A promising searching method 'Entrotaxis-Jump' for seeking hazardous gas source in a chemical cluster

A promising searching method 'Entrotaxis-Jump' for seeking hazardous gas source in a chemical cluster

연구 분야: Verification

학회: EM-GIS '19: Proceedings of the 5th ACM SIGSPATIAL International Workshop on the Use of GIS in Emergency Management

In the case of hazardous gas leakage accidents, it is significant to utilize search methods to guide a searcher to find the source and master the source term (source location, release strength) quickly. Because the essential source information can help the emergency management body to form a timely and effective accident disposal plan. This paper mainly studies the source searching problem in the chemical cluster where the searcher's movement is constrained by road network. We, firstly, introduce the Entrotaxis search algorithm, then combine it with the intermittent search strategy to propose the Entrotaxis-Jump algorithm. Besides, a prototype system of ground-based search robot is built for verifying performance of search methods and executing search tasks in a real scenario. Finally, this paper conducts a case study of a chemical cluster in Shanghai, China, and Entrotaxis-Jump algorithm proposed in this paper is verified to outperformance Entrotaxis algorithm. According to the results of 100 Monte Carlo simulations of two algorithms, the success rate of Entrotaxis-Jump is 34% higher than the Entrotaxis, and the mean search time is reduced by 374.94s.