Search

Power electronics have been an essential part of the power system, which bring difficulty to system-level real-time simulation. The calculation of network with many converters has received more attention recently. This article focused on the power electronics power grids based on H-bridge converters and presented a universal modeling and solving method of the network based on the algebraic description of switching transformation. First, the circuit structure of the H-bridge converter and its external port conversion relationship were analyzed. Next, the universal power network based on H-bridge converter units was modeled. Based on the model, the calculation method was proposed. Then, the method was extended to networks containing several groups of conventional converters other than H-bridges. Finally, the accuracy and speed of the method were analyzed by an actual project case.

The DC transmission, distribution and networking have been applied to medium and low voltage power grids, and have exhibited apparent advantages and favorable application perspective. Different from traditional AC power grids, DC power grids have shown features of low inertia, weak damping and the inclination of oscillation. The intermix of fast switching transients and control dynamics makes the DC grid much more complicated. In addition, the power fluctuation of renewable energy generation and demand-response loads imposes difficulties on the stable and smooth running of the DC grid. However, no comprehensive study has been reported on dynamic and transient patterns and characteristics of the DC power grid. In this paper, through both simulation and theoretical analysis, the transient and dynamic characteristics of medium and low voltage DC power grids are systematically tested, classified and analyzed. Finally, the mechanisms and general conclusions are extracted and refined.

A dispersion‐free scheme for stable multi‐access radio frequency dissemination in a fibre loop link is proposed and demonstrated. The phase jitter and dispersion‐induced power fading are suppressed simultaneously by transmitting different single‐band signals in opposite propagating directions. Experimentally, a 2.5 GHz radio frequency signal is steadily transmitted to different remote sites, with phase deviations less than 0.6 ps over 15,000 s.

Transient voltage instability always starts from low-voltage load areas and burgeons to high-voltage grids, where the dynamic load is one significant factor. There is an urgent demand for effective and efficient modeling of electrical loads. Thereby, this paper proposes a classification and aggregation approach to model the multi-voltage-level load clusters. Load clusters from online identification are classified and grouped according to their stabilities and transient characteristics. Then, based on the weighted average of induction motor (IM) voltage angles, an approach is proposed to aggregate models of grouped loads sharing similar dynamic features. The proposed approach can aggregate load models from a low voltage level to a higher one layer by layer. Thus, the aggregated model keeps the primary transient characteristics, as well as dramatically reduces the simulation scale and the calculation complexity.

Aiming at the problem of manual design, slow speed, and low efficiency in traditional face detection, this paper proposes a face detection algorithm based on Fast R-CNN. In this method, CNN algorithm, the Haar-Adaboost algorithm, and the candidate search algorithm are used to detect the candidate region of the face that may exist in the image. Then, the candidate region is input to the trained convolution neural network, and after a series of convolution and pooling operations, the final convolution feature (ROI) is based on the Fast R-CNN network structure, and finally, the feature is input to the two full connections (ROI), and then the characteristics of the input image are not specified. Branch, respectively, for feature classification and coordinate regression calculation. The experimental results of the proposed algorithm on the LFW test data set show that the proposed algorithm is compared with the traditional PCA + SVM algorithm, the traditional which avoids the manual design features and higher Identify the accuracy rate and speed up the detection time.

The digital control system delays the system, Considering the control delay and the resonance problem for LCL filter, a undamped control strategy of single-phase LCL type grid connected inverter is proposed. The grid connected current is controlled by network side current feedback. Not only the system unit power factor can be realized, but also no additional sensors are added, so the system control cost is reduced. The feasibility of undamped control is analyzed in detail, and the calculation procedure of controller parameters is given. An experimental platform is built. The correctness and effectiveness of the undamped control strategy are verified on the platform.

A scheme for compensation of dispersion-induced power fading is proposed and experimentally demonstrated. The system transmits two modulated laser lights via a wavelength division multiplexer and a wavelength division demultiplexer. By optimizing the power ratio and the phase difference between the two modulated signals, the dispersion-induced power fading can be adequately compensated. A key significance of our approach is that no adjustments are required when the input radio frequency signal scans over a broad bandwidth of 1 to 12 GHz. We successfully demonstrate the compensation of the power fading for both 20- and 30-km fiber links. An improvement of spurious-free dynamic range is also achieved.

A hybrid diagnostic method utilizing Extended Kalman Filter (EKF) and Adaptive Genetic Algorithm (AGA) is presented for performance degradation estimation and sensor anomaly detection of turbofan engine. The EKF is used to estimate engine component performance degradation for gas path fault diagnosis. The AGA is introduced in the integrated architecture and applied for sensor bias detection. The contributions of this work are the comparisons of Kalman Filters (KF)-AGA algorithms and Neural Networks (NN)-AGA algorithms with a unified framework for gas path fault diagnosis. The NN needs to be trained off-line with a large number of prior fault mode data. When new fault mode occurs, estimation accuracy by the NN evidently decreases. However, the application of the Linearized Kalman Filter (LKF) and EKF will not be restricted in such case. The crossover factor and the mutation factor are adapted to the fitness function at each generation in the AGA, and it consumes less time to search for the optimal sensor bias value compared to the Genetic Algorithm (GA). In a word, we conclude that the hybrid EKF-AGA algorithm is the best choice for gas path fault diagnosis of turbofan engine among the algorithms discussed.

In this article, an improved particle filter with electromagnetism-like mechanism algorithm is proposed for aircraft engine gas-path component abrupt fault diagnosis. In order to avoid the particle degeneracy and sample impoverishment of normal particle filter, the electromagnetism-like mechanism optimization algorithm is introduced into resampling procedure, which adjusts the position of the particles through simulating attraction–repulsion mechanism between charged particles of the electromagnetism theory. The improved particle filter can solve the particle degradation problem and ensure the diversity of the particle set. Meanwhile, it enhances the ability of tracking abrupt fault due to considering the latest measurement information. Comparison of the proposed method with three different filter algorithms is carried out on a univariate nonstationary growth model. Simulations on a turbofan engine model indicate that compared to the normal particle filter, the improved particle filter can ensure the completion of the fault diagnosis within less sampling period and the root mean square error of parameters estimation is reduced.