论文部分内容阅读
针对未来低压电网剩余电流保护与动作技术中,如何检测触电时刻并识别总泄漏电流中人体触电支路电流信号的难题,利用数字信号的智能处理技术和具有自适应性与最佳逼近特性的组合神经网络有机结合,提出了一种触电电流信号的自动检测方法。在对低压电网中原总泄漏电流信号进行小波消噪基础上,实现了触电时刻的自动检测,触电故障模式分类归属的决策;同时从总泄漏电流中提取触电电流幅值波形。仿真实验表明:该方法速度快且稳定,模式分类正确率达100%,提取幅值与实际值的平均相对误差为3.65%,计算时间为0.064 68 s,具有良好的适应性和实用性,对于开发新一代剩余电流保护装置具有重要的参考价值。
Aiming at the problem of how to detect the moment of electric shock and identify the current signal of the human body’s electric shock branch current in the residual current protection and operation technology of low voltage network in the future, the intelligent signal processing technology and the combination of adaptive and best approximation Neural network organic combination, put forward an electric shock current signal automatic detection method. Based on the wavelet denoising of the total primary leakage current signal in low voltage network, the automatic detection of electric shock time and the classification of electric shock failure mode are realized. At the same time, the amplitude of the electric shock current amplitude is extracted from the total leakage current. The simulation results show that the proposed method is fast and stable, the correct rate of pattern classification is 100%, the average relative error between extracted amplitude and actual value is 3.65%, and the calculation time is 0.064 68 s, which has good adaptability and practicability. Development of a new generation of residual current protection device has an important reference value.