1. UKM Learning and Research Repository
2. PHD Thesis / Tesis PHD
3. Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Title:	Development of a multiple model adaptive mechatronic suspension control system
Authors:	Wajdi Sadik Aboud (P54857)
Supervisor:	Sallehuddin Mohamed Haris, Dr.
Keywords:	Model adaptive mechatronic Mechatronics
Issue Date:	22-Jun-2014
Description:	Semasa beroperasi sesebuah kenderaan tertakluk kepada perubahan dinamik sistem secara perlahan dan mendadak. Ini mungkin berpunca daripada perubahan bebanan terpegas yang disebabkan oleh variasi jumlah penumpang dan bebanan dan daripada kehilangan komponen badan, contohnya, disebabkan kemalangan. Permukaan jalan yang memberikan input gangguan juga boleh berubah. Oleh itu sistem ampaian perlu menunjukkan prestasi yang boleh diterima di bawah keadaan yang berubah sebegini bagi memastikan keselamatan dan keselesaan penumpang. Objektif kajian ialah untuk mereka bentuk satu skim kawalan untuk sistem ampaian aktif mekatronik yang cukup lasak terhadap perubahan pada kedua bebanan terpegas dan profil permukaan jalan, serta terhadap ketidaktentuan dan tak linear yang wujud pada sistem. Sistem ini juga perlu meminimumkan penggunaan kuasa agar ia boleh direalisasikan secara praktikal. Sebuah model ampaian suku kereta dengan tiga bebanan terpegas yang berbeza telah digunakan sebagai kes ujian. Bebanan terpegas ini adalah 280 kg, 330 kg dan 380 kg masingmasing mewakili kes tiada beban, separuh beban dan beban penuh. Penggerak elektromagnet yang memberikan tindakan kawalan aktif turut dimasukkan ke dalam model ampaian aktif ini. Pecutan bebanan terpegas, daya tayar dan anjakan ampaian telah digunakan sebagai petunjuk prestasi. Sebagai rujukan, sambutan masa dan frekuensi suatu ampaian pasif telah digunakan sebagai garis dasar untuk perbandingan. Pertamanya, pengawal ampaian aktif optimum Kadaran-Kamiran-Derivatif (PID), Pengatur Linear Kuadratik (LQR) dan Linear Kuadratik Gauss (LQG) telah direka dan dianalisis. Ini diikuti dengan Pengawal Suai Penalaan-kendiri dan Pengawal Lasak Sintesis-m. Didapati bahawa pengawal Sintesis-m memberikan keputusan yang terbaik, dengan hampir tiada perbezaan prestasi apabila tertakluk kepada kes bebanan yang berbeza. Satu skim kawalan yang novel berdasarkan skim kawalan suai pelbagai model dengan input kawalan berpemberat bercampur (WMMAC) juga telah dibangunkan dan diuji. Skim baru ini telah direka untuk memberikan penyesuaian pengawal secara cepat untuk variasi pada kedua, bebanan terpegas dan input gangguan profil jalan. Empat calon model nominal telah dipilih secara a priori untuk mewakili kombinasi bebanan terpegas dan profil permukaan jalan yang berbeza. Bagi setiap calon model, satu pengawal talaan optimum telah direka bentuk. Dua variasi skim WMMAC telah dicadangkan. Dalam variasi yang pertama, calon model terdiri daripada kombinasi bebanan terpegas dan, profil jalan, sama ada berbentuk bonggol atau input langkah, menggunakan suap balik output dengan kawalan PID. Dalam variasi kedua, calon model terdiri daripada kombinasi bebanan terpegas dan input profil jalan raya stokastik, dengan pengawal suap balik keadaan LQG. Semasa kenderaan dikendalikan, keadaan operasi dikenal pasti dan dipadankan dengan calon model. Pemberat diberikan kepada setiap calon pengawal mengikut berapa hampirnya calon model yang sepadan berpadanan dengan sistem yang dikenal pasti. Jumlah input kawalan berpemberat daripada calon pengawal ini kemudiannya digunakan sebagai input kawalan keseluruhan kepada sistem. Daripada simulasi yang dijalankan, didapati bahawa skim kawalan yang dicadangkan mengatasi prestasi pengawal Sintesis-m sebanyak 10% untuk pecutan bebanan terpegas dan sebanyak 23% untuk daya tayar. Untuk anjakan ampaian, walaupun pengawal Sintesis-m memberikan prestasi yang lebih baik, prestasi WMMAC masih berada dalam had yang dibenarkan. Didapati juga bahawa keperluan kuasa skim WMMAC adalah 35% kurang berbanding pengawal Sintesis-m.,Great advancements in mechatronic devices in recent times have benefited active suspension technology. During operation, a vehicle is subjected to slow and abrupt changes in its system dynamics. This may be due to changes in the sprung mass caused by variations in the number of passengers and loads in the boot, and from the loss of body parts, for example due to an accident. Road surfaces, which impart perturbation inputs, could also vary considerably. Therefore, suspension systems need to provide acceptable performance under such changing conditions to ensure the safety and comfort of passengers. In this thesis, the objective was to design a control scheme for an active mechatronic suspension that is sufficiently robust to variations in both the sprung mass loads and road surface profiles, as well as uncertainties and nonlinearities inherent in the system. The system should also minimise power consumption so as to be practically realisable. A quarter car suspension model with three different sprung mass loads was used as the test case. The sprung mass loads were 280 kg, 330 kg and 380 kg to represent no load, half load, and full load cases respectively. An electromagnetic actuator was incorporated into the model to provide active control action. Sprung mass acceleration, tyre force and suspension deflection were used as performance indicators. As a reference, the time and frequency responses of a passive suspension was used as a baseline for comparison. Firstly, optimal Proportional-Integral-Derivative (PID), Linear Quadratic Regulator (LQR), Linear Quadratic Gaussian (LQG) active suspension controllers were designed and analysed. This was followed by Self-tuning Adaptive and robust m-synthesis controllers. It was found that the m-synthesis controllers produced the best results, with almost no performance variation when subjected to different load cases. A novel control scheme, based on multiple model adaptive control with mixed weighted control inputs (WMMAC) was also developed and tested. The new scheme was designed to provide fast controller adaptation to variations in both sprung mass loads and road profile perturbations. Four nominal candidate models were chosen a priori to represent different combinations of sprung mass loads and road profiles. For each candidate model, an optimally tuned controller was designed. Two variations of the WMMAC scheme was proposed. In the first, the candidate models comprised of different combinations of sprung mass loads and either bump or step road profile inputs, and output feedback PID controllers were used. In the second, the candidate models consisted of different combinations of sprung mass loads and stochastic road profile inputs, with state feedback LQG controllers. During vehicle operation, the operating conditions are identified and matched to the candidate models. Weights are given to each candidate controller according to how closely matched the corresponding candidate model is to the identified system. The weighted sum of control inputs from the candidate controllers is then used as the overall control input into the system. From simulations performed, it was found that the proposed control scheme produced improved performance over the m-synthesis controller by 10% for sprung mass acceleration and by 23% for tyre force. On suspension deflection, although the m-synthesis system performed better, the WMMAC performance still stayed well within the allowed limits. It was also found that the WMMAC scheme power requirement was 35% less than that of the m-synthesis controller.,PhD
Pages:	200
Call Number:	TJ163.12.A246 2014 3 tesis
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.