| The involved project is funded by National Science-Tech Tackle Key Project, ' Research on Active Power Filter for Power Harmonic Eliminating' (NO.2002BA218C) and Hunan Power Corp. Science-Tech Tackle Key Project, 'Research on Engineering Application of Power Harmonic Eliminating ( No. [2001]772) . Taking the harmonic attenuation of the 10KV power distribution network at the 110KV substation in a smeltery as object of study, this dissertation aims to do research on the control strategies and its industrial application technologies for eliminating harmonic in medium and high voltage power network, prepare theory and teclinological knowledge for its universal application in our country, and accumulate experiences for its high-voltage and high-power application. The production of the project is listed as one of the new important products in China[2003ED770013].Firstly, harmonic detection method using ip,iq algorithm, which is based on the instantaneous power theory proposed by H.Akagi, is introduced briefly and is extended to detect positive or negative sequence of arbitrary order harmonic in single-phase system, three-phase three-wire system or three-phase four-wire system.Secondly, time delay and inertia embedded in every course of the digital APF system is studied. Due to high clock rate and short sampling time, the delay is customarily assumed to have very little influence on the effectiveness of the harmonic compensation process. However, the considerations and calculations presented in this dissertation proved totally otherwise. In view of that, predictive method to calculate APF harmonic reference is proposed based on adaptive LMS algorithm. Optimum predictive strategy is also presented. Furthermore, three modified measures were introduced to improve the performance of the traditional LMS adaptive predictive algorithm in light of the characters of APF control. Simulation and experimental results confirm its effectivity. In the practice of the APF harmonic reference predictive method, some skillful measures were utilized to alleviate the detrimental influence ofthe great variation the line frequency and the indeterminate and stochastic factor in the environment on the adaptive algorithm. Consequently, the adaptivity and correctness of the predictive algorithm is enhanced, and its applied range is extended. The thought of the methods can be used for reference in other applications of adaptive predictive algorithm.In succession, deep research on the topology of APF in series with shunt passive filter through coupling transformer (NPHAPF for short) was made. Its operation principle was detailedly analyzed. Four kinds of APF control strategies were proposed and compared. And the optimum control strategy was chosen in terms of the involved project. Using the selected control strategy, the APF in NPHAPF can not only enhance the harmonic attenuating effect of the whole hybrid APF system, damp the series and parallel resonances between the passive filter capacitor and the line inductance, it can also effectively mitigate the large harmonic current caused by the harmonic voltage on the resonance circuit of the passive filter. A thorough investigation into the effect of delay time and proportion coefficient K of APF control on the stability of NPHAPF is done. The validate range of K, not causing system instability, is evaluated under different delay time. Good harmonic attenuating effect might be obtained with constant K in certain stable situation. But in case of dynamic situation, good performance cannot be hold with constant K any more. To alleviate such problems, automatic K adjusting technology was proposed. That is to regulate K adaptively on-line according to the real-time line harmonic distortion rate. It makes APF work well in terms of both harmonic attenuation and power consumption regardless of the severity of line harmonic distortion.In the end, industrial application technology was researched. Parameter designing rales of the main components were discussed in detail. Based on instantaneous variabl...
|
PDF Full Text Request