Study on the effectiveness evaluation method of formation Beyond Visual Range air combat based on genetic BP Neural Network

The Beyond Visual Range (BVR) air combat has become the most important mode of modern air combat. In this paper, a new model is set up combining situation assessment model and the formation combat capacity model. Genetic BP Neural Network is used for the effectiveness evaluation of BVR. Firstly, the main factors of the situation assessment in BVR air combat are proposed and analyzed. Secondly, Analytic Hierarchy Process (AHP) model of combat capacity assessment in BVR is established. The main factors are obtained by using Principal Component Analysis(PCA)to select input variables. A new model is presented as an AHP model integrated from the two above models, then, combine Genetic Algorithms(GA)with BP neural network, using GA's global to search the optimized BP network structure parameters, overcome the local convergence and solve other issues of BP algorithm effectively. Using the new model to get the input variables, GA-BP hybrid modeling is applied to effectiveness evaluation of BVR. Finally, a typical 2-VS-4 air combat example is presented to verify the model's availability. The results show the order of the attack of the reds that make the effectiveness evaluation maximum. The results of the numerical example show that the model can limit the artificial factors, making the solution more objective and creditable. Data link plays an important role in BVR air combat.