Fault diagnosis method without mathematical modeling of underwater robot

Abstract

The invention relates to an underwater robot technology, and aims to provide a fault diagnosis method without mathematical modeling of an underwater robot. The method includes the steps that an underwater robot executes a specified task program, and corresponding behavior responses are generated according to different preset fault types; state data in the test process is stored as a data sample; multiple tests are repeatedly carried out to collect enough data samples, and a two-dimensional matrix is obtained after preprocessing and is used for generating an original training set; the obtained training set is used to train a diagnosis model in an off-line mode; the trained diagnosis model is loaded to the underwater robot, the real-time state data of the underwater robot in the task program execution period is taken as the input of the diagnosis model, and forward calculation is performed to achieve fault prediction and diagnosis. According to the method, a complex mathematical modeling process is avoided, so that the method has higher universality. A traditional double-stage diagnosis mode can be converted into a single-stage diagnosis mode, and high diagnosis accuracy is achieved.

Description

technical field [0001] The invention relates to underwater robot technology, in particular to a fault diagnosis method (of an underwater robot) that does not require mathematical modeling of the underwater robot itself. Background technique [0002] With the exploration and development of the ocean, underwater robots are playing an increasingly important role in marine applications. At the same time, the complex and unpredictable marine environment will increase the probability of failure of underwater robots, so underwater robots must have the ability to diagnose faults. When the underwater robot fails to meet the expected state, the fault diagnosis technology can help the underwater robot to obtain accurate fault information (including whether a fault occurs and the location of the fault source) in time, so as to take appropriate measures to reduce the fault. the impact. However, underwater robot systems are often complex in structure, with strong nonlinearity and uncert...

AI Technical Summary

Problems solved by technology

However, underwater robots are complex nonlinear dynamic systems, and it is difficult to accurately describe some complex dynamic behaviors, noise, interference and other factors in the system, and it is almost difficult or even impossible to model them accurately, which means Difficult to make an accurate diagnosis

Images