*Corresponding author, e-mail:senos@itu.edu.tr Research Article GU J Sci 34(2): 517-527 (2021) DOI: 10.35378/gujs.762103 Gazi University Journal of Science http://dergipark.gov.tr/gujs Particle Swarm Optimization Method Based Controller Tuning for Adaptive Cruise Control Application Erhan OZKAYA 1 , Hikmet ARSLAN 2 , Osman Taha SEN 2,* 1 AVL Turkey Research and Engineering, 34885, Istanbul, Turkey 2 Istanbul Technical University, Department of Mechanical Engineering, 34437, Istanbul, Turkey Highlights • PID parameter tuning with PSO technique. • ACC system model embedded with PSO. • Superiority of PSO method based PID parameter tuning process. Article Info Abstract Major developments in relevant technology make the advanced driver assistance systems and autonomous driving functions more attainable. Thus, conventional practices being applied in vehicle production evolves towards highly automated, safer, and more comfortable vehicles. Although advanced driver assistance systems and autonomous driving functions have many advantages, such as enhanced driver convenience, increased comfort, and fuel economy; concerns related to safety still exist. For instance, failures related to sensors or algorithms being used can lead to critical problems. Therefore, controller algorithms should be more robust and well- optimized in order to eliminate these safety issues. This requires the implementation of automated optimization algorithms for the controller parameter tuning process. The main objective of this study is to optimize the designed controller for an adaptive cruise control system by using the particle swarm optimization method, which is a swarm intelligence optimization technique. Based on the results, it is observed that the use of automated optimization techniques for adaptive cruise control systems leads to better accuracy and safety for driving control. Furthermore, the time consumed for the controller parameter tuning process is also decreased significantly. In conclusion, the adaptive cruise control system requirements can be easily and accurately ensured by the use of particle swarm optimization algorithm. Received: 01 July 2020 Accepted: 03 Nov 2020 Keywords ADAS ACC PSO 1. INTRODUCTION According to the data provided by Turkish National Police Academy, the number of accidents that involves road vehicles is about 10 million in Turkey during the last decade [1]. Furthermore, driver fault appears as the major cause of these accidents as evident from Figure 1, where approximately 90% of accidents is caused due to driver faults. In order to minimize the effect of driver fault, enhance driving comfort and increase fuel efficiency; advanced driver assistance system (ADAS) features are being developed and integrated to modern vehicles in an increasing trend [2]. For example, Gürbüz and Buyruk [3] propose a new model that is used to calculate the safe stopping distance of a vehicle by considering factors due to driver, vehicle and environment. Authors suggest that this information can be displayed on a screen to driver as an assistant for enhancing safety [3]. Nevertheless, safety related concerns for ADAS features still persist. To avoid unexpected behaviors and eliminate these concerns for autonomous features, ADAS algorithms must be generated with robust logics in order to overcome all kinds of traffic scenarios in real life. Hence, optimization of algorithms becomes an important task which must be carefully undertaken.