Düşük Maliyetli Otonom Traktör Dönüşümüne Yönelik Pure Pursuit Tabanlı Yol Takip Algoritmasının ROS 2/Gazebo Simülasyon Çalışması

Öz

Bu çalışmada, konvansiyonel bir tarım traktörünün düşük maliyetli otonom dönüşümü için Pure Pursuit tabanlı bir yol takip algoritmasının iki katmanlı bir yöntemle geliştirilmesi ve doğrulanması sunulmaktadır. Türkiye'de yaygın olarak kullanılan, seri üretim orta sınıf bir tarım traktörünün kinematik parametreleri referans alınmıştır. Birinci katmanda, kinematik bisiklet modeline dayalı bir Python simülasyonu kontrolcü tasarımı ve grid arama ile parametre optimizasyonu (k₁ = 1,3 m; k₂ = 0,7 s) için kullanılmış; ikinci katmanda, optimize edilen kontrolcü ROS 2 Humble ve Gazebo Classic 11 tam fizik ortamında, üç bozucu senaryosunun her biri bağımsız rastgele tohumlarla n = 10 kez tekrarlanacak şekilde dört senaryo altında istatistiksel olarak doğrulanmıştır. Geliştirilen algoritma; dinamik lookahead mesafesi, kademeli headland yavaşlatma ve rate-limited direksiyon kontrolü içermektedir. Tam fizik simülasyonunda düz hat segmentlerinde ortalama çapraz hat hatası (CTE) ideal koşullarda 0,055 m, RTK-GPS senaryosunda 0,085 ± 0,023 m, standart GPS senaryosunda ise 0,446 ± 0,011 m olarak ölçülmüştür. 0,15 s'lik aktüatör gecikmesi, standart GPS koşullarında CTE'yi istatistiksel olarak anlamlı ancak sınırlı ölçüde artırmıştır (%5,6; p = 0,006). Elde edilen bulgular, düşük hızlı tarımsal operasyonlarda algoritma performansının büyük ölçüde konumlama doğruluğuyla sınırlandığını ve RTK düzeyinde konumlamanın hassas tarım uygulamaları için bir temel oluşturabileceğini düşündürmektedir. 

Anahtar Kelimeler

• Otonom traktör
• Pure Pursuit
• ROS 2
• Gazebo
• kinematik bisiklet modeli
• yol takip kontrolü

Teşekkür

Yazar, katkılarından dolayı açık kaynaklı ROS 2 ve Gazebo topluluklarına teşekkür eder.

ROS 2/Gazebo Simulation Study of a Pure Pursuit-Based Path-Following Algorithm Toward Low-Cost Autonomous Tractor Conversion

Öz

This study presents the development and validation of a Pure Pursuit-based path tracking algorithm for the low-cost autonomous conversion of a conventional agricultural tractor, using a two-layer methodology. The kinematic parameters of a mass-produced mid-size agricultural tractor widely used in Türkiye served as the reference. In the first layer, a Python simulation based on the kinematic bicycle model was used for controller design and grid-search parameter optimization (k₁ = 1.3 m, k₂ = 0.7 s). In the second layer, the optimized controller was statistically validated in the ROS 2 Humble / Gazebo Classic 11 full-physics environment under four scenarios, with each of the three disturbance scenarios repeated n = 10 times with independent random seeds. The algorithm incorporates dynamic lookahead distance, gradual headland deceleration, and rate-limited steering control. In the full-physics simulation, the mean straight line cross-track error (CTE) was 0.055 m under ideal conditions, 0.085 ± 0.023 m in the RTK GPS scenario, and 0.446 ± 0.011 m in the standard GPS scenario. A 0.15 s actuator delay produced a statistically significant, but bounded CTE increase under standard GPS conditions (5.6%, p = 0.006). The results suggest that algorithm performance appears to be largely bounded by localization accuracy at low operating speeds, and that RTK-grade positioning provides a basis for precision agriculture applications.

Anahtar Kelimeler

• Autonomous tractor
• Pure Pursuit
• ROS 2
• Gazebo
• kinematic bicycle model
• path tracking control

Teşekkür

The author thanks the open-source ROS 2 and Gazebo communities for their contributions.

Kaynakça

1. [1] TÜİK. Crop Production Statistics and Agricultural Equipment Statistics. Turkish Statistical Institute, Ankara; 2024.
2. [2] Qu J, Zhang Z, Qin Z, Guo K, Li D. Applications of autonomous navigation technologies for unmanned agricultural tractors: A review. Machines. 2024; 12: 218.
3. [3] Xu R, Li C. A modular agricultural robotic system (MARS) for precision farming: Concept and implementation. Journal of Field Robotics. 2022; 39(4): 387–409.
4. [4] Wang Q, He J, Lu C, Wang C, Lin H, Yang H, Li H, Wu Z. Modelling and control methods in path tracking control for autonomous agricultural vehicles: A review of state of the art and challenges. Applied Sciences. 2023; 13: 7155.
5. [5] Coulter RC. Implementation of the Pure Pursuit Path Tracking Algorithm. Technical Report CMU-RI-TR-92-01, Carnegie Mellon University; 1992.
6. [6] Koenig N, Howard A. Design and use paradigms for Gazebo, an open-source multi-robot simulator. IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). 2004; 3: 2149–2154.
7. [7] Hung N, Rego F, Quintas J, Cruz J, Jacinto M, Souto D, Potes A, Sebastiao L, Pascoal A. A review of path following control strategies for autonomous robotic vehicles: Theory, simulations, and experiments. Journal of Field Robotics. 2023; 40(3): 747–779.
8. [8] Liu L, Xue M, Guo N, Wang Z, Wang Y, Tang Q. Investigating the path tracking algorithm based on BP neural network. Sensors. 2023; 23: 4533.

9. [9] Willekens A, Temmerman S, Wyffels F, Pieters J, Cool S. Development of an agricultural robot taskmap operation framework. Journal of Field Robotics. 2025; 42(8): 4256–4287.
10. [10] Martelli S, Mocera F, Somà A. Autonomous driving strategy for a specialized four-wheel differential-drive agricultural rover. AgriEngineering. 2024; 6(3): 1937–1958.
11. [11] Huang P, Luo X, Zhang Z. Headland turning control method simulation of autonomous agricultural machine based on improved Pure Pursuit model. IFIP Advances in Information and Communication Technology. 2010; 317: 176–184.
12. [12] Huang P, Zhang Z, Luo X. Feedforward-plus-proportional-integral-derivative controller for agricultural robot turning in headland. International Journal of Advanced Robotic Systems. 2020; 17(1): 1729881419897678.
13. [13] Peng C, Wei P, Fei Z, Zhu Y, Vougioukas S. Optimization-based motion planning for autonomous agricultural vehicles turning in constrained headlands. Journal of Field Robotics. 2024; 41(6): 1984–2008.
14. [14] Huang P, Zhu L, Zhang Z, Yang C. Row end detection and headland turning control for an autonomous banana-picking robot. Machines. 2021; 9: 103.
15. [15] Miao H, Diao P, Xu G, Yao W, Song Z, Wang W. Research on decoupling control for the longitudinal and lateral dynamics of a tractor considering steering delay. Scientific Reports. 2022; 12: 13997.
16. [16] Çiloğlu Ç, Kutluay E. Performance improvement of Pure Pursuit algorithm via online slip estimation for off-road tracked vehicle. Sensors. 2025; 25: 4242.
17. [17] Yan J, Zhang W, Liu Y, Pan W, Hou X, Liu Z. Autonomous trajectory tracking control method for an agricultural robotic vehicle. International Journal of Agricultural and Biological Engineering. 2024; 17(1): 215–224.
18. [18] Gerdan Koç D, Vatandaş M. Development and performance analysis of an autonomous agricultural vehicle for fruit transportation. Journal of Field Robotics. 2025; 42(7): 3189–3212.
19. [19] Wang R, Li Y, Fan J, Wang T, Chen X. A novel Pure Pursuit algorithm for autonomous vehicles based on salp swarm algorithm and velocity controller. IEEE Access. 2020; 8: 166525–166540.
20. [20] Wang A, Wang Y, Ji X, Wang K, Qian M, Wei X, Song Q, Chen W, Zhang S. Fuzzy backstepping controller for agricultural tractor-trailer vehicles path tracking control with experimental validation. Frontiers in Plant Science. 2024; 15: 1513544.
21. [21] Bayar G, Bergerman M, Koku AB, Konukseven Eİ. Localization and control of an autonomous orchard vehicle. Computers and Electronics in Agriculture. 2015; 115: 118–128.