On the controller implementation in the real underactuated walking robot model*

The purpose of this paper is the experimental validation of the sensor fusion designed for the possible future use in the walking-like mechanical systems design. The validation is made using the actuated leg being the part of the four-link laboratory walking-like system and it is equipped with several sensors. The experiments aim to track the reference angular positions trajectory using the state feedback controller. The information needed by that feedback controller (angular positions, velocities and input currents of DC motors actuating rotary joints) is obtained from the sensors measurements. Two types of experiments are compared: (i) using these measurements directly; (ii) pre-processing them via the designed sensor fusion method. The fusion is based on the so-called low-level hybrid Extended Kalman Filter (HEKF) realization, which has the future potential to be implemented on the PCBs attached to the legs links only. Yet, the presented experimental results demonstrate the prevalence of the control using the fused data.

[1]  Christine Chevallereau,et al.  Models, feedback control, and open problems of 3D bipedal robotic walking , 2014, Autom..

[2]  M. Spong,et al.  CONTROLLED SYMMETRIES AND PASSIVE WALKING , 2002 .

[3]  Leonid B. Freidovich,et al.  A dynamic human motion: coordination analysis , 2015, Biological Cybernetics.

[4]  Sergej Celikovský,et al.  On the collocated virtual holonomic constraints in Lagrangian systems , 2016, 2016 American Control Conference (ACC).

[5]  Sergej Celikovský,et al.  Simple model of underactuated walking robot , 2015, 2015 10th Asian Control Conference (ASCC).

[6]  Amod Kumar,et al.  A brief review of dynamics and control of underactuated biped robots , 2017, Adv. Robotics.

[7]  E. Westervelt,et al.  Feedback Control of Dynamic Bipedal Robot Locomotion , 2007 .

[8]  Sergej Čelikovský,et al.  Sensor Fusion for simple walking robot using low-level implementation of Extended Kalman Filter , 2018 .

[9]  Christine Chevallereau,et al.  Bipedal Robots : Modeling, Design and Walking Synthesis , 2009 .

[10]  Sergej Celikovský,et al.  Advanced LMI based analysis and design for Acrobot walking , 2010, Int. J. Control.

[11]  Sergej Celikovský,et al.  Collocated virtual holonomic constraints in hamiltonian formalism and their application in the underactuated walking , 2017, 2017 11th Asian Control Conference (ASCC).

[12]  Leonid B. Freidovich,et al.  Stable Walking Gaits for a Three-Link Planar Biped Robot With One Actuator , 2013, IEEE Transactions on Robotics.

[14]  Sergej Celikovský,et al.  On the sensor fusion in the walking robots design , 2017, 2017 11th Asian Control Conference (ASCC).

[15]  Sergej Celikovsky,et al.  ADAPTIVE NONLINEAR TRACKING FOR ROBOTIC WALKING , 2012 .

[16]  R. Mehra.  On the identification of variances and adaptive Kalman filtering , 1970 .

[17]  Sergej Čelikovský,et al.  Application of the Method of Maximum Likelihood to Identification of Bipedal Walking Robots , 2018, IEEE Transactions on Control Systems Technology.

[18]  S. Celikovsky,et al.  Partial exact linearization design for the Acrobot walking , 2008, 2008 American Control Conference.

[19]  Christine Chevallereau,et al.  RABBIT: a testbed for advanced control theory , 2003 .

[20]  Koushil Sreenath,et al.  MABEL, a new robotic bipedal walker and runner , 2009, 2009 American Control Conference.