Improved Gravitional Search Algorithm-Based Adaptation Methodology Of CBR For Emergency Plan

In today's society,emergencies occur frequently.How to deal with emergencies has become an important issue.Among them,the emergency plan is of great significance for us to deal with emergencies effectively.Case-based reasoning(CBR)theory simulates human cognitive reasoning process.There are a large number of emergency cases in reality,which contain a lot of emergency management experiences and knowledge that can be used for reference.CBR has a natural advantage in emergency management decision-making.Therefore,we use case-based reasoning technology to retrieve and reuse emergency plans.However,it is too simple to retrieve and reuse the existing cases from the case base as the emergency plan,and the applicability of retrieval case for the current emergency is poor.Therefore,this paper proposes an adaptive emergency plan generation system based on CBR theory and fuzzy gravitational search algorithm(FGSA).The main research contents include the following aspects:(1)In this paper,the mechanism of gravitational search algorithm(GSA)is described,and the defect of using single function to control important parameters is analyzed.To solve this problem,a fuzzy system is proposed.The system introduces tow parameters: agent diversity and optimization process to detect the performance of the algorithm.And new control functions and input parameters are used to measure the object diversity and optimization process of the algorithm.Then we combine the important parameters (70) and in GSA with twelve kinds of fuzzy rules to balance the exploration and development ability of the algorithm.These efforts improve the convergence ability of the algorithm and avoid falling into the local optimum.(2)The fitness function is very important to evaluate the merits of the individual in the search space of FGSA,that is,the merits and demerits of the solution.In this paper,a multi-attribute fitness function is set up,which includes four indicators: disposal cost,disposal efficiency,sensitivity and hazardous area range.The weight of each indicator is calculated using the attribute hierarchy model.(3)This paper implements the process of CBR.And it focuses on the case adaptation process.First,the most similar case is extracted from the case base,which is regarded as the initial solution,that is,the initial agent.Second,set the range of each attribute value based on the initial agent,and generate the initial population randomly.Third,the fitness value of each agent is calculated,and the iteration process of FGSA is cycled according to the set number of iterations.Finally,the agent of the optimal fitness value,which is the optimal solution,is found to realize the adaptation of the emergency plan.(4)In this paper,15 benchmark functions are used to test the performance of our proposed fuzzy gravity search algorithm,and compared with GSA,CGSA,CLPSO,NFGSA,PSGSA,EKRGSA.The comparison results show that this method has good performances.More importantly,we designed the FGSA-CBR system and simulated an explosion and gas leakage accident in a certain area as a modeling study to check the system performance.The results represent that the system has good performance in the adaptive generation of emergency plan.In conclusion,the combination of CBR technology and fuzzy gravity search algorithm has achieved good results.It realizes the adaptation of emergency plan,provides guidance and help for emergency decision-makers,and offers technical support for improving emergency management system.