Research on Track Fastener Service Status Detection Based on Improved Yolov4 Model

As an important part of railway lines, the healthy service status of track fasteners was very important to ensure the safety of trains. The application of deep learning algorithms was becoming an important method to realize its state detection. However, there was often a deficiency that the detection accuracy and calculation speed of model were difficult to balance, when the traditional deep learning model is used to detect the service state of track fasteners. Targeting this issue, an improved Yolov4 model for detecting the service status of track fasteners was proposed. Firstly, the Mixup data augmentation technology was introduced into Yolov4 model to enhance the generalization ability of model. Secondly, the MobileNet-V2 lightweight network was employed in lieu of the CSPDarknet53 network as the backbone, thereby reducing the number of algorithm parameters and improving the model’s computational efficiency. Finally, the SE attention mechanism was incorporated to boost the importance of rail fastener identification by emphasizing relevant image features, ensuring that the network’s focus was primarily on the fasteners being inspected. The algorithm achieved both high precision and high speed operation of the rail fastener service state detection, while realizing the lightweight of model. The experimental results revealed that, the MAP value of the rail fastener service state detection algorithm based on the improved Yolov4 model reaches 83.2%, which is 2.83% higher than that of the traditional Yolov4 model, and the calculation speed was improved by 67.39%. Compared with the traditional Yolov4 model, the proposed method achieved the collaborative optimization of detection accuracy and calculation speed.

A Real-Time Urban Traffic Detection Algorithm Based on Spatio-Temporal OD Matrix in Vehicular Sensor Network

Currently, there are kinds of algorithms in order to detect real-time urban traffic condition. Most of these approaches consider speed of vehicles as a main metric to describe traffic situation. In this paper, we find out two important observations through several experiments. (1) In urban city, the speed of vehicles is influenced significantly by some factors such as traffic lights delay and road condition. The actual situation rarely satisfy hypothesis required for these solutions. Therefore, these traditional algorithms do not work well in practical environment. (2) Traffic volume on a road segment shows strong pattern and changes smoothly at adjacent time. This feature of traffic volume inspires us to define a metric: traffic-rate, which is used to detect traffic condition in real time. In our solution, we develop a novel traffic-detection algorithm based on original- destination (OD) matrix. We illustrate our approach and measure its performance in real environment. The performance evaluations confirm the effectiveness of our algorithm.

A reliability evaluation method for embryonic cellular array based on Markov status graph model

Due to the limitations of the existing fault detection methods in the embryonic cellular array(ECA), the fault detection coverage cannot reach 100%. In order to evaluate the reliability of the ECA more accurately, embryonic cell and its input and output(I/O) resources are considered as a whole, named functional unit(FU). The FU fault detection coverage parameter is introduced to ECA reliability analysis, and a new ECA reliability evaluation method based on the Markov status graph model is proposed.Simulation experiment results indicate that the proposed ECA reliability evaluation method can evaluate the ECA reliability more effectively and accurately. Based on the proposed reliability evaluation method, the influence of parameters change on the ECA reliability is studied, and simulation experiment results show that ECA reliability can be improved by increasing the FU fault detection coverage and reducing the FU failure rate. In addition, by increasing the scale of the ECA, the reliability increases to the maximum first, and then it will decrease continuously. ECA reliability variation rules can not only provide theoretical guidance for the ECA optimization design, but also point out the direction for further research.

Rapid and non-destructive detection method for water status and water distribution of rice seeds with different vigor

In this study,newly harvested and aged rice seeds were analyzed to determine their aging process,identify the difference between artificially and naturally aged seeds,and develop a rapid,accurate,and non-destructive detection method for water status and water distribution of rice seed with different vigor.To this end,an artificially accelerated aging test was conducted on the newly harvested rice seeds.Then,low-field nuclear magnetic resonance(LF-NMR)technology was applied to test the new(Shennong No.9816,2018),old(Shennong No.9816,2017),and artificially aged seeds(Shennong No.9816,2018).A standard germination test was conducted for three types of seeds.Finally,the differences of water status and distribution between rice seeds of different vigor were analyzed based on the standard germination test results and wave spectrometry information collected using LF-NMR.The results indicated that new seeds,old seeds,and the artificially accelerated aging rice seeds all exhibited two water phases,and the vigor of rice seeds after the artificial accelerated aging test was lower than that of new seeds.There were significant differences between the frequencies of bound water at the time of the peak and the time at the end of the peak for the three types of seeds.The two times showed an increasing trend for rice seeds with poor vigor,indicating that the ability of the water in the rice seeds having poor vigor to combine with other substances was weakened.There were significant differences between the distributions of free water peak end time for the three types of seeds.All the rice seeds with poor vigor exhibited a decreasing trend at this time,indicating that the freedom of free water inside the rice seed samples with poor vigor was weakened.The total water content of the artificially aged seeds and the aged seeds was higher than that of the new seeds,but the free water content increased from artificially aged seeds to new seeds to aged seeds.This indicates that LF-NMR technology is an effective detection method that can simply compare the differences in seed vitality with respect to water distribution as well as differentiate the seed internal water content of artificially aged and naturally aged seeds.

Pantograph-catenary electrical contact system of high-speed railways:recent progress,challenges,and outlooks

As the unique power entrance,the pantograph-catenary electrical contact system maintains the efficiency and reliability of power transmission for the high-speed train.Along with the fast development of high-speed railways all over the world,some commercialized lines are built for covering the remote places under harsh environment,especially in China;these environmental elements including wind,sand,rain,thunder,ice and snow need to be considered during the design of the pantograph-catenary system.The pantograph-catenary system includes the pantograph,the contact wire and the interface—pantograph slide.As the key component,this pantograph slide plays a critical role in reliable power transmission under dynamic condition.The fundamental material characteristics of the pantograph slide and contact wire such as electrical conductivity,impact resistance,wear resistance,etc.,directly determine the sliding electrical contact performance of the pantograph-catenary system;meanwhile,different detection methods of the pantograph-catenary system are crucial for the reliability of service and maintenance.In addition,the challenges brought from extreme operational conditions are discussed,taking the Sichuan-Tibet Railway currently under construction as a special example with the high-altitude climate.The outlook for developing the ultra-high-speed train equipped with the novel pantograph-catenary system which can address the harsher operational environment is also involved.This paper has provided a comprehensive review of the high-speed railway pantograph-catenary systems,including its progress,challenges,outlooks in the history and future.

Indoor Human Fall Detection Algorithm Based on Wireless Sensing

As the main health threat to the elderly living alone and performing indoor activities,falls have attracted great attention from institutions and society.Currently,fall detection systems are mainly based on wear sensors,environmental sensors,and computer vision,which need to be worn or require complex equipment construction.However,they have limitations and will interfere with the daily life of the elderly.On the basis of the indoor propagation theory of wireless signals,this paper proposes a conceptual verification module using Wi-Fi signals to identify human fall behavior.The module can detect falls without invading privacy and affecting human comfort and has the advantages of noninvasive,robustness,universality,and low price.The module combines digital signal processing technology and machine learning technology.This paper analyzes and processes the channel state information(CSI)data of wireless signals,and the local outlier factor algorithm is used to find the abnormal CSI sequence.The support vector machine and extreme gradient boosting algorithms are used for classification,recognition,and comparative research.Experimental results show that the average accuracy of fall detection based on wireless sensing is more than 90%.This work has important social significance in ensuring the safety of the elderly.

A deep learning method for traffic light status recognition

Real-time and accurate traffic light status recognition can provide reliable data support for autonomous vehicle decision-making and control systems.To address potential problems such as the minor component of traffic lights in the perceptual domain of visual sensors and the complexity of recognition scenarios,we propose an end-to-end traffic light status recognition method,ResNeSt50-CBAM-DINO(RC-DINO).First,we performed data cleaning on the Tsinghua-Tencent traffic lights(TTTL)and fused it with the Shanghai Jiao Tong University’s traffic light dataset(S2TLD)to form a Chinese urban traffic light dataset(CUTLD).Second,we combined residual network with split-attention module-50(ResNeSt50)and the convolutional block attention module(CBAM)to extract more significant traffic light features.Finally,the proposed RC-DINO and mainstream recognition algorithms were trained and analyzed using CUTLD.The experimental results show that,compared to the original DINO,RC-DINO improved the average precision(AP),AP at intersection over union(IOU)=0.5(AP50),AP for small objects(APs),average recall(AR),and balanced F score(F1-Score)by 3.1%,1.6%,3.4%,0.9%,and 0.9%,respectively,and had a certain capability to recognize the partially covered traffic light status.The above results indicate that the proposed RC-DINO improved recognition performance and robustness,making it more suitable for traffic light status recognition tasks.