VOL. 10, NO. 1, JANUARY 2015 ISSN 1819-6608 ARPN Journal of Engineering and Applied Sciences © 2006-2015 Asian Research Publishing Network (ARPN). All rights reserved. www.arpnjournals.com 351 STEERING CONTROL METHOD BASED ON TSL1401 LINEAR SENSOR ARRAY Mohamad Taib Miskon, Ahmad Shahran Ibrahim, Zairi Ismael Rizman and Nuraiza Ismail Faculty of Electrical Engineering, Universiti Teknologi MARA (Terengganu), Dungun, Terengganu, Malaysia E-Mail: zairi576@tganu.uitm.edu.my ABSTRACT The development of an automatic steering control system has been a major concern for most researchers towards the realization of a fully automated car in the near future. It is believed that such system could mitigate the effect of road accidents caused by human errors. Therefore, this project was carried out to develop a prototype of an intelligent car that has the capability to navigate automatically without human interference. A Freescale cup development kit was utilized in this project that consists of a development board, a servo motor, two DC drives, a driver circuit and a vision sensor. A special test track was developed using a white material with black lane along both edges of the track. A vision system, based on TSL1401 Linear Array Sensor, with 128pixels of detection resolution was developed to sense the track condition. The data captured by the sensor was sent to a Freescale FRDM-KL25Z processor and the steering angle of the car was determined. The programming code was written in C language using Freescale CodeWarrior software platform. A steering control method was introduced to navigate the prototype car autonomously. The method utilizes the linear sensor as it inputs parameters to identify current position of the car and determines the output parameters that dictate the car behavior. As a result, the car managed to steer automatically in various route conditions such as straight path, ramp, junction and also sharp turn. Keywords: steering control; automated car; FRDM-KL25Z; ARM; TSL1401 sensor. INTRODUCTION Death tolls caused by road accidents are alarmingly reaching high levels in recent years. Thus, the development of a smarter car has been a major concern in most country. Steering control is one of the basic elements in an automated car. One of the common techniques used in the navigation systems for automated car are based on the line-tracking method. Many researchers have proposed the line-detection using vision system to navigate an autonomous car (Shubra Deb Das et al., 2013), (Guo-Zu et al., 2011). A vision sensor has much spatial and optical resolutions compared to the conventional method (Norhashim and Noorfadzli, 2012). With the rapid of microelectronics technology, an intelligent car motion control system described by Xi et al. (2011) based on single-chip AT89C52 microcontroller. They proposed a system that has multi- sensor fusion technology, high flexibility, equipped with automatic obstacle avoidance and route tracing. Jian and Guangzhong (2009) and Qu et al. (2012) design a fuzzy-based steering controller by processing the route information using a miniature Charged-Coupled Device (CCD) camera. It helps the car model to run smoothly on a given raceway including ‘S’ curve and intersection at a speed of approximately 2m/s. Besides, route identification technique based on Infra- Red (IR) sensor array is implemented by Singh et al. (2010). Utilizing proportional-integral-derivative (PID) control method, the car captures and process information in real time, identifies track condition and optimizes the actual performance of the navigation (Singh et al. (2010) ( Shuan Chen et al., 2013). Median filtering algorithm with T-shaped window is proposed by Ning et al. (2011) to optimize the route identification capability and reduce unwanted noise from the camera sensor. This technique improves the edge detection by filtering noise and smoothing the whole image. However, the use of CCD camera as part of the navigation approach requires complicated data processing and also very vulnerable to external light interference (Junhua et al., 2010) (Zhang et al. 2014). This paper propose the use of TSL1401CL linear sensor as the smart car vision system and also discusses a method for steering angle and speed control calculation in order to perform autonomous navigation VISION SYSTEM AND CONTROL APPROACH Figure-1 shows the arrangement of the vision system used in this paper. A linear sensor array based on TSL1401CL was chosen in this project as it offers less pixel count than other types of sensor. Table-1 illustrates the pixel comparison between three different sensor modules.TSL1401CL linear sensor consist of 128x1 pixel array of linear image sensor with focusable 7.9mm lens. Each pixel contains a photodiode, charge amplifier circuitry, and an internal pixel data-hold function. Besides, the sensor also comes with a five way printed circuit board (PCB) connector for interfacing purposes including one analog pixel output. Compared to the sensor used by Jian and Guangzhong (2009) and Xi et al. (2011), this sensor is simpler and easy to use. Moreover,