Study On Voltage Sags And Swells Detection Method And Its Application Based On Simplified S-transform Algorithm

With the development of the smart grids, various nonlinear loads are increasing, which causes a series of power quality problems(PQ). With the high frequency and the bigger affected area of voltage sudden-change(sags and swells), the safety of electricity equipment operation is endangered, and the voltage sudden-change has been attracting more and more attention. It is a key criterion of PQ, and its real-time accurate measurement is quite important. The key to research the roots, improvement and management of the PQ is using time-frequency analysis of complex power grid signal, featuring real time and accurate. Therefore, study on real-time measurement of voltage sudden-change and explore simplified mechanisms for the time-frequency analysis method, which applies into an embedded device to form a fast and accurate detecting system, has an important theoretical significance and application value.Firstly, this thesis introduces the basic concepts, hazards and causes of the voltage sudden-change, and describes the present situation and development trend of power quality monitoring devices. It also analyses the characteristics and shortcomings of classical methods and time-frequency analysis method for voltage sudden-change interiorly and abroad, which leads to the S-transform(ST) advantages and improvement in time-frequency analysis of characteristic parameters.Secondly, this thesis researches the simplified S-transform algorithm, and a simplified S-transform algorithm beased on characteristic frequency information is proposed, which is simplified via a mechanism using secondary extreme of the signal's FFT spectrum to recognize the feature frequency points. At same time, after intensively studying the measurement of voltage sudden-change based on the simplified algorithm, an extremism width determination method using difference curve of the fundamental frequency amplitude is proposed in this thesis in order to calculate its beginning and ending time more accurate.Then the ST low time-frequency resolution of measurement, which results in large detection errors, have also been analyzed in this thesis. Thus, this thesis proposes a coefficients adaptive optimization fast method for detecting voltage sudden-change based on the simplified S-transform algorithm. The method creates adaptive modified S-transform(MST) by introducing new window width adjustment coefficients and using piecewise time interpolation algorithm to optimize the coefficients. This thesis shows the reprocess of the proposed detection method. The results of the computational amount comparing with ST are also presented, which indicates that operation is greatly reduced, and the method is proved to implement an embedded suitably. Moreover,the simulation under the circumstance of a variety of interference and white noise shows that the method is anti-noise, high reliability and provides results with less detection errors.Thirdly, according to actual project and algorithm implementation, a voltage sudden-change detection system based on ADC & DSP hardware architecture is designed. This thesis details the hardware modules design including signal conditioning and data acquisition, ADC & DSP system, power module and communication interfaces. At the same time, the method proposed is accomplished by using the TI-RTOS, which is a real-time embedded operating system. Certainly the specific program design and bootloading system design are given.Finally, the reliability and error sources of the system is analyzed and tested. The test results show that the system featuring stable and reliable can satisfy the accurate and rapid detection of voltage sudden-change in power system. Above all, it is a good platform and reference to resolve other PQ.