Classification and Recognition of Moving Targets on Sea Based on Genetic Algorithm Optimization of Probabilistic Neural Network Algorithm
In order to safeguard our country's maritime security interests,it is of great significance to study the classification and identification of moving targets at sea,and then effectively track and manage key targets.In this study,first,by extracting the characteristics of moving targets on sea,six moving characteristic indicators including moving direction velocity,average velocity,maximum velocity,standard deviation of velocity,average acceleration and track span range are obtained.Then,the characteristic indicator data is analyzed through simulated annealing algorithm fuzzy C-means clustering to obtain six cluster centers.Finally,a method for identifying moving targets on the sea based on genetic algorithm optimization probability neural network is pro-posed.Simulation results show that,compared with machine learning methods such as KNN algorithm and decision tree,this meth-od can more effectively classify and identify moving targets on the sea,providing technical support for maintaining my maritime se-curity.
maritime securitymoving targetclassification and recognitiongenetic algorithmprobabilistic neural network
万方数据
维普
