Research On The Navigation And Control System Of Autonomous Working Robot In The Field

The new energy photovoltaic power generation has gradually become an important part of the power generation system in China.The photovoltaic panels,the key unit component of solar energy collection,are exposed to harsh wind,sand,dust,rain and snow,and their surface cleanliness seriously affects power generation efficiency and the normal working state of the system.To this end,this paper designs an outdoor mobile robot control system of intelligent clean photovoltaic panels which can work all-weather / night.Aiming at the site characteristics of photovoltaic power plants with large area and regular photovoltaic panels,through controlling the trajectory and posture of outdoor mobile robots,the intelligent cleaning robots can work autonomously in photovoltaic power plants.Firstly,according to the kinematics and dynamics characteristics of the cleaning robot,such as large load,low speed and large inertia during cleaning operation,the state vector of the system is established,which takes the lateral position deviation,lateral velocity deviation,heading angle deviation and heading angle velocity deviation of the robot as the state vector,The front wheel steering angle and speed(throttle opening)of the robot are used as the mathematical model of the system control vector.By studying the integrated navigation algorithm,the GNSS / INS loose / tight integrated navigation system model based on EKF is established,and the reliability of the tight integrated navigation system is verified by simulation.Secondly,an improved LQR trajectory tracking controller based on Fuzzy preview feedforward is designed according to the mathematical model,and a feedforward compensation algorithm is designed to overcome the system disturbance and achieve no static error.The fuzzy preview controller is designed to dynamically adjust the preview distance,which solves the overshoot problem when the preview distance is zero,and improves the control accuracy and smoothness of the system.The improved LQR trajectory tracking algorithm based on Fuzzy preview feedforward is simulated and verified by using MATLAB and Car Sim professional simulation platform of mobile system.It can realize stable operation under continuous lane changing condition and trigonometric function condition,and has good tracking effect.Finally,the actual system platform of PV cleaning robot is built,and the man-machine interaction system based on ROS and QT environment is designed.Using the hardware platform NV1-TX2 of photovoltaic cleaning robot control system,the application test of improved LQR trajectory tracking algorithm based on Fuzzy preview feedforward is carried out.The results show that the path,steering and other autonomous control of photovoltaic cleaning robot can meet the application requirements.