<![CDATA[Kemacetan lalu lintas di kawasan perkotaan semakin meningkat seiring pertumbuhan jumlah kendaraan yang tidak diimbangi dengan kapasitas infrastruktur jalan. Di Indonesia, sebagian besar persimpangan masih menggunakan sistem pengendalian lampu lalu lintas konvensional berbasis waktu tetap, yang kurang adaptif terhadap fluktuasi volume kendaraan, khususnya pada jam sibuk. Oleh karena itu, diperlukan sistem pengendali lampu lalu lintas adaptif untuk meningkatkan efisiensi arus kendaraan. Penelitian ini membandingkan kinerja algoritma Fixed Priority (FP) dan Earliest Deadline First (EDF) dengan sistem konvensional melalui simulasi pada dua persimpangan yang saling terhubung di Kota Malang. Setiap skenario simulasi dijalankan selama 5 menit menggunakan Unity Engine dengan data volume kendaraan hasil observasi CCTV yang merepresentasikan kondisi lalu lintas nyata dan didominasi kendaraan roda dua. Evaluasi kinerja dilakukan menggunakan metrik average waiting time dan throughput. Hasil pengujian menunjukkan bahwa algoritma FP dan EDF mampu meningkatkan kinerja sistem secara signifikan. Algoritma FP dan EDF menurunkan average waiting time masing-masing sekitar 64,5% dan 65,1% dibandingkan sistem konvensional. Pada metrik throughput, FP meningkatkan kinerja sekitar 8,18%, sedangkan EDF mencapai peningkatan tertinggi sebesar sekitar 27,8%. Hasil tersebut menunjukkan bahwa algoritma EDF memberikan peningkatan throughput tertinggi, sedangkan kedua algoritma adaptif mampu menurunkan average waiting time dibandingkan sistem konvensional. Abstract Urban traffic congestion is getting worse, as the number of vehicles grows faster than road infrastructure capacity. In Indonesia, most intersections still employ conventional fixed-time traffic light control systems, which are less adaptive to fluctuations in traffic volume, particularly during peak hours. Therefore, an adaptive traffic light control system is required to improve traffic flow efficiency. This study compares the performance of the Fixed Priority (FP) and Earliest Deadline First (EDF) algorithms with a conventional system through a simulation of two connected intersections in Malang City. Each simulation scenario was conducted for 5 minutes using Unity Engine, with vehicle volume data obtained from CCTV observations to represent real traffic conditions dominated by two-wheeled vehicles. System performance was evaluated using two main metrics, namely average waiting time and throughput. The results show that both FP and EDF significantly improve system performance. The FP and EDF algorithms reduce average waiting time by approximately 64.5% and 65.1%, respectively, compared to the conventional system. In terms of throughput, FP improves performance by about 8.18%, while EDF achieves the highest improvement of approximately 27.8%. These results indicate that EDF provides the highest throughput improvement, while both adaptive algorithms are able to reduce average waiting time compared to the conventional system.]]>
Copyrights © 2026
