Power Quality Monitoring And Communication Scheduling In Advanced Metering Infrastructure

With the wide spread of DER (Distributed Energy Resource) and thecontinuous development of power management techniques at the demand side, AMI(Advanced Metering Infrastructure), as an important part of smart grid, is becomingone of the most popular research hotpots in power fields due to its significantadvantages in system operation, assets management and load response. The coreobjectives of AMI are to realize accurate online real-time monitoring and to ensurethe quality of two-way communication. However, the substantive research of AMIis still in its infancy. The power quality monitoring in AMI has higher requirementsthan traditional power quality monitoring, which mainly reflects in terms ofaccuracy, real-time performance and realizability. Thus, the power quality detectionalgorithm is suppposed to possess higher accuracy, faster speed and easierimplementation. Reasonable communication scheduling strategy is an effectivemeans to ensure timely, effective, and reliable information transmission as well asto improve the performance of network communication in AMI system. Therefore,the research in this dissertation can be divided into two major aspects: the“monitoring” and “communication” in the AMI system. The research of“monitoring” focuses on the detection methods of the core power quality ind icators,such as voltage sag, flicker, harmonic and inter-harmonic. The research of“communication” focuses on the optimal scheduling strategy and algorithms.The three characteristics of voltage sag are duration time, sag depth and phasejump. In order to solve the problem that the existing voltage sag detection methodis poor in real-time detection due to overly reliance on the sag depth detection whenjudging sag start-stop time, and to improve the real-time and accuracy of the phaseand amplitude detection of voltage sag, a voltage sag detection method based ongray image edge detection and improved single-phase dq transformation isproposed. First, the image edge detection operator is constructed on the basis ofmathematical morphological operations and used to judge the start and stop time ofvoltage sag. This method has a high detection speed. Meanwhile, it is independenton the sag depth because the detection is based on the geometric characteristics ofsag waveform. Second, the single-phase dq transformation is improved. The virtualthree-phase system was constructed using the delay angle of the single-phasevoltage, and then the three-phase voltage is transformed to dq rotatating coordinatesystem to compute the corresponding sag depth and phase jump. The detectionprocess and results are faster and more accurate than the traditional methods. In order to extract the fluctuant component from voltage flicker signa l quicklyand accurately, an improved Teager energy operator (ITEO) demodulation methodis proposed on the basis of the in-depth study of energy operator detection princ iple.The proposed method is simple, fast and can track the change of flicker waveform.Simulation analysis of both single frequency fluctuation source and multi-frequency fluctuation source flicker signal is carried out and compared with Hilbertdemodulation method and energy operator demodulation method. In order to obtainthe voltage flicker indicators accurately, a flicker value measurement method basedon ITEO and FFT is proposed based on the proposed demodulation method, whichoptimizes the measurement process. The flicker value is calculated using therelative voltage fluctuation value of the modulation frequency, which effectivelyreduces the measurement error of instantaneous flicker sensitivity in the low-frequency band.To resolve the frequency aliasing problem of conventional wavelet packettransform in harmonic and inter-harmonic detection, a novel detection methodbased on wavelet packet-traversal filtering is proposed. First, the origina l signal isdecomposed and reconstructed through the wavelet packet transform to get theharmonic and inter-harmonic components of each frequency band. Then, thealiasing bands are traversal filtered through a recursive filter to separate theharmonics and inter-harmonic components with adjacent ones. This approach notonly possesses high detection precision and good anti-noise performance, but alsogreatly reduces the computation amount compared with the full-band filter.In order to improve the communication efficiency of AMI system, theinformation transmission scheduling strategy is studied from two aspects: systemand specific network. At the system level, aim at the defects of conventionalinformation transmission scheduling strategy, an optimized strategy based on userimportance and device and information importance is proposed and validatedthrough network simulation software OPNET. In the wide area, the proposedscheduling strategy ensures timely processing of importance power users’information. In the local area, it ensures timely processing of information fromimportant device and important information from common device. At the specificnetwork level, the normal scheduling strategy of AMI network is analyzed, and thenan dynamic scheduling algorithm based on IPR (Initiative Priority Regulable) isproposed. This algorithm can solve the problems of inflexibility in static schedulingand large resource consumption in dynamic scheduling. It can also improve theutilization of network resources.Aim at the real-time, accuracy and network requirements for power quality monitoring in AMI, a device based on dual-processor (DSP and ARM) architectureis developed. The main unit circuit structure and the implementation of proposedpower quality detection algorithm on the DSP platform are illustrated. A dataconcentrator with various network interfaces and the measurement datamanagement software are developed, which togerther constitute an AMIcommunication system with the power quality monitoring device and otherintelligent devices. Experimental validation of the proposed power quality detectionalgorithms and communication scheduling strategies are carried out in this system.The results verify the feasibility and effectiveness of the proposed algorithms andstrategies.