Air Target Threat Assessment and Prediction Based on Improved GM(1,1) Model

Air Target Threat Assessment and Prediction Based on Improved GM(1,1) Model

연구 분야: Safety

학회: 2023 IEEE International Conference on Signal Processing, Communications and Computing (ICSPCC)

Due to non-dynamic and non-continuous problems, traditional air target threat assessment methods will result in low assessment accuracy and high threat risk. To this end, a novel threat assessment algorithm named IEICF-GM(1,1) (Improved Error and Initial Conduction and Feedback-GM(1,1)) is proposed in this paper. Firstly, based on their attribute information, the discrete threats of opposing targets that attack our detection sensors can be received. Secondly, the GM(1,1) model is employed to predict target threats, additionally, the correction factor is introduced to correct the inherent error of the background value in this model. Thirdly, the initial conditions of GM(1,1) model are optimized according to the threat's importance at the historical moment. Finally, the residual sequences between sensor measurement and target state prediction are fitted by using Fourier series method, and then the fitted results are feedbacked to the improved GM(1,1) model. Simulation results show that our algorithm has higher assessment accuracy, predicting dynamic and continuous target threats in the future.