基于差分进化的时频原子分解方法在电气化铁路谐波检测中的应用

Application of Differential Evolution-Based Time-Frequency Atom Decomposition in Harmonic Detection of Electrified Railway

将时频原子分解(time-frequency atom decomposition,TFAD)方法应用于电气化铁路谐波检测,然而计算量过大,制约了TFAD方法在信号处理中的应用。为降低其计算复杂度,利用差分进化(differential evolution,DE)算法优化原子参数,实现原子的最优匹配。通过对电气化铁路谐波电流实测数据的仿真分析,结果表明:基于DE的TFAD方法可以准确检测出相应的基波及各次谐波分量,验证了该算法的有效性。

When time-frequency atom decomposition （TFAD） method is applied to harmonic detection of electrified railway, too heavy calculation burden of TFAD restricts its application in signal processing. To reduce the computational complexity of TFAD, atom parameters are optimized by differential evolution （DE） and optimal matching among atoms is implemented. Simulation analysis on measured harmonic current data of electrified railway shows that using the DE-based TFAD method corresponding fundamental component and various harmonic components can be accurately detected, thus the effectiveness of the proposed algorithm is verified.