论文部分内容阅读
能效测评系统间谐波的实时检测的重要性日益突出。在进行间谐波分析时,常用的方法是FFT和DFT计算,但都要满足采样点数为2的N次方,而采样频率经常不是基波频率(50 Hz)的倍数。针对这一采样处理难点,在相应检测的硬件电路结构框图基础上,提出了变采样速率和重构的方法,详细阐述了变速率采样算法和用神经网络进行最大熵信号重构的理论依据,通过Matlab仿真和样机实验,验证了该方法的可行性和正确性,而且该采样方法具有计算量不大、易于由DSP实现的优点,可以应用在间谐波检测领域,具有一定的现实意义。
The importance of real-time detection of inter-harmonics in energy efficiency evaluation systems has become increasingly prominent. Commonly used methods for performing inter-harmonic analysis are FFT and DFT calculations, but all satisfy the Nth power of 2 sampling points, and the sampling frequency is often not a multiple of the fundamental frequency (50 Hz). Aiming at the difficulty of this sample processing, a method of variable sampling rate and reconstruction is proposed based on the block diagram of the hardware circuit. The theoretical basis of variable rate sampling algorithm and maximum entropy signal reconstruction by neural network are elaborated. The feasibility and correctness of this method are verified by Matlab simulation and prototype experiments. The sampling method has the advantages of low computational complexity and easy implementation by DSP, and can be used in the field of interharmonic detection. It has some practical significance.