As an essential link in fields like communication, sonar and radar, weak signal detection has been widely applied to different areas and has been becoming the focus of studies. The significance of weak signal detection technology drives people to explore and study new theories and methods that could be used to detect weak signals buried in strong background noise in a more accurate and faster way. With the development of nonlinear science technology, a new method has been put forward to solve problems in weak signal detection against the background of limitation of traditional detection methods. While achievements have been scored in bistable stochastic resonance system technology, the application of noise array bistable stochastic resonance system could further improve weak signal detection capability. Combined with theory and numerical simulation, bistable stochastic resonance method was introduced to analyze output signal-to-noise gain. Three elements-array noise, the number of array units and external noise were selected to further explore their relationship with detection performance. The simulated result showed that, compared with bistable stochastic resonance system, array bistable stochastic resonance system enjoys a better weak signal detection capability.