2A2-D18 Learning Dynamic Throwing Motion of A Wheeled Inverted Pendulum utilizing Whole Body Dynamics

2009 ◽  
Vol 2009 (0) ◽  
pp. _2A2-D18_1-_2A2-D18_4
Author(s):  
Takehiro Nishikawa ◽  
Yasutake TAKAHASHI ◽  
Takayuki NAKAMURA ◽  
Minoru ASADA ◽  
Hiroshi ISHIGURO
2011 ◽  
Vol 2-3 ◽  
pp. 74-79
Author(s):  
Luis Canete ◽  
Sunao Kimura ◽  
Takayuki Takahashi

In this paper the control of the I-PENTAR, a wheeled inverted pendulum type robot being developed by the authors, for pushing and pulling a cart is examined. To control the movement of the object is being pushed or pulled, information regarding several external parameters, eg. Mass of the object and friction components, must be considered. In most cases these parameters are not known before hand or may change. One method of compensating for these unknown or changing external parameters is to represent them as an equivalent reaction force from the object. Our first subject of this research is to design a disturbance observer to estimate and compensate the equivalent force. Another situation is of pushing and pulling a cart with the inverted pendulum type robot traversing an inclined plane. As an initial step to solving this problem in this paper, a force application method using whole body motion of the inverted pendulum type robot is proposed. The whole body motion means changing the balance of the robot to attain a certain desired force. During application of this force the robot must remain in its stabilized or balanced state. For an inverted pendulum type robot, this instantly poses a major problem. To solve the problem, a reduced order disturbance observer is used in this paper to estimate the force applied by the robot. On the other hand, I-PENTAR is targeted for environments where it can interact with humans and so safety is a major concern. For example, in the event that an obstacle bumps the robot as it is pushing the cart, a large and sudden force estimator based on the disturbance observer is also built into the controller. Simulation and experiments using the reduced order disturbance observer and evaluation of the whole body motion force control are presented.


2014 ◽  
Vol 971-973 ◽  
pp. 714-717 ◽  
Author(s):  
Xiang Shi ◽  
Zhe Xu ◽  
Qing Yi He ◽  
Ka Tian

To control wheeled inverted pendulum is a good way to test all kinds of theories of control. The control law is designed, and it based on the collaborative simulation of MATLAB and ADAMS is used to control wheeled inverted pendulum. Then, with own design of hardware and software of control system, sliding mode control is used to wheeled inverted pendulum, and the experimental results of it indicate short adjusting time, the small overshoot and high performance.


2015 ◽  
Vol 73 (6) ◽  
Author(s):  
Amir A. Bature ◽  
Salinda Buyamin ◽  
Mohamad N. Ahmad ◽  
Mustapha Muhammad ◽  
Auwalu A. Muhammad

In order to predict and analyse the behaviour of a real system, a simulated model is needed. The more accurate the model the better the response is when dealing with the real plant. This paper presents a model predictive position control of a Two Wheeled Inverted Pendulum robot. The model was developed by system identification using a grey box technique. Simulation results show superior performance of the gains computed using the grey box model as compared to common linearized mathematical model. 


Sign in / Sign up

Export Citation Format

Share Document