Evapotranspiration Modeling Using Different Tree Based Ensembled Machine Learning Algorithm

Water Resources Management - The present study investigates and evaluate the scope and potential of modern computing tools and techniques such as ensembled machine learning methods in estimating...     

Infrared study of salt free and salt treated cellulose pyrolysis under nitrogen atmosphere

The pyrolysis under nitrogen of salt free and salt treated cellulose was studied by infrared spectroscopy. The results confirm the unzipping mechanism suggested by DTA and TGA studies, and show this mechanism to apply to the untreated and salt treated samples. The results also show the unzipping to stop at the stage of char formation. The onset of this stage was found to occur after the volatilization of only 45–50% of the initial sample of the borax treated cellulose, compared with 65–70% of the KCl treated cellulose and 80–85% of the untreated cellulose.     

A closed-loop cross-correlation method for detecting model mismatch in MIMO model-based controllers

Correlation between a dithering signal and the prediction error has been used for detecting model mismatch in univariate model based control systems. This paper extends that approach to MIMO control systems. A closed-loop cross-correlation method is presented to detect which specific input-output pairings of a model-based controller are mismatched. This method may be used in screening the complete set of models and in selecting candidate models for re-identification. The method first finds the rows and columns of the transfer function matrix that contain mismatch, and then the candidates are found by the intersection of the said rows and columns. Placing the system under partial control, whereby one or more of the manipulated variables are held constant, can be used to further reduce the set of candidate models.     

Acetals as nucleating agents for polypropylene

Nucleating agents increase the impact strength, tensile strength, and tensile elasticity modulus of semicrystalline polymers. Nucleating agents also decrease product cycle times, resulting in a cost savings per product unit. We have synthesized and tested 15 compounds as nucleactors for polypropylene. Of these, trinapthylidene sorbitol, tri-(4-methyl-1-naphthylidene)sorbitol, tri-(4-methoxy-1-naphthylidene) sorbitol, and dibenzylidene xylitol are efficient nucleators of polypropylene. Trinaphthylidene sorbitol (tns) has two major diastereomers: The “S” diastereomer yields a faster crystallization rate for polypropylene than does the commercial nucleator dibenzylidene sorbitol (Millad 3905). Crystallization rates are 208 and 88, respectively (t min^?1 × 1000). The “R” diastereomer, however, is a poor nucleator and interferes with the nucleating activity of the “S” diastereomer. A 52/48 mixture of diastereomers does not nucleate polypropylene, even at twice the concentration. This is first time that the importance of stereochemistry has been demonstrated in the nucleating action. © 1994 John Wiley & Sons, Inc.     

On routing in large WDM networks

An important problem in WDM network design is to construct a logical topology and determine an optimal routing over that topology. Mixed Integer Linear Program (MILP) formulations to generate optimal solutions for this problem have been presented. Such formulations are computationally intractable, even for moderate sized networks. A standard approach is to decouple the problem of logical topology design and the problem of routing the traffic on this logical topology. Heuristics for finding the logical topology exist and a straight-forward linear program (LP), based on the node-arc formulation can be used to solve the routing problem over a given logical topology. We have found that such LP formulations become computationally infeasible for large networks. In this paper, we present a new formulation, based on the arc-chain representation, for optimally routing the specified traffic over a given logical topology to minimize the congestion of the network. We have used the revised simplex method incorporating an implicit column generation technique, and exploited the special Generalized Upper Bounding structure as well as the possibility of eta-factorization for efficiently updating the dual variables and finding the solution. Experimental results on a number of networks demonstrate the suitability of this approach.     

Generalized four-point characterization method using capacitive and ohmic contacts

In this paper, a four-point characterization method is developed for samples that have either capacitive or ohmic contacts. When capacitive contacts are used, capacitive current- and voltage-dividers result in a capacitive scaling factor not present in four-point measurements with only ohmic contacts. From a circuit equivalent of the complete measurement system, one can determine both the measurement frequency band and capacitive scaling factor for various four-point characterization configurations. This technique is first demonstrated with a discrete element four-point test device and then with a capacitively and ohmically contacted Hall bar sample over a wide frequency range (1 Hz-100 kHz) using lock-in measurement techniques. In all the cases, data fit well to a circuit simulation of the entire measurement system, and best results are achieved with large area capacitive contacts and a high input-impedance preamplifier stage. An undesirable asymmetry offset in the measurement signal is described which can arise due to asymmetric voltage contacts.     

Bang‐Bang solution of nonlinear time‐optimal control problems using a semi‐exhaustivesearch

At times, the objective is to seek a bang‐bang control policy for nonlinear time‐optimal control problems. The usefulness of iterative dynamic programming (IDP) has been shown in the literature for solving such problems. However, the convergence to the optimal solution has been obtained from about 50% of the guessed values near the optimum. In this paper, we present a semiexhaustive search method for seeking such solutions and a comparison is made with the IDP. The results show that the convergence can be obtained from a significantly higher number of guessed values chosen over a much wider region around the optimum.     

Chip-based analysis of protein-protein interactions by fluorescence detection and on-chip immunoprecipitation combined with microLC-MS/MS analysis

A new chip-based method to identify protein-protein interactions was developed using the guanine nucleotide exchange factor GRF2 and two interacting proteins, Ras and calmodulin, as model proteins. A generic immobilization strategy for FLAG-tagged bait proteins on a protein-repellent streptavidin chip surface was implemented by presentation of an oriented anti-FLAG antibody. A flow cell device, integrating different chip surfaces, was developed, and the interaction of immobilized GRF2 with the two analytes was verified by fluorescence assays. On-chip tryptic digest assays were then performed on the capture surface and analyzed by microLC-MS/MS. The interaction of GRF2 with calmodulin and Ras was demonstrated, and the lower limit of detection was determined. We also implemented an on-chip immunoprecipitation assay to identify GRF2-binding partners from complex protein mixtures. Cells overexpressing FLAG-GRF2 were lysed and then incubated with the anti-FLAG chip. In addition to detecting GRF2, we also identified calmodulin, demonstrating that this technique can successfully identify endogenous levels of proteins, bound to recombinant bait proteins. This chip-based method has the advantage that no subsequent gel separations of protein complexes prior to LC-MS analysis are required and is therefore amenable to miniaturized high-throughput determination of protein-protein interactions.     

Design and development of multicolor MWIR/LWIR and LWIR/VLWIR detector arrays

Multicolor infrared (IR) focal planes are required for high-performance sensor applications. These sensors will require multicolor focal plane arrays (FPAs) that will cover various wavelengths of interest in mid wavelength infrared/long wavelength infrared (MWIR/LWIR) and long wavelength infrared/very long wavelength infrared (LWIR/VLWIR) bands. There has been significant progress in HgCdTe detector technology for multicolor MWIR/LWIR and LWIR/VLWIR FPAs.^1–3 Two-color IR FPAs eliminate the complexity of multiple single-color IR FPAs and provide a significant reduction of weight and power in simpler, reliable, and affordable systems. The complexity of a multicolor IR detector MWIR/LWIR makes the device optimization by trial and error not only impractical but also merely impossible. Too many different geometrical and physical variables need to be considered at the same time. Additionally, material characteristics are only relatively controllable and depend on the process repeatability. In this context, the ability of performing “simulation experiments” where only one or a few parameters are carefully controlled is paramount for a quantum improvement of a new generation of multicolor detectors for various applications.     

A comparison between different windows in spectral and cross spectral analysis techniques with Kalman filtering for estimating power quality indices

Fourier based methods are still popular for calculating power quality indices (PQIs) in the technical community such as IEC and IEEE. They are robust towards noise and can be calculated efficiently using the fast Fourier transform (FFT) method. IEEE Std 1459-2010 provides formulations for calculating important PQI for single and three phase systems. The aim of this paper is to present the theory for estimating PQIs using the Welch spectral and cross spectral analysis techniques, and show experimentally the accuracy of the proposed methods when calculating the PQIs. Further it provides a comparison on the windowing techniques (WTs) that accompany the PQI estimations when using FFT based methods. This information has not been documented extensively in the technical literature when trying to calculate the PQI as specified by IEEE Std 1459-2010. Further we present a virtual instrument (VI) with a Kalman filter (KF) to estimate the PQIs. The KF is used to reduce the variance in measurements for the PQI estimation. We also compare experimentally the accuracy of the VI without the KF with the VI using the KF. Specifically we propose good and bad WTs to estimate such PQIs.