An improved SVM classifier based on double chains quantum genetic algorithm and its application in analogue circuit diagnosis

Abstract

Support Vector Machine (SVM) classifier is widely used in analogue circuit diagnosis. However, the penalty parameter C and the kernel parameter γ of SVM classifier with the radial basis function (RBF) affect the classification performance seriously. A double-chains-quantum-genetic-algorithm (DCQGA) based method is proposed to optimize C and γ. In DCQGA, each chromosome carries two gene chains, and each of gene chains represents an optimization solution, which can accelerate the search process and help to find the global solution. Thereafter, the optimal parameters C and γ are obtained by optimizing the parameter searching process with DCQGA. Two common datasets named Iris and Wine from UCI Machine Learning Repository are used to test the performance of the presented SVM classifier. The simulation results illustrate that the population's best fitness and the classifying accuracy of the proposed DCQGA-SVM are higher than that of the Particle-Swarm-Optimization based SVM (PSO-SVM), the Quantum Genetic Algorithm based SVM (QGA-SVM) and the classifier based on grid search method (GS-SVM). Finally, the proposed DCQGA-SVM is applied to analogue circuit diagnosis, a Sallen–Key bandpass filter circuit and a four-opamp biquad high-pass filter are chosen as circuits under test (CUT). Wavelet packet analysis is performed to extract the fault features before classifying. The experimental results show that the SVM parameters selected by DCQGA-SVM contribute to higher diagnosis accuracy than other methods referred in this paper.

Introduction

Analogue circuits play a vital role in ensuring the availability of industrial systems. Unexpected circuit failure in such systems during field operation can have severe implications [1]. In the past several decades, analogue circuit diagnosis has become a very difficult task, though there are many efficiency methods have been presented at the system, board, chip level, such as the fuzzy analysis method [2], neural network [3], [4], [5], support vector machine (SVM) [6] and so on. The fuzzy analysis method needs long execution time or requiring too many test nodes. In view of the nonlinear case, document [7] presents a new method for solving fuzzy differential equations based on the reproducing kernel theory which can reduce the computing time effectively. Neural network and SVM are all based on intelligent fault dictionary. Compared with neural networks, SVM has simpler structure and needs smaller training group [8].

SVM is a pattern classification technique proposed by Vapnik and his co-workers. SVM has shown a good performance on high-dimensional data classification with small training group [9], [10], [11]. For its outstanding performance, SVM has been applied to many areas and got admirable results [12], [13], [14], [15], [16]. Also, SVM classifiers are very popular in analogue circuit diagnosis [17], [18], [19], [20]. However, SVM parameters have an important influence on the classification accuracy, such as the penalty parameter C and the kernel parameter γ for SVM with RBF kernel function, and it is difficult to find out the optimal SVM parameters. In recent researches, many kinds of optimization algorithms are used to choose the SVM parameters [21], [22], [23]. Document [21] introduces genetic algorithm (GA) to search the optimal parameters of SVM. Document [23] and [24] present two SVM parameters optimization methods based on Particle Swarm Optimization algorithm (PSO) and quantum genetic algorithm (QGA), respectively. PSO-SVM has a faster convergence speed and better diagnosis accuracy compared with GA-SVM for discrete optimization problems [25], but it is easy to fall into local optimum. The local optimum will result in a poor parameter combination and a bad classification accuracy.

In this paper, double chains quantum genetic algorithm (DCQGA) is introduced to optimize the penalty parameter C and the kernel parameter γ of SVM with RBF kernel function. GA is an optimization technique based on the principles of genetics and natural selection. The stochastic behaviour of GA cannot be ignored as it affects the search efficiency greatly [26]. DCQGA can be viewed as an improved GA algorithm [27]. Document [27] states that the chromosomes of DCQGA are encoded by quantum non-gates, updated by quantum rotation gates, and mutated by quantum non-gates. Since both probability amplitudes of qubits are regarded as genes, each chromosome carries two gene chains, and each of gene chains represents an optimization solution, which can accelerate search process. These improvements evidently enhance optimization efficiency [27].

The remainder of this paper is organized as follows. In Section 2, we review the SVM principle. In Section 3, we discuss the DCQGA theory. In Section 4, the proposed SVM classifier based on DCQGA is put forward. Experiment research is in Section 5. In Section 6, the conclusions are given.