Bayesian and decision tree approaches for pattern recognition including feature measurement costs

The minimum cost classifier when general cost functions are associated with the tasks of feature measurement and classification is formulated as a decision graph which does not reject class labels at intermediate stages. Noting its complexities, a heuristic procedure to simplify this scheme to a binary decision tree is presented. The optimization of the binary tree in this context is carried out using dynamic programming. This technique is applied to the voiced-unvoiced-silence classification in speech processing.     

Steady state optimization of an ammonia reactor

A hybrid simulation technique for identification and steady state optimization of a tubular reactor used in ammonia synthesis is presented. The parameter identification program finds the catalyst activity factor and certain heat transfer coefficients that minimize the sum of squares of deviation from simulated and actual temperature measurements obtained from an operating plant. The optimization program finds the values of three flows to the reactor to maximize the ammonia yield using the estimated parameter values. Powell's direct method of optimization is used in both cases. The results obtained here are compared with the plant data.     

A Sensor Network to Monitor Process Parameters of Fermentation and Drying in Black Tea Production

Fermentation and drying process in tea production are two important processes which play a crucial role in producing good quality tea. Tea colour and flavour are formed in the fermentation and enzyme reactions are terminated and moisture contents are reduced from the leaf particles in the drying process. Ambient temperature and relative humidity (RH) are two major factors for the fermentation process. Firing temperature of dryer is also responsible for the final quality of tea. This paper presents an instrument comprising of sensor network to monitor temperature and RH for the fermentation process and firing temperature of the dryer. For dryer temperature monitoring a thermocouple based measuring system is developed. For fermentation an RH to voltage converter and temperature to voltage converter type sensor is used to develop the RH and temperature monitoring sensor node. The sensor nodes for different stages are connected in RS 485 network. Data are logged into the hard drive of a personal computer using the developed data acquisition software. The instrument will be helpful for recording these parameters and so that their influence can be determined and final quality can be improved.     

Application of baby deck initiation to reduce coal damage during cast blasting

ABSTRACT

The extra explosive energy used in cast blasting reduces the amount of overburden to be handled by the machinery but increases the risk of damage to the underlying coal. This damage will lead to reduced coal recovery and may negate much of the benefit sought from cast blasting. The Julius Kruttschnitt Mineral Research Centre in close association with the Australian open cut coal mining industry has conducted a blast monitoring program to understand and quantify the role of cast blasting in coal damage and loss. This paper attempts to explain the mechanisms of coal damage and the influence of toe confinement on coal damage during cast blasting.     

A Revision of OPTEMP-LS Model for Selecting Optimal EMP Combination for Minimizing Sediment and Water Yield from Hilly Urban Watersheds

Water Resources Management - Hilly watersheds are inherently susceptible to more sediment and water yield. Urban developments in hills cause conversion of the natural sloppy surface to bare steep...     

Conditions for entrainment into a FeOx containing slag for a carbon-containing particle in an entrained coal gasifier

When incompletely gasified coal particles impact the slag layer in an entrained slagging gasifier, a situation will arise where the particles will either get engulfed by the slag or re-entrained into the circulating gas. This will be determined by a balance of forces acting on the particle which in turn is governed by the slag and particle properties, and pertinent gasifier conditions, namely temperature, and particle impact velocity. In this paper a model is introduced that takes into account, the drag-, capillary-, and added mass-forces and predicts the behavior of spherical particles of different sizes and slags of different FeO_x contents. The model predicts that particles either get completely submerged, settle at an incompletely separated position, or oscillate at the interface. A sensitivity analysis was performed showing that for a particle of certain size and impact velocity, submersion is most strongly promoted by a low slag viscosity (largely influenced by FeO_x content) and contact angle (determined by whether or not-reactive wetting and liquid Fe formation occurs).     

Planktonic indicators of water quality: A case study in the Amacuzac River Basin (State of Morelos,Mexico)

Planktonic organisms are good indicators of water quality but are rarely used in Mexican riverine studies in spite of the high level of contamination in these lotic waters. In this study, we focused on the plankton of the river Amacuzac in Mexico and a large saline waterbody, Lake Tequisquitengo, which is located in the Amacuzac river basin. We collected samples at three sites in the River Amacuzac and one at Lake Tequisquitengo from May 2015 to March–April of 2016. We analysed the physicochemical variables, coliform bacteria, and biological oxygen demand levels. We also quantified phytoplankton and zooplankton. Our results indicate that the study sites are mesotrophic with high densities of diatoms and cyanobacteria. Among the zooplankton, the dominant organisms were rotifers. The constancy of planktonic taxa was high in the lake but low in the river. The Descy index ranged from 2.9 to 4.5 in the river and from 4.0 to 5.0 in the lake. The saprobic index shows that the system is β‐mesosaprobic. The species diversity was higher in the river than in the lake. Our study shows that plankton can efficiently be used to evaluate the water quality in rivers and lakes.     

Luminescence in Mn-doped CdS nanocrystals

We have synthesized Mn-doped CdS nanocrystals (NCs) with size ranging from 1.8–3 nm. Photoluminescence (PL) spectra of the doped NCs differ from that of the undoped NCs with an additional peak due to Mn d-d transitions. Electron paramagnetic resonance spectra along with X-ray absorption spectroscopy and PL spectra confirm the incorporation of Mn in the CdS lattice. The fact that emissions from surface states and the Mn d levels occur at two different energies, allowed us to study the PL lifetime decay behaviour of both kinds of emissions.