Binary classification fault diagnosis for octocopter using deep neural network

초록

A fault diagnosis using deep neural network for a single rotor of octocopter is proposed. The latest attitude and its command are fed to the proposed algorithm to determine the failure of the rotor in real time. Both normal cases and fault cases are considered with appropriate output pairs to form a data set. Two hidden layers are constructed with sigmoid activation functions. A generalized delta rule is adopted, and a stochastic gradient descent scheme is used to calculate the weight update of the neural network. Numerical simulations are performed to validate the performance of the proposed algorithm. The performance can be maximized by controlling the design parameters.