Positioning Algorithm of Cold Source Underwater Pollution Inspection Robot Based on Multi-Sensor Fusion

With the rapid development of robotic systems, their application areas have expanded from traditional land to underwater, and have achieved better application range. In order to expand the application scenarios, this study proposes a cold source underwater pollution detection robot positioning algorithm based on multi-sensor fusion. The algorithm aims to solve the shortcomings of existing robots in sensor fusion and positioning accuracy, and provide a reasonable path. First, this study designed a mathematical model of the six-degree-of-freedom motion of the underwater robot based on mathematical modeling. In the optimization part, the fourth-order Runge-Kutta method was used to achieve a stable solution, and it was extended to robot control, and the PID algorithm was used to optimize the motion control framework. Subsequently, the designed model improves the accuracy of robot path planning and ground point cloud removal by introducing a 3D point cloud to 2D grid conversion algorithm, thereby achieving stable positioning. Then, the model proposes a multi-sensor fusion algorithm for underwater acoustic channels, combining multiple sensor data to effectively address the challenges of signal multi-path effects and interference. The experiment is carried out in a simulated underwater environment. The test results showed that the proposed algorithm performed well in terms of positioning accuracy and operation stability. At the same time, a visualization simulation of underwater robot path was carried out to prove the universality of the proposed framework.