مجله فناوری اطلاعات در طراحی مهندسی
سال یکم شماره 2 (پاییز 1386)

In this paper, stability of Generalized Predictive Controller on control of unstable processes (having repetitive/unrepetitive unstable poles and minimum/non minimum phase property) is investigated. It is shown that the existing methods are incapable of stabilizing unstable non minimum phase process with repetitive unstable poles even thought all conditions of the related theories are satisfied. This work shows that one reason is that the optimization problem is ill-conditioned. Furthermore, for processes with repetitive unstable poles, the hard and soft constraint imply in stabilizing method not guaranties stability. First, we determine the relation of two subjects with cost function weightings and then weightings of inputs and outputs in cost function are set to solve two subjects. Different examples it show that by using the proposed method, the closed loop stability of minimum or non-minimum phase processes with repetitive unstable poles can achieved.

This paper presents a systematic procedure to mapping uniform recurrent equations on to linear systolic arrays. In this method, a bijective transformation must be computed first. Then, by applying this transformation over the N-Dim index space of a uniform recurrent equation, a new index space will be obtained that includes a 1-Dim location and a (N-1)-Dim time. By using the location and time dimensions, the location of performing each computation of uniform recurrent equations on the processor array and the time to performing each computation can be determined, respectively. Each computation time is specified by a (N-1)-Dim vector. Therefore, for reaching the scalar times correspond to logical times, we can use full trees first. Then we compute the times correspond to time vectors.

The Residue more number System is capable of supporting fast, carry-free and secure communication. One of the most important considerations when designing residue number system is the choice of the moduli set. This selection has direct effect on dynamic range of numeric representation and the number of locations. In addition there are many improvements in terms of internal computing speed, speed of converters, simplicity of circuit and hardware. In this paper a new five moduli set in radix-4 has been proposed. Applying this moduli set increases the maximum dynamic range for presenting numbers, because of co-prime property of moduli set, so internal computing and converters speed also is increased and the number of locations for numeric representation and hardware complexity and chip area is decreased.. The circuits were simulated in 0.13μm technology by suitable simulators simulation results proved our claims.

One of the most known syndromes in Parkinson's disease is increased passive stiffness in body joints. At present, clinical measurement of stiffness in one of the upper extremity joints is an index used to evaluate the level of the illness. Subjective method currently used is called Unified Parkinson's Disease Rating System, UPDRS. The subjective nature of such method makes it open for influence of the physician himself. Hence, developing a standard method based on objective indices has always been of interest to researchers this research, using a newly fabricated device, viscous and elastic indices as the range of motion were measured during passive movement of elbow joint. Relation between each index and the level of illness was analyzed. Measurements were done for 41 patients and 11 controls. Research indices were extracted using Matlab-R14 software and statistic analyses were done using Spss-13. Results showed a better correlation for viscous index than the elastic one to the level of illness. Using the results of this research may lead to a standard objective method for evaluation of Parkinson's disease.

Knowledge of desorption kinetics is essential for predicting the behavior of materials during drying process and design dryer equipment. The main objective of this study is to the drying kinetics of shelled pistachio in fluidized bed dryer. To this end, for this, thin-layer characteristics of shelled pistachios were determined experimentally as a function of air temperature and velocity. Accordingly, 6 mathematical models (Two term model, Exponential model, Logarithmic model, Page model, Modified page model and Henderson & Pabis model) for describing the thin-layer drying behavior of shelled pistachios were investigated. Tests were conducted using four air temperature (25,30,45 and 60 C), and three air velocities (6, 8 & 10m/s)Out of the 6 models considered, the two term model was found to be the most suitable for describing the drying behavior of the shelled pistachios. Also, for high air temperature and velocity, Henderson-Pabis model proved good suitable. Results showed that air velocity and operating temperature enhanced the kinetics of drying of shelled pistachio and the drying air temperature had a great effect on the drying kinetics.

Fatigue cracking and permanent deformation are two important criteria for pavement mechanistic design. In pavement mechanistic design assumes tensile strain for fatigue cracking at the bottom of the asphalt layer but for overlay mechanistic design it assumes either at the bottom of the asphalt or overlay layer. This paper determines pavement fatigue life using tensile strain at the bottom of the asphalt layer and permanent deformation life with pressure strain under subgrade layer. KenLayer software is used for analysis of pavements with different thicknesses (thin, mid and thick) on three types of subgrades (strong, moderate and weak). The results of analyses and comparisons determine critical criteria in different types of pavements permanent deflection and fatigue cracking graphs for pavement design, are also presented here.

Iron Aluminides as advanced materials have a special role in future industrial development. These materials have attractive properties and their raw materials. In this research, Synthesis of these compounds by high energy ball milling has been investigated. The results of this research showed that agglomerated Fe4Al13 with a particle size around 50 nm can be produced by 100h ball milling in a fast mill. The alumina container and balls were used in this work. No serious contamination was detected in the product. XRD analysis and SEM were used for this study.

This research addresses independent speaker’s discrete word speech recognition (DWSR) using hybrid Self-adaptive Hidden Markov Model/Support Vector Machine (SA-HMM/SVM) classifier. Our proposed method includes two main units: preprocessing unit، and classification unit. The first unit tries to frame the speech wave into proper segments and extract time-frequency relevant features in a way to maximize relative entropy of time-frequency energy distribution among segments، and the second unit classifies words within the proper classes. To fulfill this goal، SA-HMM calculates word’s likelihood to each existing class correspondently، and finally Support Vector Machine (SVM) classifies it by using all classes’ likelihood as an input vector. To validate our proposed method، we test it within our IAUM dataset which contains Persian digits uttered by Persian speakers. Comparing the results with the outcomes of a similar method based on the original HMM shows around 1. 2% improvement.