Simulation and Development of Two-Wheels Balancing Robot using State Feedback Controller and Reduced-Order Estimator

Muhammad N Mahyuddin, Mohd Rizal Arshad, S Z Ashraf, G Herrmann

    Research output: Chapter in Book/Conference proceedingConference contributionpeer-review

    Abstract

    This paper presents some of the simulation results to show that the modeled two-wheels balancing robot can be balanced using the state feedback controller coupled with a reduced-order estimator/observer. The simulated model parameters are forced to adhere to the actual parameters used to design the actual two-wheel balancing robot (already constructed). The reduced-order estimator is an essential component in the overall compensator design as there would be no encoder to read the lateral displacement. The real hardware system equipped with an inclinometer to read the tilting angle, gyroscope to read the tilt rate but without a pair of encoders to read the lateral displacement(missing state variable) travelled by the robot. The simulated result is validated through implementation of the compensator algorithm in the real hardware system (whereby the encoders are omitted to introduce the missing state variable with which the reduced-estimator is placed into use). Implementation issues of incorporating the algorithm are discussed briefly in the context of usage of suitable embedded system, the choice of sampling time, the choice of fixed point calculation versus the floating point calculations and nonlinearities. The real system consists of PIC16F877A as the robot’s 8-bit microcontroller, ADXRS300s as its single-axis gyroscope and Parallax’s Memsic 2125 dual axis accelerometer, signal conditioned to act as an inclinometer.
    Original languageEnglish
    Title of host publicationInternational Symposium on Robotics and Intelligent Sensors (IRIS2010)
    Pages203-208
    Number of pages6
    Publication statusPublished - 2010

    Fingerprint

    Dive into the research topics of 'Simulation and Development of Two-Wheels Balancing Robot using State Feedback Controller and Reduced-Order Estimator'. Together they form a unique fingerprint.

    Cite this