Performance Evaluation of a Genetically Tuned LQR Controller for an Aerial Robot

Ali Tahir Karaşahin

Makale detayları

Performance Evaluation of a Genetically Tuned LQR Controller for an Aerial Robot

Yazarlar:

Ali Tahir Karaşahin

Yayıncı:

Necmettin Erbakan University

Ana Başlıklar:

Mekatronik Mühendisliği

Dergi:

Necmettin Erbakan Üniversitesi Fen ve Mühendislik Bilimleri Dergisi Vol.6 Issue 3, pp.468-482

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

Türkçe

Document ID (Org No.):

2001326728

pISSN:

2667-7989

Anahtar kelimeler:

Özet:

 Aerial robots are widely used in many application areas due to their accessibility, simplicity and superior capabilities. Different expectations are required from aerial robots used in a wide range of applications. In order to meet these application-specific expectations, it is necessary to make some changes in the traditional controller designs used in aerial robots. In this study, a linear quadratic regulator LQR controller with genetic tuning and integral effect is developed for trajectory tracking of an aerial robot. LQR and fuzzy proportional-integral-derivative FPID are used to compare the performance of this controller. The developed and used controllers are referred to as model-based and hybrid controllers. Parrot AR. Drone 2.0. is used as an aerial robot. The aerial robot is modeled in MATLABSimulink environment with 6 degrees of freedom DOF. The performances of the developed controllers are evaluated according to the error value during the tracking of the eight curve as a reference. The error values are analyzed according to the root mean square error RMSE criterion. According to the results obtained, the standard LQR controller produced the highest RMSE error value. The LQR controller designed according to genetic tuning and integral effect obtained 43.22 better results than the standard LQR controller and 22.99 better results than the FPID controller. According to the simulation results in MATLABSimulink environment, it is observed that the genetically tuned LQR controller achieves better results in eight curve trajectory tracking.

 Aerial robots are widely used in many application areas due to their accessibility, simplicity and superior capabilities. Different expectations are required from aerial robots used in a wide range of applications. In order to meet these application-specific expectations, it is necessary to make some changes in the traditional controller designs used in aerial robots. In this study, a linear quadratic regulator LQR controller with genetic tuning and integral effect is developed for trajectory tracking of an aerial robot. LQR and fuzzy proportional-integral-derivative FPID are used to compare the performance of this controller. The developed and used controllers are referred to as model-based and hybrid controllers. Parrot AR. Drone 2.0. is used as an aerial robot. The aerial robot is modeled in MATLABSimulink environment with 6 degrees of freedom DOF. The performances of the developed controllers are evaluated according to the error value during the tracking of the eight curve as a reference. The error values are analyzed according to the root mean square error RMSE criterion. According to the results obtained, the standard LQR controller produced the highest RMSE error value. The LQR controller designed according to genetic tuning and integral effect obtained 43.22 better results than the standard LQR controller and 22.99 better results than the FPID controller. According to the simulation results in MATLABSimulink environment, it is observed that the genetically tuned LQR controller achieves better results in eight curve trajectory tracking.

 Aerial robots are widely used in many application areas due to their accessibility, simplicity and superior capabilities. Different expectations are required from aerial robots used in a wide range of applications. In order to meet these application-specific expectations, it is necessary to make some changes in the traditional controller designs used in aerial robots. In this study, a linear quadratic regulator LQR controller with genetic tuning and integral effect is developed for trajectory tracking of an aerial robot. LQR and fuzzy proportional-integral-derivative FPID are used to compare the performance of this controller. The developed and used controllers are referred to as model-based and hybrid controllers. Parrot AR. Drone 2.0. is used as an aerial robot. The aerial robot is modeled in MATLABSimulink environment with 6 degrees of freedom DOF. The performances of the developed controllers are evaluated according to the error value during the tracking of the eight curve as a reference. The error values are analyzed according to the root mean square error RMSE criterion. According to the results obtained, the standard LQR controller produced the highest RMSE error value. The LQR controller designed according to genetic tuning and integral effect obtained 43.22 better results than the standard LQR controller and 22.99 better results than the FPID controller. According to the simulation results in MATLABSimulink environment, it is observed that the genetically tuned LQR controller achieves better results in eight curve trajectory tracking.

 Hava robotları erişebilir, basit ve üstün kabiliyetlerinden dolayı birçok uygulama alanında geniş yer bulmaktadır. Geniş bir yelpazede kullanılan hava robotlarından farklı beklentiler istenmektedir. Uygulama alanlarına özel oluşan bu beklentilere cevap verebilmek için hava robotlarında kullanılan geleneksel kontrolör tasarımlarında birtakım değişikliklere gidilmesi gerekmektedir. Bu çalışmada bir hava robotunun yörünge takibi için genetik ayarlı ve integral etkisine göre çalışan doğrusal karesel regülatör LQR kontrolörü geliştirilmiştir. Bu kontrolörün performansını karşılaştırmak için LQR ve bulanık oransal-integral-türevsel FPID kullanılmıştır. Geliştirilen ve kullanılan kontrolörler model-tabanlı ve hibrit kontrolörler olarak ifade edilmektedir. Hava robotu olarak Parrot AR. Drone 2.0. kullanılmıştır. Belirtilen hava robotu MATLABSimulink ortamında 6 serbestlik derecesi DOF ile modellenmiştir. Geliştirilen kontrolörlerin performansları sekiz eğrisinin referans olarak takip edilmesi esnasında oluşan hata değerine göre değerlendirilmiştir. Hata değerleri ortalama kare hata karekökü RMSE kriterine göre incelenmiştir. Elde edilen sonuçlara göre standart LQR kontrolör en yüksek RMSE hata değerini üretmiştir. Genetik ayarlı ve integral etkisine göre tasarlanan LQR kontrolör, standart LQR kontrolöre göre 43,22 FPID kontrolörüne göre 22,99 oranında daha başarılı sonuçlar elde etmiştir. MATLABSimulink ortamında gerçekleştirilen simülasyon sonuçlarına göre sekiz eğrisi yörünge takibinde genetik ayarlı LQR kontrolörü daha başarılı sonuçlar elde ettiği gözlemlenmiştir.

Makale detayları

Performance Evaluation of a Genetically Tuned LQR Controller for an Aerial Robot

Yazarlar:

Ali Tahir Karaşahin

Yayıncı:

Necmettin Erbakan University

Ana Başlıklar:

Mekatronik Mühendisliği

Dergi:

Necmettin Erbakan Üniversitesi Fen ve Mühendislik Bilimleri Dergisi Vol.6 Issue 3, pp.468-482

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

Türkçe

Document ID (Org No.):

2001326728

pISSN:

2667-7989

Aerial Robot

LQR Controller

Genetic Algorithm

Fuzzy PID

Trajectory TrackingHava robotu

LQR kontrolör

Genetik algoritma

Bulanık PID

Yörünge takibi

Sayı

Makale detayları

Yazarlar: Ali Tahir Karaşahin

Yayıncı: Necmettin Erbakan University

Ana Başlıklar: Mekatronik Mühendisliği

Dergi: Necmettin Erbakan Üniversitesi Fen ve Mühendislik Bilimleri Dergisi Vol.6 Issue 3, pp.468-482

Belge Tipi: Makale

Yayın Tarihi: 12/2024

Belge Dili: Türkçe

Document ID (Org No.): 2001326728

pISSN: 2667-7989

Sayı

Yazarlar:

Ali Tahir Karaşahin

Yayıncı:

Necmettin Erbakan University

Ana Başlıklar:

Mekatronik Mühendisliği

Dergi:

Necmettin Erbakan Üniversitesi Fen ve Mühendislik Bilimleri Dergisi Vol.6 Issue 3, pp.468-482

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

Türkçe

Document ID (Org No.):

2001326728

pISSN:

2667-7989