This paper presents a high-precision capacitive angular position sensor (CAPS). The CAPS is designed to be excited by a single voltage to eliminate the matching errors of multi-excitations, and it is mainly composed of excitation electrodes, coupling electrodes, petal-form sensitive electrodes and a set of collection electrodes. A sinusoidal voltage is applied on the excitation electrodes, then the voltage couples to the coupling electrodes and sensitive electrodes without contact. The sensitive electrodes together with the set of collection electrodes encode the angular position to amplitude-modulated signals, and in order to increase the scale factor, the sensitive electrodes are patterned in the shape of petal-form sinusoidal circles. By utilizing a resolver demodulation method, the amplitude-modulated signals are digitally decoded to get the angular position. A prototype of the CAPS is fabricated and tested. The measurement results show that the accuracy of the sensor is 0.0036°, the resolution is 0.0009° and the nonlinearity over the full range is 0.008° (after compensation), indicating that the CAPS has great potential to be applied in high-precision applications with a low cost.

Hemispherical resonator gyroscope (HRG) is a high precision inertial sensor for measurement of rate and attitude angle, and the fused silica hemispherical resonator is key component of HRG which the ultra-low frequency split is required. During the entire fabrication process of the resonator, the imperfects of the resonators should be trimmed by evaluating the quality-factor and frequency split in real-time. Therefore, the high-precision balance method based on synchronous trimming system has been proposed in this paper, and the quality-factor, frequency split and principle-axis angle could be identified at the same time in a single test. At first, the high precision orbital parameter identification algorithm based on nonlinear least square method has been adopted, which shows the identification error of frequency split and stiffness-axis-angle is less than 3.58×10 ^-7 Hz and 0.0207° respectively. Then, the hardware of trimming system including vacuum chamber with ion mill and neutralizer, fiber optic interferometer and FPGA control board with the proposed identification algorithm has been implemented. Finally, the focus ion beam (FIB) etching experiment has been carried out to verify the proposed trimming method. The experimental result shows that the proposed method has excellent performance, which can reduce the frequency split of all resonators under 0.0005Hz and the best result can achieve 0.00005Hz.

We use ANN models to estimate the performance of latency-critical and delay-tolerant tasks when applying edge cooperation in EC-enabled EONs. Using the performance estimation, a resource allocation scheme is proposed to jointly optimize multiple performance.

Stochastic video generation methods predict diverse videos based on observed frames, where the main challenge lies in modeling the complex future uncertainty and generating realistic frames. Numerous of Recurrent-VAE-based methods have achieved state-of-the-art results. However, on the one hand, the independence assumption of the variables of approximate posterior limits the inference performance. On the other hand, although these methods adopt skip connections between encoder and decoder to utilize multi-level features, they still produce blurry generation due to the spatial misalignment between encoder and decoder features at different time steps. In this paper, we propose a hierarchical recurrent VAE with a feature aligner, which can not only relax the independence assumption in typical VAE but also use a feature aligner to enable the decoder to obtain the aligned spatial information from the last observed frames. The proposed model is named Hierarchical Stochastic Video Generation network with Aligned Features, referred to as HAF-SVG. Experiments on Moving-MNIST, BAIR, and KTH datasets demonstrate that hierarchical structure is helpful for modeling more accurate future uncertainty, and the feature aligner is beneficial to generate realistic frames. Besides, the HAF-SVG exceeds SVG on both prediction accuracy and the quality of generated frames.

Rotation modulation (RM) technique is mainly used in the high precision inertial navigation systems. With the rapid development of MEMS gyroscopes, rotation modulation systems based on the MEMS inertial devices have also been proposed and researched in recent years. Due to the commonly used high voltage (compare with 5V) DC motors in some of the present MEMS RM platforms, the power consumption and size make it unsuitable for the man-portable devices. In this paper, a low voltage, man-portable MEMS Rotation Modulation Platform (RMP) is proposed. A low voltage PMSM is customized to get a smooth torque and a high power density. The PMSM Vector Control (VC) algorithm is introduced to achieve the PMSM's better performance. Experiments demonstrate that the single position tracking error is less than 0.5" and the rate error of continuous rotation is less than 0.3%. Furthermore, the volume is only 110mm*150mm*50mm, the total power is less than 2 watts, which makes it an attracting rotation modulation platform for man-portable individual navigation equipment.