晚期瞬变电磁信号是大参数、 多频率、 且各频率分量未知的电磁信号， 特别在晚期还是强噪声背景下的微弱信号． 随机共振系统可以在极限信噪比的情况下提取微弱信号， 但仅适用于小参数、 单一频率或频率已知的高频信号的情况． 为此， 本文提出一种基于尺度变换的随机共振算法用于检测晚期瞬变电磁信号， 该方法充分利用了随机共振检测弱信号的优势， 通过引入尺度变换， 消除了随机共振系统对待测信号频率的限制， 在绝热近似理论下，实现了从强噪声中提取出微弱晚期瞬变电磁目标信号．-Transient Electromagnetic(TEM) signal is a large parameters, multi-frequency and unknown frequency with heavy noisy. Stochastic resonance (SR) can only be applied to small parameters, low-frequency or the known high-frequency signal detection. To solve this problem, a scale transformation for detecting TEM weak periodic signal of SR method is presented in this paper. The method is benefited from SR for detecting weak signals.