Bir robot manipülatör için adaptif model izleme kontrol sistemi
An Adaptive model following control system for robotic manipulators
- Tez No: 19323
- Danışmanlar: DOÇ.DR. FUAT GÜRLEYEN
- Tez Türü: Yüksek Lisans
- Konular: Makine Mühendisliği, Mechanical Engineering
- Anahtar Kelimeler: Belirtilmemiş.
- Yıl: 1991
- Dil: Türkçe
- Üniversite: İstanbul Teknik Üniversitesi
- Enstitü: Fen Bilimleri Enstitüsü
- Ana Bilim Dalı: Belirtilmemiş.
- Bilim Dalı: Belirtilmemiş.
- Sayfa Sayısı: 175
Özet
ÖZET Bir robot manipülatör yüksek hızlarda çalıştığı zaman, doğrusal olmamanın getirdiği etkiler, zamanla-değişen katsayılar ve boşluk, sürtünme gibi ilave belirsizlikler sistem hatasının büyümesinde etkili olurlar. Bu dış etkenli giriş veya bozucuların istenmeyen etkileri kompanze edilmek zorundadır.Kontrol edilen sistem parametreleri ya yetersiz derecede bilindiği ya da normal çalışma esnasında değişme gösterdiği zaman yüksek performanslı kontrol sistemlerinin tasarımı genel olarak adaptif kontrol tekniklerinin kullanımını gerektirir. Çeşitli alternatif metodlar arasında Model Referans Adaptif Sistem (MRAS) olarak bilinen tekniğin kullanımı, adaptif kontrol sistemlerinin gerçeklenmesinde mümkün olan en uygun yaklaşım olarak karşımıza çıkar. Böyle bir adaptif sistemin en önemli özelliği çeşitli yapılar altında ortaya çıkabilen ve sistemin bir parçası olarak düşünülen bir referans modelin varlığıdır. Bu çalışma, MRAS tekniğinin kullanıldığı Adaptif Model İzleme Kontrol (AMİK) probleminin iki serbestlik dereceli bir düzlemsel manipülatöre uygulanarak, taşıyabileceği maksimum yükteki büyük değişikliklere karşı manipülatörün yüksek performansını koruyabilmesini sağlamak üzere bir kontrol sistemi tasarlamayı öngörmektedir. Bu tekniğin diğer metodlar üzerindeki başlıca üstünlükleri; günümüz mikroişlemcileri ile gerçek-zamanda kontrol kuralını hesaplayabilme yeteneği sağlayan hesaplama kolaylığı ile ayrı bir kararlılık analizi gerektirmeden tüm sistemin kararlılığını otomatik olarak garanti etmesi şeklinde ifade edilebilir. Bu metodun yeteneğini göstermek üzere, iki dönel eklemli bir düzlemsel manipülatör için bir bilgisayar simülasyonu gerçekleştirilmiş ve simülasyon sonuçlarının kontrol sisteminin beklenen yüksek performansı ile uyuştuğu gözlenmiştir. IV
Özet (Çeviri)
SUMMARY AN ADAPTIVE MODEL FOLLOWING CONTROL SYSTEM FOR ROBOTIC MANIPULATORS. When a robot manipulator operates at high speed, the effects of nonli- nearities, time-varying coefficients pnd additional uncertainties such as backlash,fractlon, etc., may cause the system error to become large. These unwanted effects of various external inputs or disturbances must somehow be compensated[1].The design of high performance control systems generally requires the use of adaptive control techniques when the parameters of the controlled process either are poorly known or vary during normal operation. Among various alternative methods, the use of the technique known as model reference adaptive system (MRAS's) seems to be one of the most feasible approaches possible for the implementation of adaptive control systems. The prime characteristic of such an adaptive system is the presence of a so-called reference model as part of the system, which can appear under various forms. The basic control problems for which model reference adaptive systems might be used are (a) adaptive model-following control, (b) on-line and realtime parameter identification, and (c) adaptive state observation. In the first case, a reference model specifies the desired control performance. In the second and third cases, the process whose parameters are to be identified or whose states are to be observed represents the reference model [2], For a system with adjustable parameters the model reference adaptive method gives a general approach for adjusting the parameters so that the closed-loop transfer function will be close to a prescribed model. This is called the model-following problem. One important question is how small we can make the error e. This depends both on the model, the system, and the command signal. If it is possible to make the error equal to zero for all command signals, then perfect model-following is achieved. Model following is a simple and useful way to formulate and solve a servo control problem. The basic idea is very simple. Servo performance is specified indirectly by giving a mathematical model for the desired response. The specified model can be linear as well as nonlinear. The parameters in the system are adjusted in order to getyas close as possible to Xm for a given class of input signals. Although perfect model-following can only be obtained in idealizedsituations, an analysis of this case gives good insight into the design problem [3]- Using the adaptive model following control (AMFC) technique a control system in this study is developed to obtain high performance of a robot arm with two degrees of freedom that has wide variations in its payload and spatial configuration. The main advantage of this technique over other methods is simplicity in computation, which provides the capability of real time control calculation with today's microprocessors. Moreover, it guarantees automatically the stability of the overall.system without requring any additional stability analysis [4], In order to demonstrate the capability of this technique, I have conducted a computer simulation for a two-link planar manipulator with revolute joints. This simulation results agree well with the expected high perfor mance of the control system. A control scheme considered here provides automatic system stability regardless of variations in payload and spatial configuration. The control scheme has an advantage over [5] in that the stability of the overall system can be assured automatically using the hyperstability and positivity concepts. Thus one can eliminate the need of separate study of the system stability. The reference model in the present scheme is a general second-order linear system which may be more suitable than the pure double integrator for specifying the desired manipulator performance such as rise time, overshoot, damping, etc. Note that the manipulator dynamics are represented as a nonlinear second-order system [4]. The firs step in the development of any manipulator control system is to derive an analytical model of manipulator dynamics. The dynamics consist of a set of differential, highly coupled nonlinear and matrix oriented repre senta- tions which describe the behaviour of the robot arm. Various robot dynamic formulations have been proposed during the past few years. The lagrange-Euler (LE) method has low computational efficiency, but a very well structured and systematic representation that allows fordifferent control applications. The recursive Lagrangian (RL) method has better computation time, but destroys the structure of the equations. The Newton-Euler (NE) method has very efficient computational representation with a set of highly recursive equations. The importance and usefulness of the LE methods arise from its simple, algorithmic and highly structured equation based on lagrangian mechanics[6]. Using the Lagrangian mechanics, one can derive explicit dynamic equations which represent generalized forces/torques in terms of the joint positions, velocities, and accelerations: VID(q)q+ h (q,q) + g(q)=u (1) where u is an nxl generalized force/torque vector and q,q,*q, are vectors of generalized coordinates, velocities, and accelerations, respec tively. D(q) is an nxn inertial matrix, h(q,q) is an nxl Coriolis and centrifugal force vector, g(q) is an nxl gravitational loading vector, and n is the number of joints in the manipulator, The inertia, the gravity loading, and the Coriolis and centrifugal terms depend on the position of each joint as well as on the mass, the first moment and the inertia of each link. Also note that these terms are functions of manipulator payload (i.e., tool and parts). The inertia and the gravity terms are dominant at slower operating speeds. At faster speeds the Coriolis and centifugal effects become significant, in general, the design of manipulator controllers has been hampered by the inherent nonlinearity and the joint couplings as evidenced in (1). The dynamics equation (1) can be converted into a state-variable representation with 2n-dimensional state vector y= (ypT yvT) = [aj âJ]T (T denotes transpose) as follows: y= Apy+ Bp (y) u+ Cp(y) (2) where AP= rO 0 \) Bp(y)= [ D- 0 Cp(y> - U- (y)[c(y)+h(y)] These equations show that the manipulator dynamis are highly nonlinear coupled functions of positions and velocities of the manipulator joints. The nonlinear coupled characteristics of the manipulator dynamics cause the design of any controller to be complex and computationally demanding, thereby making its real - time impl ementation extremely difficult if not impos sible. To overcome this difficulty without appealing to complex controllers, the adaptive model following control (AMFC) method [2] has been adopted in this study. Considering the manipulator dynamics (1), the reference model is chosen to be a set of n uncoupled linear time-invariant second-order systems (i.e, one for each of n joints), which are described as follows: VIIXj+2 Ç| coj Xj + û>2i Xj = co2,n for i=l, 2 n (3) where for each joint i, *i Is the position produced by the reference model (i.e., the desired position), r, is the command input to the reference, Çi is the damping ratio, and co;]s the natural frequency. Note that the order and the structure of the reference model are the same as those of the manipulator dynamics in (1). The reference model can be regarded as a means of supplying the manipulator positions and velocities with desired characteristic of rise time, overshoot, and damping. Then the AMFC problem is to design an adaptive controller that drives the manipulator to follow the desired positions and velocities generated by the reference model as closely as possible. In such a case, the reference model can be considered as a trajectory planner that provides the manipulator controller with information on a preplanned path and its timing. Hence we can apply command inputs r, to the reference model so that the reference model generates the desired positions and velocities. The reference model equations can be converted into a state variables form as follows: x=Amx+Bmr (4) where xn + i = x for i= l,2,...,n X = *i x2n Am= [ -Cû2I -2 Ceo I BT=
Benzer Tezler
- Adaptive position control of permanent magnet synchronous motors and torque ripple compensation
Başlık çevirisi yok
O.SETA BOĞOSYAN
Doktora
İngilizce
1991
Elektrik ve Elektronik Mühendisliğiİstanbul Teknik ÜniversitesiPROF.DR. TAMER KUTMAN
- Robot manipülatör kontrolünde parametre belirsizliklerine yönelik kontrol yaklaşımları
New approaches in controlling robot manipulators with parametric uncertainty
RECEP BURKAN
Doktora
Türkçe
2002
Makine MühendisliğiErciyes ÜniversitesiMakine Mühendisliği Ana Bilim Dalı
PROF. DR. İBRAHİM UZMAY
- Manipulatör kontrolü için çok değişkenli adaptif PID regülatör
Multivariable adaptive pid regulatör for manipulator control
A.SELÇUK TEKDEMİR
- Bir robotik manipülatörün eklem ve kartezyen esaslı öngörülü kontrolu
Joint and cartesian based predictive control of a robotic manipulator
RECEP KAZAN
- Esnek imalat sistemlerinde üç serbestlikli manipülatör tasarımı ve kontrolü
Design and control of a three link manipulator in flexible manufacturing systems
SAMİ AKMERMER
Doktora
Türkçe
1997
Makine MühendisliğiErciyes ÜniversitesiMakine Mühendisliği Ana Bilim Dalı
PROF. DR. İBRAHİM UZMAY