This article considers robust fusion algorithms for linear system with multichannel stochastic multiplicative disturbance by using projection theorem and projection formula. Based on the centralized and distributed (feedback and no feedback) fusion strategies, we present optimal filtering fusion algorithms, optimal fixed‐interval smoothing fusion algorithms, and fixed‐interval deconvolution fusion algorithms. The methods of projection theorem and projection formula are used to guarantee the optimality of these derived algorithms. The presented algorithms are much less computation burden and higher accuracy than existing state‐of‐the‐art methods. Numerical simulation results including a WSN network model verify the proposed algorithms.

中文翻译：

---

基于多传感器观测的多通道随机乘法干扰鲁棒融合算法

本文采用投影定理和投影公式，考虑了具有多通道随机乘性扰动的线性系统鲁棒融合算法。基于集中式和分布式（反馈和无反馈）融合策略，我们提出了最佳滤波融合算法，最佳固定间隔平滑融合算法和固定间隔反卷积融合算法。使用投影定理和投影公式的方法来保证这些导出算法的最优性。与现有的最新方法相比，本文提出的算法具有更少的计算负担和更高的准确性。包含WSN网络模型的数值仿真结果验证了所提出的算法。