Improvement Of PSO And Its Application In Network Situation Prediction

The network situation value is an important index to measure the network security.The establishment of an effective network situation prediction model,which can prevent the occurrence of network security incidents.It plays an important role in network security protection.Through the understanding and analysis of the network situation,we can see that there are many factors that affect the network situation and the relationship is complicated.It is difficult to establish a more accurate mathematical expression to describe the network situation.Therefore,this paper uses the improved PSO algorithm to optimize the parameters of the support vector regression model.To establish the nonlinear mapping relationship between the influencing factors and the network situation.To realize the prediction and protection of the network situation.The main work as follows:Aiming at the shortcomings of particle swarm algorithm,this paper proposes a multi-swarm chaotic particle optimization algorithm.The algorithm is based on the randomness of the chaos principle.The swarm is initialized at the initial stage,and the particles are divided into three swarms.Different swarms adopt different updating strategies.Through the cooperation and information sharing among the three swarms.The convergence rate of the algorithm is accelerated.The variance of the swarm is used to determine whether the particles fall into local convergence.The chaos processing help them escape from the local convergence point,and the "premature" phenomenon of the swarm is avoided with certain efficiency,so as to improve the optimization performance of the algorithm.So as to improve the optimization performance of the algorithm.Finally,the improved algorithm is experimented by four standard test functions,and compared with PSO and LDW-PSO,it is proved that the algorithm has better performance.Statistics from 2013 to 2016 National Internet Emergency Response Center weekly published on the website of the situation report to determine the sample data set.The MSCPO-SVR network trend prediction model is established by optimizing the parameters of the vector regression model with multi-swarm chaotic particles,and the prediction results are obtained.Finally,by comparing with the BP neural network and the SVR prediction model,it isverified that the prediction result of the MSCPO-SVR prediction model is stronger fit than the real value.