Path Planning and optimization of Unmanned Ground Vehicles (UGVs) in the Field

Unmanned Ground Vehicle (UGV) can be as a replacement of human beings during a dangerous mission. It can also be applied to military and performs repetitive mechanical transportation. For the UGVs in the conventional layout and structured urban environment, the research of environment perception and path planning technology has been relatively mature. Compared with the flat ground, the control of the UGV in the field is unknown and more complex. This article revolves autonomous control of the UGV in the field, putting forward a kind of path planning and optimizing technology for the UGV under complex environment. Research content includes: Conduct field environment modeling with environment information. First, the environment is detected by the fusion of lidar and IMU. Then a more accurate environment model is obtained by training the RBF neural network. Use the obtained environmental model to conduct path planning for the UGV in the field. The path planning trajectory is obtained by introducing the distance between the starting point and target point, the environmental height and environmental gradient constraints to construct the cost function. Then Analyze different characteristics of the paths that are generated by modifying the constraint function. optimize the smoothness of the generated path for the field UGV. The path is optimized from broken lines into a curve by using a 5th order polynomial trajectory.