Research Of Two-Wheeled Self-Balancing Vehicle Based On Fuzzy PD Control

Two-wheeled self-balancing vehicle is an important branch of the study of robots. It covers electronics, machinery, automatic control and signal processing and other disciplines. Its structure is similar to the inverted pendulum and it has the property of nonlinearity, strong coupling. So it is a classic unit to test the control algorithm.The two wheels of the self-balancing vehicle were arranged in parallel. The accelerometer and gyroscope were used to detect the tilt angle and angular velocity of the forward lean and backward lean. A hall element and a magnet were used to detect the left tilt angle and the right tilt angle of the swing link. The information was transferred to the single chip microcomputer. The single chip microcomputer calculated the control value and sent it to motor. The motor worked to make the vehicle moving forward and backward and steering. The difficult point of the design lies in the accurate detection of the velocity’s attitude with cheap sensors and designing the control algorithm of the system. The signals of the accelerometer, which are of high noise, are easily disturbed by dynamic acceleration. Wavelet threshold de-noising could be used to remove the noise effectively. A kind of simple complementary filtering algorithm which was created by fusing the data of gyroscope and accelerometer was adopted to inhibit the interference of the dynamic acceleration. But the overshoot existed in the waveform. A system which was created by system identification was used to remove the overshoot. Under the combined action of the three signal processing methods, the vehicle’s attitude could be detected accurately and promptly. And the foundation of the design of control algorithm was settled.For the idealization of the vehicle’s mathematical model and the property of nonlinearity of the system, fuzzy PD control algorithm was used to control the self-balancing vehicle. With the help of the fuzzy control toolbox of MATLAB, a fuzzy controller which was suitable for the self-balancing vehicle system was designed to adjust the parameters of the PD control algorithm. For more close to actual, a three-dimensional model which was created with ADAMS was used to replace the mathematical model and ADAMS and MATLAB co-simulation was performed. The simulation results showed that the fuzzy PD control could reduce the overshoot and concussion. Finally, the fuzzy controller was transformed into the fuzzy control query table in order to be compiled into the single chip microcomputer.This research developed the physical prototype after the completion of signal processing and control algorithm design, and mainly studies the electric system design of the physical prototype. The electric system included the power circuit, the control circuit and the dynamic braking circuit. PCB electric board was designed and arranged in the vehicle.Through the experiment on the physical prototype, some disadvantages of the fuzzy PD control algorithm were found. So, the fuzzy rules were redesign to improve the algorithm. Ultimately, the self-balancing vehicle could move forward and backward and steer smoothly.