Adaptive cascade detection of weak magnetic anomalies based on marine predators algorithm-stochastic resonance

Abstract

Magnetic anomaly detection is a new technology for underground or underwater ferromagnetic targets using the physical principle that ferromagnetic targets will be long-term magnetized by the geomagnetic field to generate abnormal magnetic fields. This technology is often used in underwater security, mineral exploration, and other areas. But it is quite challenging to effectively detect weak magnetic anomaly signals in a complex underwater environment. Therefore, we propose an adaptive cascade weak magnetic anomaly detection method based on Marine Predators Algorithm-Stochastic Resonance (MPA-SR). According to the characteristics of magnetic anomaly signals, the cascade detection method of low-pass filtering, stochastic resonance, and threshold detection is designed to improve the detection probability of magnetic anomaly signals. In addition, the Marine predator algorithm with optimized initialization strategy and step size control parameters is used to improve the stochastic resonance system to adaptively detect the magnetic anomaly signal in more applications. The simulation results show that the signal-to-noise ratio (SNR) of the output signal of the MPA-SR method is 2.41 dB higher than the input signal, and the detection probability of the method is 57% higher than that of the minimum entropy (ME) method under the same low SNR environment. The method can provide the theoretical basis and empirical reference for further application of magnetic anomaly data like identifying, locating, and tracking underwater magnetic targets.