Detection And Classification Of Power Quality Based On Intelligent Algorithm

The power quality is denoted by the deviation of voltage, current or frequency, which makes electric equipments not work proper. Due to the non-liner, fluctuating and unbalanced loads, many problems such as voltage sags, voltage swells,voltage interrupts, oscillation transients and harmonics,are produced.Those problems result in inestimable economic loss to consumers and providers.At the some time, some intelligent electrical devices stipulate more rigorous requirement for the power quality. The problem of how to improve the power quality becomes hotspot in the field of power quality. It is important to detect and analyze the power quality, which is the precondition and foundation for improving and governing the power quality.The concept, classification, standards, causes and hazards of power quality are introduced in this paper, and they are followed by the traditional ways of the power quality. Aim at the problem of power quality disturbances, an approach based on S-transform and intelligent algorithm is presented in this paper, which identifies the classification of power quality disturbances.For the most serious problem in the disturbances, a method based on S-transform and GA-SVM is combined to classify the source of voltage sags.The classification system of power quality disturbances is achieved on the DSP and Matlab, the details are as follows:(1)After analyzing the characteristics of every power quality disturbance, an approach based on S-transform and intelligent algorithm is presented in this paper. Power quality disturbances mainly include harmonics, voltage sags, voltage swells, voltage interrupts, flickers, oscillation transients, voltage sags with harmonic and voltage swells with harmonic.The signal of power quality disturbances is produced by simulation models, features of every power quality disturbance extracted by S-transform are applied to relevance vector machine, support vector machine and neural network for automatic classification of the power quality disturbances.The results of experiment show that the approach based on S-transform and relevance vector machine can effectively detect and classify the power quality disturbances. This approach has the advantages of the short test time and higher correctness, and can be applied to real-time power quality detection system.(2) Voltage sags are the most serious problems in power quality disturbances. The causes for voltage sags are the short circuit, the transformer energizing and the motor starting. It is significant to identify the source of voltage sags.A new approach based on S-transform, genetic algorithm and support vector machine is presented in this paper. Some characteristics related to voltage sags are extracted by S-transform. Genetic algorithm is used for optimizing the extracted feature, which can be utilized to effectively classify the voltage sags source with support vector machine.The experiment results confirm that this approach is efficient to classify the voltage sags source.(3)A system used for classifying the power quality disturbances based on DSP and Matlab is proposed in this paper, after analyzing the technology of power quality detection system both at domestic and foreign.The system is composed of DSP and Matlab.DSP is mainly included sensor, signal processing, ad conversion, serial communication and DSP system, which is used for collecting the power data and calculating the essential parameter of power. Matlab can communication with DSP by the serial port, after receiving the power data, the algorithms such as S-transform and relevance machine are running on Matlab, which are applied to analyze and classify the power quality disturbances.