Research On The Curvature Adaptive Control Strategy Of An Aircraft Skin Detection Robot

Premier Wen Jiabao once said: "Aviation manufacturing is the will of the country, a strategic consideration, and it is also the symbol of the country's prosperity, which has great scientific,technological significance and economic benefits." So aircraft skin detection is also important. The application of the climbing robot technology to the detection of aircraft skin can make up for the shortcomings of manual detection, which is an effective way to improve the safety factor of aviation.In this paper, three modules including attitude controller, control distributor module, the robot's position and attitude integrated module are designed to perfect the robot's curvature adaptive attitude control function.Firstly, the attitude control module is designed for the better attitude control of Aircraft skin detection robot. A mathematical model of the aircraft skin detection robot is carried out, including the vacuum suction device and the cylinder device. The surface of the wall climbing robot is modeled by a dynamic model. The attitude control algorithm based on Backstepping is designed.Then, control distributor is designed to improve the adaptive function of the robot. Of wall climbing robot's stability analysis, four-pyramid method is generated to evaluate the force and moment on the stability behavior of the robot. Meanwhile, the soft switching algorithm based on control allocation principle and adsorption force control allocation algorithm is put forward, and the algorithm is analyzed by simulation.Secondly, the robot's position and attitude integrated module is designed to transmit the information of position and attitude to the robot. The positioning principle of the robot is introduced;the mathematical models of the sensors are also introduced, which are composed of the following models: five arrays of microphone and three axis accelerometer; the filter is designed by using the UKF fusion algorithm, and the real-time positioning of the robot is realized by the information filtering of different sensor data.Finally, the motion control experiment of the robot has been carried out. The feasibility of curvature adaptive control of the robot is verified by analyzing the controllability of actuator mechanism. The hardware structure of the robot actuator is introduced, and the change of the robot's curved surface is analyzed.