scholarly journals BwimNet: A Novel Method for Identifying Moving Vehicles Utilizing a Modified Encoder-Decoder Architecture

Sensors ◽  
2020 ◽  
Vol 20 (24) ◽  
pp. 7170
Author(s):  
Yuhan Wu ◽  
Lu Deng ◽  
Wei He
Keyword(s):  
Signal Reconstruction ◽  
Safety Evaluation ◽  
Large Data ◽  
Original Data ◽  
Dynamic Strain ◽  
Traffic Loading ◽  
Coupling Vibration ◽  
Bridge Model ◽  
Moving Vehicles ◽  
Decoder Architecture

Traffic loading monitoring plays an important role in bridge structural health monitoring, which is helpful in overloading detection, transportation management, and safety evaluation of transportation infrastructures. Bridge weigh-in-motion (BWIM) is a method that treats traffic loading monitoring as an inverse problem, which identifies the traffic loads of the target bridge by analyzing its dynamic strain responses. To achieve accurate prediction of vehicle loads, the configuration of axles and vehicle velocity must be obtained in advance, which is conventionally acquired via additional axle-detecting sensors. However, problems arise from additional sensors such as fragile stability or expensive maintenance costs, which might plague the implementation of BWIM systems in practice. Although data-driven methods such as neural networks can estimate traffic loadings using only strain sensors, the weight data of vehicles crossing the bridge is difficult to obtain. In order to overcome these limitations, a modified encoder-decoder architecture grafted with signal-reconstruction layer is proposed in this paper to identify the properties of moving vehicles (i.e., velocity, wheelbase, and axle weight) using merely the bridge dynamic response. Encoder-decoder is an unsupervised method extracting higher features from original data. The numerical bridge model based on vehicle-bridge coupling vibration theory is established to illustrate the applicability of this new encoder-decoder method. The identification results demonstrate that the proposed approach can predict traffic loadings without using additional sensors and without requiring vehicle weight labels. Parametric studies also show that this new approach achieves better stability and reliability in identifying the properties of moving vehicles, even under the circumstances of large data pollution.

scholarly journals Wireless Transmission Method for Large Data Based on Hierarchical Compressed Sensing and Sparse Decomposition

Sensors ◽  
2020 ◽  
Vol 20 (24) ◽  
pp. 7146
Author(s):  
Youtian Qie ◽  
Chuangbo Hao ◽  
Ping Song
Keyword(s):  
Compressed Sensing ◽  
Signal Reconstruction ◽  
Large Data ◽  
Original Data ◽  
Wireless Transmission ◽  
Wireless Sensor ◽  
Transmission Method ◽  
Widespread Application ◽  
Sparse Decomposition ◽  
Network Bandwidth

With the widespread application of wireless sensor networks, large-scale systems with high sampling rates are becoming more and more common. The amount of original data generated by the wireless sensor network is very large, and transmitting all the original data back to the host wastes network bandwidth and energy. This paper proposes a wireless transmission method for large data based on hierarchical compressed sensing and sparse decomposition. This method includes a hierarchical signal decomposition method based on the same sparse basis and different sparse basis hierarchical compressed sensing method with a mask. Compared with the traditional compressed sensing method, this method reduces the error of signal reconstruction, reduces the amount of calculation during signal reconstruction, and reduces the occupation of hardware resources. We designed comparison experiments between the traditional compressed sensing algorithm and the method proposed in this article. In addition, the experiments’ results prove that our proposed method reduces the execution time, as well as the reconstruction error, compared with the traditional compressed sensing algorithm, and it can achieve better reconstruction at a relatively low compression ratio.

Measurement of Dynamic Strain

1943 ◽  
Vol 10 (2) ◽  
pp. A85-A92
Author(s):  
C. O. Dohrenwend ◽  
W. R. Mehaffey
Keyword(s):  
High Speed ◽  
Shock Loading ◽  
Low Frequency ◽  
Dynamic Strain ◽  
Time Axis ◽  
Low Frequencies ◽  
Strain Measurements ◽  
Moving Vehicles ◽  
Machine Parts ◽  
Dynamic Strains

Abstract The measurement of dynamic strains of both high and low frequency give rise to a variety of problems in instrumentation. Two types of equipment and circuits designed and used by the authors are discussed in detail. The first type based on the amplitude-modulated method is for low frequencies from zero to about 15 per cent of the carrier frequency of 1025 cycles per sec. The equipment has application to strain measurements varying from static values to those produced in moving vehicles, various machine parts, structures such as crane bridges, in fact all strain measurements where the frequency is 150 cycles per sec or less. The second type of equipment discussed is a potentiometer type and is for high-frequency strain measurements from 100 cycles per sec to 8000 cycles per sec. This high-speed equipment is conveniently used for impact strain, such as produced in hammer blows, shock loading, forging equipment, and impact-factor determination. Both units are designed to be used with a cathode-ray oscillograph which lends itself to a variety of recording methods. The methods discussed include both the type where the time axis is obtained by sweeping the oscilloscope beam on a stationary film and where the time axis is obtained mechanically.

scholarly journals LSTM with Wavelet Transform Based Data Preprocessing for Stock Price Prediction

2019 ◽  
Vol 2019 ◽  
pp. 1-8 ◽  
Author(s):  
Xiaodan Liang ◽  
Zhaodi Ge ◽  
Liling Sun ◽  
Maowei He ◽  
Hanning Chen
Keyword(s):  
Wavelet Transform ◽  
Stock Price ◽  
Signal Reconstruction ◽  
Data Preprocessing ◽  
Original Data ◽  
Training Data ◽  
Threshold Method ◽  
Denoising Method ◽  
Stock Price Prediction ◽  
Price Prediction

For profit maximization, the model-based stock price prediction can give valuable guidance to the investors. However, due to the existence of the high noise in financial data, it is inevitable that the deep neural networks trained by the original data fail to accurately predict the stock price. To address the problem, the wavelet threshold-denoising method, which has been widely applied in signal denoising, is adopted to preprocess the training data. The data preprocessing with the soft/hard threshold method can obviously restrain noise, and a new multioptimal combination wavelet transform (MOCWT) method is proposed. In this method, a novel threshold-denoising function is presented to reduce the degree of distortion in signal reconstruction. The experimental results clearly showed that the proposed MOCWT outperforms the traditional methods in the term of prediction accuracy.

scholarly journals Operational and Safety Evaluation of Freeways with Posted Minimum Speed Limit

2010 ◽  
Author(s):  
Victor Muchuruza ◽  
Renatus Mussa
Keyword(s):  
Safety Evaluation ◽  
Maximum Speed ◽  
Speed Limit ◽  
Crash Data ◽  
Minimum Speed ◽  
Driving Speed ◽  
Moving Vehicles ◽  
Slow Moving ◽  
The Mean ◽  
The Difference

An operational and safety evaluation was conducted in relation to the posting of the minimum speed limit of 40 mph and the maximum speed limit of 70 mph on the Florida rural interstate freeway system. The results showed that 57% of the recorded vehicles exceeded the maximum speed limit. Additionally, while only 0.14% of recorded vehicles had speeds below the 40 mph posted minimum speed limit, 9% of crash-involved vehicles were estimated to have speeds below 40 mph. The overrepresentation of slow-moving vehicles in the crash data suggests that even a small proportion of under-40-mph vehicles can have negative implications on safety. Thus, regulation of vehicle speeds at the lower end of the speed distribution is equally important. The second order polynomial model developed to estimate the risk of a vehicle being involved in a crash as a function of the speed deviation from the mean speed of traffic indicated that the minimum risk occurred when the driving speed was 8 mph above the mean speed, equal to the 85th percentile speed observed in the field. Further, the Poisson regression modeling indicated that the difference between the 85th and 15th percentile speeds had a positive effect on crashes.

scholarly journals HOW MACHINE LEARNING METHOD PERFORMANCE FOR IMBALANCED DATA

TEKNOKOM ◽  
2021 ◽  
Vol 4 (2) ◽  
pp. 48-52
Author(s):  
Pardomuan Robinson Sihombing
Keyword(s):  
Machine Learning ◽  
Regression Tree ◽  
Imbalanced Data ◽  
Large Data ◽  
Original Data ◽  
Classification And Regression Tree ◽  
Support Vector ◽  
Method Performance ◽  
Survey Statistics ◽  
Working Status

This study will examine the application of several classification methods to machine learning models by taking into account the case of imbalanced data. The research was conducted on a case study of classification modeling for working status in Banten Province in 2020. The data used comes from the National Labor Force Survey, Statistics Indonesia. The machine learning methods used are Classification and Regression Tree (CART), Naïve Bayes, Random Forest, Rotation Forest, Support Vector Machine (SVM), Neural Network Analysis, One Rule (OneR), and Boosting. Classification modeling using resample techniques in cases of imbalanced data and large data sets is proven to improve classification accuracy, especially for minority classes, which can be seen from the sensitivity and specificity values that are more balanced than the original data (without treatment). Furthermore, the eight classification models tested shows that the Boost model provides the best performance based on the highest sensitivity, specificity, G-mean, and kappa coefficient values. The most important/most influential variables in the classification of working status are marital status, education, and age.

scholarly journals Deep residual detection of radio frequency interference for FAST

2020 ◽  
Vol 492 (1) ◽  
pp. 1421-1431 ◽  
Author(s):  
Zhicheng Yang ◽  
Ce Yu ◽  
Jian Xiao ◽  
Bo Zhang
Keyword(s):  
Radio Frequency ◽  
Large Data ◽  
High Sensitivity ◽  
Original Data ◽  
Training Data ◽  
Radio Frequency Interference ◽  
Data Sets ◽  
Data Set ◽  
Time Required ◽  
Key Steps

ABSTRACT Radio frequency interference (RFI) detection and excision are key steps in the data-processing pipeline of the Five-hundred-meter Aperture Spherical radio Telescope (FAST). Because of its high sensitivity and large data rate, FAST requires more accurate and efficient RFI flagging methods than its counterparts. In the last decades, approaches based upon artificial intelligence (AI), such as codes using convolutional neural networks (CNNs), have been proposed to identify RFI more reliably and efficiently. However, RFI flagging of FAST data with such methods has often proved to be erroneous, with further manual inspections required. In addition, network construction as well as preparation of training data sets for effective RFI flagging has imposed significant additional workloads. Therefore, rapid deployment and adjustment of AI approaches for different observations is impractical to implement with existing algorithms. To overcome such problems, we propose a model called RFI-Net. With the input of raw data without any processing, RFI-Net can detect RFI automatically, producing corresponding masks without any alteration of the original data. Experiments with RFI-Net using simulated astronomical data show that our model has outperformed existing methods in terms of both precision and recall. Besides, compared with other models, our method can obtain the same relative accuracy with fewer training data, thus reducing the effort and time required to prepare the training data set. Further, the training process of RFI-Net can be accelerated, with overfittings being minimized, compared with other CNN codes. The performance of RFI-Net has also been evaluated with observing data obtained by FAST and the Bleien Observatory. Our results demonstrate the ability of RFI-Net to accurately identify RFI with fine-grained, high-precision masks that required no further modification.

scholarly journals Investigation on Seismic Damage Model Test of a High Concrete Gravity Dam Based on Application of FBG Strain Sensor

Complexity ◽  
2019 ◽  
Vol 2019 ◽  
pp. 1-12 ◽  
Author(s):  
Mingming Wang ◽  
Jianyun Chen ◽  
Hai Wei ◽  
Bingyue Song ◽  
Weirong Xiao
Keyword(s):  
Strain Gauge ◽  
Safety Evaluation ◽  
Strain Sensor ◽  
High Sensitivity ◽  
Model Material ◽  
Gravity Dam ◽  
Scale Model ◽  
Dynamic Strain ◽  
Small Scale ◽  
Small Scale Model

A 203-m-high gravity dam being built in earthquake-prone areas needs to be investigated very carefully to determine its dynamic responses, damage mechanism, and safety evaluation. The dynamic characteristics, seismic responses, failure mode, and safety evaluation of the above structure are presented through dynamic fracture test for small-scale model on shaking table. Because the strength of the model material is very low, the traditional strain gauge is also not easy to be glued to the surface of model. It is difficult to measure the accurate strain data of small-scale model during testing. Therefore, Fiber Bragg Grating (FBG) strain sensor is presented to obtain the strain of small-scale model during testing, due to its high sensitivity. The dynamic strain and residual strain are obtained with the FBG sensors embedded in model. The FBG sensor is adhered to model material completely and shows advantages of ease for installation, high sensitivity, and reliability compared with traditional resistance strain gauge. The model during testing is submitted with earthquake wave from the Chinese Code. In the experiment, the peak ground acceleration (PGA) of the first crack in the model indicates the safety level of the gravity dam. The crack locations and forms determine the damageable part of gravity dam under intense earthquake. After the final analysis, the safety evaluation result of the gravity dam under strong earthquake is given in order to guide the implementation of the project.

Testing and Statistics Analysis on Dynamic Response of Expressway Bridge under Various Vehicles

2011 ◽  
Vol 90-93 ◽  
pp. 1033-1038
Author(s):  
Tao Wang ◽  
Wan Shui Han ◽  
Yan Wei Li
Keyword(s):  
Dynamic Response ◽  
Rapid Development ◽  
Bridge Engineering ◽  
Dynamic Responses ◽  
Critical Parameters ◽  
Dynamic Strain ◽  
Coupling Vibration ◽  
Vehicle Type ◽  
Traffic Lane ◽  
Video Recorder

Nowadays, with the rapid development of the traffic infrastructure construction and the growing of the traffic flowing and speed, the vehicle-bridge coupling vibration research has become the focus of the bridge engineering study. The dynamic response of the bridge under the traffic flowing is one of the vital parameters for the vehicle-bridge coupling vibration analysis. In this paper, a methodology, employing the speed radar gun, the video recorder, and the dynamic strain tester in combination with manually recording is used to continuously and detailed investigate the traffic loads on the expressway bridge within 24 hours a day. With this approach introduced by this paper, all the critical parameters, such as the vehicle type, speed, traffic lane, the arriving time of the traffic and the bridge-vehicle dynamic interaction are all recorded. In this investigation, firstly the dynamic responses of 8 pieces of girders of the bridge under 5650 individual vehicles driving through the bridge are recorded, then in conjunction with the investigated traffic flowing samples, in terms of the vehicle type, some detailed statistics study is conducted on the collected records, and finally the space-time distribution laws of the dynamic response of the bridge under the traffic flowing are studied extensively. The result of this study could provide helpful theoretic guidance and supporting data for the vehicle-bridge coupling vibration research.

scholarly journals AUTOMATIC CAL/VAL METHODS LOWERING PRODUCTION EFFORT OF UPDATING LARGE CORE SPATIAL DATA

2020 ◽  
Vol XLII-3/W10 ◽  
pp. 167-171
Author(s):  
A. Peled
Keyword(s):  
Spatial Data ◽  
Supervised Classification ◽  
Ad Hoc ◽  
Spatial Information ◽  
Value Added ◽  
Large Data ◽  
Original Data ◽  
Data Bases ◽  
Other Information ◽  
Vector Information

Abstract. There are basically two levels of calibrations and validation of digitally acquired spectral and other information via sensors carried on space-borne or airborne platforms. The basic level is carried out by the data producers executed by comparison made of results taken over test fields for example. The second level, more a part of a supervised classification effort are carried by the data users and value added spatial information users or providers to edge users. The latter is quite typical for supervised classification protocols. This is either for establishing libraries of spectral signatures for each relevant class-type or for ad-hoc classification where no previous information or specific knowledge wee kept. Such methods indicate and support even strongly the need of the basic Cal/Val step of the sensors made by the original data providers. The paper is reviewing the method of database-driven concept that allows for automatic recognition of detected features within the digital spatial 2-D (yet) realm to its identification within the digital 2.5D spatial vector information within existing large Big-data national core spatial data bases to be updated. These Large data bases are Big enough to operate the resourceful Munchhausen method of self-pulling information out of the huge abandon of data resources.

Sign in / Sign up

Export Citation Format

Share Document