Search

Pulse injection to simulate partial discharge (PD) is the main method for functional verification of gas insulated switchgear (GIS) ultra high frequency (UHF) PD monitoring device. Due to the different coupling modes of pulse injection in field verification, the equivalence law between simulated PD and actual PD is not clear, and the effectiveness of functional verification of monitoring device cannot be guaranteed. In this paper, a typical PD defect (tip, suspension, insulator bubble) and internal/external pulse injection UHF PD detection platform for 126 kV GIS are established, and the effective pulses of UHF signals are normalized and extracted. Then a signal processing method of weighted intrinsic mode functions (IMF) based on empirical mode decomposition is proposed. The signal equivalence is characterized by calculating the average and maximum values of the Euclidean distance of the signal, and finally compared with the conventional signal deviation method. The research shows that compared with the conventional signal equivalence analysis method, the weighted IMF method can effectively solve the problem of large local difference of UHF signal waveform. The internal sensor pulse injection analog PD and suspended PD have the highest equivalence. The average value Me and the maximum value Ma of Euclidean distance are 3.82% and 10.28%, respectively. Therefore, the function verification of UHF monitoring device can use constant parameter pulse injection instead of suspension defect, and the internal UHF sensor injection can be preferred for analog PD. The research in this paper can provide reference for the pulse injection method for the function verification of UHF PD monitoring device.

Ultra high frequency (UHF) partial discharge (PD) detection is a common method for transformer oil-paper insulation defect location. However, the accuracy of PD positioning in practical application is easily affected by noise and sensor layout. In order to ensure the effectiveness of PD positioning detection of transformer oil-paper insulation defect, an UHF PD positioning detection platform for oil-paper insulation defects is established. Based on the conventional K-means method, a modified clustering demarcation PD anti-interference positioning method is proposed, which effectively reduces the positioning error. Aiming at the problem of sample clustering demarcation aliasing, when the optimal correction coefficient L is 1.1, the UHF PD positioning error can be reduced to less than 0.1 m, which verifies the effectiveness of the proposed method. Finally, the variation law of positioning error of different sensor layout is analyzed, and the optimal layout scheme of UHF sensor for transformer oil-paper insulation defect detection is proposed, which can provide reference for the layout and positioning of transformer on-line monitoring sensor.

For the vertical landing phase of the reusable rocket, in order to improve the landing accuracy with consideration of multiple uncertainties, a novel strategy to calculate the ignition height online is proposed based on polynomial guidance law (PGL), particle swarm optimization (PSO), and deep reinforcement learning (DRL). Firstly, a deep neural network (DNN) is designed to describe the relationship between the state of the reusable rocket and the ignition height. To accomplish the guidance task of the vertical landing phase, PGL is modified by introducing the estimated aerodynamic acceleration. Through simulation, the output range of the DNN is estimated by the modified PSO. Then, the reward function is shaped and the parameters of the DNN are trained on a training set of simulation scenarios by the DRL algorithm. Finally, to demonstrate the effectiveness of the proposed strategy, the trained DNN is used to calculate the ignition height of 1500 unlearned simulation scenarios online. The numerical simulation results show that the proposed strategy has higher landing accuracy and lower fuel consumption than the offline strategy of fixed ignition height based on the modified PSO. [ABSTRACT FROM AUTHOR]

In this paper, a variable-structure multimodel (VSMM) filtering algorithm based on the long short-term memory (LSTM) regression-deep Q network (L-DQN) is proposed to accurately track strong maneuvering targets. The algorithm can map the selection of the model set to the selection of the action label and realize the purpose of a deep reinforcement-learning agent to replace the model switching in the traditional VSMM algorithm by reasonably designing a reward function, state space, and network structure. At the same time, the algorithm introduces a LSTM algorithm, which can compensate the error of tracking results based on model history information. The simulation results show that compared with the traditional VSMM algorithm, the proposed algorithm can quickly capture the maneuvering of the target, the response time is short, the calculation accuracy is significantly improved, and the range of adaptation is wider. Precise tracking of maneuvering targets was achieved. [ABSTRACT FROM AUTHOR]

A kind of CVT mountain bike is designed innovatively,the working principle and structure of the continuously variable transmission（ CVT） and booster are analyzed in detail. The driving force when CVT mountain bike started is calculated,all show that the bike started lightly with the reasonable transmission structure. The finite element analysis（ FEA） of frame and handlebar is carried out,the displacement contour and stress distribution figures of the frame as well as the displacement maps of the handlebar indicate that the structure could meet the requirements with a higher safety factor.