Minimum-energy trajectory planning and control on a straight line with rotation for three-wheeled omni-directional mobile robots

Abstract

The minimum-energy trajectory planning and control algorithm on a straight line with rotation is proposed for battery-powered three-wheeled omni-directional mobile robots (TOMRs). In order to lengthen the operation time of a mobile robot with carried batteries, we have chosen a practical cost function as the energy from the batteries to actuation motors. The optimal control theory using a Hamiltonian function is used to solve the minimum-energy trajectory, which gives the rotational velocity profile in analytic form and the translational velocity profile by numerical method. We also establish search algorithm to find the minimum-energy trajectory. These minimum-energy velocity trajectories are compared to the conventional trapezoidal velocity profiles via simulation to analyze actuation energy consumption, which reveal energy savings up to 4.76 %. Also a trajectory following control system is implemented to validate the actual performance.

Kinematic Modeling and Trajectory Tracking Control of a Wheeled Omni-directional Mobile Logistics Platform

“…For example, moving the vehicle in a particular direction is necessary when driving in narrow spaces or in spaces with many obstacles. In these situations, the existing cylindrical wheel structure cannot drive in a random direction because it requires a lot of space for rotating and turning the vehicle body [8][9]. An omni-directional mobile robot is a type of holonomic robots, it has the ability to move simultaneously and independently in translation and rotation [4,10].…”

Mobility Diversity-Assisted Wireless Communication for Mobile Robots

“…A TOMR is a MR with three omnidirectional wheels [20], where each wheel is driven by its own motor. The distance from each wheel to the center of the robot is denoted as L. The TOMR model described in this subsection is a version of the model presented in [21]. The robot is equipped with an antenna installed at the geometrical center of the robot (see Fig.…”

“…All four parameters (k 1 , k 2 , k 3 and k 4 ) depend on various electromechanical parameters of the MR's motors but to avoid introducing more parameters and keep the notation as simple as possible we do not present more details here. The interested reader can find the detailed expression of k 1 , k 2 , k 3 and k 4 by matching the model presented in [21] to our version.…”

“…2 to 9. If the carrier frequency used is higher than 1GHz then the wavelength is smaller than 30cm and according to the experimental results in [21], the article from which we extracted the TOMR model for this article, this particular MR can at least travel 50cm in one second. Therefore the MR can now move from q N to q opt in less than one second.…”

“…Therefore the MR can now move from q N to q opt in less than one second. Now, the time taken for the MR to estimate the channel at each stopping point will depend on the amount of data utilized for this process but in general the time required for this task can easily be assumed less than one second 21 . Considering all this information we can say that the minimum value for T max (N ) is loosely bounded by 2N seconds: N − 1 seconds to traverse all the N stopping points, N seconds to measure the channel at all the stopping points and one second to go from q N to q opt .…”