Cellular Automaton Based Personnel Evacuation Route Optimization For Ship

With the development of Ship Industry, internal structures of ships have become more complicated, ships can carry an increased number of people, which draws attention of public to the problem of their large-scale emergency evacuation. Due to the large density of people on board, narrow aisles and other features of ships, when emergencies arise, obstruction and stampede phenomenon during evacuation affects the safety of people. Therefore, optimization of ship evacuation route has great scientific value and engineering significance.The current evacuation route research is divided into two areas: route optimization and personnel motion simulation. The intelligence optimization algorithm and model of personnel motion simulation are methods to solve problems of route optimization. The intelligence optimization algorithm may obtain optimal route, but due to the crowding, obstruction and retention of people cannot be avoided during evacuation, it cannot reflect the effect of reducing evacuate efficiency caused by obstruction calculating only with ant colony algorithm.The simulate result of personnel motion simulate model is the result of random evacuation,which cannot optimize basing on optimization goals. Therefore, only by combining the two methods can we optimize the large-scale evacuation route on ships.Thus,this paper will propose the CA-ACO(Cellular Automata-ant colony optimization )model, which combines the improved ACO-S algorithm and CAOS personnel motion simulation. Comparing kinds of algorithm, we finally choose and improve ant colony algorithm as the basement of ACO-S (ant colony optimization-shipbuilding) .By improving the Ant Colony algorithm we propose ACO-S algorithm. The improved ACO-S is part of CA-ACO model, and proposed the CAOS(Cellular Automaton on ship-design) model, which added the broaden horizons and "Ground Field" by Kirchner. The CAOS is the simulation part of CA-ACO.The CA-ACO model evenly divides the deck into grids, sets the initial position, position of shunt flow and confluence flow into nodes. Determine the direction of macroscopic motion with the intelligence optimization algorithm; simulate the microscopic motion of the nodes and the range of paths connected with nodes with model of personnel motion simulation, finally achieve the fastest evacuation route and the position of obstruction during evacuation. Ship designers can make evacuation plans basing on the calculate result ofthe model, and can give out suggestions of cabin arrangement/general layout during the initial design phase.