Search

Speech feature extraction is an indispensable module in the whole speech recognition process. Especially in the energy-constrained internet of things nodes, low-power feature extraction greatly improves the working time of the system. This paper optimizes a complete energy-efficient speech feature extraction architecture in the mixed-signal domain for speech recognition. The speech feature extraction architecture extracts acoustic features in the mixed-signal domain, which significantly reduces the cost of Analog-to-Digital Converter (ADC) and computational complexity. Moreover, the noise robustness of the mixed-signal MFCC feature has been investigated to adapt to the real scene. In order to evaluate the performance of the proposed optimized architecture, we fabricate a chip about the proposed feature extraction architecture in 180nm CMOS process, the post-simulation results show that the core of mixed-signal MFCC feature extraction achieves 70% power saving and enhanced noise robustness than state of the art.

The repeatability of muzzle velocity of electromagnetic railgun has been identified as a potential development issue for several years. The muzzle velocity is not only sensitive to initial parameters as charged voltage of capacitors, mass of armature, starting position, but also dynamic parameters as electric contact resistance and friction between armature and rails. This paper analyzes the effects of initial and dynamic parameters on the muzzle velocity. It is found that the most sensitive parameter is the charged voltage of capacitors. A series of 2 twenty shot bursts were performed with two sets of rails, respectively. The details of the experimental set-up and velocity measurement are described. The maximum deviation of the muzzle velocity is 3.8% with 8.47 g armatures at a mean velocity of 1913 m/s in the first twenty shot burst, and 0.92% with 10.50 g armatures at a mean velocity of 2373 m/s in the second twenty shot burst. It indicates that the repeatability of the high muzzle velocity railgun with heavier armature is better. The effects of rail current and muzzle voltage on the repeatability are discussed. The controllable key parameters affecting the repeatability of muzzle velocity are identified.

A coilgun design code is programmed according to the formulas described the inductive coupling between the thrust coils and the armature, the dynamic response of the armature, and the ohmic heating in the thrust coils and armature. The algorithms determined the operation statuses of the driven pulse forming network are presented. The convergence of the code is analyzed. The maximum time-step and mesh size of the armature are dependent on the rise time of the driven pulse current. The code is validated by comparing the simulated results with the experimental ones. The simulated current wave is agreement with the experimental results. However, the simulated velocity is greater than the measured one about 6.5%. The convergence and validation of the code indicate that it can be used for the primary design for a coilgun system.

Through the literature search, the paper summarizes the experience of information capacity training of European and other developed countries. Combining the fact that information awareness of college students in China is not strong and the actual information capacity is not high, the paper looks for the existing problems and shortcomings, and conducts in-depth analysis and discussion. On this basis, countermeasures and suggestions of the process of designing and building model are put forward to help train information capacity of college students.