Donghua Zhou

Variable structure control for a class of nonlinear systems with mismatched uncertainties

A scheme to stabilize nonlinear time-varying systems with both matched and mismatched uncertainties is proposed in this paper by switching between two control laws: a first-order sliding-mode control and a second-order sliding-mode control. Based on this idea, a variable structure control algorithm is designed for a class of second-order systems. The closed-loop system is globally or locally asymptotically stable. It has been proven that the stability region has relation with the order of the boundary function and the region can be obtained by solving an inequality. The uncertainty considered in this work is also more general than those in the existing works.

Adaptive Filtering for Stochastic Systems With Generalized Disturbance Inputs

This letter presents a new class of discrete-time linear stochastic systems with the statistically-constrained disturbance input, which can represent an arbitrary linear combination of dynamic, random, and deterministic disturbance inputs to generalize the complicated modeling error encountered in actual applications. An adaptive filtering scheme is proposed for such systems by recursively constructing and adaptively minimizing the upper-bounds of covariance matrices of the state predictions, innovations, and estimates. The minimum-upper-bound filter is then obtained via online scalar convex optimization. The experiment on maneuvering target tracking shows that the proposed filter can significantly reduce the peak estimation errors due to maneuvers, compared with the well-known IMM method.

Sensor fault reconstruction for a class of 2-D nonlinear systems with application to fault compensation

This paper considers the problem of sensor fault reconstruction and compensation for a class of two dimensional (2-D) nonlinear systems. The 2-D nonlinear system is described by the Fornasini–Marchesini local state-space second model with Lipschitz nonlinearity. The sensor fault considered in this study could be of arbitrary form and its size can be even unbounded. An integrated fault/state observer is proposed to obtain the asymptotic estimation of sensor faults and system states at the same time. A sufficient condition for the existence of the integrated observer is given in terms of linear matrix inequalities.

Estimation of systems with statistically-constrained inputs

H_\infty

Iterative learning model predictive control for multi-phase batch processes

H∞ sensor fault estimation/reconstruction is also considered for the 2-D nonlinear system when there are both sensor faults and input disturbances. Based on the estimation of sensor faults, a sensor compensation scheme can be performed by subtracting the fault term from the measurement output, and the existing output feedback controller can run normally without the switchover of sensors or reconfiguration when sensor faults occur. An example is provided to illustrate the effectiveness of the proposed method for both sensor fault reconstruction and compensation.

Iterative Learning Fault-Tolerant Control for Batch Processes

Abstract This paper deals with control of multi-phase batch processes. The process in each cycle is formulated as a switched system with internally forced switching instants. In this formulation, there are two kinds of switching sequence: dynamics-switching-sequence and control-switching-sequence. The control problem is transformed into finding of the control-switching-sequence and the control signal between any two consequential points in the sequence. This problem is proposed to be decomposed into two subtasks: determining the control-switching-sequence by detecting the dynamics-switching-sequence and designing the control law by using iterative learning scheme. Challenges within this framework are discussed, and some possible solutions to these challenges are suggested.

Robust fault-tolerant control of a class of non-minimum phase nonlinear processes

Abstract This paper discusses the estimation of a class of discrete-time linear stochastic systems with statistically-constrained unknown inputs (UI), which can represent an arbitrary combination of a class of un-modeled dynamics, random UI with unknown covariance matrix and deterministic UI. In filter design, an upper bound filter is explored to compute, recursively and adaptively, the upper bounds of covariance matrices of the state prediction error, innovation and state estimate error. Furthermore, the minimum upper bound filter (MUBF) is obtained via online scalar parameter convex optimization in pursuit of the minimum upper bounds. Two examples, a system with multiple piecewise UIs and a continuous stirred tank reactor (CSTR), are used to illustrate the proposed MUBF scheme and verify its performance.

Iterative learning reliable control of batch processes with sensor faults

Most processes can be divided into two classes: batch processes and continuous processes. In general, batch processes, which run intermittently, are best suited to low-volume and high-value products. Due to the high value of products, much more requirements are proposed for the control performance of batch processes. Hence, the plants become more and more complicated to achieve the high requirements. Consequently, the process complication exposes the possibility of system faults. Therefore, there is a trade-off between high performance and reliability. Fault-tolerant control (FTC) should be a good choice to handle this trade-off. Unfortunately, the reported work on FTC for batch processes is scarce. Trying our best, only four papers in the subject area were found by the authors. Therefore, both great challenges and opportunities exist in this field. Compared to continuous processes, batch processes have mainly three features: repetitive nature, finite duration, and nonlinear property. Generally, faults can be divided into three classes based on their location; FTC methods include two classes according to whether fault detection and diagnosis (FDD) is used; in addition, the proposed scheme could be linear or nonlinear. By means of these criteria, the existing papers are categorized. Thanks to these categorizations, some promising directions will be presented in this paper.