X.Z. Gao

A hybrid optimization method of harmony search and opposition-based learning

The harmony search (HS) method is an emerging meta-heuristic optimization algorithm. However, like most of the evolutionary computation techniques, it sometimes suffers from a rather slow search speed, and fails to find the global optimum in an efficient way. In this article, a hybrid optimization approach is proposed and studied, in which the HS is merged together with the opposition-based learning (OBL). The modified HS, namely HS-OBL, has an improved convergence property. Optimization of 24 typical benchmark functions and an optimal wind generator design case study demonstrate that the HS-OBL can indeed yield a superior optimization performance over the regular HS method.

An Introduction to Harmony Search Optimization Method

This brief provides a detailed introduction, discussion and bibliographic review of the nature1-inspired optimization algorithm called Harmony Search. It uses a large number of simulation results to demonstrate the advantages of Harmony Search and its variants and also their drawbacks. The authors show how weaknesses can be amended by hybridization with other optimization methods. The Harmony Search Method with Applications will be of value to researchers in computational intelligence in demonstrating the state of the art of research on an algorithm of current interest. It also helps researchers and practitioners of electrical and computer engineering more generally in acquainting themselves with this method of vector-based optimization.

Power control for mobile DS/CDMA systems using a modified Elman neural network controller

It is well recognized that power control is an important technique to combat the near-far effect and increase the maximum capacity in direct sequence code division multiple access (DS/CDMA) cellular systems. We propose a modified Elman (1990) neural network (MENN) based multiple-model power control scheme to keep the received power almost constant at the base station. Unlike the conventional bang-bang control and fuzzy control, the MENN can identify the inverse dynamical characteristics of the channel by adaptive on-line learning. The inverse channel model is then used for reverse link power control. Hence, large overshoot and long rise time can be effectively avoided. Simulations show that the fluctuation of controlled received power is satisfactory. The overshoot and channel tracking error are small.

The Overview of Harmony Search

When musicians compose the harmony, they usually try various possible combinations of the music pitches stored in their memory, which can be considered as a optimization process of adjusting the input (pitches) to obtain the optimal output (perfect harmony). Harmony search draws the inspiration from harmony improvisation, and has gained considerable results in the field of optimization, although it is a relatively NIC algorithm. With mimicking the rules of various combining pitches, harmony search has two distinguishing operators different from other NIC algorithms: harmony memory considering rate (HMCR) and pitch adjusting rate (PAR) that are used to generate and further mutate a solution, respectively. This candidate generation mechanism and single search memory involved decide its excellence in structure simplicity and small initial population. This chapter presents the discussions of the inspiration of harmony search, the basic harmony search optimization algorithm, and an overview of different application areas of the harmony search.

Hybrid Swarm Algorithms for Parameter Identification of an Actuator Model in an Electrical Machine

Efficient identification and control algorithms are needed, when active vibration suppression techniques are developed for industrial machines. In the paper a new actuator for reducing rotor vibrations in electrical machines is investigated. Model-based control is needed in designing the algorithm for voltage input, and therefore proper models for the actuator must be available. In addition to the traditional prediction error method a new knowledge-based Artificial Fish-Swarm optimization algorithm (AFA) with crossover, CAFAC, is proposed to identify the parameters in the new model. Then, in order to obtain a fast convergence of the algorithm in the case of a 30u2009kW two-pole squirrel cage induction motor, we combine the CAFAC and Particle Swarm Optimization (PSO) to identify parameters of the machine to construct a linear time-invariant(LTI) state-space model. Besides that, the prediction error method (PEM) is also employed to identify the induction motor to produce a black box model with correspondence to input-output measurements.

Delay-Dependent Stability Analysis of Uncertain Fuzzy Systems with State and Input Delays under Imperfect Premise Matching

This paper discusses the stability and stabilization problem for uncertain T-S fuzzy systems with time-varying state and input delays. A new augmented Lyapunov function with an additional triple-integral term and different membership functions of the fuzzy models and fuzzy controllers are introduced to derive the stability criterion, which is less conservative than the existing results. Moreover, a new flexibility design method is also provided. Some numerical examples are given to demonstrate the effectiveness and less conservativeness of the proposed method.

Knowledge-based artificial fish-swarm algorithm

Abstract In this paper, a knowledge-based Artificial Fish-Swarm (AFA) optimization algorithm with crossover, CAFAC, is proposed to enhance the optimization efficiency and combat the blindness of the search of the AFA. In our CAFAC, the crossover operator is first explored. The knowledge in the Culture Algorithm (CA) is next utilized to guide the evolution of the AFA. Both the normative knowledge and situational knowledge is used to direct the step size as well as direction of the evolution in the AFA. Ten high-dimensional and multi-peak functions are employed to investigate this new algorithm. Numerical simulation results demonstrate that it can indeed outperform the original AFA.

The Harmony Search in Context with Other Nature Inspired Computational Algorithms

Inspiration drawn from nature and modeling of natural processes are the two common characteristics existing in most NIC algorithms. These methodologies, therefore, share many similarities, e.g., adaptation, learning, and evolution, and have a general flowchart including candidate initialization, operation, and renewal. On the other hand, mimicking various natural phenomena leads to their different generation, evaluation, selection, and update mechanisms, which may result in individual inherent distinctive properties, advantages, as well as drawbacks in the performances of dealing with different optimization problems. For example, the CSA on the basis of modeling the clonal selection principle of the artificial immune system performs well in the local search but suffers from a long convergence time. This chapter compares three typical evolutionary optimization methods, GA, CSA, and HS, with regard to their structures and performances using illustrative examples.

Negative Selection Algorithm-Based Motor Fault Diagnosis

In this paper, a Negative Selection Algorithm (NSA)-based motor fault diagnosis scheme is proposed. The hierarchical fault diagnosis scheme takes advantage of the feature signals of the healthy motors so as to generate the NSA detectors, and uses the analysis of the activated detectors for fault diagnosis. It can not only efficiently detect incipient motor faults but also correctly identify the corresponding fault types. Our method has been investigated using one practical motor fault diagnosis example.

Fault detection in ink jet printers using neural networks

We explore the feasibility of using both feedforward and Elman neural networks to detect assembly faults in ink jet printers. The method is an extension of the motor fault detection scheme proposed by Gao and Ovaska (2002). Two types of cartridge faults are studied here: encoder belt misalignment and encoder strip error. These two faults are detected from the characteristics variants in the neural networks-based prediction of cartridge velocity signals. Results of experiments with real-world data demonstrate that neural networks can be trained to effectively detect the inherent encoder faults. Some discussions on the selection of appropriate fault detection criteria are also given.