A method for diagnosing faults of UAV gyroscopes

Abstract

The invention discloses a fault diagnosis method for a UAV gyroscope. Based on the fuzzy technology, the degree of proximity between the sample and the expected value is calculated, and the probability of the degree of proximity is used to design the sample membership factor, and the membership factor is configured to indicate support In the mathematical model of the vector machine, the influence of sample distribution imbalance and noise characteristics is eliminated. Support vector machine training is carried out offline, and the decision function parameters of the optimal interval classification surface for distinguishing positive and negative samples are obtained. When the drone is flying, the attitude signal output by the airborne sensor gyroscope is collected in real time, and the db4 wavelet basis decomposition method is used to extract it. The eigenvector of the signal is normalized, and then input to the decision function of the optimal interval classification surface to calculate the category label of the signal to be divided in real time and identify the fault condition of the gyroscope. The diagnosis method of the invention significantly reduces the bias degree of the fault classification hyperplane, improves the fault diagnosis accuracy, has a small calculation amount, and realizes the online real-time performance of the fault diagnosis algorithm.

Description

technical field [0001] The invention relates to a fault diagnosis of a sensor, in particular to a fault diagnosis method for a gyroscope of an unmanned aerial vehicle. Background technique [0002] With the advantages of UAV 3D (Dull: boring, Dirty: harsh environment, Dangerous: dangerous) mission execution advantages and low price advantages, UAVs are widely used in military reconnaissance, target positioning, electronic warfare, missile interception, etc. and civilian fields such as emergency rescue and disaster relief, inspection of power lines and oil and gas pipelines, anti-terrorism and stability maintenance, and aerial photography. [0003] UAV platforms in the field of military applications are often equipped with multi-type jamming equipment such as TACAN and radar; The navigation signal, the radiation signal generated by the engine operation, and the complex airborne electromagnetic environment produce uncertain interference on the output signal of the gyroscope, ...

AI Technical Summary

Problems solved by technology

However, this method does not discuss the influence of sample data set imbalance and noise singularity on the accuracy of fault diagnosis.

During the training process of support vector machine, the imbalance of positive and negative sample data sets and noise interference will lead to the hyperplane bias of the optimal classification decision function and reduce the accuracy of fault diagnosis

Images