Zoran D. Jelicic

Generalized particle swarm optimization algorithm - Theoretical and empirical analysis with application in fault detection

A generalization of the particle swarm optimization (PSO) algorithm is presented in this paper. The novel optimizer, the Generalized PSO (GPSO), is inspired by linear control theory. It enables direct control over the key aspects of particle dynamics during the optimization process. A detailed theoretical and empirical analysis is presented, and parameter-tuning schemes are proposed. GPSO is compared to the classical PSO and genetic algorithm (GA) on a set of benchmark problems. The results clearly demonstrate the effectiveness of the proposed algorithm. Finally, an application of the GPSO algorithm to the fine-tuning of the support vector machines classifier for electrical machines fault detection is presented.

Second-order sliding mode approaches to disturbance estimation and fault detection in fractional-order systems

Abstract This paper outlines some results concerning the application of second-order sliding-mode techniques to address estimation and fault detection problems involving fractional order (FO) dynamics. Perturbed and switched FO systems are dealt with throughout the paper. Simple tuning formulas for the suggested schemes are constructively developed along the paper by means of appropriate Lyapunov analysis. Simulation and experimental results confirm the expected performance.

On-line adaptive clustering for process monitoring and fault detection

An adaptive clustering procedure specifically designed for process monitoring, fault detection and isolation is presented in this paper. The key feature of the proposed procedure can be identified as its underlying capability to detect novelties in the systems mode of operation and, thus, to identify previously unseen functioning modes of the process. Once a novelty is detected, relevant informations are used to enrich the knowledge-base of the algorithm and as a result the proposed clustering procedure evolves and learns the new features of the monitored process in accordance with the available process data. The suggested clustering procedure is theoretically illustrated and its effectiveness has been investigated experimentally. Particularly, the on-line implementation of the algorithm and its integration with a fault detection expert system have been considered by making reference to a pneumatic process.

Nonlinear fractional PI control of a class of fractional-order systems

Abstract This paper deals with the design of nonlinear PI control techniques for regulating a class of fractional-order dynamics governed by a commensurate-order model, possibly nonlinear, perturbed by an external disturbance. The suggested control algorithm is the combination between a fractional-order PI controller and a nonlinear robust version of it, namely a second-order sliding mode control algorithm called “super-twisting” controller in the literature. A key feature of the approach is the use of ad-hoc sliding manifolds whose construction involves fractional order derivatives. A constructive Lyapunov based synthesis is illustrated, which leads to simple tuning rules for the controller parameters guaranteeing the asymptotic rejection of the external disturbance under appropriate smoothness restrictions. Computer simulations illustrate the effectiveness of the proposed technique.

Trapezoidal rule for numerical evaluation of fractional order integrals with applications to simulation and identification of fractional order systems

This paper presents an extension of the well-known trapezoidal (bilinear) integration rule, that in the present work is applied to the numerical evaluation of fractional-order integrals. Particularly, this approximation is exploited to derive viable numerical algorithms addressing two distinct problems: i) simulation of Linear Time-Invariant (LTI) Commensurate Fractional Order Systems (CFOS); ii) non-recursive parameter estimation in LTI-CFOS. More precisely, the problem of non-recursive parameter estimation is addressed in two different scenarios. The first one is when the commensurate order of the CFOS is known in advance, while the second, more general, one is that in which the commensurate order is unknown and is to be estimated. The effectiveness of the proposed methods is illustrated by numerical examples.

On second-order sliding-mode control of fractional-order dynamics

A second-order sliding mode control scheme is developed to stabilize a class of linear uncertain fractional-order dynamics. After making a suitable transformation that simplifies the sliding manifold design, a chattering-free second order sliding mode approach that accomplishes the control task by means of a continuous control action is developed. Simple controller tuning formulas are constructively developed along the paper by Lyapunov analysis. The simulation results confirm the expected performance.

Adaptive identification of the commensurate order in fractional processes by means of variable-order operators

A gradient-based algorithm for the on-line continuous estimation of the commensurate order in linear fractional order processes is presented. A key aspect of the proposed methodology is the use of appropriate variable-order fractional filters, and linear Laplace operators of logarithmic type, within the estimation mechanism. A Lyapunov based analysis will be provided for deriving appropriate sufficient conditions guaranteeing the parameter convergence property. Realization issues associated to the involved variable order operators are discussed, and a fully developed analysis and design example, accompanied by relevant simulation results, is provided to support the presented theory.

Optimization of distributed order fractional PID controller under constraints on robustness and sensitivity to measurement noise

This paper describes a novel approach towards optimal tuning of distributed order fractional PID controller parameters. A distributed order fractional PID controller (DPID) is approximated by a compound fractional controller with multiple fractional differintegrators connected in parallel. Orders of these differintegrators have been equally spaced, with the first one being the classical integrator of order 1, and the last one being the classical differentiator of order 1. A classical noise cancellation filter is considered as a part of controllers structure. The controller parameters, being the gains of all differintegrators, have been tuned. Proposed tuning procedure maximizes gain of classical integrator term under constraints on Maximum sensitivity, Ms, Maximum sensitivity to measurement noise, Mn and Maximum complementary sensitivity, Mp. Results are presented via a number of numerical simulations.

Induction motor broken rotor bar detection using vibration analysis — A case study

Early fault diagnosis and condition monitoring can reduce the consequential damage and breakdown maintenance, prolong the machine life and increase the performance of industrial systems. This paper describes a real fault detection problem of a high-power (3.2 MW) induction motor driving pumps in a heating plant. Steady-state vibration signals were collected and some characteristic low- and high- frequency features were observed, resulting in broken rotor bar diagnosis. The obtained results were verified by disassembling the motor, which proved that this particular technique can be successfully applied in induction motor fault detection.

Generalized PSO algorithm — an application to Lorenz system identification by means of neural-networks

In this paper a new, generalized PSO (GPSO) algorithm is presented and analyzed, both theoretically and empirically. The new optimizer enables direct control over the properties of the search process. In addition, PSO is addressed in conceptually different manner, revealing further aspects of the algorithm behavior. GPSO is applied for training radial basis function neural network (RBF-NN) to identify dynamics of a nonlinear system. The target system is chosen to be of Lorenz type, known for its complex, chaotic behavior. Results presented in this paper clearly demonstrate effectiveness of the proposed algorithm.