Font Size: a A A

Kinematics Of Redundant Manipulator And Dead-reckoning And Path Planning Research Of Mobile Platform

Posted on:2016-08-20Degree:DoctorType:Dissertation
Country:ChinaCandidate:H DongFull Text:PDF
GTID:1108330503469734Subject:Mechanical and electrical engineering
Abstract/Summary:PDF Full Text Request
As a significant subdomain of the manipulator, redundant mobile manipulator includes the mobile platform and the redundant mobile manipulator is fixed on a platform.Redundant mobile manipulator not only can movement in the infinite space like mobile robot, and also can realize flexible operation as mechanical arm which can substitute human labor in working conditions with high complexity and dangerousness. Two systems are included in a redundant mobile manipulator, namely the redundant manipulator and the mobile platform. Since the inherent high redundancy, size optimization and inverse kinematics solution are critical to redundant mobile manipulator study. While, as the mobile platform is employed to avoid the obstacles and reach targeted locations, the accurate positioning and path planning are of great significance. Kinematics of redundant manipulator includes forward kinematics and inverse kinematics. Solutions of forward kinematics are the end joint positions of manipulator that are solved by the joint variables.Forward kinematics is commonly used to build the workspace of manipulator that is the reference of size optimization. Solutions of inverse kinematics are the values of joints that are solved by the end joint positions of manipulator. Inverse kinematics is commonly used to control of manipulator.Meanwhile, caused by the big amount of bars in a redundant manipulator, size optimization is essential for offering adequate workspace and flexibility. In this work, following the workspace size and the optimizing criterion of flexible area, a size optimization algorithm has been proposed. Base on the structural parameters of the manipulator, a workspace density function has been set up which generates workspace and flexible area distribution effectively. The workspace density function is employed as the optimization objective function in the process of size optimization. The calculation quantity of the presented approach is much less than the traditional Monte-Carlo method.Furthermore, the approach generates the workspace which is reluctant for Monte-Carlo method while the links of manipulator increases.In calculation, non-monolithic solutions of inverse kinematics exist as the redundant manipulator has high redundancy, and thus inverse kinematics of redundant manipulator is important. This paper presents an inverse kinematics iterative algorithm based on the workspace density function with improved precision and convergence. Obstacles in workspace have also been involved in the algorithm to acquire an effective path.Mobile robot positioning is the basis of control algorithm. However, high precision positioning sensors are prone to being affected by the working condition with high complexity and dangerousness.Dead-Reckoning error algorithm relies on the structural parameters of the platform and physical output to obtain the position of mobile robot, and is independent of position sensor. However, the algorithm is open-loop and the deviation caused by noise will be increased with time duration. Here, the effects on wheel output by noise are estimated by Dead-Reckoning error algorithm.To improve positioning precision, the information from multi-sensor has been fused.Also, we use the collecting information and set up corresponding mathematical models by employing Bayesian theorem followed by information fusing by Fourier transform to improve the accuracy of positioning. Similarly, the information collected by a small number of sensors have been fused with the error calculated by the Dead-Reckoning algorithm.In parallel, path planning is a main aspect in mobile robot research and this work presents a path planning algorithm based on the motion differential equation of mobile platform. The algorithm generates an optimal path in the obstacle-environment by balancing obstacle avoidance and targeted position, and is capable of providing guidance for the mobile manipulator’s next execution.In summary, this work focuses on key techniques of the redundant manipulator and the mobile platform, and advances the existing theoretical studies. The results from theoretical analysis have been validated by experimental results.
Keywords/Search Tags:Workspace Density Function, Inverse Kinematics, Dead-Reckoning Error, Sensor Fusion, Path Planning
PDF Full Text Request
Related items