• 제어로봇시스템학회:학술대회논문집
• /
• 1989.10a
• /
• pp.653-662
• /
• 1989

Cerebellar Model Arithmetic Controller (CMAC) has been introduced as an adaptive control function generator. CMAC computes control functions referring to a distributed memory table storing functional values rather than by solving equations analytically or numerically. CMAC has a unique mapping structure as a coarse coding and supervisory delta-rule learning property. In this paper, learning aspects and a convergence of the CMAC were investigated. The efficient training algorithms were developed to overcome the limitations caused by the conventional maximum error correction training and to eliminate the accumulated learning error caused by a sequential node training. A nonlinear function generator and a motion generator for a two d.o.f. manipulator were simulated. The efficiency of the various learning algorithms was demonstrated through the cpu time used and the convergence of the rms and maximum errors accumulated during a learning process. A generalization property and a learning effect due to the various gains were simulated. A uniform quantizing method was applied to cope with various ranges of input variables efficiently.

• 제어로봇시스템학회:학술대회논문집
• /
• 1990.10a
• /
• pp.7-12
• /
• 1990

In this paper, we propose a new manipulator control scheme based on the CMAG neural network. The proposed control consists of two components. The feedforward component is an output of trained CMAC neural network and the feedback component is a modified sliding mode control. The CMAC accepts the position, velocity and acceleration of manipulator as input and outputs two values for the controller : One is the nominal torque used for feedforward compensation(M1 network) and the other is the inertia matrix related information used for the feedback component(M2 network). Since the used control algorithm guarantees the robust trajectory tracking in spite of modeling errors, the CMAC mapping errors due to the memory limitation are little worth consideration.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.3
• /
• pp.329-337
• /
• 2000

This study describes a pressure tracking control of hydroforming process which is used for precision forming of sheet metals. The hydroforming operation is performed in the high-pressure chamber strictly controlled by pressure control valve and by the upward motion of a punch moving at a constant speed, The pressure tracking control is very difficult to design and often does not guarantee satisfactory performances be-cause of the punch motion and the nonlinearities and uncertainties of the hydraulic components. To account for these nonlinearities and uncertainties of the process and iterative learning controller is proposed using Cerebellar Model Arithmetic Computer (CMAC). The experimental results show that the proposed learning control is superior to any fixed gain controller in the sense that it enables the system to do the same work more effectively as the number of operation increases. In addition reardless of the uncertainties and nonlinearities of the form-ing process dynamics it can be effectively applied with little a priori knowledge abuot the process.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.14 no.3
• /
• pp.290-297
• /
• 2004

The control laws for cart-pole system are studied to see the schemes on which the control laws are made. Also, the odd symmetry of the relation between the output of the control laws and the system state vector is observed. Utilizing the symmetry in quantizing the system state variables and implementing the control laws into look-up table is discussed. Then, a CMAC is trained for a nonlinear control law for a cart-pole system such that the symmetry is conserved and its learning performance is evaluated. It is found that utilizing the symmetry is to reduce the memory requirement as well as the training period while improving the learning quality in terms of preserving the symmetry.

• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10b
• /
• pp.936-939
• /
• 1996

In this paper, two specially designed associative mapping memories, called Associative Mapping Elements(AME) and Multiple-Digit Overlapping AME(MDO-AME), are presented for learning of nonlinear functions including kinematics and dynamics of robot manipulators. The proposed associative mapping memories consist of associative mapping rules(AMR) and weight update rules(WUR) which guarantee generalization and specialization of input-output relationship of learned nonlinear functions. Two simulation results, one for supervised learning and the other for unsupervised learning, are given to demonstrate the effectiveness of the proposed associative mapping memories.