Multiple Bee Colony Algorithm And Its Application In Crowd Evacuation Simulation

In today's social life,fires,earthquakes,and stampede accidents often occur.Therefore,the simulation of emergency evacuation in public places where people gather is getting more and more attention,and the simulation of crowd evacuation behavior in real scenes has received more and more attention.Research and analysis of the evacuation movement of the crowd have fundamental significance.Moreover,simulating the behavior of evacuated people has primary social relevance.For example,it can help the emergency department to formulate similar emergency plans and assist in the design guidance of evacuation scenarios.This method can minimize the occurrence of disasters and reduce casualties.For traditional evacuation drills,this form not only consumes a lot of workforces and material resources but also there may also be casualties,so crowd simulation has become a hot trend in the study of population evacuation.By using computer technology to simulate real scenes,not only can real-time simulations be performed on actual stages,but also more reasonable and valid opinions and suggestions can be proposed for crowd evacuation in real scenes.At present,there are some technical deficiencies in the existing crowd evacuation simulation model,which can not wholly and effectively simulate the group aggregation phenomenon and pedestrian movement law in the real crowd evacuation process.To solve the existing deficiencies in this field,this paper proposes a multi-swarm artificial bee colony algorithm and applies it to the crowd evacuation simulation movement.The method combines the macro evacuation simulation method with the micro evacuation simulation method to improve the shortcomings of the two evacuation simulation methods used in large-scale crowd evacuation simulation.The macro-level evacuation method is to use the multi-swarm artificial colony algorithm to perform path planning at the macro level after the original synthetic bee colony algorithm population initialization step.The bees that select the same food source are divided into one group,the selection of the dead bees and the role conversion of the remaining bees are performed in groups,and the information of the best path is conveyed and transmitted in the same group.The convergence rate of the bee population after grouping and the diversity of the bee population are reduced,so to increase the global communication ability of the bee colony.Adding the auxiliary community accelerates the transmission speed of the optimal solution,which increases the convergence speed of the proposed algorithm and improves the diversity of the city.The micro-distribution method is to use the improved social force model to guide the micro-motion of the individual.The combination of the two approaches can effectively improve the efficiency of the evacuation of people in public places,ensure the safety of evacuated peoplein emergencies,facilitate the design and implementation of evacuation programs,and provide guidance and assistance for real evacuation drills.The main work and innovations are summarized as follows:1.A multi-swarm artificial bee colony algorithm(MABC)is proposed.The theory of the K-means algorithm is used to divide the bee colony into multiple sub-bees based on the original ABC algorithm.Combining the global communication mode and the local communication mode in the sub-bee colony,the fitness function based on the local communication mode is extended to understand the diversity of the scheme,and the global communication mode of the food source location update speeds up the convergence of the algorithm and avoids falling into local parts.The algorithm proves that the algorithm has better performance.2.An evacuation simulation method is proposed,which combines the MABC algorithm with the improved social force model.In the real evacuation scenario,there is often a group phenomenon of group aggregation.This paper uses the MABC algorithm to plan the macro-level path and divides the sub-group to achieve the collection phenomenon in the real scene process.At the same time,this paper proposes an improved social force.The model guides the individual's microscopic level of motion behavior and completes the crowd evacuation simulation in simple scenes and complex scenes.3.A crowd evacuation simulation system was constructed to combine the MABC algorithm with the improved social force model.The system design includes six modules: scene modeling,semantic extraction,group division,path planning,crowd movement,and data export.The simulation environment of the system and the simulation results of the system are introduced in detail,and the practicability and effectiveness of the proposed method are verified.The traditional artificial bee colony algorithm has many advantages,such as strong global optimization ability,simple operation,and easy implementation.In this paper,a multi-swarm artificial bee colony algorithm is proposed,which overcomes the shortcomings of traditional artificial bee colony algorithm,which is premature convergence and easy to fall into local optimum.The algorithm is tested by test functions,and the algorithm shows good performance.At the same time,the crowd evacuation simulation system constructed in this paper is used to simulate the path of the proposed multi-swarm artificial bee colony algorithm.It has a good evacuation effect for the group phenomenon in the real scene evacuation process,and has better evacuation effect than other evacuation methods.