Comparison of PID and fuzzy logic to control   the motions of robotic prosthetic limbs

1. INTRODUCTION

Prosthetic limbs are one of the important devices or elements of functional restoration. Significant improvement in the study of prosthetics has highly accelerated during the past few decades. However prosthetic limbs in Indonesia have a relatively cheaper price and are easily manufactured, generally, the users still have some trouble and spend a lot of energy to move it. It has become an important challenge for society and the government to rehabilitate these people with disabilities. These problems can be reduced by making bionic legs or robotic prosthetic limbs. Bionic legs can be used as an alternative tool for persons with lower limb disabilities to facilitate the movement of the person in carrying out activities. Bionic is a study about replacing anatomic structures or physiological processes with electronic and mechanical components. Bionic is a term that refers to the flow of ideas from biology to engineering and vice versa. The application of bionic aims to replace or repair the function of organs or parts of the body with a mechanical approach. Bionic limbs have progressed greatly today, shown in a previous study, a review has been made to discuss how main functions of the human hand are replicated artificially and the key areas of research that could conduct to vast advancement in bionic limbs functionality.

Bionic legs movements can be regulated by the implementation of a control system on the microcontroller. The control system is a tool to control and regulate the condition of a system so that the movement is more optimal. The complication of the dynamic system makes it urgent to develop process control technology. Recently, a study shows that using microprocessor-controlled systems allows dynamic controls of the flexion and extension behavior of the knee joint, resulting in potential benefit to the amputee. Researchers focus on four elements through improving prosthetic limbs: mechanisms,

actuators, sensing, and control.  In a previous study conducted by HualongXie, bionic legs with a PID control system could produce normal human movements using a trial and error method. Another study shows that the implementation using fuzzy logic on a prosthetic hand’s thumb offers flexibility in the hand movement and is similar to the movement of an actual human hand. The real-time implementation of non-linear PID controller was developed to control the servomechanism drive system based on MATLAB/ Simulink In this study, a PID and Fuzzy Logic control system was designed in MATLAB / Simulink based bionic leg to be compared to determine the optimal and accurate bionic leg prototype movements and fast response. The bionic leg’s characteristic is shown in the transient response control. One of the most important requirements for every control system is good transient response. The desired result is

done by reducing signal errors, maximum overshoot, and settling time.

2. METHODOLOGY

At the beginning of the research, a 3D model design as a simulation and visualization of the bionic leg system is needed. The 3D model of the bionic foot system is drawn using the software Inventor-Auto Desk. The measurement of the bionic foot system is matched with one side of the prosthesis. Matlab/Simulink the software provides the facility to export the 3D model in Inventor to the XML file in Matlab/Simulink. To avoid being upset with the design of the robot, it must first be simulated with a control system. Figure 1 shows the bionic robot leg. The bionic robot leg has one degree of freedom with installed a dc motor as an actuator of this robot. The DC motor used is the DC motor of car wiper with 12 Voltage, 85 ± 25 rpm, current ≤ 3 A, and torque 30 kg.cm. Other components used are Arduino Mega 2560, L298N motor driver, Myoware sensor, and potentiometer. DC motors that have been installed as actuators are connected to the L298N motor driver which is connected to the Arduino Mega 2560 and battery. Arduino functions as a microcontroller to adjust the rotating speed and direction of rotation of a DC motor. The L298N motor driver is used so that the DC motor can be adjusted like a servo motor. The potentiometer and Myoware sensor are connected to the Arduino Mega as an analog input. Myoware sensor is used as input of muscle movements and potentiometers are used as feedback and output in the form of angles. The problems are how to increase the power of the EMG signal when the muscle in the fatigue condition. The signal of EMG was collected muscle of biceps to perform flexion and extension.

3. RESULT

Testing is done by interface with Matlab / Simulink on the prototype using the PID control system and Fuzzy Logic control system with a time of 10s at a reference angle of 90o as a comparison. The results obtained in the form of the transient response to the bionic leg prototype motion to get signal error parameters and maximum overshoot of less than 5% and settling time is less than 1s. These parameters will be the point of reference for comparing the two control systems.