In networked control systems, time delay and data dropout can degrade the performance of the control system and even destabilize the system. In the present paper, the Extended Kalman filter is employed to compensate the effects of time delay and data dropout in feedforward and feedback paths of networked control systems. In the proposed method, the extended Kalman filter is used as an observer ...

This paper presents two feedforward control algorithms to attenuate the tension error in digital tape systems. Tape systems are time-varying because not only do the radii of both reels change as a result of tape winding from one reel to the other but reels also have eccentricities. We address the time-varying nature of the reel radii due to tape winding by introducing a time-varying feedforward...

Acceptable yields for nanofabrication will require significant improvement in CD control. One method to achieve better runto-run CD control is through inter-process feedforward control. The potential benefits of feedforward control include reduced run-to-run post-etch CD variance, rework, and scrap. However, measurement noise poses a significant threat to the success of feedforward control. Sin...

Iterative Learning Control (ILC) is an effective control technique for motion systems that perform repetitively the same trajectory (setpoint), e.g. a wafer stage. The result of the learning procedure is a feedforward signal that perfectly compensates all deterministic dynamics in the system for the learned setpoint performed at a specific start position. For other setpoints and start positions...

d (t) u(t) y(t F((J,q) P(q) ~ I .. ... I I I I I I I I Figure 2: A general architecture of tracking systems using feedforward control I I I I I I I I I I I I I I I I I I I I I ~ Feedforward _ ............. : . Closed LooP--": ; Controller ~ Plant ! In Figure 2, the feedforward controller appears as a pre-filter before the closed-loop system comprising the prevalently adopted is to feed forward ...

A feedforward neural network based on multi-valued neurons is considered in the paper. It is shown that using a traditional feedforward architecture and a high functionality multi-valued neuron, it is possible to obtain a new powerful neural network. Its learning does not require a derivative of the activation function and its functionality is higher than the functionality of traditional feedfo...

The intrinsic hysteresis behavior of piezoelectric material limits the tracking control accuracy of the actuators. This paper describes a tracking control method for piezoelectric actuators based on the combination of feedforward and feedback loops. The hysteresis of piezoelectric actuators is linearized in feedforward loop with an inverse hysteresis model based on a neural network. The number ...

We introduce a new structure for memory neural networks, called feedforward sequential memory networks (FSMN), which can learn long-term dependency without using recurrent feedback. The proposed FSMN is a standard feedforward neural networks equipped with learnable sequential memory blocks in the hidden layers. In this work, we have applied FSMN to several language modeling (LM) tasks. Experime...

The present study tested two hypotheses about apraxia of speech (AOS), framed in the DIVA model (Guenther, Ghosh, & Tourville, 2006). The DIVA model assumes that speech targets are regions in auditory space, and combines two mechanisms to reach those targets: feedback control and feedforward control. The Feedforward System Deficit (FF) hypothesis states that feedforward control is impaired in A...

Matching the brain’s flexibility and ability to quickly incorporate new information remains difficult using recognition algorithms. In addition, recognition scenarios in which humans require more time to recognize do not inherently require more time in recognition algorithms. This work proposes a recognition algorithm that uses dynamics in an opposite way from traditional recognition algorithms...