基于相对波动的自适应信号融合算法

针对传统随机加权算法在目标信号为变量时,总均方误差远大于信号为常量时的情况,提出了一种改进的自适应随机加权算法.该算法利用所采信号的相对波动值自适应地调整当前采集信号与历史采集信号间的比例关系,得到更加接近真值的估计值.由于相对波动值可以根据信号变化情况自主调整,因此可以和传统的随机加权算法进行良好的结合.数值模拟证明...

Adaptive Signal Fusion Algorithm Based on Relative Fluctuation

An improved adaptive random weighting algorithm was proposed to solve the problem that in the traditional random weighting algorithm, when the target signal is variable, the total mean square error is much larger than the constant signal. In order to make the estimated value closer to the true value, the mentioned algorithm used the relative fluctuation value of the collected signal to adaptively adjust the proportional relationship between the currently acquired signal and the historically acquired signal. Since the relative fluctuation value can be adjusted autonomously according to the signal variation, it can be well combined with the traditional random weighting algorithm. The effectiveness of the proposed fusion method had been demonstrated by numerical simulations.