As information development has progressed in the field of architecture in China, owing to the lack of unified information exchange standards and information integration mechanisms, it is difficult to exchange and share information between different stages and application systems during the construction lifecycle. The formation of information islands and faults hinders the application of information technology in the field of construction, thus affecting the production efficiency of the industry. In this study, construction engineering management and computer building information modeling (BIM) are deeply integrated, and an international standard for the construction industry is introduced. Moreover, a BIM information integrated building management platform is developed that combines BIM technology with the construction engineering management to realize the exchange of engineering information and shared and integrated management, in addition to providing theories, methods, technologies, and platforms.
In the traditional method of monitoring the state of electronic voltage transformer, there are problems of large monitoring error and weak robustness. Therefore, a new state monitoring method of electronic voltage transformer based on L-M algorithm is proposed. The relationship between input voltage and output voltage of capacitor voltage divider in electronic voltage transformer is obtained by using Laplasse transform. The transfer function model of electronic voltage transformer is constructed based on the relationship result and L-M algorithm. The transfer function model is used to analyze the frequency characteristics of the electronic voltage transformer and the range of normal measurement frequency, and then the partial pressure ratio of the electronic voltage transformer under the high frequency condition is derived. On this basis, by calculating the over voltage amplitude on the two sides of acquisition card in the electronic voltage transformer, the capacitance value between the two adjacent coaxial cylindrical cylinders of the capacitance divider in the electronic voltage transformer is obtained, thus the monitoring of the state of the electronic voltage transformer is completed. The experimental results show that the proposed method has low detection error and strong robustness, and can effectively improve the reliability of electronic voltage transformer.
Three dimensional virtual vision technologies improves the realistic effect of interior design, improves the rationality of interior design, and guides interior decoration and graphic design. In this paper, MultiGen Creator 3D modeling technology is used to reconstruct the indoor landscape and form the plane image of interior design. At the same time, the edge matching method is used to divide and decompose the features of the interior design image, and the 3D vision of the interior design is reconstructed with the key points of the interior design. In addition, through the adaptive tracking and rendering technology, the fidelity and space utilization efficiency of the room plane design are improved. The simulation results show that 3D virtual vision technology has better guidance for interior design, visual effect and landscape color fusion.
The numerical control separation in the Software-Defined Network (SDN) allows the control plane to have the absolute management rights of the network. As a new management plane of the SDN, once it is attacked, it will cause the entire network to face flaws. For this reason, this paper proposes a SDN control plane attack detection scheme based on deep learning, which can detect and respond to attacks on the SDN control plane in time. In this scenario, we propose a new pooling scheme that uses the TF-IDF idea to weight the characteristics of network traffic. Ultimately, our method achieved an accuracy of 99.8% in the SDN network’s traffic data set including 24 attack types.
One type of the partial periodic pattern is known as recurring patterns, which exhibit cyclic repetitions only for particular time period within a series. A key property of the patterns is the event can start, stop, and restart at anytime within a series. Therefore, the extracted meaningful knowledge from the patterns is challenging because the information can vary across patterns. The mining technique in recurring patterns plays an important role for discovering knowledge pertaining to seasonal or temporal associations between events. Most existing researches focus on discovering the recurring patterns in transaction. However, these researches for mining recurring patterns cannot discover recurring events across multiple transactions (inter-transaction) which often appears in many real-world applications such as the stock exchange market, social network, etc. In this study, the proposed algorithm, namely, CP-growth can efficiently perform in discovering the recurring patterns within inter-transaction. Besides, an efficient pruning technique to reduce the computational cost of discovering recurring patterns is developed in CP-growth algorithm. Experimental results show that recurring patterns can be useful in multiple transactions and the proposed algorithm, namely, CP-growth is efficient.
This paper proposes a method using joint classification of monogenic components with discrimination analysis for target recognition in synthetic aperture radar (SAR) images. Three monogenic components, namely, phase, amplitude, and orientation, are extracted from the original image and classified by joint sparse representation for target recognition. Considering that the three components may have different discrimination capabilities for different operating conditions, the discrimination analysis is incorporated into the classification scheme. The components with low discriminability are not used in the joint classification. Afterwards, those discriminative components for a certain condition are classified to determine the target type. Experiments are conducted on the moving and stationary target acquisition and recognition (MSTAR) to evaluate the performance of the proposed method.
The paper presents the correlation coefficient of refined-single valued neutrosophic sets (Refined-SVNSs) based on the extension of the correlation of single valued neutrosophic sets (SVNSs), and then a decision making method is proposed by the use of the weighted correlation coefficient of Refined-SVNSs. Through the weighted correlation coefficient between the ideal alternative and each alternative, we can rank all alternatives and the best one of all alternatives can be easily identified as well. Finally, to prove this decision making method proposed in this paper is useful to deal with the actual application, we use an example to illustrate it.
This paper proposes a compensation technique for the global navigation satellite system (GNSS)/real-time kinematic (RTK) course angle data using an electronic compass for an unmanned system. Additionally, the proportion, integral, and derivative control based on a back-propagation neural network (BP-PID) is introduced to improve the steering safety and riding comfort. The course angle jitter was determined. Because the GNSS/RTK receiver cannot offer stable heading data under specific conditions, including but not limited to susceptibility to obstacles, complex electromagnetic environment, and fewer satellites. The compensation algorithm is based on the determination of the GNSS course angle variance ratio and the asynchronous characteristic between the GNSS and an electronic compass. The combined data provide accurate and robust navigation information for an outdoor unmanned system. To address the limitation of the in-system parameter adjustment, a back-propagation (BP) neural network is adhibited to a conventional proportion, integral, and derivative (PID) lateral control system. The BP-PID control module updates the incremental PID parameters through self-learning, and results in the smoother operation of the vehicle. The flowchart of the learning algorithm and method of calculating the parameters are presented. A typical measurement was conducted and the obtained results were compared with typical RTK navigation results. Thus, the effectiveness of the proposed compensation method was confirmed.
This study describes a sound modulation system based on the use of a neural network model. The inputs to the model are a) a basic, original sound wave, and b) the degree of Kansei, while the output of the model is modulated sound depending on the degree of Kansei. The degree of Kansei is the numerical value that expresses the modulation level based on a Kansei linguistic expression, such as hardness or brilliance. In the experiment, the models are constructed for the sounds of piano and Marimba. Three types of training data are used for each sound, and the degree of Kansei is assigned manually for each dataset. By changing the degree of Kansei at the input of the model, we have validated that each model could appropriately modulate the basic sound. In addition, the modulation results are illustrated for one octave of piano sounds. The potential of our proposed model and future work are also discussed.
The efficiency of facial expression recognition (FER) is important for human-robot interaction. Detection of the facial region, extraction of discriminative facial expression features, and identification of categories of facial expressions are all related to the recognition accuracy and time-efficiency. An FER framework is proposed, in which 2D Gabor and local binary pattern (LBP) are combined to extract discriminative features of salient facial expression patches, and extreme learning machine (ELM) is adopted to identify facial expression categories. The combination of 2D Gabor and LBP can not only describe multiscale and multidirectional textural features, but also capture small local details. The FER of ELM and support vector machine (SVM) is performed using the Japanese female facial expression database and extended Cohn-Kanade database, respectively, in which both ELM and SVM achieve an accuracy of more than 85%, and the computational efficiency of ELM is higher than that of SVM. The proposed framework has been used in the multimodal emotional communication based humans-robots interaction system, in which FER within 2 seconds enables real-time human-robot interaction.
Lake Hachiroko, Japan, has many water quality issues, evident from phenomena such as green algae blooms. Understanding the details of the surface water quality of the lake, and the effect of seasons on the quality, is important. In our previous studies, we conducted fuzzy regression analysis of remote sensing data and direct measurements of water quality. The results showed that estimation maps of water quality were well created, using only five data points of the water quality parameters. To obtain maps that are in good agreement with the experimental data, remote sensing data and water quality values should be acquired simultaneously. However, performing such simultaneous observations can affect the preparation of the water quality estimation maps. We overcame this obstacle by using fuzzy c-means clustering (FCM), and considered the effect of specific disturbances and uncertainties on the remote sensing data. Furthermore, FCM using only remote sensing data creates estimation maps in which relative water surface conditions are classified. Therefore, determining the relationship between FCM results and water quality facilitates the creation of low-cost, high-frequency water quality estimation maps. Our results indicated that FCM was particularly effective in determining the presence of suspended solids (SS) during water quality analysis. However, the relationship between FCM results and water quality has not been determined in detail. In this study, we analyzed the water quality conditions of Lake Hachiroko with FCM using the data collected by the Advanced Space-borne Thermal Emission and Reflection Radiometer on Terra and, the Operational Land Imager on Landsat-8. In addition, FCM results were compared with the maps created by fuzzy regression analysis and the actual conditions of water pollution. The results indicated that (i) the maps created using FCM are effective in determining the water surface conditions, (ii) the FCM maps using data obtained during August and September have a strong relationship with biochemical oxygen demand (BOD) and SS, and (iii) the FCM maps using data obtained during May and June have a strong relationship with chemical oxygen demand (COD), SS, and total nitrogen (T-N).
Complex illumination condition is one of the most critical challenging problems for practical face recognition. However, numerous studies have had no effective solutions reported for full illumination variation of face images in the facial recognition research field. In order to effectively solve full illumination variation problem, we propose a novel approach for illumination normalization for facial images based on the enhanced contrast method of histogram equalization (HE) and fusion of illumination estimations (FOIE). Then, feature extraction is applied with consideration of both Gabor wavelet and principal component analysis methods to process illumination normalization. Next, a support vector machine classifier (SVM) is used for face classification. Experimental results show that superior performance can be obtained in the developed approach by comparisons with some state-of-the-arts.
This paper proposes a method to estimate the posterior distribution of a Boltzmann machine. Due to high feature extraction ability, a Boltzmann machine is often used for both of supervised and unsupervised learning. It is expected to be suitable for multimodal data because of its bi-directional connection property. However, it needs a sampling method to estimate the posterior distribution, which becomes a problem during an inference period because of the computation time and instability. Therefore, it is usually converted to feedforward neural networks, which means to lose its bi-directional property. To deal with these problems, this paper proposes a method to estimate the posterior distribution of a Boltzmann machine fast and stably without converting it to feedforward neural networks. The key idea of the proposed method is to estimate the posterior distribution using a simulated annealing on non-uniform temperature distribution. The advantage of the proposed method against Gibbs sampling and conventional simulated annealing is shown through experiments with artificial dataset and MNIST. Furthermore, this paper also gives the mathematical analysis of Boltzmann machine’s behaviour with regard to temperature distribution.
This study shows that a general regularized fuzzy c-means (rFCM) clustering algorithm, including some conventional clustering algorithms, can be constructed if a given regularizer function value, its derivative function value, and its inverse derivative function value can be calculated. Furthermore, the results of the study show that the behavior of the fuzzy classification function for rFCM at an infinity point is similar to that for some conventional clustering algorithms.
In this study, a collaborative filtering method that uses fuzzy clustering and is based on q-divergence is proposed for categorical multivariate data. The results of experiments conducted on an artificial dataset indicate that the proposed method is more effective than the conventional one if the number of clusters and the initial setting are adequately set. Furthermore, the results of the experiments conducted on three real datasets indicate that the proposed method outperforms the conventional method in terms of recommendation accuracy as well.
For spinal curvature measurements, because of the anatomical complexity of the spine CT image, developing an automated method to avoid manual landmark is a challenging task. In this study, we propose an intelligent framework that integrates the cascade AdaBoost classifier and region-based distance regularized level set evolution (DRLSE) with the vertebral centroid measurement. First, the histogram-of-oriented-gradients based cascade gentle AdaBoost classifier is used to detect automatically and localize vertebral bodies from computer tomography (CT) spinal images. Considering these vertebral pathological images enables us to produce a diverse training dataset. Then, the DRLSE method introduces the local region information to converge the vertebral boundary quickly. The located bounding box is regarded as an accurate initial contour. This avoids the negative impact of manual initialization. Finally, we perform vertebral centroid extraction and spinal curve fitting. The spinal curvature angle is determined by calculating the angle between two tangents to the curve. We verified the effectiveness of the proposed method on 10 spine CT volumes. Quantitative comparison against the ground-truth centroids yielded a detection accuracy rate of 98.3% and a mean centroid location error of 1.15 mm. The comparative results with existing methods demonstrate that the proposed method can accurately detect and segment vertebral bodies. Furthermore, the spinal curvature can be automatically measured without manual landmark.
This paper proposes an analytical model that clarifies the relationship between specific place and human emotions as well as the cause of the emotions using tweet data with location information. In addition, Twitter data with location information are analyzed to show the effectiveness of our proposed model. First, geotags are provided to collect Twitter data and increase the number of data for analysis. Second, training data with emotion labels based on the emotion expression dictionary are created and used, and supervised learning is done using fastText to obtain the emotion estimates. Finally, by using the result, topic extraction is performed to estimate the causes of the emotions. As a result, the transition of emotion in time and space as well as its cause is obtained.
Human posture recognition has been a popular research topic since the development of the referent fields of human-robot interaction, and simulation operation. Most of these methods are based on supervised learning, and a large amount of training information is required to conduct an ideal assessment. In this study, we propose a solution to this by applying a number of unsupervised learning algorithms based on the forward kinematics model of the human skeleton. Next, we optimize the proposed method by integrating particle swarm optimization (PSO) for optimization. The advantage of the proposed method is no pre-training data is that required for human posture generation and recognition. We validate the method by conducting a series of experiments with human subjects.
In recent years, population aging has become an important social issue. We hope to achieve an elderly health care system through technical means. In this study, we developed an elderly health care system. We chose to use environmental sensors to estimate the behavior of older adults. We found that traditional methods have difficulty solving the problem of excessive indoor environmental differences in different households. Therefore, we provide a fuzzy spike neural network. By modifying the sensitivity of input using a fuzzy inference system, we can solve the problem without additional training. In the experiment, we used temperature and humidity data to make an estimation of behavior in the bathroom. The results show that the system can estimate behavior with 97% accuracy and 78% sensitivity.
The improvement of a country’s technological innovation level is influenced by the technology spillover of inward foreign direct investment (IFDI) and outward foreign direct investment (OFDI). Based on the Coe and Helpmen’s theory of international capital flow model and one-way causality measure model, this study examines the similarities and dissimilarities between the dynamic effects of IDFI and OFDI on technological innovation in China and Japan to enumerate the differences in the utilization effect of FDI between developed and developing countries. The empirical results show that the one-way causality intensity of IFDI to technological innovation in China is weaker than that in Japan, but the FDI volatility in China is stronger than that in Japan. The one-way causality intensity of OFDI to technological innovation are low both in China and Japan, and the patterns of long-term and short-term effects are not identical. According to the results of our empirical research, we draw the conclusions and proposed suggestions for using IFDI and OFDI in China and Japan.
Increasing the credit of small and medium-sized enterprises (SMEs) is the key to solving SMEs’ financing difficulties. Because of their small size and fixed assets, it is not easy for SMEs to get loans from mortgage or private guarantee institutions. Therefore, to alleviate the credit rationing faced by SMEs and reduce financing cost, the key is to increase corporate credit score. This study uses small and micro-businesses’ data from Taizhou city to identify the key factors affecting corporate default. The results show that enterprise scale, enterprise operation status, financial environment, and credit-increasing means are the key factors affecting enterprise default, and credit protection funds do not play a significant role. Therefore, it can be argued that at present, the credit growth of SMEs still relies mainly on fixed asset mortgage, while the role of credit protection funds needs further refinement to effectively assist SMEs to solve difficult and expensive financing.
Utility functions on two-dimensional regions are demonstrated for decision makers’ risk averse behavior by weighted quasi-arithmetic means. For two utility functions on two-dimensional regions, a concept is introduced that decision making with one utility is more risk averse than decision making with the other utility. A necessary condition and sufficient conditions for the concept are demonstrated by their utility functions.
This study presents a fuzzy clustering algorithm for classifying spherical data based on q-divergence. First, it is shown that a conventional method for vectorial data is equivalent to the regularization of another conventional method using q-divergence. Next, based on the knowledge that q-divergence is a generalization of Kullback-Leibler (KL)-divergence and that there is a conventional fuzzy clustering method for classifying spherical data based on KL-divergence, a fuzzy clustering algorithm for spherical data is derived based on q-divergence. This algorithm uses an optimization problem built by extending KL-divergence in the conventional method to q-divergence. Finally, some numerical experiments are conducted to verify the proposed methods.
The fuzzy c-regression models are useful for datasets with various correlations. To deal with nonlinear datasets, a kernel fuzzy c-regression (KFCR) method was previously proposed. However, this method is weak for outliers because its objective function is based on the least square principle. We introduce the least absolute deviation (LAD) method with a modified Huber function into the KFCR (LAD-KFCR) to overcome the abovementioned problem. We verify the usefulness of the proposed LAD-KFCR method through numerical examples.
The Louvain method is a method of agglomerative hierarchical clustering (AHC) that uses modularity as the merging criterion. Modularity is an evaluation measure for network partitions. Cluster validity measures are also used to evaluate cluster partitions and to determine the optimal number of clusters. Several cluster validity measures are constructed considering the geometric features of clusters. These measures and modularity are considered to be the same concept in the viewpoint of evaluating cluster partitions. In this paper, cluster validity measures based agglomerative hierarchical clustering (CVAHC) is proposed as a novel clustering method for network data. The cluster validity measures are used as a merging criterion and an evaluation measure for network data in the proposed method. Numerical experiments show that Dunn’s and Xie-Beni’s indices for network partitions are useful for network clustering.
The research and development of robot partners have been actively conducted to support human daily life. Human-robot interaction is one of the important research field, in which verbal and nonverbal communication are essential elements for improving the interactions between humans and robots. Thus, the purpose of this research was to establish a method to adapt a human-robot interaction mechanism for robot partners to various situations. In the proposed system, the robot needs to analyze the gestures of humans to interact with them. Humans have the ability to interact according to dynamically changing environmental conditions. Therefore, when robots interact with a human, it is necessary for robots to interact appropriately by correctly judging the situation according to human gestures to carry out natural human-robot interaction. In this paper, we propose a constructive methodology on a system that enables nonverbal communication elements for human-robot interaction. The proposed method was validated through a series of experiments.
The interval type-2 fuzzy possibilistic C-means clustering (IT2FPCM) algorithm improves the performance of the fuzzy possibilistic C-means clustering (FPCM) algorithm by addressing high degrees of noise and uncertainty. However, the IT2FPCM algorithm continues to face drawbacks including sensitivity to cluster centroid initialization, slow processing speed, and the possibility of being easily trapped in local optima. To overcome these drawbacks and better address noise and uncertainty, we propose an IT2FPCM method based on granular gravitational forces and particle swarm optimization (PSO). This method is based on the idea of gravitational forces grouping the data points into granules and then processing clusters on a granular space using a hybrid algorithm of the IT2FPCM and PSO algorithms. The proposed method also determines the initial centroids by merging granules until the number of granules is equal to the number of clusters. By reducing the elements in the granular space, the proposed algorithms also significantly improve performance when clustering large datasets. Experimental results are reported on different datasets compared with other approaches to demonstrate the advantages of the proposed method.
A repetitive controller contains a pure-delay positive-feedback loop that makes it difficult to stabilize a strictly proper system. A low-pass filter is inserted in a repetitive controller to relax the stability condition of the modified repetitive-control system at the cost of degrading the tracking performance. In this study, a modified repetitive-control approach is developed, which reaches a balance between the stability and tracking performance for a class of affine nonlinear systems based on the Takagi–Sugeno fuzzy model. First, a 2D model is established to adjust continuous control and discrete learning actions preferentially induced by exploiting the 2D property in a repetitive-control process. Then, the Lyapunov stability theory and 2D system theory are used to derive a sufficient stability condition in the form of linear matrix inequalities to design parallel-distributed-compensation-based state-feedback controllers. Finally, an application-oriented example is used, and a comparison is performed to show that an extra variable is introduced such that the developed method has a better tracking performance.