The extensive deployment of smart meters in millions of households provides a huge amount of individual electricity consumption data for demand side analysis at a fine granularity. Different from traditional aggregated system-level data, smart meter data is more irregular and unpredictable. As a result, probabilistic load forecasting, which can provide a better understanding of the uncertainty and volatility in future demand, is critical to constructing energy-efficient and reliable smart grids. In this paper, a recently developed technique called Bayesian deep learning is employed to solve this challenging problem. In particular, a novel multitask probabilistic load forecasting framework based on Bayesian deep learning is proposed to quantify the shared uncertainties across distinct customer groups while accounting for their differences. Further, a clustering-based pooling method is designed to increase the data diversity and volume for the framework. This not only addresses the problem of overfitting but also improves the predictive performance. Numerical results are presented which demonstrate that the proposed framework provides superior probabilistic forecasting accuracy over conventional methods.

Non-intrusive load monitoring (NILM) is a technique that infers appliance-level energy consumption patterns and operation state changes based on feeder power signals. With the availability of fine-grained electric load profiles, there has been increasing interest in using this approach for demand-side energy management in smart grids. NILM is a multi-label classification problem due to the simultaneous operation of multiple appliances. Recently, deep learning based techniques have been shown to be a promising approach to solving this problem, but annotating the huge volume of load profile data with multiple active appliances for learning is very challenging and impractical. In this paper, a new semi-supervised multi-label deep learning based framework is proposed to address this problem with the goal of mitigating the reliance on large labeled datasets. Specifically, a temporal convolutional neural network (CNN) is used to automatically extract high-level load signatures for individual appliances. These signatures can be efficiently used to improve the feature representation capability of the framework. Case studies conducted on two open-access NILM datasets demonstrate the effectiveness and superiority of the proposed approach.

This communication presents new designs of dual-circularly polarized (CP) planar aperture antenna and array for 60 GHz applications. First, a four-port planar aperture antenna with its feeding network is developed that exhibits wide impedance bandwidth, dual-CP radiation as well as high gain. Then, based on the proposed antenna, a new scheme of building a dual-CP array is proposed. With this new scheme that combines the power splitting network and series feeding method, four sequentially fed antenna elements are successfully combined to form a CP radiation array. The experimental results show that the -10 dB impedance bandwidths of the antenna and the array are more than 18.2% (55-66 GHz). While the 3 dB axial ratio bandwidths are from 54.2 to 64.3 GHz (17.2%) for the antenna and from 54.8 to 65 GHz (17%) for the array. The maximum gains achieved are 13.7 and 17.85 dBic for the antenna and array, respectively.

A wideband highly isolated MIMO patch antenna with an artificial magnetic conductor (AMC) ground plane is proposed in this letter. The antenna, which is rotationally symmetrical, is constructed by V-shaped patches coupled-fed by a bowtie dipole. The bowtie dipole and the V-shaped parasitic patches produce two resonant modes that are coupled together. Therefore, a wide impedance bandwidth is obtained. Thanks to the orthogonal feeding and the rotational symmetrical geometry, the antenna exhibits high port isolation as well as symmetrical radiation patterns with a compact size of 40${\times}$40${\times}$1.524 mm3. By placing an artificial magnetic conductor (AMC) ground plane with a distance of 10.5 mm below the radiating element, the antenna can achieve a unidirectional high gain with a lower height of the profile compared to that with PEC ground. Measured results show the antenna has an impedance bandwidth of 31.0% (3.0-4.1 GHz) with port isolation as low as -25 dB. The peak gain can reach 7.1 dBi with stable and symmetrical radiation patterns.

This paper presents a new dual-band dual circularly polarized (CP) high gain patch antenna array for vehicle satellite communications. The array consists of 16 linearly polarized dual-band elements backed with a new frequency-selective-surface-integrated artificial magnetic conductor (FAMC) ground. The FAMC is located underneath the array, which not only acts as a reflector to increase the array boresight gain but also enables the array's dual CP radiation. The reflected electromagnetic (EM) waves by the FAMC retard either 90° or 270° with respect to the EM waves of the array directly radiated to the upper sphere at two frequency bands, leading to the dual-band dual CP radiation. This is totally different from many of the former counterparts which are based on either the perturbation on the stacked patch or external feeding network. Simulated and measured results show that the -10-dB impedance bandwidth is from 7.8 to more than 8.5-GHz for X-band and from 14 to 15.3-GHz for Ku-band. The 3-dB axial ratio is from 8.15 to 8.35-GHz for the lower band (right-hand CP) and from 14.2 to 14.8-GHz for the higher band (left-hand CP), respectively.