A multi-sensor fusion data processing method for underwater robots

Abstract

The invention discloses a multi-sensor fusion data processing method of an underwater robot, which estimates attitude, position and speed information of the underwater robot based on multiple sensors; by fusing all the sensor measurement data, the sensor measurement with obvious errors can be better eliminated Data, so that the underwater robot is not easily affected by a single sensor failure, and is more suitable for nonlinear systems such as underwater robots; when constructing the state vector, the angle deviation and speed deviation are added to the state vector, and the angle deviation and speed are considered The impact of bias on state updates is used to accurately estimate the attitude, position and velocity information of the underwater robot. By assuming that the process noise is obtained from the filtering result and the observation result, the differential evolution algorithm is used to optimize the variance of the obtained process noise to improve the filtering precision. The invention is suitable for nonlinear systems of underwater robots to accurately estimate the attitude, position and speed information of the underwater robots.

Description

technical field [0001] The invention belongs to the technical field of underwater robot data processing, and in particular relates to an underwater robot multi-sensor fusion data processing method based on an extended Kalman filter algorithm of a differential evolution algorithm. Background technique [0002] There is huge economic potential in the ocean, which has attracted extensive attention from all over the world. In order to better detect and develop deep-sea resources, scholars from various countries have stepped up research and development of marine engineering detection devices and mining equipment. Among them, underwater robots have been widely used in military and civilian applications such as underwater target search, marine resource detection, and marine military missions. In practical applications, how to improve the maneuverability and maneuverability of underwater robots in complex underwater environments is an urgent problem to be solved. Research on multi-...

AI Technical Summary

Problems solved by technology

[0003] At present, the extended Kalman filter has been widely used in practical fields such as unmanned aerial vehicles, but due to the complexity and diversity of the underwater environment, the application of multi-sensor fusion data processing methods in underwater robots is still in a blank stage.

The extended Kalman filter assumes that the process noise and the observation noise are a constant matrix. However, in actual engineering, due to external force or magnetic interference, the working environment of the filter is not constant, and different environments correspond to the best noise. The matrix is ​​not the same, and the filter cannot correct the noise matrix in real time for these transformed environments, which brings errors to the estimation of attitude, position and speed, and the accuracy is not high.

Images