Collision Avoidance Warning System for Deep-Sea Underwater Operation Platforms

As the utilization of deep-sea underwater operation platforms continues to expand in deep-sea monitoring and resource exploration, the complex and unpredictable underwater environment presents considerable safety challenges to these long-term operating underwater experimental platforms, particularly in the context of collision threats posed by non-cooperative targets, such as autonomous underwater vehicles (AUVs). Conventional underwater collision avoidance systems rely on optical cameras and multi-beam imaging sonars. However, these systems are prone to limitations in deep-sea environments due to light absorption and high energy consumption. In response, this study proposes a collision avoidance warning system for deep-sea underwater operation platforms based on mechanical scanning sonar The system incorporates both mechanical scanning sonar and a remotely wake-up recording system, which not only fulfils the requisite target detection criteria but also exhibits evidence retention capabilities. The fundamental component of this system is the ISS360-HD sonar, which employs Chirp signals in conjunction with a motor to facilitate comprehensive 360-degree acoustic detection signal coverage. Given that the deep-sea underwater operation platform is often stationary during extended underwater operations, this study employs a background subtraction method as the foundation for the target detection algorithm of the collision avoidance warning system. This decision is driven by the necessity to minimize power consumption while achieving the perception of underwater targets. The background subtraction method has been enhanced in the actual processing to mitigate the impact of "impure" backgrounds, a prevalent challenge with traditional methods, thereby enhancing the visibility of moving targets. The results of the pool test demonstrate that the system is capable of accurately identifying targets despite the presence of complex background interference, and that it exhibits power consumption characteristics that are well-suited to long-term monitoring applications.