Learning the dynamics of embedded clauses

Recent work by Siegelmann has shown that the computational power of recurrent neural networks matches that of Turing Machines. One important implication is that complex language classes (infinite languages with embedded clauses) can be represented in neural networks. Proofs are based on a fractal encoding of states to simulate the memory and operations of stacks. In the present work, it is shown that similar stack-like dynamics can be learned in recurrent neural networks from simple sequence prediction tasks. Two main types of network solutions are found and described qualitatively as dynamical systems: damped oscillation and entangled spiraling around fixed points. The potential and limitations of each solution type are established in terms of generalization on two different context-free languages. Both solution types constitute novel stack implementations - generally in line with Siegelmann's theoretical work - which supply insights into how embedded structures of languages can be handled in analog hardware.

Digitally-implemented naturally sampled PWM suitable for multilevel converter control

For dynamic closed loop control of a multilevel converter with a low pulse number (ratio of switching frequency to synthesized fundamental), natural sampled pulse-width modulation (PWM) is the best form of modulation. Natural sampling does not introduce distortion or a delayed response to the modulating signal. However previous natural sampled PWM implementations have generally been analog. For a modular multilevel converter, a digital implementation has advantages of accuracy and flexibility. Re-sampled uniform PWM is a novel digital modulation technique which approaches the performance of natural PWM. Both hardware and software implementations for a five level multilevel converter phase are presented, demonstrating the improvement over uniform PWM.

The synthesis and X-ray crystal structure of 9-carboxyhexahydro-7-methoxy-4a,7-ethano-benzopyran-5-en-1-one

9-Carboxyhexahydro-7-methoxy-4a,7-ethano-benzopyran-5-en-1-one (1) was prepared and examined by X-ray crystallography to probe its potential as a new peptide scaffold/template. The crystal structure of the anhydride precursor 7-(2-acetoxyethyl)-4-methoxy-3a,4,7,7a-tetrahydro-4,7-ethanoisobenzofuran-1,3-dione (6) is also reported.