ISTANBUL KÜLTÜR UNIVERSITY DEPARTMENT OF INDUSTRIAL ENGINEERING SEMINAR SERIES
A novel fuzzy clustering-based task allocation method for location and routing of multi robots in the response phase of disasters
by

Abdullah OSMAN M.Sc.

Abstract:
Between 1998 and 2017, 1.3 million people died, 4.4 billion were wounded, or were displaced due to natural disasters, according to CRED (2018). This study considers the problem of maximizing the survival probability of survivors in severe and large-scale natural disasters emergency response operations using multi robots. We present a novel two-stage probabilistic clustering then routing strategy called (P-MWFCM-ACO) to solve location-routing problems from the perspective of emergency response operations. In the first phase of the suggested method, we used a modified weighted fuzzy c-means clustering algorithm to determine relief depot locations based on weights related to a survival probability function. In the second phase, a standard ACO algorithm was applied to identify the optimal paths for multi-robots between depots and survivors. We presented two objective functions to solve the problem and verify our proposed novel objective function's superiority. The first objective’s goal is to minimize route’s length within each cluster which is the most often used objective function in multi robot task allocation (MRTA) problems. The second one is a novel objective function based on survivor survival probability that maximizes survivor survival probability while decreasing total number of deaths and total distance travelled by robots. The proposed method was com-pared against two clustering techniques: k-means and fuzzy c-means clustering algorithm and the first objective function. According to simulation results, our strategy outperforms others in terms of reducing the overall number of deaths, improving the probability of survival.

Keywords: Emergency response, MRTA, Task clustering-based allocation, Fuzzy c-means clustering, Ant colony optimization

Biography: Mr. Abdullah Osman received his B.Sc. degree in Industrial Engineering from Aleppo University in 2017 and M.Sc. degree in Industrial Engineering from Istanbul University – Cerrahpaşa in 2022. He is an Optimization Engineer and an active user of Python/Matlab who is currently working on multiple projects to optimize multi robots task allocation (MRTA) using clustering-based optimization methods