A coevolutionary algorithm for rules discovery in data mining

One of the major challenges in data mining is the extraction of comprehensible knowledge from recorded data. In this paper, a coevolutionary-based classification technique, namely COevolutionary Rule Extractor (CORE), is proposed to discover classification rules in data mining. Unlike existing approaches where candidate rules and rule sets are evolved at different stages in the classification process, the proposed CORE coevolves rules and rule sets concurrently in two cooperative populations to confine the search space and to produce good rule sets that are comprehensive. The proposed coevolutionary classification technique is extensively validated upon seven datasets obtained from the University of California, Irvine (UCI) machine learning repository, which are representative artificial and real-world data from various domains. Comparison results show that the proposed CORE produces comprehensive and good classification rules for most datasets, which are competitive as compared with existing classifiers in literature. Simulation results obtained from box plots also unveil that CORE is relatively robust and invariant to random partition of datasets.     

Observer-based control of multi-agent systems under decentralized information structure

This paper addresses the problem of decentralized implementation of a global state feedback controller for multi-agent systems. The system is assumed to be under the constraint of a complete decentralized information structure. The decentralization of the control task is achieved through the construction of low-order decentralized functional observers with the purpose of generating the required corresponding control signal for each local control station. A design procedure is developed for obtaining an approximate solution to the design of the observers. Stability analysis is provided for the global system using the proposed observer-based approach. A numerical example is given to illustrate the design procedure and cases when the observers’ order increases from the lowest value.     

Design of linear functional observers for linear systems with unknown inputs

This paper presents a reduced-order linear functional state observer for linear systems with unknown inputs. A simple observer construction procedure is provided. A numerical example is given to illustrate the properties of the observer. The example deals with a linear system comprising of 20 states, 2 inputs, 10 outputs and 5 unknown inputs for which a fourth-order observer is designed to estimate two linear functions of the states.     

Asymptotic normality analysis of the estimation error of steady-state models for industrial processes: SISO case

The asymptotic normality of the estimation error of steady-state models for industrial processes is investigated under quite mild conditions. The estimate is formed from the estimated parameters of an approximate linear model which is strongly consistent with the steady-state gain of slow time-varying linear SISO systems. In the parameter estimation, the weighted least-squares method is employed. The input signal (the system set point) is the usual step change din the optimization procedure. The rate of convergence is given. The stationarity and the distribution of the stochastic process are not demanded. It is also worth mentioning that, under some acceptable conditions, robustness to the structure of the approximate linear model is achieved. A simulation study shows that, for limited length of the sampled data, the best choice for the structure of approximate models as regards estimation precision is dependent upon the realization of the stochastic noise.     

Optimal control for large-scale systems: a recursive approach

A recursive fixed-point-type method is presented for the study of the optimal control problem of large-scale systems. The control is obtained by decomposition of the system to ‘ε coupled’ subsystems so that only low-order systems are involved in algebraic computations. It is shown that the developed reduced-order parallel algorithms converge to the desired solution with the rate O(ε). Owing to its recursive nature, the presented method produces a considerable saving of computation. An illustrative numerical example is given to verify the proposed approach.     

Comments on ‘Robust stabilization of uncertain linear dynamical systems’

Based on the stabilizability of a nominal system, by making use of the Lyapunov stability criterion and combining with the algebraic Riccati equation, Wu and Mizukami (1993) present a new approach to designing a robust linear state feedback controller for uncertain linear dynamical systems. The robust stabilization conditions given by Wu and Mizukami (1993), however, seem to be conservative and can be easily improved upon. In this note we design another robust linear feedback controller to render the uncertain system asymptotically stable. An example is presented to demonstrate our new control law and to compare it with the existing one.     

On the existence of solutions of the general algebraic Riccati equation

This paper addresses the problem of real symmetric solutions of the general algebraic Riccati equation (GARE) with an indefinite quadratic term. The GARE arises in linear quadratic differential games, in the stabilization of uncertain systems, robust optimal control and disturbance attenuation problems. Using the properties of the solutions of the differential equation corresponding to the GARE and the related conclusions of differential games, we have established the main results of this paper. Theorems 1 and 2     

Fault detection of NCS based on eigendecomposition,adaptive evaluation and adaptive threshold

A method for fault detection (FD) of networked control systems with unknown network-induced delay and unknown input based on eigendecomposition, adaptive evaluation and adaptive threshold is proposed in this paper. In the method, a known structure matrix of network-induced delay which is supposed to be smaller than one sampling period is extracted by eigendecomposition and an observer based residual generator is constructed. Then an adaptive evaluation function is proposed to ensure the optimal robustness to network-induced delay and unknown input at the evaluation stage based on the Parity Space approach. Finally, based on the estimation of the maximal influence of unknown random delay and unknown input on the evaluation function, adaptive threshold is designed to enhance robustness of the FD system and reduce the miss detection rate as much as possible.     

Non-linear output frequency response functions for multi-input non-linear Volterra systems

The concept of non-linear output frequency response functions (NOFRFs) is extended to the non-linear systems that can be described by a multi-input Volterra series model. A new algorithm is also developed to determine the output frequency range of non-linear systems from the frequency range of the inputs. These results allow the concept of NOFRFs to be applied to a wide range of engineering systems. The phenomenon of the energy transfer in a two degree of freedom non-linear system is studied using the new concepts to demonstrate the significance of the new results.