Accuracy analysis of satellite antenna panel expansion based on BP neural network

Abstract

AbstractLarge deployable space mechanisms are widely used in the field of aerospace and have been paid increasingly high attention recently. The satellite antenna expansion system is the classic large deployable space mechanism. However, during the expansion of the satellite antenna deployable mechanism, the expansion accuracy is affected by the existing various uncertain factors which even result in scraping the satellite. For example, the hinge locking error has the significant influence on the deployment accuracy of the satellite antenna panel. In term of this issue, considering the advantage of the back‐propagation (BP) neural network for the high dimensional nonlinear problem, this paper mainly adopts it to analyze the impact of hinge locking error on the expansion accuracy of the antenna panel. The results show that the probability of the actual flatness deviation of the satellite antenna panel falling in the required accuracy area is 99.56%, and the probability of the actual pointing angle deviation falling in the required accuracy area is 99.84%.