Advanced Computational Methods in Life System Modeling and Simulation

In view of some shortcomings of the existing rectal function diagnosis method, we propose that use the artificial anal sphincter system to collect the human rectal pressure signal, and then achieve the diagnosis of human rectal status through the rectal function diagnosis model. Since the collected signal is not pure rectal pressure signal, the single-dimensional pressure signal is extended to a multidimensional time series by phase space reconstruction. And then preprocessing of the reconstructed signal is carried out by the improved fifteenth order Newton iteration Fast ICA algorithm. The improved algorithm is simulated and the better separation effect is realized, proving the feasibility of the algorithm.

Heart rate measurement is important for monitoring people’s physiological and body state. In this paper, a heart rate measurement methodology based on PhotoPlethysmoGraphy (PPG) signal is proposed. Human face positions are detected and tracked in real time by using facial color videos taken from cameras by non-contact shooting. Signals containing pulse components are extracted from images of the forehead skin area for the purpose of calculating blood volume pulse waves via wavelet filtering. Hence, heart rates are calculated after energy spectrum analysis using Fourier transform. The method realizes non-contact measurement, which avoids potential discomfort caused by direct skin contact, and has the advantages of simple operation and low costs. The result indicates that it is sensible to apply this method to daily family heart rate monitoring and remote medical monitoring equipment.

Mechanomyography (MMG) signal is the sound from the surface of a muscle when the muscle is contracted. The traditional filtering algorithms for the processing of MMG signal would make most useful signal filtered when they are used to remove noise. According to MMG signal’s characteristics, a new signal filtering method is presented in this paper based on combining empirical mode decomposition with digital filter, which has a better performance on MMG signal filtering processing in experimental analysis. With extracting the energy feature of wavelet packet coefficient as the feature of classifier, the BP neural network classifier gets a better classification results. The average classification results showed that the best performance for recognizing hand gestures with the energy feature of wavelet packet coefficient features was achieved by BP neural network with the accuracy of 86.41%. This work was accomplished by introducing the new signal filtering method for the recognition of different hand gestures; And suggesting basing on combining empirical mode decomposition with digital filter as a new filtering method in MG-based hand gesture classification.

Single trial electroencephalogram classification is indispensable in online brain–computer interfaces (BCIs) A classification method called adaptive Kernel Fisher Support Vector Machine (KF-SVM) is designed and applied to single trial EEG classification in BCIs. The adaptive KF-SVM algorithm combines adaptive idea, SVM and within-class scatter inspired from kernel fisher. Firstly, the within-class scatter matrix of a feature vector is calculated. And to construct a new kernel, this scatter is incorporated into the kernel function of SVM. Ultimately, the recognition result is calculated by the SVM whose kernel has been changed. The proposed algorithm simultaneously maximizes the discrimination between classes and also considers the within-class dissimilarities, which avoids some disadvantages of traditional SVM. In addition, the within-class scatter matrix of adaptive KF-SVM is updated trial by trail, which enhances the online adaptation of BCIs. Based on the EEG data recorded from seven subjects, the new approach achieved higher classification accuracies than the standard SVM, KF-SVM and adaptive linear classifier. The proposed scheme achieves the average performance improvement of 5.8%,5.2% and 3.7% respectively compared to other three schemes.

The detection of face region can be divided into two kinds: frontal and non-frontal faces. This thesis focuses on the detection of human face region in non-frontal cases. A method of separating face and neck region is presented to extract the non-frontal face in the image. Facial features are usually used in frontal face detection, such as eyes, mouth and etc. With complete facial features, the frontal face can be easier to detected with high accuracy now. However, the research on non-frontal face detection is just beginning. Since the non frontal face image can not provide complete facial features information, it is necessary to develop a new method. Skin color is the most prominent facial feature in the non-frontal cases. It is found that the skin color has better clustering capability in YCbCr color space. According to the skin color characteristics and illumination conditions in the YCbCr color space, the Gaussian model and the Otsu method are used to segment the skin color to extract the non-frontal face region in the images. But the segmented skin color area often contains the neck region. In this paper, the contour line of the chin is fitted by illumination intensity and position information, remove the neck area and get a face region without the neck. Simulation results show the effectiveness of the proposed method for the detection of non-frontal face region.

Based on B-spline neural network, the analysis of membrane potential conduction has been presented for peripheral nerve fibres whereby the effects of the interactions between axons have been taken into account. In particular, the modelling problem is investigated firstly with the vector-valued weight transformation and parameter identification. Using the presented model, the control design is proposed to reproduce the membrane potential along nerve fibres. The algorithm procedure and interaction characterization for coupled axons are given while the numerical simulation illustrates the effectiveness of the presented algorithm.

A computational fluid dynamics (CFD) model based on realistic voxelized representation of human brain tumor vasculature is presented. The model utilizes dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) data to account for heterogeneous porosity and permeability of contrast agent inside the tumor. Patient specific arterial input function (AIF) is employed in this study. Owing to higher accuracy of Leaky Tracer Kinetic Model (LTKM) in shorter duration human imaging data, the model is employed to determine perfusion parameters and compared with General Tracer Kinetic Model (GTKM). The developed CFD model is used to simulate and predict transport, distribution and retention of contrast agent in different parts of human tissue at different times. In future, a patient specific model can be developed to forecast the deposition of drugs and nanoparticles and tune the parameters for thermal ablation of tumors.

The cerebrospinal fluid (CSF) flow in the spinal cord is important in maintaining the stability of the central nervous system. However, the interaction between CSF and spinal cord is not well understood. A three-dimensional (3D) simplified finite element model (FEM) of a sheep CSF and spinal cord segment was developed, verified using clinical experimental data, and used to investigate the effect of deformations and stress distributions on spinal cord in normal physiological conditions. The commercial software ANSYS Workbench was adopted to simulate the unidirectional CSF flow along the coaxial tube, which considered the bi-directional fluid-solid coupling. It was demonstrated that CSF had a slight impact on the spinal cord, which was transmitted to the white and gray matter through the pia mater. The pia mater protected the normal physiological function of the white and gray matter while the spinal dura mater ensured the regular rate and pressure of CSF. It was also showed that the CSF flow in the spinal cord was laminar. This model might help us to better understand the mechanism of interaction between CSF and spinal cord and provide a baseline for mechanical comparisons in spinal cord injury.

The use of customized reconstruction plate is an effective method for reconstruction of mandibular continuity defects. Plate fracture is one of the most common postoperative complications. The aim of this study was to investigate the biomechanical behavior of the customized reconstruction plate by finite element method. The geometry model was created from computed tomography (CT) data of a patient. The muscle forces for the defected mandible under two common static biting tasks were estimated by a numerical optimization strategy with the objective function of minimization of overall muscle force. The simulation results revealed that changing bite from molar region to incisor region increased the maximum stress in the plate. The position of stress concentration, the upper-inner edge of the plate near ramus-end, was in agreement with that of fracture, which indicated that stress concentration regions were critical regions for fracture failure.

Cerebral aneurysm is known to initiate at the cerebral artery bifurcation. The pathological mechanism of cerebral aneurysm awaits further understanding especially on its initiation. This study sought to elucidate the three-dimensional structure of cerebral vascular bifurcations with and without aneurysms using human cadavers. The two cases had aneurysmal initiations out of total 7 cases. The studied structure was intimal hyperplasia, tunica media and internal elastic lamina, which were recognized by elastica masson staining. The results showed that the non-existence of tunica media and internal elastic lamina was found in the lesion without aneurysm. The non-existence of intimal hyperplasia was only found in the lesion with aneurysm. These data suggest that the formation of intimal hyperplasia may be related with the initiation of aneurysm. We regarded the boundary of existence arteriosclerosis as the position for new arteriosclerosis occurs and thought the direction of new arteriosclerosis grows would influence whether the cerebral aneurysm initiates or not.

Objective: To study the biomechanical stability of a new screw-setting technique, we used cortical bone trajectory (CBT) in injury vertebra relative to the traditional pedicle screw-setting technique.Methods: We used thoracolumbar spine CT data of a healthy adult male volunteer and engineering data of internal fixation system of spine to simulate intact state, burst fracture state and combination of three kinds of internal fixation state of the spine: (1) 4 pedicle screws cross segment and 2 rods (P4); (2) 4 pedicle screws, 2 CBT screws at injured vertebrae and 2 rods (P4C2); (3) 6 pedicle screws and 2 rods (P6). Then we compared differences of the stability of the corresponding fixed system and stress distribution of fixation models of three groups above.Results: The total deformation of all nodes of the fracture spine model of P4C2 was less than the fracture spine model node group of P4 and larger than the fracture spine model node group of P6 during normal weight status, rotation(right), bending forward, stretch and lateral bending(right) state. The equivalent stress of all nodes of internal fixation system of P4C2 was smaller than the fixation model node group of P4 and bigger than the fixation model node group of P6 during normal weight status, rotation(right), bending forward, stretch and lateral bending(right) state.Conclusion: CBT technology for injured vertebra fixation could provide more stability of the vertebral body and reduce stress concentration of internal fixation system compared to the traditional P4 fixation.

Radiofrequency ablation (RFA) as an alternative treatment to the conventional open surgery is the most popular minimally invasive thermal therapy, and it is widely used in clinic today. One of the most important limits for the RFA in clinic is the difficulty to deal with the heat-sink effect of blood vessels, as it causes the difficulty of control the RFA process and consequently the coagulation size of RFA is decreased considerably (empirically, the coagulation size is less than 3 cm with a single RFA electrode). This paper reviews the literature of the current solution for the heat-sink effect due to large blood vessels and suggests future work for finding more effective solutions.

To avoid the deformation of spicules surrounding pulmonary nodules caused by the classic rubber band straightening transform (RBST), we propose a novel RBST technique to extract spicules-based features. In this paper, the run-length statistics (RLS) features are extracted from the RBST image, in which a smooth circumference with a suitable radius inside the nodule is proposed as the border of transformed object. An experimental sample set of 814 images of pulmonary nodules was used to verify the proposed feature extraction technique. The best accuracy, sensitivity and specificity achieved based on the proposed features were 79.4%, 66.5%, 89.2%, respectively, and the area under the receiver operating characteristic curve was 87.0%. These results indicate that the proposed method of feature extraction is promising for classifying benign and malignant pulmonary nodules.

It has been commonly believed that concomitant meniscus injuries may alter the dynamical condition of knee joint. The aim of this study was to analyze dynamical characteristics of ACLD knees with or without meniscus deficiency during level walking. The results indicated that meniscus plays an important role in bearing knee rotation moment. Additionally, the deficiency of meniscus could affect the dynamical condition of ACLD knees, especially during mid-stance phase and mid-swing. Future studies should focus on dynamical characteristics during those phases and related muscles.

With a growing number of elderly people in the UK, more and more of them suffer from various kinds of cognitive impairment. Cognitive impairment can be divided into different stages such as mild cognitive impairment (MCI) and severe cognitive impairment like dementia. Its early detection can be of great importance. However, it is challenging to detect cognitive impairment in the early stage with high accuracy and low cost, when most of the symptoms may not be fully expressed. This survey paper mainly reviews the state of the art techniques for the early detection of cognitive impairment and compares their advantages and weaknesses. In order to build an effective and low-cost automatic system for detecting and monitoring the cognitive impairment for a wide range of elderly people, the applications of computer vision techniques for the early detection of cognitive impairment by monitoring facial expressions, body movements and eye movements are highlighted in this paper. In additional to technique review, the main research challenges for the early detection of cognitive impairment with high accuracy and low cost are analysed in depth. Through carefully comparing and contrasting the currently popular techniques for their advantages and weaknesses, some important research directions are particularly pointed out and highlighted from the viewpoints of the authors alone.

This paper presents a method for automated measurement of vascular angle in immunohistochemical images of liver cancer. Firstly, Colour Deconvolution is used to conduct staining separation on a H&E-stained immunohistochemical image, and then blood vessels are segmented using an improved Otsu algorithm. Then the standard SURF algorithm is used to select feature points of the image, and then these feature points are divided into two equal groups according to the distance between individual feature points and the far left (or right) feature point. Finally, a standard least squares method is used to fit two lines using the two groups of points. When the linear deviation of the fitting result based on the two groups of feature points is significant, it is necessary to adjust the belonging of the points of the two groups, and then the two sets are fitted again respectively till the correlation coefficients of the two fitted lines are greater than the predefined threshold, meaning that the measurement of the blood vessel angle in the immunohistochemical map is completed. Compared with the experts’ results, our proposed technique results in better accuracy. It is worthy to point out that, to our knowledge, our system is the first one that conducts automated measurement of blood vessel angle of immunohistochemistry.

In this paper, we present a novel segmentation framework for glandular structures in Hematoxylin and Eosin stained histology images, choosing poorly differentiated colon tissue as an example. The proposed framework’ target is to identify precise epithelial nuclei objects. We start with staining separate to detect all nuclei objects, and deploy multi-resolution morphology operation to map the initial epithelial nuclei positions. We proposed a new bag of words scheme using sparse random feature to classify epithelial nuclei and stroma nuclei objects to adjust the rest nuclei positions. Finally, we can use the boundary of optimized epithelial nuclei objects to segment the glandular structure.

Congenital tracheal stenosis (CTS) has a high clinical mortality in neonates and infants. Although the procedure of slide tracheoplasty (STP) applied over the years, it is still a challenge for clinicians to predict the surgical timing of the CTS correction. In the present study, we studied on three-dimensional (3D) aerodynamic analysis of an original tracheal model from a specific patient with CTS and bridging bronchus (BB) and four new reconstructed models. We constructed a 3D patient-specific tracheal model based on CT images and applied computer-aided design (CAD) to reconstruct four models to imitate the stenosis development of CTS. Average pressure drop (APD), wall shear stress (WSS) and velocity streamlines were calculated to analyze local aerodynamic characteristics for the evaluation of airflow at the inspiration phase and expiration phase, respectively. We found APD, WSS and AEL decreased during the respiration with the decrease of stenosis. Three abnormal gradients in APD were observed between the main stenosis of trachea arrived at 80% and 60%. This implied the surgical correction may be required when the main stenosis reached 60%. The combination of CAD and aerodynamic analysis is a potential noninvasive tool for surgical timing prediction in the management of patient-specific correction of CTS.

Prosthetic hand is usually made by rigid body mechanism with ropes and pulleys. Such a hand is not “soft” to patients or to objects to be manipulated by the hand. In this paper, the concept of compliant mechanism is applied to prosthetic finger. The main challenge in designing and constructing such a finger lies in the design of flexure hinge. First, a fully compliant finger with a monolithic structure and flexure hinge was built. Then, finite element analysis for the compliant finger was implemented, and the results were compared with the experimental result to verify the design. Finally, the complaint finger was applied in a prosthetic hand design and worked excellent with the hand.

It is found that small particles can be successfully manipulated by the acoustic tweezers. This paper presents a method to improve the acoustic trapping capability by punching specific round holes on two vibrating V-shaped metal strips of the acoustic tweezers. A particle is trapped under the sharp edges of metal strips with some specific round holes. Its trapping capability is improved under certain conditions compared with the original acoustic tweezers. A finite element model is developed to calculate the acoustic radiation force. The effects of the radius, the number and the arrangement of the round holes on the acoustic radiation force on the top surface of the particle are discussed. It is found that the acoustic radiation force increases obviously when the radius of the hole is more than a certain magnitude by changing the vibrational mode of the acoustic tweezers. With the increase of number and the row in vertical direction of the round holes, the acoustic radiation force acting on the particle increases correspondingly.

Grasping and moving objects is a natural behavior in human daily life, whereas it turns into an enormous challenge with robots. To analyze the difficulty of grasping and moving target objects, a arm-hand system is performed with 7-DOF dual arms robot and bionic hand in this paper. A numerical method is proposed to solve the problem of arm motion planning. And a novel grasping strategy is proposed for enabling bionic hand to grasp efficiently. Finally, the effectiveness of the proposed methodology is demonstrated using both computer simulation and physical experiment.

This paper presents a control method of grasping force of prosthetic hand. Firstly, the correlated features of surface electromyogram (sEMG) signal that collected by MYO are calculated, and then principal component analysis (PCA) dimension reduction is processed. According to pattern classification model and sEMG-force regression model which based on support vector machine (SVM) to gain the force prediction value. In this approach, force is divided into different grades. The predicted force value is used as the given signal, and grasping force of prosthetic hand is controlled by a fuzzy controller, and combined with vibration feedback device to feedback grasping force value to patient’s arm. The test results show that the method of prosthetic hand grasping force control is effective.

Tendon-driven robot manipulators are often used to actuate distal joints. The tendons allow the actuators to be located outside the fingers. Conventionally, the use of the tendons of the fingers allows for the significant reduction to the size and weight, in this case, which approximately similar to that of the human. To achieve the interaction with unstructured environments, a torque control system is presented based on the single neuron networks (SNN) in this paper. The torque control allows the system maintain proper torques on the joints. Meanwhile, this controller calculates actuator positions based on the error measured by the actual joint torques and desired joint torques. Simulations have been conducted on a tendon-driven finger model to demonstrate that the proposed controller can achieve the faster response, and then decrease overshoot comparing to a PI controller.

In this paper, Human Learning Optimal (HLO) algorithm is presented to solve the scheduling problem. HLO is a meta-heuristic search algorithm which is inspired by the process of human learning. Three learning operators are developed to generate new solutions and search for the optima by mimicking the learning behaviors of human. This new algorithm has been proved to be very effective in solving optimization problems. HLO is applied to solve an actual production scheduling problems in a dairy factory and the performance of HLO is compared with that of two other meta-heuristics algorithms, BSO-PSO and HGA. Comparison results demonstrate that HLO is a promising optimization algorithm.

Aiming at the dependence of the traditional indoor clustering positioning accuracy on the initial center and clustering number selection, an improved WKNN indoor fingerprint localization algorithm based on adaptive H clustering algorithm is proposed in this thesis. Specifically, an adaptive hierarchical clustering combined with positioning environment and fingerprint information without initial clustering center is introduced. At the same time, a RSSI information compensation method based on cosine similarity is proposed aiming at the problem of RSSI information packet loss for test nodes in complicated indoor location environment, with the result of positioning error decrease at test node by using cosine similarity between test nodes and fingerprint points to approximately compensate the missing RSSI information. The experimental results indicate that the proposed adaptive hierarchical clustering algorithm can divide the experimental area adaptively according to fingerprint information, meanwhile the proposed fingerprint information compensation method can decrease the positioning error of the test node with incomplete information, by which the average positioning error in the experimental environment is decreased to 0.78 m compared with other indoor positioning algorithms.

With the rapid development of smart grid, the importance of power load forecast is more and more important. Short-term load forecasting (STLF) is important for ensuring efficient and reliable operations of smart grid. In order to improve the accuracy and reduce training time of STLF, this paper proposes a combined model, which is back-propagation neural network (BPNN) with multi-label algorithm based on K-nearest neighbor (K-NN) and K-means. Specific steps are as follows. Firstly, historical data set is clustered into $$ K $$ clusters with the K-means clustering algorithm; Secondly, we get $$ N $$ historical data points which are nearest to the forecasting data than others by the K-NN algorithm, and obtain the probability of the forecasting data points belonging to each cluster by the lazy multi-label algorithm; Thirdly, the BPNN model is built with clusters including one of $$ N $$ historical data points and the respective forecasting load are given by the built models; Finally, the forecasted load of each cluster multiply the probability of each, and then sum them up as the final forecasting load value. In this paper, the test data which include daily temperature and power load of every half hour from a community compared with the results only using BPNN to forecast power load, it is concluded that the combined model can achieve high accuracy and reduce the running time.

Filtered-x Least Mean Square (FxLMS) algorithm is a meaningful adaptation algorithm used in the field of Active Vibration Control (AVC). Hybrid FxLMS algorithm, which is the combination of the feedforward structure and the feedback structure of FxLMS, has a better stability and could get the same performance with a lower filter order. In order to get a faster convergence speed, this paper adopts Normalized LMS (NLMS) algorithm to replace of LMS algorithm in the hybrid AVC system. To verify the Hybrid Fx-NLMS algorithm, this paper developed a simulation platform for active vibration control of a flexible beam with piezoelectric stack actuator using ADAMS and MATLAB SIMULINK. Simulation results show that the convergence speed and vibration suppression performance of the Hybrid Fx-NLMS algorithm are better than other traditional algorithms.

Differential evolution algorithm is a heuristic global search technology based on population, which has received extensive attention from the academic community. Evolution algorithm is applied to the identification and optimization of double-tank system in this article. Firstly, the paper introduces the basic principle of the system identification and differential evolution algorithm. Secondly, design the identification scheme of double-tank system based on differential evolution algorithm. Identify the system according to the data measured in the experiment. Based on the commonly used three models and combined with DE/rand/1/bin, the model structure which best complies with the original experimental data is selected, and the improved form of the difference algorithm is further studied on the basis of the model structure. A large number of experiments have been carried out, the algorithm in other references may only improve one of CR or F, and the two will be all compared in this paper. The results of comparative analysis show that the improved differential evolution algorithm is, to some extent, superior to the basic differential evolution algorithm on identification accuracy of double-tank.

To minimize the losses of multi-variety perishable agricultural products, a mathematical model considering time sensitive feature of each perishable agricultural product is proposed. Meanwhile, a cuckoo search algorithm (CSA) is introduced to minimize the total losses of agricultural products. In view of poor exploration and exploitation ability of CSA, adaptive adjusting discovery probability and dynamic step-length is imposed to form an improved cuckoo search algorithm (ICSA). Finally, to verify the performance of the proposed algorithm, it is compared with cuckoo search and genetic algorithm (GA). Simulation results prove that the feasibility and superiority of the proposed algorithm.

The location of entities in a smart indoor environments is an important context information. To this end, several indoor localization algorithm have been proposed with the received signal strength fingerprint (RSS-F) based algorithm being the most attractive due to the higher localization accuracy. However, RSS-F based localization accuracy is highly degraded on account of non-line-of-sight (NLOS) propagation in indoor or harsh environment. This thesis proposes an approach for NLOS self-monitoring and autonomous compensation. Firstly, the localization area is regionalized according to Voronoi Diagram. Then, the self-monitoring and autonomous compensation is realized by propagation environment similarity represented by the dynamic path attenuation index between the domains. The verification experiment results show that the proposed algorithms can adaptively identify the NLOS interference and accomplish compensation. Compared with other localization algorithm, the maximum error is reduced from 3.04 m to 1.71 m, the average error is reduced to 0.90 m, and the localization time is reduced to 2.113 s (contain 10 test point) compared with other tracking algorithm.

Co-simulation using ADAMS and MATLAB is implemented for active vibration control of flexible beam with piezoelectric stack actuator. The virtual prototype of flexible beam with piezoelectric actuator is created in ADAMS, and the implement of prototype provides an approach for acquiring the information of dynamic and kinematic properties. When the properties analysis is finished, the controller based on FXLMS algorithm is established in MATLAB. The controller calculates the signals of acceleration which are measured from virtual prototype, then the force is generated to suppress the vibration of flexible beam. The results and analysis prove that active vibration control for flexible beam has a great suppression performance.

As the security accidents in public places frequently emerge, the research based on crowd evacuation gets more and more people’s attention. Now, the crowd evacuation research has shifted from the traditional live exercise to computer simulation. This paper chose five kinds of crowd evacuation simulation platform and summarized Cellular Automata, Agent-based model, network model they involved. Then, the thesis introduced the software, analyzed the performance of them and stated the respective advantages and disadvantage in order to help user choose proper platform to achieve fast and efficient results of crowd evacuation simulation.

Document priors that encode our prior knowledge about the importance of different documents are essential to an expert finding system. This study proposed a TopicRank-based document priors model for expert finding. TopicRank algorithm is an extension of the DocRank algorithm. Latent dirichlet allocation was used to extract topics of the documents. We assumed there was a link between two documents that share common topics. Link analysis techniques were then used to obtain document priors. The proposed model was evaluated using the CSIRO Enterprise Research Collection and the results showed that the performance of the expert finding system was dramatically improved by introducing TopicRank-based document priors. In particular, Mean Average Precision increased 19.9% while Mean Reciprocal Rank rose as much as 23.4%.

The models of airway tree designed in this paper are different from the general visual models. The model preserves all the information of the space data when it is created. The space data is mainly consisted by the coordinates of the boundary pixels and the spatial functions of the model surfaces. The algorithm consists of three main steps. Firstly, the boundaries of the airway tree are extracted by Sobel operator. Then, the boundary pixels are ring-likely sorted according to the distance between each other. Finally, each three pixels belong to the adjacent layers form a surface. An airway tree model can be eventually created by iterating the main steps. What’s more, the algorithm has also been optimized, we can mostly get a model in 50 s.

With the increasing demand of light-weight alloys, such as magnesium (Mg) and aluminum (Al), the need for joining these two alloys is unavoidable. In this study, AZ31B Mg and 1060 Al alloys were joined by continuous wave laser micro-welding using a 0.05 mm thick Cu/Zn interlayer. The microstructure and phases constituent of the weld seam were examined by optical microscope, SEM and EDS. The formation and distribution of the intermetallic compounds (IMCs) and the relationship between these structures and the micro-hardness of the weld were discussed in detail. The effect of Cu/Zn interlayer on the performance of Mg/Al joint was also analyzed. The results showed that Mg/Al IMCs were formed in the weld, which indicates that the Cu/Zn foil could not prevent the reaction between Mg and Al. However, the addition of Cu and Zn into the weld pool refined the microstructure by improving the number of eutectic structures. The micro-hardness of Mg/Al IMCs in the middle of the weld was very high which can be detrimental to the toughness of the Mg/Al joints.

Recent advancement in the integration of power systems and information communication technology has brought the key concerns towards security operation of cyber physical power system. This paper focuses on realizing the unified system modeling under intrusion attacks and refining the attack effects on communication network by simulation research. We start this survey with an overview of the system operation and crucial intrusion attacks associated with operational security from fusion system perspective. A novel limited stochastic Petri net (LSPN) graph theory is introduced to establish the unified firewall protection system model of intelligent substation network. By proposing quantitative computational methodology of communication throughput variation, the potential consequence on the communication network is determined with information transmission constraints. The final test on IEEE-30 node power system illustrates the usefulness of the proposed model analysis. The research work would raise awareness of the cyber intrusion threats and provide the basis for security defense.

Aiming at the problem that the low heat dissipation of the safety helmet has an impact on the workers which can easily lead to safety accidents under high temperature and high strength, this paper analyzed the temperature field and gas flow field inside the safety helmet by using Finite Element Analysis, modeled and meshed the real safety helmet, obtained the distribution nephogram of temperature and heat flux by steady state and transient thermal analysis, calculated the results of the hot gas flow inside safety helmet and the gas flow nephogram by turbulence model. The results show that the temperature inside the safety helmet decreases gradually from bottom to top, the heat transfer rate of the contact part with the head is faster, only a small amount of hot gas can be vented from the vent while most of the hot gas is still concentrated inside the safety helmet difficult to vent. Therefore, it is necessary to improve the design of safety helmet and the comfort of the operation to ensure the safety of the workers.

Due to the increasingly usage of the heavy-load train, the effect of the moving axial load on the noise radiation of elevated bridge structure should be concerned. In this paper, high speed train-track-bridge coupled model is established, and the train axle load, which is taken as the boundary condition of the load, is applied to the finite element model of elevated box bridge structure to calculate the vibration response of the surface of a box bridge. In this model, the vibration response is taken as the acoustic boundary conditions and is added to the boundary element model of elevated box bridge structure to study its sound radiation. The results show that the plate-shell unit can well reflect the overall and local vibration characteristics of the bridge structure, and under the effect of moving axial load, the vibration frequencies of box bridge structure concentrate in 0–300 Hz and the main peaks are in 10–160 Hz; the finite element-boundary element method can effectively analyze the low frequency noise radiation of box bridge caused by the moving axial load; the most of structure noise induced by moving axial load is below the audible range of which the noise is greatly harmful to human body, thus it must be taken seriously.

Through an analysis of the design contents of different architectural disciplines under the requirements of sustainable and low-carbon development, this study analyzes the low-carbon architectural design process and the makeup of relevant information from the various disciplines using a building information modeling (BIM) system. Based on BIM, we have constructed a carbon emission budgeting platform that captures the whole building life cycle, and have set forth evaluation criteria for the quantitative analysis of low-carbon buildings. In light of the above research, evaluation and optimization of low-carbon building designs, as well as the subsequent reduction of carbon emissions during the lifecycle of newly constructed buildings, can be achieved using BIM.

The quality of measurement data can be improved by data reconciliation. More accurate data will be provided for chemical process industry. However, the reconciliation results may be affected by gross errors. The influence of gross errors cannot be reduced effectively by classical method. Aimed at this problem, an improved robust NT-MT steady-state data reconciliation method is proposed in the paper. NT-MT method is used to detect suspicious nodes and variables with gross error. The suspicious variables are detected by critical value of adjustment detection. Robust estimator is used in data reconciliation. Finally, the measurement data is reconciled by the proposed robust estimator. The advantages of robust estimator and NT-MT method is combined together in this method. The simulation results show that the influence of gross error can be reduced effectively by the method proposed in the paper, thereby a better reconciliation results can be obtained.

3D mapping is a difficult problem due to real-world places whose appearance and scale can be various. Owing to the rapid development of computer and robot system, remarkable improvements of performance are achieved in 3D map technology, which in turn contribute to the significant advances in SLAM. This paper presents the state-of-the-art 3D map technology and system, which is classified into topological maps, metric maps and semantic maps. Additionally, the advantages and disadvantages of various 3D map technologies are analyzed in different aspects, including navigation performance, localization performance, visual perception, scalability, computation cost and mapping difficulty. In order to better understand them, the key performance parameters of the 3D map technologies are compared in a table. Finally, the paper ends with a discussion on the open problems and future of 3D map technology.

From security perspective, cyber physical system (CPS) security issue is investigated in this note. Based on a double-loop security control structure, the typical cyber attack called information disclosure, denial-of-service (DoS), deception attack and stealth attack are analyzed from uncertainty perspective. The performance of these attacks are formulated, meaningful models are proposed meanwhile. According to aforementioned attacks, security control scenarios are obtained via the character of each kind cyber attack. And some novel results are obtained via a well designed double closed loop structure. At last, from traditional standpoint, uncertainties of a separately excited DC motor is taken as an example to demonstrate the problem.

For Hammerstein-Wiener model with colored process noise, this paper derives an identification approach. The correlation function between input and output data points is derived by using separable signal to realize that the unmeasurable internal variable is replaced by the correlation function of input, and then correlation analysis method is used to estimate the parameters of the output nonlinear part and linear part. Furthermore, a correction term is added to least square estimation to compensate error caused by process noise, and then to derive an error compensation recursive least square method for the observed data from Hammerstein-Wiener model. Therefore, the parameters of the input nonlinear part can be estimated by error compensation method. Finally, the advantages of proposed algorithm are shown by simulation example.

Lamb wave based active and passive monitoring technologies are both hot points in structural health monitoring (SHM). However, active and passive monitoring methods were usually worked independently. The interaction and complementarity between Lamb wave active and passive monitoring techniques was analyzed. According to the advantages and disadvantages of active and passive monitoring methods, the active and passive cooperative working mechanism was proposed which combined the active and passive monitoring approaches. In the new method, active scanning and monitoring was set to be trigged by passive acoustic emission event, and the scanning interval could be greatly extended to save the energy consumption and improving monitoring efficiency as the evolution of damages caused by service and external erosion were usually very long. Meanwhile, the results and diagnosis information of active and passive monitoring method could be fused to improve the monitoring accuracy. In addition, the hardware implementation and software frame of the new integrated system were given. The experimental validation showed that the new approach combined the advantages of active and passive monitoring methods, and improved the damage monitoring efficiency and accuracy.

The embedded driver fatigue detect system is a real-time system can detect driver fatigue. In order to improve the performance of embedded driver fatigue monitor system, we propose a new system on chip (SOC) structure for accelerating the fatigue estimate. The new SOC consists two parts including the main processor and support vector machine IP core. An embedded Linux was transplanted and run the main algorithm which consists Haar-Adaboost classifier to locate the face and eyes. The SVM IP core accomplished the task of classifying the eyes’ statues. At last the system will estimate the state with PERCLOS standard. The results show that the system can content the need of real-world.

Feature representation is usually a key point in image recognition. The recognition performance can be potentially improved if the data distribution information is exploited. In this paper, we propose an image recognition approach based on generative score space. Specifically, we first leverage probabilistic latent semantic analysis (pLSA) to model the distribution of images. Then, we derive the mid-level feature from the model in a generative feature learning manner. At last, the derived feature is embedded into a discriminative classifier for image recognition. The advantages of our proposed approach are two folds. First, the probabilistic generative modeling allows us exploiting information hidden in data and has good adaptation to data distribution. Second, the discriminative learning process can utilize the information of label effectively. To confirm the effectiveness of our method, we perform image recognition on three datasets. The results demonstrate its advantages.

Multi-channel feature for pedestrian detection is proposed to solve problems of real-time and accuracy of pedestrian detection in this paper. Different from traditional low level feature extraction algorithm, channels such as colours, gradient magnitude and gradient histogram are combined to extract multi-channel feature for describing pedestrian. Then classifier is trained by AdaBoost algorithm. Finally the performance of the algorithm is tested in MATLAB. The result demonstrates that the algorithm has an excellent performance on both detection precision and speed.

Laser lines emitted by the laser level are mostly detected manually and laser particle and optical effects also bring difficulties on measurement. In this paper, we design a detection system for the five-line laser level and propose a laser line measurement method based on ma- chine vision. Image processing is divided into two stages: in the first stage, we use random sample consensus (RANSAC) algorithm combined with Hough transform to fit the laser axis, which can get its position information. In the second stage, a laser edge extraction method based on conditional random fields (CRFs) is proposed, and the sub-pixel width of laser line is obtained by spline interpolation algorithm. The results confirm that the laser level detection method proposed in this paper can realize the corresponding detection precision and requirement.

In this paper, we proposed a novel method of geometric calibration and synchronization for a multi camera system. Traditional calibration methods of visible cameras can’t be applied to thermal cameras. According to the imaging characteristics of thermal cameras, we designed a new calibration board using materials with different emissivity for calibration. Our calibration board can accurately calibrate RGBD cameras and thermal cameras. In general, thermal cameras have regular non-uniformity corrections, which will result in camera interruption about 1.5 to 2 s and can impact synchronization. In this respect, we adopted the method named nearest adjacent time using timestamp to solve the problem of non-uniformity corrections and synchronization. We evaluated our methods and the experiments showed that our methods had an ideal result for camera calibration and synchronization.

Image segmentation is a very important phase in automatic image analysis. Of the developed techniques for image segmentation, iterative methods have been proven to be one of the most effective algorithms in the literature. Mean shift algorithms is one of the iterative approaches which have been successfully deployed to many applications. However, despite its promising performance, mean shift has shown its weaknesses in convergence in some of the application areas. In this paper, an improved version of the standard mean-shift algorithm using a memory gradient method is proposed and implemented in order to achieve fast convergence rates by integrating mean shift and memory gradient. Experimental results on real images demonstrate that our proposed algorithm not only improves the efficiency of the classical mean shift algorithm, but also achieves better segmentation results.

The traditional method for on-line detection of cigarette filter stick packing is manual sampling, which has low efficiency, high labor cost and can’t detect all products. Therefore, it is necessary to establish a set of image processing system [10] based on machine vision. This system uses CCD image sensor to get the filter rod section image, through the image smoothing, edge detection, Binarization and feature extraction, the region of interest is analyzed, and finally get the number of filter rods. The filter rod arrangement on the production line is not completely flat. Using the dynamic area threshold method to calculate the number of filter rod in the flat area. 3D reconstruction of a single image is used to calculate the number of filter rod in the uneven area. The accuracy and practicability of the system are verified by theoretical analysis and experimental comparison.

Corners are the key feature of image. Stable corners are particularly important in the industrial pipelining of beer cap surface defects detection, greatly affecting the efficiency of image matching and detection precision. To find a stable algorithm for the cap surface defects detection, Stable Corner and Stable Ration are proposed to evaluate the stability of corner detectors, which are able to give an intuitive and unified stability description of various corner detection algorithm. After comparing the stability with Difference of Gaussian (DOG) and Features from Accelerated Segment Test (FAST), Harris is selected as the detector of cap surface images due to its high stability. To eliminate the redundant corners detected by Harris, Circular Mask and Harris (CMH) corner detection is proposed. In CMH, a circular mask with an adaptive threshold is adopted to remove the redundant corners, whereby comparing the intensity between the center pixel and others on the mask in a rapid way, more stable corners are obtained eventually. The effectiveness and robustness of CMH are verified in this paper, and the Stable Ratio increased by 16.7% relatively.

In recent years, magnetoencephalography (MEG) has played a prominent role on neocortical epilepsy preoperative evaluation. However, its clinical utility with locating deeper sources may be more challenging such as the mesial temporal structures. We proposed a new source imaging algorithm for finding the epileptogenic zone in mesial temporal lobe epilepsy (mTLE). Since the localization results using the Elekta MEG method are very sensitive to some MEG noises, the source modeling was modified by spatial filtering in wavelet domain and cortex constraint. Two surgical patients randomly selected with medically refractory mTLE, which were diagnosed based on a comprehensive preoperative evaluation, had been studied in this manuscript. The localization results using proposed method on individual MRI showed that the deeper regions had been exactly found in the mesial temporal lobe. Yet, the results using the Elekta Neuromag Software only appeared in the lateral temporal lobe. Thus, the proposed algorithm maybe become an effective method in detecting deeper epileptogenic zone.

The traditional MeanShift algorithm cannot obtain accurate tracking results in some complex situations where tracking targets have scale changes or similar color with background. In this paper, a new MeanShift target tracking algorithm, namely DEPTH & SIFT-MeanShift algorithm, is proposed by using a depth camera and SIFT (Scale Invariant Feature Transform) feature metric. The algorithm firstly combines feature points extracted from gray and depth images respectively, and then represents tracked objects with Modulus, i.e. Direction distribution histogram of feature points in the tracking object field, so that targets can be effectively tracked. Experimental results show that the proposed algorithm can achieve good tracking performance when the tracking target changes its scale, and have the strong adaptability to occlusion. Moreover, it is very robust to illumination changes, and able to discriminate targets from background very well.

The existing body detection methods based on depth images mostly depend on the extraction of image gradient features, which is the evolution of the traditional 2D plane image processing method for human body detection. Although their detection accuracy is high, the algorithms consume a large amount of computing and storage resources. Aiming at the real-time demand of safe-driving of forklift trucks in industry, this paper presents a novel 3D expansion and corrosion method for human detection by using depth information. A depth image of human body is detected based on the characteristics of human Head-Shoulder-Body Density (HSBD), which can reduce the error and loss of the depth information caused by changing light conditions, complex background scenes and various distances from objects. Experimental results show that the recognition rate of the proposed method is over 96%, and the recognition speed is over 15 frames per second. This can satisfy the safe-driving demands of forklift truckers in factory.

Edge detection is the key to image processing and has a significant impact on the high level of description, classification and matching of subsequent images. The traditional Canny algorithm requires human intervention in the selection of Gaussian function and its fixed parameters. To solve these problems, an improved algorithm based on Canny algorithm is proposed in this paper. The approach introduces the edge preserving filter to replace the original Gaussian filter, and calculates the magnitude and direction of image gradient with a new designed templates from x direction, y direction, and two oblique directions (45°, 135°). Meanwhile, the Otsu algorithm is used to calculate the thresholds, which avoids the problem that the thresholds need to be set repeatedly. The proposed method is successfully applied to the metal plate detection system. Experimental results show that the algorithm has good performance in bright and dark domains.

An algorithm based on HOG (Histograms of Oriented Gradients) and SVM (Support Vector Machine) is developed for traffic sign recognition on Android platform, and the dynamics link library is used as the native layer of Android end by employing Android NDK (Native Development Kit) technology. The test results show that the algorithm can be successfully applied to the Android platform, and Android NDK technology can implement the cross-platform and portability of the programs, while improving the detection and recognition speed.

Generally, plant grows slowly and is difficult to be observed, the apical growing points can reflect the changes of plant, such that the extraction of apical growing points is helpful for the analysis of plant growth. In this paper, a new digital visual-based method of tomato apical growing points segmentation is proposed, which is depended on depth segmentation, color segmentation and position histogram statistic. First of all, use the depth image captured by KinectV2 to remove complex background through depth segmentation. Then, position histogram of the two value image after depth segmentation has been obtained to get the column position of the apical growing points. Using the KinectV2 coordinate mapping mechanism to restore the color information of the two value image, and then the RBG-D image can be color segmented. Finally, the region of the apical growing points is segmented by coordinate mapping, and the apical growing point is extracted by the contour detection. The experimental results show that the method to segment the growth environment is effective.

In recent years, visual human tracking of quadcopter has become a topic of interest to many research institutions. To overview the recent research status of visual human tracking based on quadcopter, firstly, the problem of human tracking is divided into quadcopter control and vision based human tracking which are discussed separately. The present controlling means and the latest applications of quadcopter are summarized systematically. The advantages and disadvantages of each human tracking method are compared and the tracking strategies are summarized. Then, the difficult issues on visual human tracking are discussed specifically. Finally, the future research directions of visual human tracking based on quadcopter are prospected by summarizing related literatures.