A novel approach to multiparametric quadratic programming

Multiparametric (mp) programming pre-computes optimal solutions offline which are functions of parameters whose values become apparent online. This makes it particularly well suited for applications that need a rapid solution of online optimization problems. In this work, we propose a novel approach to multiparametric programming problems based on an enumeration of active sets and use it to obtain a parametric solution for a convex quadratic program (QP). To avoid the combinatorial explosion of the enumeration procedure, an active set pruning criterion is presented that makes the enumeration implicit. The method guarantees that all regions of the partition are critical regions without any artificial cuts, and further that no region of the parameter space is left unexplored.     

Bandit algorithms to personalize educational chatbots

Machine Learning - To emulate the interactivity of in-person math instruction, we developed MathBot, a rule-based chatbot that explains math concepts, provides practice questions, and offers...     

Influence of spray drying and dispersing agent on surface and dissolution properties of griseofulvin micro and nanocrystals

The purpose for the current research is to compare and evaluate physiochemical properties of spray-dried (SD) microcrystals (MCs), nanocrystals (NCs), and nanocrystals with a dispersion agent (NCm) from a poorly soluble compound. The characterization was carried out by performing size and surface analysis, interfacial tension (at particle moisture interface), and in-vitro drug dissolution rate experiments. Nanosuspensions were prepared by media milling and were spray-dried. The SD powders that were obtained were characterized morphologically using scanning electron microscopy (SEM), polarized light microscopy (PLM), and Flowchem. Solid-state characterization was performed using X-ray powder diffraction (XRPD), Fourier transfer infrared spectroscopy (FT-IR), and differential scanning calorimetry (DSC) for the identification of the crystalline nature of all the SD powders. The powders were characterized for their redispersion tendency in the water and in pH 1.2. Significant differences in redispersion were noted for both the NCs in both dissolution media. The interfacial tension for particle moisture interface was determined by applying the BET (Braunauer–Emmett–Teller) equation to the vapor sorption data. No significant reduction in the interfacial tension was observed between MCs and NCs; however, a significant reduction in the interfacial tension was observed for NCm at both 25?°C and 35?°C temperatures. The difference in interfacial tension and redispersion behavior can be attributed to a difference in the wetting tendency for all the SD powders. The dissolution studies were carried out under sink and under non-sink conditions. The non-sink dissolution approach was found suitable for quantification of the dissolution rate enhancement, and also for providing the rank order to the SD formulations.     

An innovative approach for Steam Generator Pressure Control of a nuclear power plant

The main function of the Steam Generator Pressure Control (SGPC) Program is to match the power (heat) generation in the reactor core with the heat removal in the steam generators (SGs). For most of the designs these programs have been over simplified to cater to the limitation of the instrumentation and control, hardware and software. The main objective of balancing the heat generation, transfer and removal gets lost in the process, which leads to reduction in the availability of the nuclear power plant. This is reflected in under utilization of the process and control system provisions to avoid reactor trips on low/high pressure. Most of the transients/accidents have their origin in the mismatch among the heat generated in the reactor core and the heat removal in the SGs. A new Advanced Process Control (APC) based supervisory controller is proposed to ride over the existing SGPC to achieve the goal. This APC makes use of the estimated/measured heat generation-removal error to alter the SGPC set point to tide over the transients after detection. The transients are detected based on the magnitude of this error to activate the APC. After tiding over the transient successfully the control switches back to the existing SGPC. For evaluation of this error additional instrumentation is proposed in the SG tube walls and other structures of the primary system.The Reactor Regulating System (RRS) which manipulates the reactor power, PHT pressure controller and the SGPC must work in a synchronized and optimized fashion so as to avoid a reactor trip following the PCP trip and other transients. With the existing design it is often observed that these control loops conflict with each other, as they are not designed with a multivariable integrated approach. The objective is being fulfilled by developing new supervisory APC logics to ride over the existing Proportional Integral Derivative (PID) controllers of various process loops. The new APC concept/logics formulation and supervisory control details along with performance tests are presented in this paper.     

Hydrogen evolution reaction and electronic structure calculation of two dimensional bismuth and its alloys

We present a systematic ab initio study of atomic hydrogen and oxygen adsorption on bismuthene monolayer and its alloys with arsenic and antimony through electronic structure calculations based on density functional theory within generalized gradient approximation. We systematically investigated the preferable adsorption site for hydrogen and oxygen atom on 2D Bi, BiAs and BiSb. It was found that the hydrogen atom prefers top site of bismuth atom and oxygen atom prefers to reside in the hexagonal ring of these 2D bismuth alloys. The free energy calculated from the individual adsorption energy for each monolayer subsequently guides us to predict the best suitable catalyst among the considered 2D monolayers. The 2D BiSb serves better for hydrogen evolution reaction (HER) with hydrogen adsorption energy as ?1.384 eV while 2D BiAs is suitable for oxygen evolution reaction (OER) with oxygen adsorption energy as ?1.092 eV. We further investigated the effects of the adsorbate atom on the electronic properties of 2D Bi, BiAs and BiSb. The adsorption of oxygen on 2D BiAs and BiSb was shown to reduce the bulk band gap by 40.56 and 67.79% respectively which will be advantageous for the observation of Quantum Spin Hall effect at ambient conditions.     

In Situ Formation of Conductive Metal Sulfide Domain in Metal Oxide Matrix: An Efficient Way to Improve the Electrochemical Activity of Semiconducting Metal Oxide

A new effective way to improve the electrochemical activity of semiconducting metal oxide is developed by the in situ formation of conductive metal sulfide domain in the metal oxide matrix. The Li_0.96Ti_1.08S₂?Li₄Ti₅O₁₂ nanocomposites with tunable compositions and electrical properties are synthesized by the reaction of Li₄Ti₅O₁₂ with CS₂ at elevated temperature. The resulting incorporation of conductive Li_0.96Ti_1.08S₂ domain in the Li₄Ti₅O₁₂ matrix is effective in enhancing the electrical conductivity and electrode activity of semiconducting lithium titanate. As anode materials for lithium ion batteries, the obtained Li_0.96Ti_1.08S₂?Li₄Ti₅O₁₂ nanocomposites show much greater discharge capacity and better rate characteristics than does the pristine Li₄Ti₅O₁₂. The usefulness of the present method is further evidenced from the improvement of the electrochemical activity of semiconducting CsTi₂NbO₇ after the reaction with CS₂. The present study clearly demonstrates the in situ formation of conductive metal sulfide domain using CS₂ liquid can provide an efficient and universal way to improve the electrode functionality of semiconducting metal oxide.     

Blockage of KHSRP-NLRP3 by MCC950 Can Reverse the Effect of Manganese-Induced Neuroinflammation in N2a Cells and Rat Brain

Manganese neurotoxicity has been reported to cause a neurodegenerative disease known as parkinsonism. Previous reports have shown that the expression of the KH-type splicing regulatory protein (KHSRP), a nucleic acid-binding protein, and NLRP3 is increased upon Mn exposure. However, the relation between these two during Mn toxicity has not been fully deduced. The mouse neuroblastoma (N2a) and SD rats are treated with LPS and MnCl₂ to evaluate the expression of KHSRP and NLRP3. Further, the effect of the NLRP3 inhibitor MCC950 is checked on the expression of NLRP3, KHSRP and pro-inflammatory markers (TNFα, IL-18 and IL-1β) as well as the caspase-1 enzyme. Our results demonstrated an increment in NLRP3 and KHSRP expression post-MnCl₂ exposure in N2a cells and rat brain, while on the other hand with LPS exposure only NLRP3 expression levels were elevated and KHSRP was found to be unaffected. An increased expression of KHSRP, NLRP3, pro-inflammatory markers and the caspase-1 enzyme was observed to be inhibited with MCC950 treatment in MnCl₂-exposed cells and rats. Manganese exposure induces NLRP3 and KHSRP expression to induce neuroinflammation, suggesting a correlation between both which functions in toxicity-related pathways. Furthermore, MCC950 treatment reversed the role of KHSRP from anti-inflammatory to pro-inflammatory.     

Constructing table-of-content for videos

A fundamental task in video analysis is to extract structures from the video to facilitate user's access (browsing and retrieval). Motivated by the important role that the table of content (ToC) plays in a book, in this paper, we introduce the concept of ToC in the video domain. Some existing approaches implicitly use the ToC, but are mainly limited to low-level entities (e.g., shots and key frames). The drawbacks are that low-level structures (1) contain too many entries to be efficiently presented to the user; and (2) do not capture the underlying semantic structure of the video based on which the user may wish to browse/retrieve. To address these limitations, in this paper, we present an effective semantic-level ToC construction technique based on intelligent unsupervised clustering. It has the characteristics of better modeling the time locality and scene structure. Experiments based on real-world movie videos validate the effectiveness of the proposed approach. Examples are given to demonstrate the usage of the scene-based ToC in facilitating user's access to the video.