Data-driven methods have shown promising performance in power converter fault diagnosis. However, the existing methods are bothered by the inadequacy of accuracy and robustness, due to the feature variation in the different operating points, especially in nonstationary fault signals. This article proposes a robust data-driven fault diagnosis method combining automatic feature extraction (AFE) and improved random forests (IRFs). In AFE, this method uses optimal discrete wavelet transform, optimized by the defined information component function, to extract multiresolution time-frequency features from nonstationary fault signals, and the improved ReliefF algorithm is used to select important fault features considering both the independent and interactive weights to overcome nonstationary influence. To enhance the robustness, the IRFs adjusts the weight of decision trees by robustness testing based on out-of-bag data, which extracts the mapping relationship between fault features and fault labels for fault location. The comparative experimental results show that the proposed method achieves 99.94% fault diagnosis accuracy, and has superior fault diagnosis robustness compared to existing methods.

Reinforcement learning (RL) has proven to be an effective approach for achieving intelligence in Cognitive Radio (CR). Through interactions with the environment, RL enables a CR to optimize in an efficient and flexible manner. The vast majority of studies, however, are carried out in a spectrum environment with prefixed user access rules, typically with a constant transition probability and reward distribution. In fact, in a real-world spectrum environment, changes in access rules are common, which has a significant impact on the effectiveness of RL, while few studies have been conducted on this topic. This paper demonstrates how changes in primary user’s (PU) access rules affect RL strategies. To improve the secondary user’s (SU) performance for the dynamic spectrum environment, a transfer Deep Q-Network (DQN) is proposed, this method screens out knowledge from historical experience while avoiding interference from irrelevant information with an experience playback mechanism. Experiments show that this method outperforms traditional RL methods in terms of conflict rate, spectrum utilization rate, and convergence rate in the dynamic spectrum. Given the scarcity of studies on this topic, this study is expected to serve as a benchmark for the future research.

We present a novel TE $_{\text{20}}$ -mode 90 $^{\circ}$ waveguide twist concept for millimeter-wave (mm-wave) and terahertz (THz) applications. This waveguide twist has the unique characteristic of mode filtering; that is, only the desired TE $_{\text{20}}$ mode can be rotated and output, the fundamental TE $_{\text{10}}$ mode, and higher order modes will be reflected. In the design, the input/output (I/O) channels are orthogonal oversized waveguides, the TE $_{\text{20}}$ mode is converted into a cross-waveguide (cross-WG) with gradually changed arm lengths. Broad bandwidth (BW) and high efficiency can be easily achieved in a compact size. To verify the proposed concept, two 160–270-GHz prototypes based on rectangular waveguides (RWGs) and groove gap waveguides (GGWs) were designed, fabricated, and measured. The RWG-based prototype was milled in one metal block, while the GGW-based one was fabricated in two metal split blocks. The measured results reveal that the two prototypes achieve average insertion losses of 0.21 and 0.36 dB for the interested TE $_{\text{20}}$ mode, and minimum suppressions of 45 and 35 dB for the unwanted TE $_{\text{10}}$ mode over the desired frequency band. The features of polarization rotation and mode filtering make the proposed TE $_{\text{20}}$ -mode waveguide twist be a candidate for compact, broadband, and efficient waveguide systems.

In the high-resolution analog circuit, the performance of chips is an important part. The performance of the chips needs to be determined by testing. According to the test requirements, stimulus signal with better quality and performance is necessary. The main research direction is how to generate high-resolution and high-speed analog signal when there is no suitable high-resolution and high-speed digital-to-analog converter (DAC) chip available. In this paper, we take the high-resolution analog-to-digital converter (ADC) chips test as an example; this article uses high-resolution DAC chips and multiplexers to generate high-resolution high-speed signals that can be used for testing high-resolution ADC chips based on the principle of time-alternating sampling. This article explains its method, analyzes its error and proposes a digital pre-processing method to reduce the error. Finally, the actual circuit is designed, and the method is verified on the circuit. The test results prove the effectiveness of this method for generating high-resolution ADC test signals.

Image registration plays a vital role in the mosaic process of multiple UAV (Unmanned Aerial Vehicle) images acquired from different spatial positions of the same scene. Aimed at the problem that many fast registration methods cannot provide both high speed and accuracy simultaneously for UAV visible light images, this work proposes a novel registration framework based on a popular baseline registration algorithm, ORB—the Oriented FAST (Features from Accelerated Segment Test) and Rotated BRIEF (Binary Robust Independent Elemental Features) algorithm. First, the ORB algorithm is utilized to extract image feature points fast. On this basis, two bidirectional matching strategies are presented to match obtained feature points. Then, the PROSRC (Progressive Sample Consensus) algorithm is applied to remove false matches. Finally, the experiments are carried out on UAV image pairs about different scenes including urban, road, building, farmland, and forest. Compared with the original version and other state-of-the-art registration methods, the bi-matching ORB algorithm exhibits higher accuracy and faster speed without any training or prior knowledge. Meanwhile, its complexity is quite low for on-board realization.

White Rabbit (WR) time synchronization has an accuracy up to a sub-nanosecond level. However, the current application scenario of WR is limited to wired transmission links. In this paper, we have proposed a time synchronization technique over a free space link using WR. In the WR-based free space synchronization scheme, we replace the original WDM (Wavelength Division Multiplexing) with single-wavelength transmission to reduce the asymmetry of the path and design a high-power optical transceiver module to improve the transmission power. With the scheme, a free space synchronization experiment with a transmission distance of 50 m is conducted. The experimental results show that the RMS (root mean square) time drift of this free space synchronization system is 20.5 ps over a 24 h period, and the TDEV (Time Deviation) of the time synchronization is 14.3 ps at 1 s and 3.9 ps at 20,000 s. The experiment proves that it will be convenient to complete the free space time synchronization network between clock sites with the proposed technique in the future application of complex environments.