An Innovative Magnetic Anomaly Detection Method Based on Adaptive Triple-Stable Stochastic Resonance

Magnetic anomaly detection (MAD) is a passive method for detecting magnetic targets. Stochastic resonance (SR)-based MAD is a widely utilized technique because it detects magnetic signals at low signal-to-noise ratios (SNRs). In practice, the performance of the SR system is influenced by the waveform of the magnetic signal, which complicates achieving consistent detection performance across different systems. In addition, the conventional SR system is constrained by a narrow range of adjustable parameters, limiting its detection capabilities. Addressing these limitations, we propose a novel adaptive parallel triple-stable SR (APTSR) system that integrates the conventional bistable potential function with the Woods-Saxon (WS) function. This system enhances the performance by tuning parameters through the gray wolf optimizer (GWO) and identifies magnetic anomalies via threshold detection. Finally, through simulation and experimental analysis, this system demonstrates a higher probability of successful detection at lower SNR compared to the other two parallel stochastic systems. In a colored noise environment, the detection probability is 91% at an input SNR of -6 dB and approximately 80% even at an input SNR of -8 dB.