Geun Hyeong Lee

Position control of a mobile inverted pendulum system using radial basis function network

This article presents the implementation of position control of a mobile inverted pendulum(MIP) system by using the radial basis function network(RBF). The MIP has two wheels to move on the plane and to balance the pendulum. The MIP is known as a nonlinear system whose dynamics is non-holonomic. The goal is to control the MIP to maintain the balance of the pendulum while tracking a desired position of the cart. The reference compensation technique (RCT) scheme is used as a neural network control method to control the MIP. The back propagation learning algorithm for the RBF network is derived for on-line learning and control. The control algorithm has been embedded on a DSP 2812 board to achieve real-time control. Experimental results are conducted and show successful control performances of both balancing and tracking the position of the MIP.

Control of inverted pendulum system using a neuro-fuzzy controller for intelligent control education

This article presents the implementation of control of an inverted pendulum system by using the neuro-fuzzy network. The inverted pendulum system has been built in the educational kit whose purpose is to educate control engineers in both undergraduate students and graduate students. The inverted pendulum system is known as a nonlinear system whose goal is to maintain the balance of the pendulum while tracking a desired position of the cart. The Takagi-Sugeno (T-S) neuro-fuzzy control scheme is used to control the system. The back-propagation learning algorithm for the T-S neuro-fuzzy network is derived for on-line learning and control. The control algorithm has been embedded on a DSP 2812 board to achieve the real-time control performance. Experimental results show that successful control performances of both balancing and tracking the position for the inverted pendulum system.

Neural Network Compensation Technique for Standard PD-Like Fuzzy Controlled Nonlinear Systems

In this paper, a novel neural fuzzy control method is proposed to control nonlinear systems. A standard PD-like fuzzy controller is designed and used as a main controller for the system. A neural network controller is added to the reference trajectories to form a neural-fuzzy control structure and used to compensate for time-varying effects. We study two neural-fuzzy control schemes based on two well-known neural network control schemes such as the FEL scheme and the RCT scheme. Those schemes are tested to control the angle and the position of the inverted pendulum and their performances are compared.

Robust control of a mobile inverted pendulum robot using a RBF neural network controller

This article presents robust control of the mobile inverted pendulum system(MIPS) whose structure is a combination of a wheeled mobile robot and an inverted pendulum with two arms. The MIPS navigates on the horizontal plane while balancing the pendulum. Control of the MIPS is difficult since the system is non-holonomic and nonlinear so that simple linear controllers may have poor performances for the system. The radial basis function(RBF) network is used as an auxiliary controller to help the primary PID controllers to perform better. The back propagation algorithm has been developed for the RBF function network. Real time control of the RBF network has been achieved by embedding the learning algorithm onto the DSP board. The performance of the RBF network controller has been tested for the remotely controlled MIPS by conducting experiments of climbing the slanted surface while balancing.

Motion control of a mobile pendulum system using neural network

The motion of the mobile inverted pendulum system (MIPS) is controlled by a neural network controller along with PID controllers. The MIPS is required to follow the circular trajectory while maintaining the balance. The radial basis function (RBF) network is trained with the back-propagation algorithm in on-line fashion. The DSP 2812 board is developed and used for the massive calculation of neural parallel processing to achieve an on-line learning and control. Experimental studies are conducted and confirm the performance of the RBF neural network controller.

Design and control of an inverted pendulum system for intelligent mechatronics system control education

This paper presents a mechatronics system for intelligent control education. The inverted pendulum system is designed and built to fit an educational kit as an intelligent mechatronics system. The neuro-fuzzy control method whose structure is the Tagaki-Sugeno model is introduced to control the pendulum. The Takagi-Sugeno(T-S) neuro-fuzzy control structure is implemented on the DSP board built in our Lab. Parameters of the neuro-fuzzy controller are updated based on the back-propagation algorithm derived in this paper. Performances of the neuro-fuzzy controller are evaluated by experimental studies of testing desired position tracking control of the cart while balancing.

Implementation and Experiment of Neural Network Controllers for Intelligent Control System Education

This paper presents the implementation of an educational kit for intelligent system control education. Neural network control algorithms are presented and control hardware is embedded to control the inverted pendulum system. The RBF network and the MLP network are implemented and embedded on the DSP 2812 chip and other necessary functions are embedded on an FPGA chip. Experimental studies are conducted to compare performances of two neural control methods. The intelligent control educational kit(ICEK) is implemented with the inverted pendulum system whose movements of the cart is limited by space. Experimental results show that the neural controllers can manage to control both the angle and the position of the inverted pendulum systems within a limited distance. Performances of the RCT and the FEL control method are compared as well.

Experimental Studies of Neural Compensation Technique for a Fuzzy Controlled Inverted Pendulum System

This article presents the experimental studies of controlling angle and position of the inverted pendulum system using neural network to compensate for errors caused due to fuzzy controller. Although fuzzy control method can deal with nonlinearities of the system, fixed fuzzy rules may not work and result in tracking errors in some cases. First, a nominal Takagi-Sugeno (TS) type fuzzy controller with fixed weights is used for controlling the inverted pendulum system. Then the neural network is added at the reference input to form the reference compensation technique (RCT)control structure. Neural network modifies the input trajectories to improve system performances by updating internal weights in on-line fashion. The back-propagation learning algorithm for neural network is derived and used to update weights. Control hardware of a DSP 6713 board to have real time control is implemented. Experimental results of controlling inverted pendulum system are conducted and performances are compared.

Reference compensation technique of a fuzzy controlled inverted pendulum system for control education

This article presents the implementation of a neuro-fuzzy like controller design for an inverted pendulum system. The inverted pendulum system is controlled by a nominal Takagi-Sugeno-Kang (TSK) type fuzzy controller whose outputs are linear. Then the neural network is added to improve system performances by compensating signals at the reference input. Shaping input signals forms an inverse dynamics control scheme of the closed loop system whose scheme is called the reference compensation technique. The back-propagation learning algorithm for the neural network is derived for on-line learning and control. The learning algorithm has been implemented on a DSP 6713 board to achieve real time control. The proposed controller has been tested to control both balancing and tracking the position of the inverted pendulum system.

Implementation of a Neural Network Controller on a DSP for Controlling an Inverted Pendulum System on an X-Y Plane

Abstract This paper presents the hardware implementation of a neural network controller for controlling an inverted pendulum system on an x-y plane robot. The inverted pendulum system can move on an x-y plane while balancing the angle of the pendulum. Neural network algorithm is implemented on a cost effective DSP board in association with an FPGA chip. The reference compensation technique of neural network control scheme is used for on-line learning and control of the inverted pendulum system. Experimental results of tracking the circular trajectory while balancing the pendulum demonstrate to confirm the successful performance of the neural network hardware.