No abstract

A thin film lead zirconate titanate Pb(Zr,Ti)O 3 (PZT), energy harvesting MEMS device is developed to enable self-supportive sensors for in-service integrity monitoring of large social and environmental infrastructures at remote locations. It is designed to resonate at specific frequencies of an external vibrational energy source, thereby creating electrical energy via the piezoelectric effect. Our cantilever device has a PZT/SiN x bimorph structure with a proof mass added to the end. The Pt/Ti top electrode is patterned into an interdigitated shape on top of the sol-gel-spin coated PZT thin film in order to employ the d 33 mode. The base-shaking experiment at the first resonant frequency of the cantilever (170 × 260 μm) generates 1 μW of continuous electrical power to a 5.2 M resistive load at 2.4 V DC. The effect of proof mass, beam shape and damping on the power generating performance are modeled to provide a design guideline for maximum power harvesting from environmentally available low frequency vibrations. A spiral cantilever is designed to achieve compactness, low resonant frequency and minimum damping coefficient, simultaneously.

We report the design, device fabrication, and measurements of tunable silicon photonic band gap microcavities in optical waveguides, using direct application of piezoelectric-induced strain to the photonic crystal. We show, through first-order perturbation computations and experimental measurements, a 1.54 nm shift in cavity resonances at 1.56 m wavelengths for an applied strain of 0.04%. The strain is applied through integrated piezoelectric microactuators. For operation at infrared wavelengths, we combine x-ray and electron-beam lithography with thin-film piezoelectric processing. This level of integration permits realizable silicon-based photonic chip devices, such as high-density optical filters, with active reconfiguration.

Abstract-This paper presents the design, fabrication, and characterization of a piezoelectrically actuated MEMS diffractive optical grating, whose spatial periodicity can be tuned in analog fashion to within a fraction of a nanometer. The fine control of the diffracted beams permits applications in dense wavelength-division multiplexing (DWDM) optical telecommunications and high-resolution miniaturized spectrometers. The design concept consists of a diffractive grating defined on a deformable membrane, strained in the direction perpendicular to the gratings grooves via thin-film piezoelectric actuators. The tunable angular range for the first diffracted order is up to 400 rad with 0.2% strain ( 8 nm change in grating periodicity) at 10 V actuation, as predicted by device modeling. The actuators demonstrate a piezoelectric 31 coefficient of 100 pC/N and dielectric constant r of 1200. Uniformity across the tunable grating and the out-of-plane deflections are also characterized and discussed.[1036]

The problem of power demand forecasting for the effective planning and operation of smart grid, renewable energy and electricity market bidding systems is an open challenge. Numerous research efforts have been proposed for improving prediction performance in practical environments through statistical and artificial neural network approaches. Despite these efforts, power demand forecasting problems remain to be a grand challenge since existing methods are not sufficiently practical to be widely deployed due to their limited accuracy. To address this problem, we propose a hybrid power demand forecasting model, called (c, l)-Long Short-Term Memory (LSTM) + Convolution Neural Network (CNN). We consider the power demand as a key value, while we incorporate c different types of contextual information such as temperature, humidity and season as context values in order to preprocess datasets into bivariate sequences consisting of <Key, Context[1, c]> pairs. These c bivariate sequences are then input into c LSTM networks with l layers to extract feature sets. Using these feature sets, a CNN layer outputs a predicted profile of power demand. To assess the applicability of the proposed hybrid method, we conduct extensive experiments using real-world datasets. The results of the experiments indicate that the proposed (c, l)-LSTM+CNN hybrid model performs with higher accuracy than previous approaches.