Research On Stability Control Technology Of UAV Flight Attitude

With the rapid development of electronic technologies and intelligent control technologies,UAVs have been widely used in fields such as agricultural plant protection,security monitoring,power inspection,resource detection,and logistics express delivery.The attitude stability of UAVs occupies a pivotal position in the field of UAVs technology research.It is related to whether UAVs can maintain the desired flight attitude according to a certain degree of precision during the flight process and directly determine the working efficiency of UAVs.Therefore,how to stably and autonomously control the UAVs flight attitude has become one of the key issues that need to be solved in the UAVs field.In this dissertation,we have deeply studies the attitude stability control technology of UAVs,analyzed the advantages of deep learning in nonlinear control problems,and aims at the multivariable and non-linear attitude control system of quadrotor UAVs to optimize the control process.A method of stability control for UAVs flight attitude based on deep learning is presented.The convolutional neural network is used to adjust the attitude control parameters of UAVs in real time,and the autonomous stability control of UAVs flight attitude is achieved,and it solves the deficiencies of traditional control technology in complex nonlinear systems.In this dissertation,aiming at the problem of UAVs attitude prediction,a convolutional neural network is used to design the attitude prediction model.Through this model,the predicted output of the UAVs attitude angle is obtained.Aiming at the UAVs attitude control processing,the convolutional neural network is used to design the attitude.The stability controller,which corrects the attitude angle of the UAVs in real time through a convolutional neural network algorithm,obtains a stable control of the attitude of the UAVs.During the training process of attitude prediction network and attitude control network,the convolutional neural network algorithm is adopted respectively,and the attitude stability control of the UAVs is performed through the "offline training and on-line correction" method.In this dissertation,the convolutional neural network and the traditional BP neural network were used to simulate the aircraft attitude stability control performance and compare the results.The simulation results show that the proposed UAVs flight attitude stabilization control method based on convolutional neural network can realize the autonomous stability control of UAVs attitude angle,and the stability control efficiency and attitude error correction ability are obviously higher than the traditional neural network.This method has a fine effect on improving the attitude control accuracy and control capability of UAVs.