Route Optimization Of Fire Evacuation Based On Improved Ant Colony Algorithm

Building fire has become the most serious casualties and direct loss disaster. Thecomplexity and integrated of modern architecture have brought new challenge toefficient evacuation.The Ant Colony Algorithm(ACA) is anevolutionary algorithmwith good inspiration mechanism, strong robustness,parallelism and Easy toassociative, can meet the flexible complex needs and goals of evacuation problems,and easier to converge to the optimal solution. Various types of factors affect the fireevacuation time were studied,the dynamic parameters of firewere introducedtoimprovethe selection strategy and the pheromone update strategy of ACA. Theimproved adaptive ant colony algorithmwas applied to thedynamic path optimizationin fire evacuation,it show that the convergence speed and global search capability ofthe algorithm were improved. Main conclusions are as follows:(1)Based on the research of the environment factorimpact to evacuate personnel, thecalculation methodof impact parameters to personnel speed oftemperature, carbonmonoxide and visibility was determined, personnel activity index in fire wascalculated,to determine the calculation methodof the passage difficulty coefficient,personnel activity index and theequivalentlength.(2)Heuristic function was calculated by using the equivalent length instead ofgeometric length to optimize the personnel transfer strategy;the pheromonevolatilization factor was adaptive changed to improve the pheromone update strategy,thusimprove the convergence speed and global search capability.(3)The space model was established through the grid identification method, the staticand dynamic attributes of nodes and evacuation routes was defined and described.The parameters’ impact to improved adaptive ant colony algorithm was testedinthe30*30grid simulation environment, theoptimal parameters weredetermined.Differentsizes of grid environmentwith different export quantity werecreated to simulate using the testingparameters. It showed that we should set differentparameters combination in different sizes of grid environment.(4) Different grid environmentsof three fire scenarios were created, differentfireparameters were setting to simulate dynamically.The results show that the optimal evacuation path change dynamically with the development of fire: the greater theimpact on personnel speed of fire，the longer the evacuation equivalent length，thelonger the evacuation time，As a result, the optimal path algorithm output alsodynamic change with the change of the equivalent length.The dynamic parameters of firewere introduced to improve the ACA，thedynamicpath optimization algorithm for fire evacuation was build. The effectiveness of thealgorithm was verifiedthrough simulation.