Huajing Fang

Takagi-sugeno fuzzy-model-based fault detection for networked control systems with Markov delays

A Takagi-Sugeno (T-S) model is employed to represent a networked control system (NCS) with different network-induced delays. Comparing with existing NCS modeling methods, this approach does not require the knowledge of exact values of network-induced delays. Instead, it addresses situations involving all possible network-induced delays. Moreover, this approach also handles data-packet loss. As an application of the T-S-based modeling method, a parity-equation approach and a fuzzy-observer-based approach for fault detection of an NCS were developed. An example of a two-link inverted pendulum is used to illustrate the utility and viability of the proposed approaches

Weighted Average Prediction for Improving Consensus Performance of Second-Order Delayed Multi-Agent Systems

In this paper, the weighted average prediction (WAP) is introduced into the existing consensus protocol for simultaneously improving the robustness to communication delay and the convergence speed of achieving the consensus. The frequency-domain analysis and algebra graph theory are employed to derive the necessary and sufficient condition guaranteeing the second-order delayed multi-agent systems applying the WAP-based consensus protocol to achieve the stationary consensus. It is proved that introducing the WAP with the proper length into the existing consensus protocol can improve the robustness against communication delay. Also, we prove that for two kinds of second-order delayed multi-agent systems: 1) the IR-ones with communication delay approaching zero and 2) the ones with communication delay approaching the maximum delay, introducing the WAP with the proper length into the existing consensus protocol can accelerate the convergence speed of achieving the stationary consensus.

Quasi T-S Fuzzy Models and Stable Controllers for Networked Control Systems

Networked control systems (NCS) are systems with at least one loop closed through data networks. NCS have many peculiarities, such as network-induced time delay, data packet dropouts, asynchronous clock among network nodes etc. It is necessary to develop new theory and technology for control system over networks. Many researchers are interested in NCS recently. In this paper, we emphasized on the NCS with random time delays. We presented a new modeling method for linear and nonlinear NCS with random time delays, and named these models as quasi T-S fuzzy models. Based on these new descriptions for NCS, two theorems for the stability of closed-loop NCS and stable controller design were presented. All the discussions and results in this paper are not only applicable to NCS but also applicable to normal nonlinear system with multiple time delays

Networked control system: state of the art

Networked control systems (NCS) are one type of distributed control systems where serial communication networks is used to exchange system information and control signals between various physical components of the systems that may be physically distributed. The existence of real time network in the feedback control loop makes analysis and design of an NCS complex, and conventional control theories such as synchronized control and non-delayed sensing and actuation must be reevaluated prior to application to networked control systems. Numerous research papers have investigated analysis and design of NCS from different views. This paper aims to introduce state of art NCS and to point out the field of further research.

Kalman filter based FDI of networked control system

Networked control system (NCS) deals with how to realize the close-loop control through network. Fault diagnosis for NCS is a new research region. In this paper, network channel connects the sensor and the controller. So NCS was regarded as a system with output time-variant delay and white noise. The iterative method was used to compensate the time-delay. Then different Kalman filters were constructed for NCS according to different system faults. So the respective residues were generated and a fault detection and isolation (FDI) approach was introduced. At last, an example was given to show the feasibility of the approach.

Takagi–Sugeno Model Based Analysis of EWMA RtR Control of Batch Processes With Stochastic Metrology Delay and Mixed Products

In many batch-based industrial manufacturing processes, feedback run-to-run control is used to improve production quality. However, measurements may be expensive and cannot always be performed online. Thus, the measurement delay always exists. The metrology delay will affect the stability and performance of the process. Moreover, since quality measurements are performed offline, delay is not fixed but is stochastic in nature. In this paper, a modeling approach Takagi-Sugeno (T-S) model is presented to handle stochastic metrology delay in both single-product and mixed-product processes. Based on the Markov characteristics of the delay, the membership of the T-S model is derived. Performance indices such as the mean and the variance of the closed-loop output of the exponentially weighted moving average (EWMA) control algorithm can be derived. A steady-state error of the process output always exists, which leads the output deviating from the target. To remove the steady-state error, an algorithm called compensatory EWMA run-to-run (COM-EWMA-RtR) algorithm is proposed. The validity of the T-S model analysis and the efficiency of the proposed COM-EWMA-RtR algorithm are confirmed by simulation.

ARMAX model based run-to-run fault diagnosis approach for batch manufacturing process with metrology delay

In batch manufacturing process, the product quality data are often gained after several runs owing to the existence of the metrology delay. The metrology delay will affect the stability, performance and reliability of the process. In this paper, the batch processes with exponentially weighted moving average/double EWMA (EWMA/DEWMA) controller and metrology delay are represented by data-based autoregressive moving average with exogenous inputs (ARMAX) models instead of the mechanical models. A parameter-resetting recursive extended least square (PRELS) algorithm is proposed to identify the coefficients of this model. Comparing with traditional RELS algorithm, PRELS have faster convergence speed and higher accuracy when the coefficients greatly vary. The relationship of the process parameters and ARMAX model coefficients is derived. On basis of this, a statistical online fault diagnosis scheme is presented to detect and quickly identify a fault. Dynamic principal component analysis is used on the coefficients of ARMAX model which is stationary instead of the non-stationary process data to detect the process fault. Furthermore, the deviation of the faulty parameter is derived using a least-squares estimation, and the influence matrix algorithm is applied to achieve the online fault isolation. The validity and effectiveness of the proposed approach are illustrated through some simulation results in general semiconductor manufacturing processes.

Modeling and Control of Scalable Engineering Swarm

In this paper we propose a two-layer emergent model for flocking behavior of intelligent engineering swarm system. The first emergent layer describe the flocking behavior of swarm resulting from the individual motion to its local goal position, the center of minimal circumcircle which is decided by the neighbors in the positive visual set of individual; the second one describe the individual flocking behavior to the local goal position resulting from the individual motion to one or two farthest neighbors in its positive visual set. The scale of the swarm will not be limited because only local individual information is used for modeling in the two-layer topology. We study the implementation method for engineering swarms with given vehicle dynamics based on sliding mode control. The simulations show that the swarm of intelligent system can converge to goal regions while maintaining cohesiveness

Stability analysis of local swarms in an environment with a quadratic profile

This paper first constructs an model of local swarms with a quadratic profile. Then study the stability properties of the collective behavior of the swarm and provide conditions for collective convergence to more favorable regions of the profile. It is shown that the individuals of the swarms will aggregate and eventually enter into a bounded hyperball around the swarm center or objective position. Aimed at removing the local trap, this paper gives the bound of parameter Aδ.

Constrained Kalman Filtering with observation losses

In this paper, we consider networked constrained Kalman filtering with observation losses. The observation losses of communication network is modeled as an i.i.d. Bernoulli process. Based on physical consideration, at each time step through projecting the unconstrained Kalman filter solution onto the state constraint surface, the constrained estimation can be derived, which significantly improves the prediction accuracy of the filter. We study the statistical convergence properties of the error covariance matrix, showing the existence of a critical value for the arrival rate of the observation, beyond which a transition to an unbounded state error covariance occurs. Simulations are provided to demonstrate the effectiveness of the theoretical results.