Distributed Kalman filtering for cascaded systems

Abstract

The Kalman filter provides an efficient means to estimate the state of a linear process, so that it minimizes the mean of the squared estimation error. However, for naturally distributed applications, the construction and tuning of a centralized observer may present difficulties. Therefore, we propose the decomposition of a linear process model into a cascade of simpler subsystems and the use of a Kalman filter to individually estimate the states of these subsystems. Both a theoretical comparison and simulation examples are presented. The theoretical results show that the distributed observers, except for special cases, do not minimize the overall error covariance, and the distributed observer system is therefore suboptimal. However, in practice, the performance achieved by the cascaded observers is comparable and in certain cases even better than the performance of the centralized observer. A distributed observer system also leads to increased modularity, reduced complexity, and lower computational costs.

A cascaded Kalman filter-based GPS/MEMS-IMU integration for sports applications

“…The cascaded structure simplifies the mathematical model by removing the calculations related to orientation states in the position/velocity Kalman filter. Thus, the cascade formulation has much less computational overhead than a global Kalman filter [26]. Even though the cascaded filter is suboptimal, the accuracy is comparable to the global Kalman filter [26].…”

Data-based fault detection in chemical processes: Managing records with operator intervention and uncertain labels

“…The current value of the state variables can be recursively estimated based on the attributes up to time t. In a DLM, the Kalman filter allows updating the current inference on the state variables as new data become available (Lendek, Babuška, & De Schutter, 2008). Passing from the filtering density π(θt|Y1:t) to π(θt+1|Y1:t+1) is fully detailed described in appendix A.2.…”

Adaptive neuro-fuzzy fusion of sensor data

“…However, for naturally distributed applications, the construction and tuning of a centralized observer may present difficulties. Therefore, the decomposition of a linear process model into a cascade of simpler subsystems and the use of a Kalman filter to individually estimate the states of these subsystems was proposed in [6]. A new algorithm that combines a fuzzy adaptive fusion and wavelet analysis to form an efficient data fusion technique for the target tracking system was presented in [7].…”

“…Multisensory Kalman filter (KF) is suitable to integrate a high number of sensors, without rebuilding the whole structure of the filter. The Kalman filter provides an efficient means to estimate the state of a linear process, so that it minimizes the mean of the squared estimation error [6]. However, for naturally distributed applications, the construction and tuning of a centralized observer may present difficulties.…”