Publications

Development and experimental verification of an IoT sensing system for drive-by bridge health monitoring

Vehicles equipped with various types of sensors have the great potentials to effectively evaluate the health conditions of a population of bridges at a low cost. However, existing drive-by structural health monitoring (SHM) methods acquire vehicle vibration responses offline and export them to a computer for postprocessing. Furthermore, the vehicle trajectory information on the bridge is important for scaling up the drive-by SHM for in situ applications, which is not synchronously measured by existing systems. Therefore, a single-board computer-based IoT sensing system for continuous and real-time drive-by bridge health monitoring is developed in this study. The developed IoT sensing system integrates a triaxial microelectromechanical system (MEMS) accelerometer, temperature sensor, GPS and 4G module on Raspberry Pi 4 Model B. The sensor node can be mounted on a moving vehicle to collect the triaxial acceleration responses, temperature and GPS information. A graphical user interface (GUI) is developed based on the Python Tkinter package to remotely control the sensor node and visualise the collected data in real time. The fast Fourier transform of the measured acceleration responses is performed on the sensor node inboard processor. The raw data are sent to both the cloud server and remote terminal computer through a 4G module. The goal is to provide a low-cost, accurate and scalable sensing system for easy implementation of drive-by bridge health monitoring. The system architecture and workflow of the developed IoT sensing system are presented in detail. A series of experimental tests are conducted to validate the accuracy of the measured acceleration responses and feasibility of using the developed IoT sensing system for drive-by SHM applications.

Research on the Cooling Performance and Economical Efficiency of Heat Reflective Coating and Heat Resistance Adhesive Layer

The Optimization Design Analysis of Loukouni Bridge in the Republic of Congo

Loukouni bridge is a deck type arch bridge with box section which is a part from the second stage of the 1st national road project of the Republic of Congo. This paper focuses on the optimization design analysis of the Loukouni bridge to make it more economical and reliable. The optimization analysis is implemented by ANSYS and contains two parts. One part focuses on the whole shape of the arch and try to optimize the arch axial coefficient. The other part considers the optimization of the size of the box section. The analysis results could provide useful suggestions for the design of Loukouni bridge and related steel arch bridges.

The antecedents and innovation outcomes of firms’ absorptive capacity in global buyer–supplier relationships

Temporal variability of hydro-chemical characteristics and water quality assessment of collapse pond in Zhuxianzhuang coal mining area, China.

Simulation on wellbore integrity at casing shoe during fracturing for shale gas wells

Wellbore integrity is significant to maintain and improve the production performances of shale wells. In Duvernay Canada, casing deformation near the top of Ireton with a few natural faults and cracks is severe during multi-fracturing. It is urgent to reveal the mechanism to reduce the risk of wellbore failure. In this paper, casing deformation and micro-seismic signal at casing shoe is analysed. The maximum deformation of the casing can reach to be 44.4mm. Based on the focal mechanism, it is easy to get the formation slip displacement. Under the condition of geology and wellbore geometry, a three-dimensional stage finite element method considering the whole drilling process is established to simulate the influence of fault on casing deformation. The results indicate that cement sheath at casing shoes intend to be failure during fracturing, where the fracturing fluid will immerse into the formation of Ireton thorough the micro-annulus of cement sheath. When the pore pressure is large enough to activate the natural fault, the micro-seismic signal at the casing shoe is frequent with the magnitude up to 3. Under this condition, the faults slippage can be 55 mm, and the casing deformation will be 34.9 mm. This is consistent with the actual deformation of casing. Along easy-slip formation position, cement property and wellbore structure should be optimized to prevent fracturing fluid entering the formation. Fracturing operation should be optimized to avoid generating high-magnitude seismic signals during the fracturing process, thereby reducing the possibility of casing shear deformation.

Tensile characteristics of strain-hardening cement-based composites with different curing ages

The Design and Application of SVG for Low Voltage Distribution Networks

SVG can compensate reactive power deficiency, suppress harmonics, improve three-phase imbalance and power quality more flexibly. There are very few small volume SVG products available for low voltage distribution network in the past. The generic SVG products are very expansive, thus not suitable for low voltage distribution network. Therefore, it is an urgent task to design a new generation distribution network SVG product that offers good value for money. This paper studied a SVG digital controller based on TMS320F28335 DSP chip. The fast and powerful computing and parallel operation capability of TMS320F28335 can satisfy the real-time, multifunction and multiple objective coordination control of SVG. Applied the instantaneous reactive power theory and adopted current direct control mode, an enhanced filtering algorithm to filter instantaneous sampling value is proposed. Automatic bi-directional compensation control strategy effectively reduced voltage variation at the user side. Its effectiveness is verified by an engineering project.

Construction stability analysis of large section highway tunnel portal section based on numerical analysis

Abstract the entrance portal of No.8 tunnel of F3 bid section of E60 highway in Georgia is shallow buried and under unsymmetrical pressure, which leads to high risk and difficulty in tunnel construction. Tunnel 8 is designed based on ADECO-RS (analysis of controlled deformation in rock and soil). In order to understand the stability characteristics of tunnel portal in the whole section construction process and determine reasonable and feasible engineering countermeasures, the numerical simulation analysis of tunnel construction process is carried out. We not only studied the relationship between the stability of the slope at the entrance and the stability of the whole tunnel body under the actual stratum conditions, but also studied the mechanical and deformation characteristics of the surrounding rock and supporting structure of the tunnel.The numerical analysis results show that the stability characteristics of the portal slope directly affect the overall stability of the tunnel portal section; due to the disturbance of tunnel excavation, a large range of plastic zone appears at the interface of rock and soil at the portal section, resulting in large settlement deformation of vault. The collapse and deformation of the upper soil layer with poor engineering properties will cause excessive additional load on the supporting structure of the tunnel.Therefore, advance support measures should be taken to control the excessive deformation of vault. At the same time, due to the influence of asymmetric pressure, the supporting structure of the shallow buried side and the deep buried side of the tunnel appears obvious asymmetric pressure state. It is suggested to strengthen the stiffness of the shallow buried side support structure and adopt the asymmetric support structure to ensure the safety of the support structure.

Deep residual network framework for structural health monitoring

Convolutional neural networks have been widely employed for structural health monitoring and damage identification. The convolutional neural network is currently considered as the state-of-the-art method for structural damage identification due to its capabilities of efficient and robust feature learning in a hierarchical manner. It is a tendency to develop a convolutional neural network with a deeper architecture to gain a better performance. However, when the depth of the network increases to a certain level, the performance will degrade due to the gradient vanishing issue. Residual neural networks can avoid the problem of vanishing gradients by utilizing skip connections, which allows the information flowing to the next layer through identity mappings. In this article, a deep residual network framework is proposed for structural health monitoring of civil engineering structures. This framework is composed of purely residual blocks which operate as feature extractors and a fully connected layer as a regressor. It learns the damage-related features from the vibration characteristics such as mode shapes and maps them into the damage index labels, for example, stiffness reductions of structures. To evaluate the efficacy and robustness of the proposed framework, an intensive evaluation is conducted with both numerical and experimental studies. The comparison between the proposed approach and the state-of-the-art models, including a sparse autoencoder neural network, a shallow convolutional neural network and a convolutional neural network with the same structure but without skip connections, is conducted. In the numerical studies, a 7-storey steel frame is investigated. Four scenarios with considering measurement noise and finite element modelling errors in the data sets are studied. The proposed framework consistently outperforms the state-of-the-art models in all the scenarios, especially for the most challenging scenario, which includes both measurement noise and uncertainties. Experimental studies on a prestressed concrete bridge in the laboratory are conducted. The proposed framework demonstrates consistent damage prediction results on this beam with the state-of-the-art models.

The new method of seismic attributes extraction and fault detection based on the rotating rhombus method

The conventional methods of attribute extraction and calculation have many limitations for complex low level faults. Therefore, we propose a new method of seismic attribute extraction based on the rhombus that can be arbitrarily rotated and regulated its axis which achieves a satisfactory result on the application of fault identification. Research shows that: 1) the attributes extracted by rotating rhombus have better ability of direction identification, and it is particularly suit to identify the fault and fracture zone; 2) the attribute extracted by rotating rhombus its major to minor axis ratio is 5:2 has the best resolution; 3) we can extract a variety of amplitude and energy attributes by using the method, among which the total absolute amplitude attribute or total energy attribute is better for fault description; 4) our method has a good noise immunity, which can extract accurate direction attribute if the SNR is greater than 1, which can completely fulfill the require of data interpretation; 5) the method can also be used to quantitatively identify the shattered fault zones.

The Application of Triangular FAHP to Evaluate MOOC Quality

In order to improve the quality of massive online open courses (MOOCs), we proposed an evaluation method applying a triangular fuzzy analytical hierarchy process (FAHP)-based comprehensive evaluation model. In the model, FAHP was used to determine the weights of indexes at each level, which proved effective in avoiding the fuzziness of expert evaluations in index assignment. In addition, we also illustrated how to construct the evaluation model, taking maritime-related vocational education as an example. The research results showed that the proposed model is scientific and rational, and can be well-applied to the reality.

Structural damage identification by sparse deep belief network using uncertain and limited data

The accuracy of structural damage identification is affected by the uncertainties in the vibration measurements and the finite element modeling. This paper proposes a novel approach based on sparse deep belief network (DBN) for structural damage identification with uncertain and limited data. Vibration characteristics, that is, natural frequencies and mode shapes, are extracted as the input to the network, while the output are the damage locations and severities of the structure. DBN is chosen to train the generated data sets and identify structural damages. Restricted Boltzmann Machines (RBMs) are used as building blocks to composite a DBN. To further enhance the capacity of the RBMs, an arctan-based sparse constraint is utilized to enable the hidden units to become sparse. This is achieved by adding an arctan norm constraint on the whole of the hidden units' activation probabilities. Numerical and experimental studies are conducted to verify the accuracy and performance of the proposed method. Undetermined damage identification is conducted, in which the quantity of input modal data is less than that of the system parameters to be identified. The identification results show that the proposed sparse DBN based on arctan can identify the damage effectively, and its accuracy is better than those obtained by other methods, even when the modeling uncertainty and the measurement noise exist and only limited data is available.

Design of Sliding Mode Variable Structure Control System for Resonant Fiber Optic Gyroscope

Research on Semi-Supervised Gearbox Fault Diagnosis Method Using Masked Contrastive Learning Under Time-Varying Operating Conditions

SHPB試験とシミュレーションに基づくTi‐6Al‐4V合金加工のための構成モデル化【Powered by NICT】

Modes recognition in high-power fiber laser by convolutional neural networks

In this paper, mode distribution in large-mode-area (LMA) 25/400 fiber was investigated while attempts to recognize and sort different modes with their combination were carried out on a CNN net via Tensorflow. VGG16 model was chosen as the backbone net through several test to increase precision. The model was trained on a dataset including 6000 pictures in 15 categories. And the final accuracy was up to 0.98. It indicates that recognizing modes in high power fiber laser system based on a CNN net was a feasible plan in the mode control assignment.

Structural damage identification with power spectral density transmissibility: numerical and experimental studies

This paper proposes a structural damage identification approach based on the power spectral density transmissibility (PSDT), which is developed to formulate the relationship between two sets of auto-spectral density functions of output responses. The accuracy of response reconstruction with PSDT is investigated and the damage identification in structures is conducted with measured acceleration responses from the damaged state. Numerical studies on a seven-storey plane frame structure are conducted to investigate the performance of the proposed damage identification approach. The initial finite element model of the structure and measured acceleration measurements from the damaged structure are used for the identification with a dynamic response sensitivity-based model updating method. The simulated damages can be identified accurately without and with a 5% noise effect included in the simulated responses. Experimental studies on a steel plane frame structure in the laboratory are performed to further verify the accuracy of response reconstruction with PSDT and validate the proposed damage identification approach. The locations of the introduced damage are detected accurately and the stiffness reductions in the damaged elements are identified close to the true values. The identification results demonstrated the accuracy of response reconstruction as well as the correctness and efficiency of the proposed damage identification approach.