Jiao-Min Liu

Research for speed controller of switched reluctance motor using F-PID control

In this paper, the system structure of switched reluctance motor drive is studied. On the basis of knowledge about mid-speed motor and its converter, some profound research for designing and implementing has been made. Lastly, the speed controller with TMS320C240 as core controller is designed. The results provided in this paper show that with the designed controller SRM can operate in mid-speed with good performance.

Selection of Parameters Based on Fuzzy Extension Matrix

Fuzzy extension matrix (FEM) inductive learning is an important method that generates knowledge from cases. Compared with conventional extension matrix techniques, it is more powerful and practical to handle with ambiguities in classification problems. Rule extraction from fuzzy extension matrix involves three parameters alpha, beta and gamma. These parameters play an importation role in the entire process of rule extraction based on FEM. They greatly affect the computation of fuzzy entropy and extract rules, however those important parameter value are usually estimated based on users by domain knowledge, personal experience and requirements. This paper introduces an approach to optimization of the three parameters based GA, and provides some theoretical support of directly selection of the parameter values through experiment. The main contributions of this paper are as follows: by combining GA and local search methods, we can get the reasonable parameters. Five data sets from the UCI machine learning database are employed in the study. Experimental results and discussions are given

Improvement of rotor position angle estimation in SRD using fuzzy logic based motor model

In this paper, a novel method of sensorless rotor position angle estimation based on the knowledge of motor model and the relation of /spl Psi/-I-/spl theta/ is developed on using fuzzy logic based motor model. Due to electromagnetic interference and measurement error, there are more errors in result. In order to improve the precision of the method, a predictive module is added in system. Through comparing predicted value with estimated value, more accurate value is gotten in test. The result shows that, with improved estimation system, rotor position angle can be accurately estimated in different regions, including startup, low-speed and mid-speed region.

A wavelet approach for rms and harmonics measurements

This paper discusses the subject of root mean square (rms) of voltage or current and harmonic ratio value measurement problem in low voltage electrical apparatus testing system. A wavelet packet transform (WPT) method is introduced as a useful tool for detecting, classifying and quantifying the rms of testing waveform and harmonic ratio value. The proposed approach can measure the distribution of the rms of testing waveform with respect to individual frequency bands directly from the wavelet transform coefficients. The harmonic ratio value relative to basic frequency band can be calculated as well. The method is evaluated by its application to both analytical waveform and actual testing waveform data. The experimental results are offered to illustrate the performance of the proposed method.

A New Naïve Style Possibilistic Network Classifier

This paper proposes a new approach of classification under the possibilistic network (PN) framework with Tree Augmented Naive Bayes Network classifier (TAN), which combines the advantages of both PN and TAN. The classifier is built from a training set where instances can be expressed by imperfect attributes and classes. A new operator, the possibilistic mean is designed to estimate the conditional possibility distributions of each attribute with imperfection, and the weight between two attributes given the class is determined by the conditioning specificity gain. Experiment has shown the efficiency of the new classifier in imperfect cases.

Mechanical fault diagnosis for L-V circuit breakers based on energy spectrum entropy of wavelet packet and Naive Bayesian classifier

The paper has extracted the energy spectrum entropy of wavelet packet as the eigen vector of fault patterns, through analyzing the vibration signal in the decomposition of wavelet packet when Low-Voltage (LV) Circuit Breaker broke down. Based on the concept of Clustering Center, a Naïve Bayesian classifier has been constructed. By using the weight of probability measure, the correlations between the eigen vector has been described. Thus the simulated fault diagnosis of the LV circuit breaker has been achieved. Through simulating, the efficiency of the method has been verified, which could fasten the computing speed, optimize the real-time performance and classification precision comparing with the neural network which uses black-box modeling.

A New Method of Eliminating Noise Based on Clustering

Clustering is often constructed on noise-free datasets. In real-world applications, it is inevitable that the datasets contain noises, which may result in unsatisfactory results of the clustering algorithms. In this paper, several methods of reducing noises are systemic introduced, and at the first time we propose a heuristic algorithm of reducing noises in clustering theory (GK-means). The empirical results show that GK-means is simpler and more precise, and can handle noises in the real-world database effectively. Some samples are used to prove the validity of this algorithm.

Edge detection on arc image of low voltage apparatus

The switching arc image of low voltage apparatus has been captured with a high frame rate CCD camera system and composed filter technology. In this paper, the discussion is focused on the theory of edge detection. An efficient image edge detection operator, Canny operator, is introduced to the processing of arc image. The performance of Canny operator and the traditional operators in the practical edge detection of arc image is analyzed. Experimental results prove that Canny operator is better than Roberts, Sobel, Prewitt, Log operators by analyzing weak edge region in the switching arc image. And Canny operator is suitable for the edge detection of arc image of low voltage apparatus.

Study on integrated monitor and control of the apparatus quality based on the multi-parameter analysis

Electric arc is a especial switching apparatuss phenomenon in switching loop current, reliability and using life of apparatus can be directly influenced by time of arcing, which is considered as the important index of switching apparatus. According to the time of apparatus arcing, we can divide it into different grades. Factors that influence the time of apparatus arcing include material, design, techniques, operation, power, equipment and circumstance. Selecting all these factors as sample, using the time of apparatus arcing to determine the sample type, we can extract the key factor using FICSEM in the same type samples, and can obtain the factor combination that achieve the same apparatus arcing time. We can improve the product quality by improving the key factor, and can reduce the product cost by using the optimal factor combination. So we can evade producing risks and can improve economy benefits.

Rotor position angle estimation and speed control in SRD using fuzzy logic based motor model

A method of sensorless rotor position angle estimation for switched reluctance motor drives (SRDs) is developed based on using a fuzzy logic based motor model. The real-time experimental results given show that this method can overcome the drawbacks of previous sensorless techniques. In addition, a digital signal processor is used as the control unit in this scheme to lessen the effects due to error.