Research Of Application Of Machine Learning In Effectiveness Evaluation Of Electromagnetic Gun

Effectiveness evaluation of weaponry plays an important role in the development of weaponry.High-altitude weapon is an important component of terminal air defense force and modern war,and electromagnetic gun is the focus of research and development of new antiaircraft weapon system,so it is necessary to evaluate the effectiveness of electromagnetic gun system.In this paper,the effectiveness evaluation of electromagnetic gun is mainly divided into three parts.Firstly,the construction of the comprehensive efficiency index system of electromagnetic gun is studied,and the solving methods and ideas of the five comprehensive index values are studied.Then,the hit effectiveness related to damage is studied,and then the comprehensive effectiveness evaluation is continued.Firstly,based on the project of "Demonstration of a certain type of electromagnetic gun scheme",this paper studies the problems related to the calculation of hit probability of electromagnetic gun,such as target determination,element solution,error analysis,etc.The calculation program of hit probability of electromagnetic gun is compiled with MATLAB,and the calculation example is given and the hitting probability is compared with that of common antiaircraft guns..Then,using the adaptive particle swarm optimization and support vector regression machine and BP neural network combined machine learning method,the comprehensive performance of the electromagnetic gun is studied.Machine learning is the key technology in the current era of big data and artificial intelligence.Effectiveness evaluation technology needs to keep pace with the times,introduce new ideas,and conduct innovative research on evaluation methods.Combining with the optimization characteristics of adaptive particle swarm optimization algorithm,this paper optimizes the support vector machine regression model and BP neural network,establishes a new machine learning model,and uses it as an evaluation model for the comprehensive effectiveness of the electromagnetic gun system,and the comprehensive effectiveness of the electromagnetic gun system is evaluated and analyzed.This paper also systematically studies the basic theories of performance evaluation and machine learning,and introduces the sources of data used in machine learning.The basis of machine learning is data.Learn rules from data and determine the undetermined parameters of the model,and obtaining the model for evaluation analysis.In this paper,the main data of the service probability and the smashing probability of the electromagnetic gun are obtained by a simulation system.Combined with the classical evaluation model ADC method of WSEIAC,the comprehensive performance of the electromagnetic gun is obtained,which is used as the complete data set used for supervised learning training.Finally,the program implementation of the model.This paper uses the most widely used language Python in the field of artificial intelligence,and writes the main program of the evaluation model and the data processing program of the feature engineering,and imports and trains the data.The solution function of the machine learning performance metric and the image output of the evaluation result are written for more intuitive observation of the model's comprehensive performance evaluation results of the electromagnetic gun and the performance of the model.