A Novel Self-Creating Neural Network for Learning Vector Quantization

This paper presents a novel self-creating neural network scheme which employs two resource counters to record network learning activity. The proposed scheme not only achieves the biologically plausible learning property, but it also harmonizes equi-error and equi-probable criteria. The training process is smooth and incremental: it not only avoids the stability-and-plasticity dilemma, but also overcomes the dead-node problem and the deficiency of local minimum. Comparison studies on learning vector quantization involving stationary and non-stationary, structured and non-structured inputs demonstrate that the proposed scheme outperforms other competitive networks in terms of quantization error, learning speed, and codeword search efficiency.     

Optimization of sandblasting process of complex 3D surface polishing using variable viscoelastic diamond particles abrasive

This article reports a novel and efficient diamond particles abrasive with tunable viscoelasticity for sandblasting polishing. Controlling the rust inhibitor content can change its viscoelasticity to adhere diamond particles on polymer materials. Using the sandblasting mechanism, the abrasive deform and slide on the workpiece surface, so that the diamond particles can cut onto the surface peaks of the workpiece. Thus, the complicated morphology can be rapidly and precisely polished. The friction generated by the abrasive on the surface of the workpiece will cause the rust inhibitor solution to evaporate, resulting in reduced viscosity, which makes the diamond particles gradually fall off from the abrasive. Applying Taguchi method, the optimal parameters for viscosity and injection angle were identified. The surface roughness was found to decrease from Ra?=?1.47?μm to Ra?=?0.2?μm in 3?min. The two kinds of complex concave surfaces of different materials were polished by this innovative composite diamond particles abrasive with the optimal parameter values, which has been verified to have 40 times higher efficiency than the traditional manual polishing.     

Fluid-kinetics enhanced selective etching process of NiPt film by piranha chemistry in silicide formation for complementary metal oxide semiconductor fabrication

NiPtSi is a prime candidate for the complementary metal-oxide-semiconductors CMOS self-aligned silicidation process beyond the 22 nm node. The formation of NiPt silicide in smaller geometries demands more Platinum (Pt) additive to control the silicide quality and a more capable NiPt selective etching process to remove surface residual metals for complementing the formation of silicide. Both higher Pt selective etch rate and lower surface material loss are desired in NiPt selective etching process.High temperature (> 150 °C) sulfuric acid base piranha chemistry in fresh dispensing on wafer can etch Pt with less damage to the exposed wafer surface. By using (1) a larger mass-to-charge density Pt redox reaction zone in the electrochemistry spectrum of the Pt redox behavior, (2) stronger chemical fluid kinetics and (3) intensified voltammetric cycles, the Pt selective removal rate can be boosted. Two types of wet chemical processors are used to examine the fluid-chemical kinetics effect on the Pt selective etching rate. It is shown that higher chemical flow rates and stronger fluid-kinetics can enhance the Pt transport behavior. The collateral wafer surface material loss rate also increases by higher chemical flow rates, but the amount of total material loss actually reduces due to a greater reduction in the required process time. The fluid-kinetics enhanced selective etching process can cover a wider range of NiPt film conditions (5% Pt, up to 200 Å, 10% Pt up to 180 Å).     

Histogram-based fuzzy filter for image restoration

In this paper, we present a novel approach to the restoration of noise-corrupted image, which is particularly effective at removing highly impulsive noise while preserving image details. This is accomplished through a fuzzy smoothing filter constructed from a set of fuzzy membership functions for which the initial parameters are derived in accordance with input histogram. A principle of conservation in histogram potential is incorporated with input statistics to adjust the initial parameters so as to minimize the discrepancy between a reference intensity and the output of defuzzification process. Similar to median filters (MF), the proposed filter has the benefits that it is simple and it assumes no a priori knowledge of specific input image, yet it shows superior performance over conventional filters (including MF) for the full range of impulsive noise probability. Unlike in many neuro-fuzzy or fuzzy-neuro filters where random strategy is employed to choose initial membership functions for subsequent lengthy training, the proposed filter can achieve satisfactory performance without any training.     

Improved Representation-burden Conservation Network for Learning Non-stationary VQ

In a recent publication [1], it was shown that a biologically plausible RCN (Representation-burden Conservation Network) in which conservation is achieved by bounding the summed representation-burden of all neurons at constant 1, is effective in learning stationary vector quantization. Based on the conservation principle, a new approach for designing a dynamic RCN for processing both stationary and non-stationary inputs is introduced in this paper. We show that, in response to the input statistics changes, dynamic RCN improves its original counterpart in incremental learning capability as well as in self-organizing the network structure. Performance comparisons between dynamic RCN and other self-development models are also presented. Simulation results show that dynamic RCN is very effective in training a near-optimal vector quantizer in that it manages to keep a balance between the equiprobable and equidistortion criterion.     

On the Internal Representations of Product Units

This paper explores internal representation power of product units [1] that act as the functional nodes in the hidden layer of a multi-layer feedforward network. Interesting properties from using binary input provide an insight into the superior computational power of the product unit. Using binary computation problems of symmetry and parity as illustrative examples, we show that learning arbitrary complex internal representations is more achievable with product units than with traditional summing units.     

A fuzzy DEA and knapsack formulation integrated model for project selection

Project selection has become crucial in the fields of science and engineering. This paper discusses the specific problem of selecting a portfolio of projects that achieves an organization's objectives without exceeding limited capital resources, especially when each project possesses vague input and output data in the selection. In this paper, a data envelopment analysis (DEA), knapsack formulation and fuzzy set theory integrated model is proposed to deal with the problem, and the model is demonstrated via a case study problem in engineering-procurement-construction (EPC) industry. Moreover, this paper applies three constraint handling techniques, which are factor-free penalty function based, to transform a constrained optimization problem into an unconstrained problem, and for the first time adopts the artificial bee colony (ABC) algorithm to search the solutions. The performances of these three constraint handling techniques with respect to the ABC algorithm are compared for the first time in this paper.     

Nonlinear adaptive motion control for manipulators with compliantjoints

How to perform control and achieve stability of robotic manipulators with joint flexibility forms a problem of profound practical and theoretical interest. This paper is to investigate and to solve this problem without strict assumption on the joint stiffness. Here, an adaptive control scheme of a flexible-joint manipulator, which takes into account its full nonlinear dynamics, is presented. Without the knowledge of the system model, the developed control lams, requiring only the position and velocity information of the actuators and links, is capable of driving the link tracking errors asymptotically to zero, while maintaining the uniform boundedness of all signals in the closed-loop system. To demonstrate the effectiveness of the proposed control law, an example of a two-link flexible-joint manipulator is constructed and a number of computer simulations are performed which show quite satisfactory results.     

An efficient analytical approach for extracting the emitter inductance of collector-up HBTs

An efficient analytical parameter-extraction approach for the emitter inductance in a hybrid-/spl pi/ equivalent circuit of collector-up heterojunction bipolar transistors (HBTs) is developed for the first time. A full set of elements is derived unambiguously from impedance and admittance formulation. The good agreement between measured and simulated S-parameters ensures the accuracy of this method.