Abstract
The one-wheeled robot has a nonlinear and strongly coupled complex dynamic system. It
is a great challenge to control theory and dynamics research, and it is a very good research
model. This paper will design a one-wheeled robot, focusing on the analysis of the movement
of the one-wheeled robot, and establish a dynamic model and write a control program.
This paper analyzes the working principle of the one-wheel robot and selects the
appropriate electronic components for the one-wheel robot. In this paper, the mpu9250
electronic gyroscope is selected to capture the motion posture of the one-wheel robot. The
GA12-N20 motor provides the driving force for the one-wheeled robot. L298N motor drive
module is used for motor drive. Two 18650 batteries powered; Arduino uno board is used as
the controller of one-wheel robot.
In order to solve the problem of deviation and noise in the data of the electronic gyroscope,
this paper corrected the data and designed a filtering algorithm inspired by the filtering
characteristics of the capacitance. In this paper, from the capacitor filter circuit, analysis,
capacitance mathematical modeling. According to the characteristics of computer data
processing, this paper discretizes the model and uses the recursive formula to replace the
summation formula. In this paper, the capacitance filtering algorithm was preliminarily tested
and analyzed with MATLAB software, and the actual arduino was used for practical
verification. The experimental results show that the algorithm can effectively filter the noise of
the signal. In order to make the data more readable, more convenient for data analysis. In this
paper, the corresponding data processing, so as to obtain the angular velocity, angular
acceleration, Angle three physical quantities.
The dynamic system of one-wheeled robot is characterized by strong coupling of multiple
variables, which makes the design simple. In this paper, the dumping process of one-wheel
robot is disassembled into a combination of front and back dumping and left and right dumping.
The dumping motion of each axis is analyzed independently, the dynamic model is established,
and the control system is designed and simulated. In the design of left-right dumping control,
the PD control system is obtained by analyzing the change of force, Angle and angular velocity
in the process of dumping. However, in the simulation of left and right dumping, the control
system could not distinguish the dumping and backswing of the one-wheeled robot.
In order to solve the problem that the control system cannot recognize the difference
between backswing and dumping, this paper analyzes the dumping process again. The integral
term, which is the Angle parameter, is added to the original control system. The Angle
parameter allows the control system to know where the one-wheeled robot needs to be restored.
In this paper, the control system has been tested again, and it is found that the control system
has been able to distinguish back swing from toppling, and can make the one-wheeled robot
return to the upright state after several wobbles.
With the control experience of right and left toppling, the control system of front and back
toppling directly adopts the simultaneous control of three sets of data including Angle, angular
velocity and angular acceleration. As the forward and backward dumping will lead to
displacement, the simulation program used in this paper is animated to view the posture of the
one-wheeled robot in real time. It can be seen from the simulation data curve that the designed