基于多尺度逼近的多维传感信息解耦新方法

针对多尺度插值解耦方法的尺度特征计算和尺度阈值优化求解问题，提出了一种基于多尺度逼近的多维传感信息解耦新方法，该方法通过函数多尺度逼近快速获取尺度特征，避免了特征方程的构建，根据尺度取值顺序确定尺度阈值和相应的插值方法，实现准确度目标下的快速解耦。结果表明，尺度阈值等于 时，多尺度插值解耦方法的解耦误差约为0.4575%,计算时间为80ms。

Novel approach of multi-dimensional sensing information decoupling based on multi-scale approximation

Aiming at scale calculation and scale threshold optimization solution problems of multi-dimensional sensing information decoupling, a novel approach of multi-dimensional sensing information decoupling based on multi-scale approximation is proposed. The approach uses multi-scale approximation of function to calculate scale rapidly, it avoids the constitution of eigenfunction, according to the scale order of sensing information, scale threshold and corresponding interpolation methods are selected, the selected scale threshold can ensure the desirable decoupling accuracy in the best time. Experimental results show that, the error of multi-scale interpolation decoupling can reach to 0.4575% and calculation time is 80ms when scale threshold equals to 2-4.