Research And Design On Manned Self-balancing Robot

Compared with conventional vehicles, manned self-balancing robot is a new means oftransport with portable, low-carbon environmental, flexible and simple operationcharacteristics, with a wide range of applications. Meanwhile, the self-balancing robotcombines mobile wheeled robot and inverted pendulum structure, it can be used as typicalcontrol object and an ideal platform in the field of automatic control for experimental research.Therefore, this paper developed a manned self-balancing robot, expanded research from thesystem model, sensor fusion, physical production and experimental test as well as otheraspects.Based on the movement principle of manned self-balancing robot, we use Lagrangeequations to establish three degrees of freedom mathematical model to determine the systemparameters based on system dynamics, thus we can work out the state-space model, andanalyze the controllability and observability of system, as well as derive the maximumcontrollable angle.For the simplicity shortcoming of inertial sensor, we put forward Kalman filteringalgorithm to fuse position information calculated by accelerometer and gyroscope to obtainthe angle data of high accuracy and reliability. Proposing piecewise PID control algorithm toachieve the robot’s self-balancing control, setting different control parameters to achievesegmentation and time-share self-balancing effect in the state of no-load or manned. Throughbuilding MATLAB/Simulink simulation model to validate designed control method, theresults indicate that piecewise PID control algorithm has better robustness and rapidity, ableto adapt to different working environments.A manned self-balancing robot system is established, we choose the appropriatecontroller, perform motor and sensors, and design manned self-balancing robot hardware,software and mechanical architecture. Hardware system including a STM32microcontrolleras the core control system, an accelerometer and a gyro compose angle detection system, DCgeared motor and motor drive system modules. For software part, we use C language to writeprograms in KEIL software platform, mainly achieve the fusion processing of the collectedangle data, piecewise PID control and motor drive functions.The manned self-balancing robot prototype is manufactured, and related tests such as ADconversion, Kalman filter and manned performance were promoted. The best value of processcovariance matrix Q and measuring covariance matrix R in Kalman filtering was obtained bychanging parameters constantly. Self-balancing and anti-interference tests show that the datafusion effectively eliminated the interference noise and vibration, the results which have small errors and good dynamic effects are accurate. Manned test verified the robot have the basicability to work, which contained the motions of starting, stoping, advancing, recession,up-down slope, turning,etc. under manned cases.The research emphasis in this paper includes design and manufacture of mannedself-balancing robot prototype, using Kalman filtering algorithm to fuse angle data which arecalculated by accelerometer and gyroscope, achieving self-balancing in the state of no-load ormanned by piecewise PID control algorithm. Therefore, the manned self-balancing robot canrun safely and stationary. This work lays the foundation for the further study of self-balancingcontrol theory and self-balancing vehicles.