Adaptive signal fusion based on relative fluctuations of variable signals

摘要

To solve the problem of the total mean square error of the traditional random weighting algorithm being much larger when the target signal is variable than when the signal is constant, an improved adaptive random weighting algorithm is proposed. To bring the estimated value closer to the true value, the algorithm adjusts the proportional relationship between the current acquisition signal and the historical acquisition signal adaptively by using the relative fluctuation value of the collected signal. Because the relative fluctuation value can be adjusted autonomously according to the signal variation, it can be combined with the traditional random weighting algorithm very successfully. Numerical simulations demonstrate the effectiveness of the proposed fusion method. (C) 2019 Elsevier Ltd. All rights reserved.