Research On Path Planning And Trajectory Tracking Of Mobile Robot In Complex Environment

With the rapid development of economy,science,and technology in various countries around the world,the application fields of mobile robots have become more and more extensive.And their application scenarios have also expanded from indoor environments to various complex environments,such as the field,underwater,sky,and even outer space.At present,there are still many unresearched topics for robot systems in complex environments.In this dissertation,the path planning and trajectory tracking control technology of wheeled robots under complex environmental conditions with known global information and uneven terrain are studied.The uneven terrain is modeled,and a multi-scale technique is proposed for the improvement of the time complexity of the curved surface path planning algorithm.Aiming at the problem of improving the accuracy of the curved surface path planning algorithm,a two-dimensional equivalent expansion technology of curved surfaces is presented.Based on the planned optimal path on the uneven terrain,the surface trajectory tracking control research is carried out.Under the constraint of torque limitation,the optimization improvement method based on the gradient descent parameter optimization method is presented to optimize the adjustment time index and the maximum torque derivative.Besides,the trajectory tracking control algorithm is researched and analyzed on the unevenness of the rugged ground in the complex environment and the influence of the uncertainty interference caused by the robot turning.The main work is as follows:First of all,for the path planning of wheeled robots with curved surfaces,a multiscale graph method that reduces the time complexity of the path planning algorithm is proposed.The time complexity of common path planning algorithms will increase exponentially with the increase of the node size of the planning scene.For rugged terrain planning in complex and large-scale scenarios,the time cost of using this type of path planning algorithm is too great.The multi-scale method presented in this dissertation reduces the influence of the increase of the number of nodes on the increase of time complexity by finding the optimal path on the coarse-fine scale graph,thereby reducing the time cost of the path planning algorithm.The larger the scale of the node on the rugged ground,the better the improvement effect of the multi-scale method on the time complexity.This dissertation also extends the application of the multi-scale method to multi-robot path planning.Secondly,for the path planning of wheeled robots with curved surfaces,a twodimensional equivalent expansion method is proposed to reduce the path length error of the path planning algorithm.The traditional path planning algorithms usually calculate the optimal path based on the Euclidean distance formula between nodes.When the traditional algorithms are applied to optimal path planning on curved surfaces,there is an error between the distance between nodes on the curved surface and space Euclidean distance between nodes.The two-dimensional equivalent expansion method reduces the optimal path length error of the path planning algorithm by calculating the path length on the two-dimensional equivalent curved surface.The greater the unevenness of the rugged ground or the larger the scale of the nodes on the rugged ground,the better the error improvement effect of the two-dimensional equivalent expansion method in calculating the optimal path length of the curved surface.In this dissertation,the two-dimensional equivalent expansion method is also extended to the path planning of multi-robot.Then,for the trajectory tracking of surface wheeled robots,an optimization method of adjustment time with torque limitation is proposed.Since the input torque obtained by the usual trajectory tracking control algorithm design often exceeds the maximum saturation value in practical applications,which makes the practicability of the trajectory tracking algorithm worse,the maximum input torque should be limited during the algorithm design.Reducing the torque usually sacrifices the system adjustment time index.The design should not make the adjustment time too long due to the control torque to make the algorithm unable to be practically applied.The adjustment time optimization method with torque limit controls the maximum input torque at the desired value,and at the same time obtains the smallest possible adjustment time,which ensures the practicability of the algorithm.This optimization method saves the energy required by the mobile robot by reducing the maximum input torque to an ideal desired value.The reduction of the maximum input torque also reduces the output power required by the robot motor,which can reduce the size of the robot's motor and other mechanical structures.Finally,for the surface trajectory tracking of wheeled robots,an optimization method for the maximum torque derivative with torque limitation is proposed.At the initial moment of the application of force,the usually designed input torque will have a sudden change of a large shock load in a very short time.This impact load will cause harm to the electrical equipment and mechanical structure of the mobile robot,and reduce the lifespan of the robot.The optimization method of the maximum torque derivative is based on limiting the maximum torque and the system adjustment time and minimizes the maximum torque change rate of the system so that the start process of the robot is smooth,and the shock load during the start is reduced.It reduces the load requirements and mechanical size requirements of the robot's electrical equipment and mechanical equipment.The optimization of the maximum value of the input torque derivative makes the required torque input small,which effectively save energy.