Design Of Two-wheel Self-balancing Trolley System Based On Fuzzy Sliding Mode Control

In the field of wheeled robots,two-wheeled self-balancing trolleys are one of the most important branches.Miniaturization and simplification are its main advantages,and they generally work in small and dangerous spaces.At the same time,it is a nonlinear,underactuated and strongly coupled system,which is suitable as a scientific experimental platform for the research and verification of various control algorithms.This thesis mainly studies the control method of two-wheel self-balancing trolley,and also elaborates and verifies the filter design and software implementation.Firstly,Newton's mechanics method is used to analyze the force of the car wheel and body.The dynamic model and mathematical model of the car system are established.The system is linearized near the equilibrium point and decoupled into two subsystems.controllability,observability and Stability were analyzed.Secondly,based on the mathematical model,a sliding mode variable structure controller based on exponential approach law is designed.The simulation results show that chattering phenomenon occurs in the system.Combined with the advantages of fuzzy control,an adaptive fuzzy sliding mode controller is designed,and the fuzzy gain algorithm is used to adaptively adjust the switching gain term of sliding mode control.Simulations show that the improved controller can weaken chattering very well.Thirdly,the data measured by the accelerometer and the gyroscope are processed,then the Kalman filter algorithm and the complementary filter algorithm are compared and analyzed,and the Kalman filter algorithm is selected to process the data,so that more accurate dip angle data can be obtained.Finally,based on the integrated development environment of AVR Studio and GCCAVR,the functional modules of the car are designed,the physical platform of the car system is built,and the car balance experiment is carried out to obtain the PWM voltage of the motor terminal voltage when the car balances.Thereby verifying the effectiveness of the control algorithm.