UCAV situation assessment based on SAE-ELM

Abstract

A Stacked Autoencoder Extreme learning machine (SAE-ELM) assessment model is proposed to study the weight of an Unmanned Combat Air Vehicle (UCAV) situation assessment. On the basis of the existing air combat situation description parameters and their optimized situation advantage function weights, the mapping relationship between the air combat situation description parameters and the situation advantage function weights is established by using the SAE-ELM to achieve a more accurate evaluation. From the four error indicators of MAE, RMSE, MAPE, and R-Squared, the simulation experiment shows that the SAE-ELM evaluation error is the smallest and can effectively achieve accurate situation assessment.