Point Stabilization Research Of Two-wheeled Robot Based On Human-simulated Intelligent Control

Two-Wheeled Mobile Robot (TWMR) is widely used in the robot system. BecauseTWMR has the kinematic nonholonomic constraint,dynamic nonlinear saturation, twodrive system which are mutual coupling, strong interference and time-variantcharacteristics of the system in the real-time movement process. So the motion controland optimization problem of TWMR is always the key and difficult point of theresearch at home and abroad. In addition, point stabilization problem is the mostimportant and basic question in the TWMR motion control.Human-Simulated Intelligent Control is imitating human behavior to designcontroller.The main objective of the study is not the object itself, but controllersimulates the control structure and control behavior of human, especially suitable forsolving complicated nonlinear problems. In this paper, based on the theory of humansimulated intelligent control, what had done aiming at point stabilization control of theTWMR is as follows:①Put forward a kind of three layers structure of human-simulated intelligent(TL-HSIC) TWMR point stabilization controller and parameter tuning method. Thecontroller is composed of three layers: bottom executive layer of Human-Simulatedintelligent control, dynamic parameter adaptive adjustment layer and task coordinationlayer. The bottom executive layer aims at the movement control effect of kinematicnonholonomic constraint of the robot and the dynamics nonlinear, makes drivingmotor’s acceleration as control parameter into the controller system and useshuman-simulated intelligent multimode controller constituted the basic control unit; thedynamic parameter adaptive adjustment layer aims at strong interference andtime-variant characteristics of the system in the real-time movement process, employsthe neural networks on-line real-time to change control parameter of the basic controlunit of the executive bottom layer and achieve the control effect of the optimization ofall the basic control unit; the top layer is the task coordination layer, according to thesystem and the state of the environment, planning and controlling all the basic controlunit and work state of dynamic parameter adaptive adjustment. Taking geneticalgorithm and using the time that the robot reaches target as fitness function to realizecontrol parameter tuning of the entire control systems.②Establish TWMR simulation experiment simulation based on equivalent state space model and the actual experimental platform based on the new mode DC motordouble closed-loop drive, and carry on the motion control experiment point stabilizationof TWMR. Based on equivalent state space model which is put forward by laboratoryuses matlab to establish simulation platform and carries on the parameter tuning andoptimization experiments to verify the validity of the controller; based ondynamicparameter adjustment DC motor human-simulated intelligent control methoddesigns new mode DC motor double closed-loop drive which can carry on motor todrive current limit control and uses the actual TWMR system to complete the actualsystem experiments.Through the simulation experiment and real experiment confirmed TWMR threelayers of human simulated intelligent point stabilization controller structure which is putforward in this paper is effective. Compared with the effect of multimodehuman-simulated intelligent controller (MP-HSIC) and improved multimodehuman-simulated intelligent controller (MP-HSIC), it showed that the control ofTWMR three layers of human simulated intelligent point stabilization controller wasquicker and motion trajectory was smoother. In addition, it obtained better controleffect.