A Fuzzy Expert System for Damage Assessment Using Genetic Algorithms and Neural Networks

Abstract: An efficient knowledge-acquisition support method is required for improvement and maintenance of the knowledge base in durability evaluation of an RC bridge deck. Such a method is proposed in this paper to automatically acquire fuzzy production rules. This method makes joint use of genetic algorithms and a neural network. Using a neural network as a subsystem, the evaluation function of genetic algorithms can be provided with the weights of the neural network. Introducing a neural network into genetic algorithms, it is possible to acquire new knowledge so that the method is useful when it is difficult to acquire knowledge in the field.     

Calculation of second virial coefficients and gaseous viscosities of the refrigerants HFC-32 (CH2 F2), HFC-23 (CHF3), and HCFC-22 (CHC1F2)

The second virial coefficients of refrigerants HFC-32 (CH₂F₂), HFC-23 (CHF₃), and HCFC-22 (CHC1F₂) have been correlated on the bisis of site site model potential and have been compared with experimental results. The molecular interactions consisted of repulsion dispersion and electrostatic parts. From the site site potentials adjusted to the experimental second virial coefficients, spherically averaged potentials have been determined and a subsequent calculation of gaseous viscosity has been carried out. Agreement between measured and calculated values of second virial coellicients and gaseous viscosity is satisfactory. Calculated values of second virial coefficients and gaseous viscosity beyond available experimental data, therefore. can be assumed as a reliable extrapolation to lower and higher temperatures.     

Solution synthesis of crystallized AMO4 (A=Ba, Sr, Ca; M=W, Mo) film at room temperature

A well-crystallized AMO₄ (A=Ba, Ca, Sr; M=W, Mo) films have been prepared at room temperature through a simple solution reaction in respective alkaline solution at higher pH ranging from 12–14. Adopting the corrosion principle for oxidation of metal substrate, these double oxide films were carried out in presence of chemical driving force without any special apparatus or devices. Hydrogen peroxide was used to enhance the dissolution rate of metal substrates. The driving force for the film formation and growth were high concentration of A²⁺, MO₄²⁻ ions with high pH conditions. Average grain sizes of 8–10 μm with bipyramidal shaped particle were grown to the thickness of about 10–14 μm after 3–6 hours treatment. The crystallization of AMO₄ was characterized by three-dimensional nucleation. This work demonstrates the possibility of fabrication of functional ceramic films directly from the aqueous solution in a single step by solution reactions.     

Effect of rapid sample cooling on efficiency of multiple impurity-atom doping

Shallow and efficient doping of wide band-gap semiconductors has remained one of yet unresolved problems to date. A possible solution to this problem is doping with complexes of a few impurity atoms at a quasi-equilibrium state, which is introduced by controlled cooling of a sample after doping. In this work, (1) we first define a global and quasi-equilibria of our interest based on a simple thermodynamic model for a doped crystal, and then (2) we discuss how the cooling rate affects the probability of impurity-complex formation at a quasi-equilibrium as defined. Our main message is that one should design impurity complexes as small in size as possible which have as large a binding energy as possible. This is a required condition for complex designs when it is difficult to tune the cooling rate.     

Three-dimensional pin power reconstruction for the axially heterogeneous region in BWR

A three-dimensional pin power reconstruction method was proposed and verified by applying to the axially heterogeneous region problem of the BWR core calculation. Because the production assembly calculation has been carried out by two-dimensional deterministic calculation methods like current coupling collision probability or the method of characteristics, it has been difficult to predict the detailed three-dimensional heterogeneous pin power distribution of the axially heterogeneous region. Consequently, only radial intranodal homogeneous power distributions have been considered, and axial intranodal homogeneous power distributions have not been considered in the estimation of linear-heat-generation-ratio at common BWR core calculations.     

Magnetohydrodynamic stability of boundary layers along a flat plate with pressure gradient

Summary The purpose of the paper is to consider theoretically the characteristics of steady two-dimensional magnetohydrodynamic boundary layer flow past a flat plate with pressure gradient in the presence of transverse magnetic field and its stability problem. The boundary layer equation is transformed into a non-similar one and numerical solutions are obtained by the difference-differential method. The neutral stability curves for wavelike disturbances of Tollmien-Schlichting type are then presented for the velocity profiles obtained above. Numerical results for the magnetic parameter and pressure gradient parameter are given for the velocity profiles, coefficient of skin friction, displacement thickness and the critical Reynolds number.With 10 Figures     

Pulse-Width Dependence in Current-Driven Magnetization Reversal Using GaMnAs-Based Double-Barrier Magnetic Tunnel Junction

We have investigated the current pulse width dependence on current-driven magnetization reversal in double-barrier structures using GaMnAs-based magnetic tunneling junctions (MTJ) in order to clarify the origin of low threshold current density for current-driven magnetization reversal. Comparing with the case of single-barrier MTJ, the pulse-width dependence reveals that threshold current density is reduced by double-barrier MTJ. We confirmed that the threshold current density in the order of 10⁴ A/cm² is estimated considering the effect of current pulse width.     

An all-digital analog-to-digital converter with 12-/spl mu/V/LSB using moving-average filtering

A compact, high-resolution analog-to-digital converter (ADC) especially for sensors is presented. The basic structure is a completely digital circuit including a ring-delay-line with delay units (DUs), along with a frequency counter, latch, and encoder. The operating principles are: (1) the delay time of the DU is modulated by the analog-to-digital (A/D) conversion voltage and (2) the delay pulse passes through a number of DUs within a sampling (= integration) time and the number of DUs through which the delay pulse passes is output as conversion data. Compact size and high resolution were realized with an ADC having a circuit area of 0.45 mm/sup 2/ (0.8-/spl mu/m CMOS) and a resolution of 12 /spl mu/V (10 kS/s). Its nonlinearity is /spl plusmn/0.1% FS per 200-mV span (1.8-2.0 V), for 14-b resolution. Sample holds are unnecessary and a low-pass filter function removes high-frequency noise simultaneously with A/D conversion. Thus, the combination of this ADC and a digital filter that follows can eliminate an analog prefilter to prevent the aliasing before A/D conversion. Also, both this ADC can be shrunk and operated at low voltages, so it is an ideal means to lower the cost and power consumption. Drift errors can be easily compensated for by digital processing.     

Prediction of spectral reflectance factor distribution of automotive paint finishes

It is necessary to determine the accurate reflectance of painted surfaces for the review of paint finishes by computer graphics (CG) before actual painting of the exterior color of automobiles, and for quality control during production and inspection processes. We have optimized a method for measuring reflectance by using a statistical technique. We have found that the reflectance of a painted surface is best measured at an incident angle of 60° and at five aspecular angles of 10°, 18°, 28°, 40°, and 90°. Our method makes it possible to accurately reproduce reflection characteristics of paint finishes containing special flake pigments, such as pearl mica. Also it was proved that our method can apply not only to solid and metallic coatings but to all painted surfaces. © 2005 Wiley Periodicals, Inc. Col Res Appl, 30, 275–282, 2005; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/col.20125