Protocol‐based extended Kalman filtering with quantization effects: The Round‐Robin case
Protocol‐based extended Kalman filtering with quantization effects: The Round‐Robin case
复制标题
DOI：
10.1002/rnc.5205
复制
发表时间：
2020-10
影响因子：
3.9
通讯作者：
Shuai Liu;Zidong Wang;Jun Hu;G. Wei
中科院分区：
文献类型：
--
作者：
Shuai Liu;Zidong Wang;Jun Hu;G. Wei
本文研究了一类具有量化效应的随机非线性系统的扩展卡尔曼滤波问题，并给出了系统的循环通信协议。均匀量化被认为是一个加性白色噪声序列服从一定的时间间隔上的均匀分布的特点和由此产生的量化误差。为了减少网络的通信量以及减轻数据冲突，RR机制被引入到调度从传感器到过滤器的数据传输。通过结合RR协议的周期性和零阶保持器策略，滤波器的输入信号由延迟量化测量序列建模。本文的主要目的是设计一个扩展卡尔曼滤波器的随机非线性系统，同时存在量化误差，随机非线性，随机噪声，这样的最佳上限的滤波误差协方差推导。通过求解两个耦合的Riccati-like差分方程，滤波器增益矩阵被显式地公式化。基于RR协议的递归滤波算法被开发用于在线实现。此外，还建立了滤波误差在均方意义下一致有界的充分条件。最后，通过仿真实例验证了所设计滤波算法的有效性.
This article investigates the extended Kalman filtering problem for a class of stochastic nonlinear systems with quantization effects and Round‐Robin (RR) communication protocols. The uniform quantization is considered and the resulting quantization error is characterized as an additive white noise sequence obeying the uniform distribution over certain intervals. For the sake of reducing communication traffic of the network as well as alleviating data collisions, the RR mechanism is introduced to schedule the data transmission from the sensors to the filter. By combining the periodic property of the RR protocol and the zero‐order holder strategy, the input signal of the filter is modeled by a sequence of delayed quantized measurements. The main purpose of this article is to design an extended Kalman filter for the stochastic nonlinear systems, in the simultaneous presence of quantization errors, stochastic nonlinearities, and stochastic noises, such that an optimized upper bound for the filtering error covariance is derived. By solving two coupled Riccati‐like difference equations, the filter gain matrix is explicitly formulated. An RR protocol‐based recursive filtering algorithm is developed for the online implementation. Furthermore, a sufficient condition is established to ensure the uniform boundedness of the filtering error in the mean‐square sense. Finally, a simulation example is given to demonstrate the practical validity of the designed filter algorithm.