Underwater robot multi-sonar data fusion method

Abstract

The invention provides an underwater robot multi-sonar data fusion method. The invention aims to improve the measurement accuracy based on the fusion of sonar data and an attitude sensor, and performsoptimization based on an improved SH adaptive filtering algorithm. The invention performs time alignment on different sensors based on a ranging error model of the ranging sonar, fuses the measurement data of sensor signals such as an AUV ranging distance, a heading angle and a roll angle, corrects the data by using the SH self-adaptive Kalman filter, and adaptively adjusts the system noise and appropriately reduces the unbiasedness of the measurement noise estimation for the case of poor algorithm robustness, so as to realize accurate estimation of the sonar ranging data, improve a forgetting factor algorithm, and enhance the system robustness. The underwater robot multi-sonar data fusion method has great advantages compared with traditional methods, can be used for various underwater robots, and has far-reaching application prospects.

Description

technical field [0001] The invention relates to a multi-sensor data processing method of an underwater robot, which belongs to the technical field of underwater robots. [0002] technical background [0003] Underwater robots are more and more widely used in underwater detection and underwater operation tasks. The task of the robot to detect the environmental state of the open water channel has high requirements for the measurement accuracy of the robot's ranging sonar. [0004] When the robot is working, it sails parallel to the wall at a specified distance from the wall under the action of the propeller. This has high requirements on the accuracy of sonar measurement data. However, due to the influence of water flow and propeller action, the robot always has certain yaw and roll during navigation, which increases the data error measured by sonar and increases At the same time, due to the different operating frequencies of the sonar and the attitude sensor, the generated d...

AI Technical Summary

Problems solved by technology

[0005] The traditional adaptive Kalman filter uses measurement data for real-time estimation and adjustment of noise statistical characteristic parameters, which solves the problem that the system process noise and measurement noise are difficult to determine, but its algorithm itself cannot guarantee the positive definiteness of the measurement noise covariance matrix. It is easy to cause a large amount of calculation and filter divergence

Images