基于人工免疫算法的小电流接地故障选线

Artificial immune algorithm based faulty line detection method for indirectly grounding power system

小电流接地系统的单相接地故障选线问题一直没有得到很好的解决,通常采用的单一故障选线方法如小波变换法、五次谐波电流法和零序电流有功分量法等只能对部分故障信息进行处理,选线可靠性不高,因此提出了基于人工免疫算法的小电流接地故障选线方法。利用小波包变换和FFT算法提取零序电流的暂态、五次谐波和有功分量故障特征作为免疫系统的抗原,利用训练样本对该免疫系统训练后得到记忆集,利用抗原和记忆集的亲合度实现故障选线。ATP仿真结果表明该方法得出的选线结果具有较高的精度和较强的鲁棒性。

The faulty line detection of single-phase-to-earth fault in indirectly grounding power system has not been well solved. The commonly used single faulty line detection methods, such as wavelet transform method, the fifth harmonic current method and zero sequence current active components method, etc, can only process partial fault information, their reliability of faulty line detection is not satisfied, so artificial immune algorithm based faulty line detection method for indirectly grounding power system is proposed. The extracted transient characteristic, fifth harmonic characteristic and active component characteristic of zero sequence current using wavelet packet transform and fast Fourier transform （FFT） method, are looked as antigens. After training the immune system using training samples, the memory set is obtained. Based on affinities between antigens and the memory set, the results of faulty line detection are achieved. Simulation results by ATP show that the faulty line detection result by the proposed method is more precise and possesses stronger robustness.