Learning multiple experiences useful visual features for active maps localization in crowded environments

Crowded urban environments are composed of different types of dynamic and static elements. Learning and classification of features is a major task in solving the localization problem in such environments. This work presents a gradual learning methodology to learn the useful features using multiple experiences. The usefulness of an observed element is evaluated by a scoring mechanism which uses two scores – reliability and distinctiveness. The visual features thus learned are used to partition the visual map into smaller regions. The robot is efficiently localized in such a partitioned environment using two-level localization. The concept of active map (AM) is proposed here, which is a map that represents one partition of the environment in which there is a high probability of the robot existing. High-level localization is used to track the mode of the AMs using discrete Bayes filter. Low-level localization uses a bag-of-words model to retrieve images and accurately localize the robot. The pose of the robot is the one retrieved from the AM that has maximum a posteriori. Experiments have been conducted on a unique highly crowded data-set collected from Indian roads. The results support the proposed method due to speed and localization accuracy.     

Applicability of neutron noise technique for diagnostics of in-core components for heavy water reactors

Neutron flux signal is composed of a steady or mean component resulting from the flux produced by power operation of the reactor and a very small fluctuating component called ‘noise’ component. Analysis of neutron noise from suitably located sensors is a proven technique to monitor the in-core components of light water reactors (LWRs). However, the use of neutron noise has been rare, if any, for heavy water reactors (HWRs) as it was generally felt that the unfavourable transfer function characteristics of the reactors would limit its applicability. To assess the applicability of technique in pressurised heavy water reactors (PHWRs), experiments were carried out using in-core and out-of-core neutron sensors in a research reactor. This paper discusses the measurement details and results of the experiment. This paper concludes that the neutron noise technique can be effectively utilised for diagnostics/characterisation of the in-core components of heavy water reactors.     

Kinetics of the Hydrolysis of some alkyl halides in aqueous ethyl alcohol

A comparison of the rates of substitution and elimination reactions occuring simultaneously in alkaline hydrolysis of n-amyl chloride, isoamyl chloride, n-amyl iodide and n-hexyl iodide at different temperatures and in varying alcohol-water mixtures, has been made. Under the same conditions the rates of solvolysis of these halides have also been reported. In the case of n-amyl halides there is a much larger rate difference between chloride and bromide than between bromide and iodide.     

Recursive Technique For Computing System Reliability

This paper presents a simple recursive formula for calculating source-to-sink congestion, and hence reliability, in a network. This congestion is derived from individual blocking probabilities of each link. The method has an advantage of not requiring cut-sets. Such a knowledge is a prerequisite in most other methods of reliability analysis. We have programmed the algorithm using Pascal as it allows recursive procedure calls. An example illustrates the method.     

Meta-cognitive RBF Network and its Projection Based Learning algorithm for classification problems

‘Meta-cognitive Radial Basis Function Network’ (McRBFN) and its ‘Projection Based Learning’ (PBL) algorithm for classification problems in sequential framework is proposed in this paper and is referred to as PBL-McRBFN. McRBFN is inspired by human meta-cognitive learning principles. McRBFN has two components, namely the cognitive component and the meta-cognitive component. The cognitive component is a single hidden layer radial basis function network with evolving architecture. In the cognitive component, the PBL algorithm computes the optimal output weights with least computational effort by finding analytical minima of the nonlinear energy function. The meta-cognitive component controls the learning process in the cognitive component by choosing the best learning strategy for the current sample and adapts the learning strategies by implementing self-regulation. In addition, sample overlapping conditions are considered for proper initialization of new hidden neurons, thus minimizes the misclassification. The interaction of cognitive component and meta-cognitive component address the what-to-learn, when-to-learn and how-to-learn human learning principles efficiently. The performance of the PBL-McRBFN is evaluated using a set of benchmark classification problems from UCI machine learning repository and two practical problems, viz., the acoustic emission signal classification and the mammogram for cancer classification. The statistical performance evaluation on these problems has proven the superior performance of PBL-McRBFN classifier over results reported in the literature.     

In situ synthesis of gold–polyaniline composite in nanopores of polycarbonate membrane

In situ one-step chemical synthesis route for the preparation of a gold–polyaniline composite in nanopores of polycarbonate (PC) membrane is reported. PC membrane, which was placed in a specially designed two-compartment cell, separated the aqueous solution of aniline from HAuCl₄ solution. Concentration gradient across the membrane caused movement of AuCl₄ ⁻ and anilinium ions in the pores of polycarbonate membrane. Nanopores in PC membrane acted as reaction vessels where aniline and HAuCl₄ were allowed to mix together, and the redox reaction between aniline and HAuCl₄ led to the formation of gold–polyaniline composite. The gold–polyaniline composite in PC membrane was characterised by EDXRF, XRD, UV–Vis spectroscopy, FTIR and TEM. Peak broadening in XRD suggests that Au particles formed in the membrane are nanocrystallites and average crystallite size is (24 ± 4) nm. TEM studies show that gold nanoparticles are randomly dispersed in polyaniline clusters formed in the nanopores of PC membrane. Characterisation results show that the surfaces of the PC membrane exposed to HAuCl₄ and aniline have significantly higher concentrations of Au nanoparticles and polyaniline, respectively.     

Investigation on natural radiation level and its hazardous nature of river sediments using γ-ray spectroscopy

The content of natural radionuclides (²³⁸U, ²³²Th, ⁴⁰K) in sediment samples of Ponnaiyar River, taken from different depths, were determined using a γ-ray spectroscopic system with a NaI(TI) detector with an aim of evaluating the radiation hazard. The results are compared with Indian and world average values. To know the complete radiological characteristics, the radiological indices such as absorbed dose rate, annual effective dose equivalent (indoor and outdoor), radium equivalent activity, hazard index (H _ex and H _in), γ-index, activity utilization index, and excess lifetime cancer risk were calculated. The results obtained were discussed and compared with the recommended values. The outdoor γ-dose rates were measured at each sampling site using an environmental radiation dosimeter. The measured results were compared with literature values and correlated with the calculated absorbed dose rate. Averages of all the radiation hazard indices and concentrations of all the measured radionuclides in all the layers are lower than the recommended level. The origin of the detected radionuclides is assessed by comparing the present results with the corresponding values for the earth crust and other rock formations.