With the continuous development of aerospace production and tasks, the actual demand for environmental monitoring in aerospace field, especially the monitoring and early warning of pollutants, is increasing. The tradi...
With the continuous development of aerospace production and tasks, the actual demand for environmental monitoring in aerospace field, especially the monitoring and early warning of pollutants, is increasing. The traditional aerospace environment monitoring technology has certain shortcomings, such as not enough method, weak adaptability and difficulty in fully mining the potential value of aerospace data acquired. A data-driven intelligent monitoring technology for aerospace environment is designed based on the time series neural network LSTM with automatic feature extraction. The simulation results show that the intelligent monitoring technology for aerospace environment based on time series neural network could automatically extract the sequence features, and the prediction error is reduced by about 10% compared with the traditional monitoring technology, and has higher data utilization and adaptability.
To improve the performance of the cooperative vehicular network (CVN) system, hybrid decode-amplify forward (HDAF) protocol, antenna selection (AS), and equal gain combining (EGC) techniques are applied to the N-Nakag...
To improve the performance of the cooperative vehicular network (CVN) system, hybrid decode-amplify forward (HDAF) protocol, antenna selection (AS), and equal gain combining (EGC) techniques are applied to the N-Nakagami fading channel. The relays apply HDAF protocol, and the destination employs the EGC. Optimal antenna selection (OAS) and sub-optimal antenna selection (SAS) are proposed based on the received signal-to-noise ratio, and the outage probability expressions are derived. The simulation results show that the theoretical values are consistent with the simulation values. In addition, the number of antennas, channel fading coefficients and channel cascade number significantly impact the system performance.
Improving the prediction accuracy of energy consumption in office buildings is necessary to achieve high energy efficiency in smart buildings. The existing forecasting methods rarely analyze the periodic characteristi...
Improving the prediction accuracy of energy consumption in office buildings is necessary to achieve high energy efficiency in smart buildings. The existing forecasting methods rarely analyze the periodic characteristics of energy consumption independently. In this paper, a short-term office building energy consumption prediction model (DLnet) is proposed to address the problem of inefficiency in the utilization of periodic energy consumption ***, the period component of the energy consumption data is decomposed using STL, and the optimal period of the energy consumption data is searched for by a grid search algorithm, and then the Periodic block is constructed based on the optimal period; Secondly, the Time-series block data is constructed according to the data shape of the Periodic block; then the Time-series block data and the Periodic block data are trained and learned using LSTM; Finally, the prediction results of the Time-series block data and the Periodic block data are fused by linear *** four prediction accuracy indicators of the proposed model have been demonstrated to be 7%, 21%, 25% ,and 26% higher than those of the LSTM model.
Face recognition technology has been studied since the 1960s, and face detection is the most critical step in face recognition. In the early days of computer vision, face detection was a classical problem that deeply ...
Face recognition technology has been studied since the 1960s, and face detection is the most critical step in face recognition. In the early days of computer vision, face detection was a classical problem that deeply studied machine vision. It has important applications in security surveillance, human-witness comparison, human-computer interaction, social networks, etc. However, it is slightly lacking in the accuracy of detection. The current streaming media is growing, and the characters of movies and TV shows are gradually being replaced. Many people want to know the chain in the videos but do not know the information related to the people playing them. To realize the detection work of face recognition in streaming media, this paper adopts the Faster-RCNN model and adopts the method of fusing multiple loss functions in the calculation of object loss, which reduces the loss of feature object framing caused by the overlap rate and the loss of not including feature objects, to solve the problem of missing the overlapping faces and thus improve the accuracy of face detection. The model in this paper is compared with the previously used Faster-RCNN model to determine the stability and reliability of the model. Finally, the trained face detection model is put into the face detection of streaming media to frame the face accurately and provide a convenient way for the subsequent face recognition development. The overall recognition accuracy of this experiment reaches 93.2%, and the reliability of the accuracy is high throughout the results.
This paper proposes a human motion twin system for real-time motion capture and reconstruction, which has the following advantages: low cost, high environmental adaptability, high degree of motion restoration, and str...
This paper proposes a human motion twin system for real-time motion capture and reconstruction, which has the following advantages: low cost, high environmental adaptability, high degree of motion restoration, and strong versatility. Through sensor initialization calibration, the estimated motion angle error has been minimized. Test experiments are conducted to investigate the accuracy of limb movement twinning, thus verifying the motion twinning effectiveness of the proposed system. The results show that the proposed system can accurately capture and reproduce limb motion, realistically replicating human activities in real-time. A satisfactory level of subtle differences in sensor-measured motion data between real and mannequins in the virtual domain has been found. During locomotion, the variations between the rotation angles obtained from the three-axis ZYX sensor and the virtual rotation angle of a randomly chosen limb are minimized, with values of only 0.128, 0.027, and 0.031, respectively. In addition, the system enables visualization, preservation, rewinding and further analysis of limb movement data, which means it can provide a new and practical technology for rehabilitation medicine, physical therapy, ergonomics research and the creation of special effects in film and television, etc.
As a national musical instrument, the sounding mechanism and acoustic characteristics of the lute is worthy of our indepth study, and there is a lack of relevant standards for quality evaluation. The main work of this...
As a national musical instrument, the sounding mechanism and acoustic characteristics of the lute is worthy of our indepth study, and there is a lack of relevant standards for quality evaluation. The main work of this thesis is to firstly study the string vibration of Lute, analyze its error hull formula according to the principle of lute string vibration, and study the influence of different plucking positions on the fundamental frequency and harmonic frequency of Lute, and to discuss the relationship of the fundamental error of the lute musical instrument and the influence of different plucking states on the occurrence frequency ; secondly, calculate the vibration equation of Lute strings to get. We have recorded a large number of Lute signals and analyzed them in professional numerical signal analysis software, using a series of signal processing methods and MATLAB programs edited by ourselves to analyze lute audio signals. First, a series of preprocessing such as sampling, noise removal, predicted weighting, etc. are carried out for the collected pipa tone signal. Then the time domain analysis, frequency domain analysis, pitch detection and spectral centroid detection are carried out for the preprocessing sound signal. We can quantify the amount of information of these acoustic features to a certain extent, and preliminarily study the change of the acoustic characteristics of Lute, which is the follow-up theoretical research of Lute Lay the groundwork with process improvements. From these analyses and calculations, the objective physical parameters related to the sound quality of the pipa are found to establish a comprehensive discrimination system for the quality of the lute.
Telecommunication networks have become increasingly important in our society. To best satisfy the needs and interests of customers, the operators must be able to offer high quality services at the best price. However,...
Telecommunication networks have become increasingly important in our society. To best satisfy the needs and interests of customers, the operators must be able to offer high quality services at the best price. However, users are often faced with the problems of network saturation and congestion, which considerably limits the transfer of their calls to destination. To deal with this, we propose in this paper a new approach of cooperation between the antennas evolved Node Base station (eNBs) of a Long Term Evolution (LTE) network in order to minimize and reduce the saturation of these eNBs and thus avoid as much as possible congestion in the network. This approach, called Approach based Defensive Alliance for Reducing ENBs Saturation (ADARES), is based on the concept of defensive alliances in graphs that ensures an effi-cient collaboration between the eNBs of an operator or several operators in order to best transmit the calls made by the users. For validation, we propose two analytical models based on Markov chains to compare our “ADARES” approach with a concurrent approach Load Balancing via Coalition Formation (LBCF). The obtained analytical comparison results are favorable to “ADARES” by showing its performance compared to “LBCF” in terms of successful call transmission. This is mainly made possible in our approach thanks to the collaboration between the eNBs of the network which is carried out based on the principle of the defensive alliances in graphs.
Roshal Archive (RAR) format is one of the most widely used data archive formats, enabling users to reduce the size of data and protect it with the desired password before the data is transferred to its intended recipi...
Roshal Archive (RAR) format is one of the most widely used data archive formats, enabling users to reduce the size of data and protect it with the desired password before the data is transferred to its intended recipients over the network. This work focuses on the security of encrypted RAR archives and various different approaches for their decryption. Two different datasets composed of randomly generated and real-world user passwords were used for deploying brute force and dictionary attacks on password-protected RAR archives. Two available and widely used tools, John the Ripper and Hashcat, were used for cracking passwords of encrypted RAR3 and RAR5 archives. Experimental results indicate that both brute force and dictionary attacks were unsuccessful for RAR archives protected with randomly generated passwords, even of very small length. Real-world user passwords were successfully cracked only partially by brute force attacks, whereas dictionary attacks were very successful. The success rate for RAR5 archives was only slightly lower than for RAR3 archives and processing times were similar, indicating that this new version of the RAR format does not significantly improve data security. Instead, the security of RAR archives can be increased by using longer passwords more similar to randomly generated data, which are not present in commonly used dictionaries, as indicated by the experimental results.
Various controllable units in micro-grid make it possible to provide ancillary services for power grid under grid connection mode. In this paper, a new energy management strategy using peak time rebate strategy for mi...
Various controllable units in micro-grid make it possible to provide ancillary services for power grid under grid connection mode. In this paper, a new energy management strategy using peak time rebate strategy for micro-grid is proposed to reduce the peak load of the grid. In this strategy, real-time data of controllable load, battery storage, diesel generator and other controllable units in the micro-grid is collected, transmitted and analyzed through the Internet of Things technology, and these units voluntarily participate in the demand response to maximize revenue and optimize energy management of the micro-grid. The results show that this strategy performs well in shaving peak load of the main grid as well as ensuring optimal operation of the micro-grid.
In the radar automatic recognition of human targets, the extracted human micro-Doppler features are commonly used to identify human movement states or action attitudes. Affected by human posture and noise, human micro...
In the radar automatic recognition of human targets, the extracted human micro-Doppler features are commonly used to identify human movement states or action attitudes. Affected by human posture and noise, human micro-Doppler features are sometimes weak and vague, which is difficult to extract stably and use for classification. In this paper, a new behavior recognition method for micro-Doppler spectra is proposed based on FMCW (Frequency-modulated continuous wave) radar. Firstly, MSR (Multi-Scale Retinex) algorithm is used to enhance the constructed micro-Doppler map information. Finally, a deep convolutional neural network (ResNet50) is used to extract the micro-Doppler features of the spectrum and classify six actions such as walking and drinking. The average classification accuracy of the six actions reached 88.22%, which verified that the method of using MSR to enhance the original atlas proposed in this paper is feasible and effective for human action classification.
暂无评论