SQP-Based Mobile Manipulator Motion Planning With Controlled Infeasibility for Physically Valid Task Failure

Volume 6A: 37th Mechanisms and Robotics Conference ◽

10.1115/detc2013-13377 ◽

2013 ◽

Author(s):

Chang B. Joo ◽

Joo H. Kim

Keyword(s):

Quadratic Programming ◽

Motion Planning ◽

Sequential Quadratic Programming ◽

Merit Function ◽

Mobile Manipulator ◽

Nonlinear Functions ◽

Unified Approach ◽

Optimal Motion ◽

Sqp Algorithm ◽

Optimal Motion Planning

Since anticipating or recovering infeasibility in optimal motion planning is not always possible, infeasibilities occur frequently and are not completely avoidable. We introduce an enhanced sequential quadratic programming (SQP) based framework of controlled infeasibility for physically valid solutions, based on our previous study. A priority weight function is incorporated into an SQP algorithm combined with constraints and objective function normalization to ensure strict satisfaction of high-priority constraints. These are embedded in the SQP algorithm through its merit function and composite cost function, in which general nonlinear functions can be incorporated in a unified approach. Several simple mobile manipulator examples demonstrate the advantages of the proposed method.

Download Full-text

Task-Constrained Optimal Motion Planning of Redundant Robots Via Sequential Expanded Lagrangian Homotopy

Journal of Mechanisms and Robotics ◽

10.1115/1.4039395 ◽

2018 ◽

Vol 10 (3) ◽

Cited By ~ 1

Author(s):

Audelia G. Dharmawan ◽

Shaohui Foong ◽

Gim Song Soh

Keyword(s):

Motion Planning ◽

Real Time ◽

Dynamic Environment ◽

Computational Time ◽

Mobile Manipulator ◽

Optimal Motion ◽

Task Constraints ◽

Moving Obstacles ◽

Time Motion ◽

Optimal Motion Planning

Real-time motion planning of robots in a dynamic environment requires a continuous evaluation of the determined trajectory so as to avoid moving obstacles. This is even more challenging when the robot also needs to perform a task optimally while avoiding the obstacles due to the limited time available for generating a new collision-free path. In this paper, we propose the sequential expanded Lagrangian homotopy (SELH) approach, which is capable of determining the globally optimal robot's motion sequentially while satisfying the task constraints. Through numerical simulations, we demonstrate the capabilities of the approach by planning an optimal motion of a redundant mobile manipulator performing a complex trajectory. Comparison against existing optimal motion planning approaches, such as genetic algorithm (GA) and neural network (NN), shows that SELH is able to perform the planning at a faster rate. The considerably short computational time opens up an opportunity to apply this method in real time; and since the robot's motion is planned sequentially, it can also be adjusted to accommodate for dynamically changing constraints such as moving obstacles.

Download Full-text

Optimal Motion Planning for the Teaching Experimental Mobile Manipulator

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.138-139.56 ◽

2011 ◽

Vol 138-139 ◽

pp. 56-61

Author(s):

Huai Ping Zhou ◽

Ping Ge ◽

Yong Fang

Keyword(s):

Optimal Control ◽

Motion Planning ◽

Minimum Principle ◽

Teaching Experiment ◽

Mobile Manipulator ◽

Optimal Motion ◽

Control Scheme ◽

Experimental Apparatus ◽

Optimal Motion Planning ◽

Optimal Control Scheme

An optimal motion planning based on minimum principle is presented to address the motion problem of the mobile manipulator in a sort of experimental system. In view of the characteristic of the practical experimental apparatus, the model of the manipulator is deduced based on the kinetic analysis and mathematic method. An optimal control scheme is then investigated to deal with the optimization problem of the motion planning for the manipulator, so as to guarantee the demand of the teaching experiment. Simulation verifies the control performance of the optimal control scheme for the optimal motion planning of the manipulator, and it helps improve the teaching experiment effect.

Download Full-text

Controlled Infeasibility for Physically Feasible Optimal Motion Planning of Manipulators

Volume 4: 36th Mechanisms and Robotics Conference, Parts A and B ◽

10.1115/detc2012-71342 ◽

2012 ◽

Author(s):

Joo H. Kim ◽

Chang B. Joo

Keyword(s):

Motion Planning ◽

Merit Function ◽

Programming Algorithm ◽

Weight Functions ◽

Optimal Motion ◽

Elastic Mode ◽

Priority Weights ◽

Optimal Motion Planning ◽

Three Degree Of Freedom ◽

Planning Problems

In this presentation, infeasible solutions of the optimal motion planning problems are treated through constraint prioritization and associated weight functions. The merit function and elastic mode of a sequential quadratic programming algorithm are used as main driving formulations. In the proposed multiple-loop iterative algorithm, constraints are first normalized according to the current values at every certain number of iterations and then the priority weights are assigned. The algorithm was demonstrated using a three-degree of freedom planar manipulator for two problems such as obstacle avoidance and excessive external load for static configuration and a dynamic motion. The results of those two examples show reliable and physically consistent manipulator configurations which demonstrate the valid formulation of the prioritized constraints.

Download Full-text