Dynamic Control Platform Of Mobile Robot

In the design of control system of a mobile robot,control engineers often find that simulation results are difficult to apply to system directly due to large difference between theory and practice.Thus,an advanced control platform which can observe the performance of the algorithm is needed.To meet the requirements of designing mobile robot,a wireless miniaturization platform will be desired.Existing solutions such as dSPACE real-time system cannot meet the requirements because of their large size and inability to wireless control.This thesis presents a ballbot control platform base on rapid control prototyping.First,we finish the configuration of software,hardware and the code generation modules.Then,the simulation models of the robot are established,and the simulation results are used as references for the actual debugging.Finally,we adjust the sensors,actuators and debug self-balanced algorithm.The main contents of this thesis are as follows:The The design of hardware platform.Firstly,we design electrical and mechanical system according to system requirement.Secondly,the design and the assembly of mechanical structures are completed.Thirdly,the electrical system is optimized to establish a solid foundation of total system.The design of software platform.Firstly,the design idea and instructions of software platform are introduced.Secondly,the development of sensors/actuators code is finished.Thirdly,these driver code is wrapped into code generation module with the utilization of S-Function.Software platform connect simulation and reality by combing the hardware and the MATLAB/Simulink environment.In this way,engineer can obeseve the real-time simulation results and modify system parameters conveniently.The creation of simulation model.Firstly,we obtain the first principle model of the ballbot by analyzing the Lagrange's equation.Secondly,the procedure of physical modeling in Simulink environment is provided.Thirdly,some tricks about the procedure of physical modeling will be mentioned.Experiment.Firstly,we show the process of real-time interactions between control platform and reality by three experiments:attitude estimation,motor modification,self-balance test.Secondly,we make a careful comparison between the simulation results and the actual data.Thirdly,an idea is proposed:Theoretical simulations provide the approximate range while practical experience determine the specific details.