Autonomous vehicle is a vehicle that can automatically navigate both long-range and short-range navigation. In this paper, we designed a motion and navigation control system of an autonomous vehicle that is implemented on a mobile robot. The remote navigation process involves using the global positioning system (GPS) sensor as a reference to navigate, and for short-range navigation using the proximity and force sensor. The robot task is to reach several coordinate points that are determined and marked with a traffic-cone at each point. In the control of short-range navigation, the robot uses five ultrasonic and three switch sensors to read the environment as input to the description of the distance between the robot and the object. The micro-controller then compares it to the robot’s safe distance instructions in navigating with surrounding objects. GPS sensor is used to provide remote navigation path information. To compensate for the accuracy of the wide GPS sensor readings, several waypoints are designed to guide the robot to the specified traffic-cone. From the implementation, the robot movement can go along the navigation area and can avoid short-range objects in the robot trajectory and get to the traffic-cone point with the percentage of success reaching the final destination above 86% both in sunny and cloudy weather.
Applying a speed control system to a mobile robot needs to be done to make the robot move stably in an uneven environment with the variation of load. This study purpose is to designed a speed control system for a mobile robot to follow setpoints with a fast response to reach a predetermined speed. The speed controller used the fuzzy logic control method. The input is a speed error and the change in speed error is known from an encoder sensor and the output of the system is pulse width modulation (PWM) in the range of 8-bit converted to fuzzy forms. The rule formation and rule analysis with its implication function use the MIN function and defuzzification use the weighted average method. By implementing this fuzzy control the results of the robot move at the speed corresponding to the setpoint both on flat or uneven terrain, with load or without load. In experiments carried out on robots with a setpoint of 9,000 rpm from rest, the response time (in rise-time) is less than 47ms and steady-state error close to zero.
A long-range and short-range navigation systems of an autonomous vehicle are the most important thing to develop because it is very complex and related to safety in the driving of an automatic vehicle. The purpose of this research is to develop a short-range navigation system for autonomous vehicles focused to detect and avoid humans which implemented on an autonomous vehicle prototype. Image Processing and deep learning algorithm used for human detection and ultrasonic sensor used for distance calculation. Decision making for human avoidance system using the Fuzzy algorithm based on the position of detected human and human distance. The implemented system has a success rate of 85.71% to avoid human which ideal distance more than 2 meters. The implemented prototype of an autonomous vehicle can be implemented in the real vehicle with some adjustments in the sensor and hardware system, however, the systems are expected to reduce traffic accidents caused by human error.
Peningkatan jumlah kendaraan terjadi pada semua jenis kendaraan setiap tahunnya. Hal ini akan menimbulkan kemacetan yang memiliki dampak negatif. Kemacetan tersebut dapat dikurangi dengan melakukan rekayasa lalu-lintas. Sistem yang dibangun dapat mengambil data kondisi lalu-lintas seperti jumlah dan kecepatan kendaraan menggunakan kamera dan pengolahan citra. Dengan memanfaatkan vanishing point, sebuah gambar yang memiliki distorsi perspektif dapat direktifikasi, Dengan menghitung jarak perpindahan kendaraan setiap frame nya pada gambar rektifikasi, maka kecepatan kendaraan dapat terdeteksi. Hasil estimasi kecepatan kendaraan oleh sistem dibandingkan dengan hasil pengamatan langsung memiliki selisih sebesar 8.22 km/jam.
Kata Kunci - Estimasi Kecepatan Kendaraan; Pengolahan Citra; Rektifikasi Gambar; Vanishing Point;
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.