Trapezoidal) and MATLAB fuzzy logic toolbox all have coherent risk predictions. Furthermore, a preliminary expert system for risk assessment of mountain climbing executable on android smart phones and platforms has been developed and the results are successful and effective for practical applications.