The improved algorithm based on entropy information to multi-sensor fusion
Kalman filter has good ability about signal tracing and valuations, But because of the signal noise, the signal estimate by single sensor kalman filter will produce certain deviation, the signal is against us to predict and estimate, multi-sensor filter can weaken the adverse effects and improve the accuracy of the estimation. In this paper give kalman filtering algorithm based on the minimum entropy of the error estimation. Through the new rate changes to fixed estimate and introducing new assignment function to measure the real-time weighted value. Finally, through multi-sensor fusion, the results achieved very good signal filtering and estimation.
information entropy kalman filter new interest assignment function
Kang Jian Xie Hong Li Yi-bing
Harbin Engineering University Harbin, China
国际会议
桂林
英文
9-12
2010-11-17（万方平台首次上网日期，不代表论文的发表时间）