基于人工免疫算法的输电线路故障类型识别新方法

摘要

According to the negative selection mechanism of the human immune system, an improved negative selection algorithm is presented. The improved algorithm can not only detect the “No-self" successfully, but also be used as classifier to identify different fault types. And the flaws of the traditional artificial negative selection algorithm, such as poor knowledge expression and noise tolerance ability, unable to set the matching threshold and slow search efficiency, are overcome. The improved algorithm is proposed to apply to the fault type recognition of transmission lines based on analyzing transmission line fault characteristic, and a new recognition scheme using faulty transient current energy is put forward. Simulations in PSCAD/EMTDC show that this method can identify the fault type rapidly and accurately and it won' t be affected by faulty time, transition resistance, faulty location, and system capacity, etc. So it will have good adaptability.%根据人体免疫系统阴性选择机理,提出一种改进的阴性选择算法,能准确地识别"自我"与"非我",还能识别出不同的故障类型,并克服了传统阴性选择算法知识表达能力差、无法设定匹配阈值、抗噪能力差和搜索效率低等缺点.在分析输电线路故障特征的基础上,首次将阴性选择算法应用于输电线路故障类型识别中,并提出了一种基于故障暂态电流能量的故障类型识别新方案.基于PSCAD/EMTDC的仿真试验结果表明:该故障类型识别方案能快速准确地识别各类故障,并且不易受到故障时刻、过渡电阻、故障位置、系统容量等因素的影响,具有较好的适应性.