Mixed-frequency testing of induction machines using inverters

This work has concentrated on the testing of induction machines to determine their temperature rise at full-load without mechanically coupling to a load machine. The achievements of this work are outlined as follows. 1. Four distinct categories of mixed-frequency test using an inverter have been identified by the author. The simulation results of these tests as well as the conventional 2-supply test have been analysed in detail. 2. Experimental work on mixed-frequency tests has been done on a small (4 kW) squirrel cage induction machine using a voltage source PWM inverter. Two out of the four categories of test suggested have been tested and the temperature rise results were found to be similar to the results of a direct loading test. Further, one of the categories of test proposed has been performed on a 3.3 kW slip-ring induction machine for the conformation of the rotor values. 3. A low current supply mixed-frequency test-rig has been proposed. For this purpose, a resonant bank was connected to the DC link of the inverter in order to maintain the exchange of power between the test machine and the resonant bank instead of between the main supply and the test machine. The resonant bank was then replaced with a special electro-mechanical energy storage unit. The current of the main power supply was then reduced in amplitude. 4. A variable inertia test for full load temperature rise testing of induction machines has been introduced. This test is purely mechanical in nature and does not require any electrical connection of the test machine to any other machine. It has the advantage of drawing very little net power from the supply.

Direct current ripple compensation for multi-phase fault-tolerant machines

A second-harmonic direct current (DC) ripple compensation technique is presented for a multi-phase, fault-tolerant, permanent magnet machine. The analysis has been undertaken in a general manner for any pair of phases in operation with the remaining phases inactive. The compensation technique determines the required alternating currents in the machine to eliminate the second-harmonic DC-link current, while at the same time minimising the total rms current in the windings. An additional benefit of the compensation technique is a reduction in the magnitude of the electromagnetic torque ripple. Practical results are included from a 70 kW, five-phase generator system to validate the analysis and illustrate the performance of the compensation technique.

Computational capabilities of multilayer committee machines

We obtained an analytical expression for the computational complexity of many layered committee machines with a finite number of hidden layers (L < 8) using the generalization complexity measure introduced by Franco et al (2006) IEEE Trans. Neural Netw. 17 578. Although our result is valid in the large-size limit and for an overlap synaptic matrix that is ultrametric, it provides a useful tool for inferring the appropriate architecture a network must have to reproduce an arbitrary realizable Boolean function.

Load losses in synchronous machines:a study of the pole face loss caused by armature reaction M.M.F. harmonics

DUE TO COPYRIGHT RESTRICTIONS ONLY AVAILABLE FOR CONSULTATION AT ASTON UNIVERSITY LIBRARY SERVICES WITH PRIOR ARRANGEMENT

Learning in ultrametric committee machines

The problem of learning by examples in ultrametric committee machines (UCMs) is studied within the framework of statistical mechanics. Using the replica formalism we calculate the average generalization error in UCMs with L hidden layers and for a large enough number of units. In most of the regimes studied we find that the generalization error, as a function of the number of examples presented, develops a discontinuous drop at a critical value of the load parameter. We also find that when L>1 a number of teacher networks with the same number of hidden layers and different overlaps induce learning processes with the same critical points.

A novel framework of training hidden Markov support vector machines from lightly-annotated data

Natural language understanding (NLU) aims to map sentences to their semantic mean representations. Statistical approaches to NLU normally require fully-annotated training data where each sentence is paired with its word-level semantic annotations. In this paper, we propose a novel learning framework which trains the Hidden Markov Support Vector Machines (HM-SVMs) without the use of expensive fully-annotated data. In particular, our learning approach takes as input a training set of sentences labeled with abstract semantic annotations encoding underlying embedded structural relations and automatically induces derivation rules that map sentences to their semantic meaning representations. The proposed approach has been tested on the DARPA Communicator Data and achieved 93.18% in F-measure, which outperforms the previously proposed approaches of training the hidden vector state model or conditional random fields from unaligned data, with a relative error reduction rate of 43.3% and 10.6% being achieved.

Steps and bumps:precision extraction of discrete states of molecular machines

We report statistical time-series analysis tools providing improvements in the rapid, precision extraction of discrete state dynamics from time traces of experimental observations of molecular machines. By building physical knowledge and statistical innovations into analysis tools, we provide techniques for estimating discrete state transitions buried in highly correlated molecular noise. We demonstrate the effectiveness of our approach on simulated and real examples of steplike rotation of the bacterial flagellar motor and the F1-ATPase enzyme. We show that our method can clearly identify molecular steps, periodicities and cascaded processes that are too weak for existing algorithms to detect, and can do so much faster than existing algorithms. Our techniques represent a step in the direction toward automated analysis of high-sample-rate, molecular-machine dynamics. Modular, open-source software that implements these techniques is provided.

A comparison of compacting and caking behaviour of carbonate-based washing powders

Two types of sodium carbonate powder produced by spray drying (SD) and dry neutralisation (DN) were studied for their compaction properties using a uniaxial compression tester. Dry neutralised sodium carbonate showed a greater resistance to compression and also produced a weaker compact when compressed to 100kPa. Differential Scanning Calorimetry (DSC) showed that both types of powder were predominantly amorphous in nature. Moisture sorption measurements showed that both powders behaved in a similar way below 50% RH. However, dry neutralised sodium carbonate had a high moisture affinity above this RH. On examining the particle structures using Scanning Electron Microscopy (SEM), the most likely explanation for the increased tendency of spray dried sodium carbonate to form strong compacts was the hollow particle structure.

A comparison of compacting and caking behavior of carbonate-based washing powders

Two types of sodium carbonate powder produced by spray drying (SD) and dry neutralization (DN) were studied for their compaction properties using a uniaxial compression tester. A comparison was also made with Persil washing powder. Dry neutralized sodium carbonate showed greater resistance to compression and also produced a weaker compact when compressed to 100 kPa. Spray-dried sodium carbonate had an absence of fine particles but compacted easily. Differential scanning calorimetry (DSC) showed that both types of powder were predominantly amorphous in nature. Moisture sorption measurements showed that both powders behaved in a similar way below 50% relative humidity (RH). However, dry neutralized sodium carbonate had a high moisture affinity above this RH. Particle structures were also examined using scanning electron microscopy, showing the heterogeneous interior of the spray-dried particles. © 2013 Copyright Taylor and Francis Group, LLC.

The characterisation and emulation of rotary mechanical machines

DUE TO COPYRIGHT RESTRICTIONS ONLY AVAILABLE FOR CONSULTATION AT ASTON UNIVERSITY LIBRARY AND INFORMATION SERVICES WITH PRIOR ARRANGEMENT

A study of stray load losses in core clamp plates of salient pole synchronous machines

DUE TO COPYRIGHT RESTRICTIONS ONLY AVAILABLE FOR CONSULTATION AT ASTON UNIVERSITY LIBRARY SERVICES WITH PRIOR ARRANGEMENT

The use of electronic techniques in the construction of machines which can learn

DUE TO COPYRIGHT RESTRICTIONS ONLY AVAILABLE FOR CONSULTATION AT ASTON UNIVERSITY LIBRARY AND INFORMATION SERVICES WITH PRIOR ARRANGEMENT

Storage capacity of ultrametric committee machines

The problem of computing the storage capacity of a feed-forward network, with L hidden layers, N inputs, and K units in the first hidden layer, is analyzed using techniques from statistical mechanics. We found that the storage capacity strongly depends on the network architecture αc ∼ (log K)1-1/2L and that the number of units K limits the number of possible hidden layers L through the relationship 2L - 1 < 2log K. © 2014 IOP Publishing Ltd.

Learning higher-level features with convolutional restricted Boltzmann machines for sentiment analysis

In recent years, learning word vector representations has attracted much interest in Natural Language Processing. Word representations or embeddings learned using unsupervised methods help addressing the problem of traditional bag-of-word approaches which fail to capture contextual semantics. In this paper we go beyond the vector representations at the word level and propose a novel framework that learns higher-level feature representations of n-grams, phrases and sentences using a deep neural network built from stacked Convolutional Restricted Boltzmann Machines (CRBMs). These representations have been shown to map syntactically and semantically related n-grams to closeby locations in the hidden feature space. We have experimented to additionally incorporate these higher-level features into supervised classifier training for two sentiment analysis tasks: subjectivity classification and sentiment classification. Our results have demonstrated the success of our proposed framework with 4% improvement in accuracy observed for subjectivity classification and improved the results achieved for sentiment classification over models trained without our higher level features.