【摘要】 噪声对变压器局部放电信号干扰强烈,严重影响最终监测效果。为进一步提高稀疏分解方法在变压器局部放电去噪过程中的收敛速度和去噪效果,文中提出了一种基于改进双链量子遗传算法与正交匹配追踪方法相结合的局部放电稀疏分解去噪新方法。利用该方法对仿真局部放电信号进行去噪,并与传统的小波阈值去噪法和EEMD去噪法的去噪效果进行对比,结果表明,新方法的去噪效果在各项评价指标上都明显优于后两者,同时新方法的收敛速度明显快于传统的匹配追踪方法。通过对实测变压器内部典型局部放电信号的去噪,进一步验证了该方法的有效性和实用性。更多还原

【Abstract】 The interference of noise to partial discharge signal of transformer is very strong and affects seriously the final monitoring result. For improving the convergence rate and denoising effect of the sparse decomposition method in the denoising process of the partial discharge of the transformer,a kind of partial discharge sparse decomposition denoising method based on the combination of improved double chains quantum genetic algorithm and orthogonal matching tracking method is proposed. The method is used to de-noise the simulation partial discharge signal and compare with the denoising result which is obtained by the traditional wavelet threshold denoising method and EEMD de-noising method. The result shows that the de-noising result of the new method is superior to the latter two methods in terms of each evaluation index and,at the same time,the convergence speed of the new method is obviously faster than the traditional matching tracking method. Both effectiveness and applicability of the method is verified further by way of denoising of typical actually measured discharge signal inside the transformer.更多还原