A Research Of Quadrotor Identification And Control Algorithm

In recent years,rotor UAV has been widely used in agriculture,aerial photography,patrol inspection and many other fields.However,in the complex external environment,it is difficult to realize the accurate control of UAV,so it is necessary to improve the control robustness of UAV.A highly robust control method relies on high identification accuracy.In this paper,by studying and improving the deep echo state network,the UAV wind speed identification method is designed,and the identified wind speed information is used to improve the robustness of UAV control.The main research is as follows:（1）The reason for the deterioration of the output prediction of the deep echo state network is studied.Based on the idea of the weight and consistency method of the network in the field of biology,a method of input weight matrix selection is proposed,which avoids the internal the saturation of the neuron state enables the network to take advantage of its depth,improve the output prediction ability of the network,and verify the generalization ability of the network.（2）On the basis of studying the control model of UAV,an improved wind speed identification method based on deep echo state network is proposed.The input data is preprocessed,and the parameters such as the number of layers and neurons of the deep echo state network are optimized and adjusted to realize the prediction of wind speed.（3）Under the premise of wind speed identification,the UAV position controller of PID,backstep control and sliding mode control is designed and improved,and the control effect is compared and analyzed to verify the feasibility and effectiveness of the wind speed identification method.Finally,it can be seen from the simulation verification that the prediction ability of the improved DESN is significantly better than the original one.In the case of wind,compared with the original controller,the trajectory tracking errors of the three types of controllers considering the wind speed are reduced by more than 20%,which improves the wind resistance of the UAV,which proves the wind speed identification method designed in this paper.feasibility.