Selection of network and learning parameters for an adaptive neural robotic control scheme

This paper presents the results of a study in the design of a neural network based adaptive robotic control scheme. The neural network used here is a two hidden layer feedforward network and the learning scheme is the well-known backpropagation algorithm. The neural network essentially provides the inverse of the plant and acts in conjunction with a standard PD controller in the feedback loop. The objective of the controller is to accurately control the end position of a single link manipulator in the presence of large payload variations, variations in the link length and also variations in the damping constant. Based on results of this study, guidelines are presented in selecting the number of neurons in the hidden layers and also the parameters for the learning scheme used for training the network. Results also indicate that increasing the number of neurons in the hidden layer will improve the convergence speed of learning scheme up to a certain limit beyond which the addition of neurons will cause oscillations and instability. Guidelines for selecting the proper learning rate, momentum and fast backpropagation constant that ensure stability and convergence are presented. Also, a relationship between the r.m.s. error and the number of iterations used in training the neural network is established.     

基于一类T-模的模糊Hopfield网络的稳定性分析

本文建立了基于一类T-模的一种动态模糊神经网络-(∨,T)Hopfield网.文章首先证明了系统本身的稳定性及系统平衡态的Lyapunov稳定性,然后建立了吸引子的一个非平凡吸引域,使系统具有良好的容错性.最后,通过例子验证了得到的结论.     

Low-loss three-phase frequency conversion using a TDM-PWM technique

A strategy for the realisation of a product-modulation frequency conversion technique is presented. For maximum efficiency and spectral purity, a pulse-width modulation (PWM) circuit is used. The addition of pulse-width-modulated signals from the three input phases to each of the three output phases is accomplished by time-division multiplexing (TDM). The need for the product modulator to accommodate multiplication by negative numbers is eliminated by a modification to the basic equations.     

3D polymeric microfluidic device fabrication via contact liquid photolithographic polymerization (CLiPP)

In this contribution, a new method for the fabrication of complex polymeric microfluidic devices is presented. The technology, contact liquid photolithographic polymerization (CLiPP), overcomes many of the drawbacks associated with other rapid prototyping schemes, such as limited materials choices and time-consuming microassembly protocols. CLiPP shares many traits with other photolithographic methods, but three distinct features: (i) liquid photoresists in contact with the photomask, (ii) readily removed sacrificial materials, and (iii) living radical processes, enable multiple polymeric chemistries and mechanical properties while simultaneously enabling facile fabrication of 3D geometries and surface chemistry control. This contribution details fabrication techniques and methods for the fabrication of high aspect ratio posts covalently bonded to a polymeric substrate, an array of independently stacked bars on top of perpendicular bars, multiple undercut structures fabricated simultaneously, and a complex 3D geometry with intertwined channels.     

Influence of SiO2 and CaO additions on the microstructure and magnetic properties of sintered Sr-hexaferrite

The effect of simultaneous addition of CaO and SiO₂ on the microstructure and magnetic properties of sintered SrO-excess Sr-hexaferrites was studied. Both additives markedly affect the grain growth behavior and the magnetic properties. CaO-additions promote densification, which results in increased remanence, but due to simultaneuous grain growth the coercivity drops to <100 kA/m. SiO₂ additions are known to suppress grain growth. Simultaneous addition of CaO and SiO₂ is shown to be very beneficial in tailoring a dense microstructure with relatively small grains. The ratio of CaO/SiO₂ was found to be optimum at about 1, and magnets with a remanence of 430 mT and a coercivity of 300 kA/m were obtained. Transmission electron microscopy (TEM) studies and investigations by energy-dispersive analysis of X-rays (EDX) in the scanning TEM (STEM) mode show that both CaO and SiO₂ are concentrated at grain boundaries and grain junctions forming an amorphous secondary phase.     

Goal-directed imitation for robots: A bio-inspired approach to action understanding and skill learning

In this paper we present a robot control architecture for learning by imitation which takes inspiration from recent discoveries in action observation/execution experiments with humans and other primates. The architecture implements two basic processing principles: (1) imitation is primarily directed toward reproducing the outcome of an observed action sequence rather than reproducing the exact action means, and (2) the required capacity to understand the motor intention of another agent is based on motor simulation. The control architecture is validated in a robot system imitating in a goal-directed manner a grasping and placing sequence displayed by a human model. During imitation, skill transfer occurs by learning and representing appropriate goal-directed sequences of motor primitives. The robustness of the goal-directed organization of the controller is tested in the presence of incomplete visual information and changes in environmental constraints.     

Efficient and stable field emission from the oxidized porous polysilicon using Pt/Ti multilayer electrode

The field emission characteristics of an oxidized porous polysilicon (OPPS) were investigated with a Pt/Ti multilayer electrode, which showed highly efficient and stable electron emission characteristics compared with those of conventional Au/NiCr electrodes. The thin Ti layer played an important role in promoting the adhesion of Pt to SiO₂ surface and the distribution of the electric field on the OPPS surface. Additionally, the Ti layer efficiently blocked the diffusion of emitter metal, which resulted in more reliable emission characteristics.     

Optimal and self‐tuning fusion Kalman filters for discrete‐time stochastic singular systems

Based on the optimal fusion estimation algorithm weighted by scalars in the linear minimum variance sense, a distributed optimal fusion Kalman filter weighted by scalars is presented for discrete‐time stochastic singular systems with multiple sensors and correlated noises. A cross‐covariance matrix of filtering errors between any two sensors is derived. When the noise statistical information is unknown, a distributed identification approach is presented based on correlation functions and the weighted average method. Further, a distributed self‐tuning fusion filter is given, which includes two stage fusions where the first‐stage fusion is used to identify the noise covariance and the second‐stage fusion is used to obtain the fusion state filter. A simulation verifies the effectiveness of the proposed algorithm. Copyright © 2008 John Wiley & Sons, Ltd.