Search

The fundamental mission of a distribution network is to satisfy the customer demand by providing sufficient delivery capacity. However, the capacity of a distribution network is practically stochastic because of unexpected events, and moreover, the commodities may rot or be spoilt during delivery owing to inclement weather, traffic accidents, collisions, and so on, such that the intact commodity flow may not meet market demand. This paper focuses on the reliability of a multi-state distribution network (MSDN) with cost and spoilage characteristics, defined as the probability that the MSDN is able to distribute a sufficient quantity of goods to meet the market demand under delivery spoilage and budget limit considerations. A specific spoilage rate associated with each route is adopted to characterize the perishability of commodity flows, and the critical routes whose spoilage rate change has the biggest impact on network reliability are identified with the use of sensitivity analysis method. Apart from delivery cost, the cost involved with the disposal of spoilt goods is also incorporated into the reliability indicator. A minimal paths based algorithm is presented to calculate network reliability, together with an example to illustrate the procedure. A real fruit distribution network is accordingly discussed to demonstrate the utility of the algorithm and the managerial implication of network reliability.

The search for a bona fide robustness index to quantify the system performance upon unexpected disturbances has been a continuous research effort. Previous studies have largely neglected to reflect...

Deep neural network (DNN) based face recognition system has become one of the most popular modalities for user identity authentication. However, some recent studies have indicated that, the malicious attackers can inject specific backdoors into the DNN model of a face recognition system, which is known as backdoor attack. As a result, the attacker can trigger the backdoors and impersonate someone else to log into the system, while not affecting the normal usage of the legitimate users. Existing studies use the accessories (such as purple sunglasses or bandanna) as the triggers of their backdoor attacks, which are visually conspicuous and can be easily perceptible by humans, thus result in the failure of backdoor attacks. In this paper, for the first time, we exploit the facial features as the carriers to embed the backdoors, and propose a novel backdoor attack method, named BHF2 (Backdoor Hidden in Facial Features). The BHF2 constructs the masks with the shapes of facial features (eyebrows and beard), and then injects the backdoors into the masks to ensure the visual stealthiness. Further, to make the backdoors look more natural, we propose BHF2N (Backdoor Hidden in Facial Features Naturally) method, which exploits the artificial intelligence (AI) based tool to auto-embed the natural backdoors. The generated backdoors are visually stealthy, which can guarantee the concealment of the backdoor attacks. The proposed methods (BHF2 and BHF2N) can be applied for those black-box attack scenarios, in which a malicious adversary has no knowledge of the target face recognition system. Moreover, the proposed attack methods are feasible for those strict identity authentication scenarios where the accessories are not permitted. Experimental results on two state-of-the-art face recognition models show that, the maximum success rate of the proposed attack method reaches 100% on DeepID1 and VGGFace models, while the accuracy degradation of target recognition models are as low as 0.01% (DeepID1) and 0.02% (VGGFace), respectively. Meantime, the generated backdoors can achieve visual stealthiness, where the pixel change rate of a backdoor instance relative to its clean face image is as low as 0.16%, and their structural and dHash similarity score are high up to 98.82% and 98.19%, respectively.

Most of modern technological networks that can perform their tasks with various distinctive levels of efficiency are multistate networks, and reliability is a fundamental attribute for their safe operation and optimal improvement. For a multistate network, the two-terminal reliability at demand level d, defined as the probability that the network capacity is greater than or equal to a demand of d units, can be calculated in terms of multistate minimal paths, called d-minimal paths (d-MPs) for short. This paper presents an efficient algorithm to find all d-MPs for the multistate two-terminal reliability problem. To advance the solution efficiency of d-MPs, an improved model is developed by redefining capacity constraints of network components and minimal paths (MPs). Furthermore, an effective technique is proposed to remove duplicate d-MPs that are generated multiple times during solution. A simple example is provided to demonstrate the proposed algorithm step by step. In addition, through computational experiments conducted on benchmark networks, it is found that the proposed algorithm is more efficient.

A major performance index for analyzing a stochastic transport network is the two-terminal capacity reliability Rd, defined as the probability that d units of goods can be successfully transported via stochastic arc capacities from the source to the destination. This paper presents an efficient d-minimal path method to calculate Rd based on some newly obtained results. The proposed method uses a simple method to check d-minimal path candidates and a more efficient approach to remove duplicate d-minimal paths that are the biggest obstacle in solving d-minimal paths, along with an indication of the advantage over the existing methods. Besides, sensitivity analysis is adopted to explore the most important arc whose reliability change affects the network reliability most significantly, which helps supervisor identify and enhance the critical arcs for improving the network reliability more effectively. Computational and application examples demonstrate the efficiency and utility of the method, respectively.

Disproportionate progressive collapse is a structural failure mode with low probability and high consequences. Due to this nature, progressive collapse design is usually treated as a secondary design; namely, a design check after the design for primary loadings. Various design procedures have been recommended, notably the alternate load path method through nonlinear pushdown analyses. However, most of these design procedures are iterative by nature to satisfy the design acceptable criteria. To improve the design efficiency, this study presented an innovative design technique that allows to directly determine the proper amount of reinforcement in beams and slabs of a reinforced concrete (RC) structure in order to withstand the gravitational loads under column removal scenarios. Inspired from fire-induced progressive collapse research, the proposed method employs a virtual thermal pushdown analysis, in which the temperature increase affects only the strength of reinforced steel. With carefully developed strength-temperature relationships of the rebars, the virtual thermal analysis produces a displacement-temperature curve that represents the genuine nonlinear relationship between the amount of reinforcement and structural performance (often represented by vertical displacement). This curve is then used to directly determine the amount of reinforcement at the prescribed performance target. Two beam-column sub-assemblage examples and the Alfred P. Murrah Federal Building were analyzed and redesigned to demonstrate the effectiveness of the proposed method. All three examples showed that the proposal virtual thermal pushdown method can accurately determine the appropriate amount of reinforcements to meet the performance target.

The training process of deep learning model is costly. As such, deep learning model can be treated as an intellectual property (IP) of the model creator. However, a pirate can illegally copy, redistribute or abuse the model without permission. In recent years, a few Deep Neural Networks (DNN) IP protection works have been proposed. However, most of existing works passively verify the copyright of the model after the piracy occurs, and lack of user identity management, thus cannot provide commercial copyright management functions. In this paper, a novel user fingerprint management and DNN authorization control technique based on backdoor is proposed to provide active DNN IP protection. The proposed method can not only verify the ownership of the model, but can also authenticate and manage the user's unique identity, so as to provide a commercially applicable DNN IP management mechanism. Experimental results on CIFAR-10, CIFAR-100 and Fashion-MNIST datasets show that the proposed method can achieve high detection rate for user authentication (up to 100% in the three datasets). Illegal users with forged fingerprints cannot pass authentication as the detection rates are all 0 % in the three datasets. Model owner can verify his ownership since he can trigger the backdoor with a high confidence. In addition, the accuracy drops are only 0.52%, 1.61 % and -0.65% on CIFAR-10, CIFAR-100 and Fashion-MNIST, respectively, which indicate that the proposed method will not affect the performance of the DNN models. The proposed method is also robust to model fine-tuning and pruning attacks. The detection rates for owner verification on CIFAR-10, CIFAR-100 and Fashion-MNIST are all 100% after model pruning attack, and are 90 %, 83 % and 93 % respectively after model fine-tuning attack, on the premise that the attacker wants to preserve the accuracy of the model.

The popularity of various social platforms has prompted more people to share their routine photos online. However, undesirable privacy leakages occur due to such online photo sharing behaviors. Advanced deep neural network (DNN) based object detectors can easily steal users' personal information exposed in shared photos. In this paper, we propose a novel adversarial example based privacy-preserving technique for social images against object detectors based privacy stealing. Specifically, we develop an Object Disappearance Algorithm to craft two kinds of adversarial social images. One can hide all objects in the social images from being detected by an object detector, and the other can make the customized sensitive objects be incorrectly classified by the object detector. The Object Disappearance Algorithm constructs perturbation on a clean social image. After being injected with the perturbation, the social image can easily fool the object detector, while its visual quality will not be degraded. We use two metrics, privacy-preserving success rate and privacy leakage rate, to evaluate the effectiveness of the proposed method. Experimental results show that, the proposed method can effectively protect the privacy of social images. The privacy-preserving success rates of the proposed method on MS-COCO and PASCAL VOC 2007 datasets are high up to 96.1% and 99.3%, respectively, and the privacy leakage rates on these two datasets are as low as 0.57% and 0.07%, respectively. In addition, compared with existing image processing methods (low brightness, noise, blur, mosaic and JPEG compression), the proposed method can achieve much better performance in privacy protection and image visual quality maintenance.

Deep neural network (DNN) based face recognition systems have been widely applied in various identity authentication scenarios. However, recent studies show that the DNN models are vulnerable to backdoor attacks. An attacker can embed backdoors into the neural network by modifying its internal structure or poisoning the training set. In this way, the attacker can login into the system as the victim, while the normal use of the system by legitimate users will not be affected. However, the backdoors used in existing attacks are visually perceptible (black-frame glasses or purple sunglasses), which will arouse humans' suspicions thus lead to the failure of the attacks. In this paper, we propose a novel backdoor attack method, BHF2 (Backdoor Hidden as Facial Features), where the attacker can embed the backdoors as the inherent facial features. The proposed method can greatly enhance the concealment of the injected backdoor, which makes the backdoor attacks more difficult to be discovered. Besides, the BHF2 method can be launched under the black-box conditions, where the attacker is completely unaware of the target face recognition system. The proposed backdoor attack method can be applied in those rigorous identity authentication scenarios where the users are not allowed to wear any accessories. Experimental results show that the BHF2 method can achieve high attack success rate (up to 100%) on the state-of-the-art face recognition model, DeepID1, while the normal working performance of the system has hardly been affected (the recognition accuracy of the system has only dropped by 0.01% at the lowest).

Irregular structures are known to be more prone to disproportionate collapse (DC) than structures with regular horizontal and vertical layouts. Despite this, many public, commercial and institutional buildings have to be designed with various irregularities in structural layout due to architectural and aesthetic considerations. While extensive experimental and numerical studies have been carried out to advance the basic understanding of DC behaviors of regular structures, limited research has been carried out on irregular structures. On the other hand, alternate load path method has been recommended by several design codes and standards for DC design, but the large number of potential column removal scenarios involved poses a real challenge for the wide use of the method. This paper presents a quantitative measure referred to as irregularity index, or Π , that can be used to quickly identify the worst column removal scenarios for design check of DC resistance of reinforced concrete (RC) frame structures. Based upon a solid understanding of various competing failure mechanisms and DC resistance of RC frame structures, the irregularity index was derived from the maximum load factors between the yielding load and the limit load at the end of tensile membrane action stage. For verification and validation of the proposed Π , a 3D nonlinear finite element model was developed in OpenSees and calibrated by beam-slab sub-assemblage experiments. Various numerical experiments were carried out to verify the effectiveness of the Π . The very high correlation coefficient between the calculated irregularity indices and DC resistance obtained from finite element analyses showed that the proposed irregularity index can quickly identify the worst column removal scenarios in various irregular structures including setback buildings and large-bottom-space structures. Sensitivity analysis further verified the reliability of the proposed irregularity index.

Modular integrated Construction (MiC) is a game changing construction approach which could significantly increase construction efficiency, quality, and sustainability. For steel MiC structures, modules are manufactured in the factories and assembled at the construction sites through inter-module connections. The concern of the vulnerability of these connections under abnormal hazards leading to disproportionate progressive collapse will impede industrial applications of MiC structures in the construction community. This study investigates the structural robustness of corner-supported modular steel buildings with the focus on different inter-module connections. A sub-structure extracted from a five-storey modular building is analysed by high fidelity finite element analysis under a corner column removal scenario. The load redistribution mechanism and failure modes are investigated thoroughly, based on which the simplified macro model is proposed where connections are modelled with rotational springs. Good consistence is found for the simplified model and the detailed FE analysis in terms of the pushdown curve. It is also found that the inter-module connection types will affect the beam-column joint properties which dominate the progressive collapse resistance of MiC structures. Therefore, attention should be paid when selecting different inter-module connections. The current rigid beam-column joint connection may lead to nonconservative evaluation.

Recently, many studies show that deep neural networks (DNNs) are susceptible to adversarial examples, which are generated by adding imperceptible perturbations to the input of DNN. However, in order to convince that adversarial examples are real threats in real physical world, it is necessary to study and evaluate the adversarial examples in real-world scenarios. In this paper, we propose a natural and robust physical adversarial example attack method targeting object detectors under real-world conditions, which is more challenging than targeting image classifiers. The generated adversarial examples are robust to various physical constraints and visually look similar to the original images, thus these adversarial examples are natural to humans and will not cause any suspicions. First, to ensure the robustness of the adversarial examples in real-world conditions, the proposed method exploits different image transformation functions (Distance, Angle, Illumination and Photographing), to simulate various physical changes during the iterative optimization of the adversarial examples generation. Second, to construct natural adversarial examples, the proposed method uses an adaptive mask to constrain the area and intensities of the added perturbations, and utilizes the real-world perturbation score ( R P S ) to make the perturbations be similar to those real noises in physical world. Compared with existing studies, our generated adversarial examples can achieve a high success rate with less conspicuous perturbations. Experimental results demonstrate that, the generated adversarial examples are robust under various indoor and outdoor physical conditions, including different distances, angles, illuminations, and photographing. Specifically, the attack success rate of generated adversarial examples indoors and outdoors is high up to 73.33% and 82.22%, respectively. Meanwhile, the proposed method ensures the naturalness of the generated adversarial example, and the size of added perturbations is as low as 29361.86, which is much smaller than the perturbations in the existing works (95381.14 at the highest). Further, the proposed physical adversarial attack method can be transferred from the white-box models to other object detection models. The attack success rate of the adversarial examples (generated targeting Faster R-CNN Inception v2) is high up to 57.78% on the SSD models, while the success rate of adversarial example (generated targeting YOLO v2) on SSD models reaches 77.78%. This paper reveals that physical adversarial example attacks are real threats in the real-world conditions, and can hopefully provide guidance for designing robust object detectors and image classifiers.

The simulation of electromagnetic fields distribution around high-speed railway catenary on the platform by the finite element method. The research is based on simulation models as classification material properties of human body parts, and presents subdivision mesh sub-model method to calculate and simulate the frequency electromagnetic fields around the human, and giving a comprehensive evaluation on the safety of electromagnetic environment around high-speed railway according to national standard limit.

This paper presents a novel description for factor logic, which puts state description into a factor space, and sets up truth sets for formulae. Along this way, factor space may provide a suitable platform for the development of quantitative logic. Factor logic can be also described as a derivative system of Boolean logic with added hypotheses \(\varGamma \), which can be used in data mining and function-structure analysis in switch systems.

© Springer International Publishing Switzerland 2015. Fourier has been a powerful mathematical tool for representing a signal into an expression consist of sin and cos. Recently a new developed signal decomposition theory is proposed by Pro. Tao Qian named Adaptive Fourier Decomposition, which has the advantage in time frequency over Fourier decomposition and without the need for a fixed window size problem such as short-time frequency transform. Studies show that AFD can fast decompose signals into positive-frequency functions with good analytical properties. In this paper we apply AFD into image decomposition and reconstruction area first time in the literature, which shows a promising result and gives the fundamental prospect for image compression.

Optimal sensor placement is an important part in the structural health monitoring of bridge structures. However, some defects are present in the existing methods, such as the focus on a single optimal index, the selection of modal order and sensor number based on experience, and the long computation time. A hybrid optimization strategy named MSE-AGA is proposed in this study to address these problems. The approach firstly selects modal order using modal participation factor. Then, the modal strain energy method is adopted to conduct the initial sensor placement. Finally, the adaptive genetic algorithm (AGA) is utilized to determine the optimal number and locations of the sensors, which uses the root mean square of off-diagonal elements in the modal assurance criterion matrix as the fitness function. A case study of sensor placement on a numerically simulated bridge structure is provided to verify the effectiveness of the MSE-AGA strategy, and the AGA method without initial placement is used as a contrast experiment. A comparison of these strategies shows that the optimal results obtained by the MSE-AGA method have a high modal strain energy index, a short computation time, and small off-diagonal elements in the modal assurance criterion matrix.

As a general purpose system in optical computing,optical vector-matrix multiplier (OVMM) can be applied in a variety of fields. After recalling Goodman’s classic principle of OVMM we proposed a new method taking advantage of liquid crystal array, which can solve the flaws of the former. Here we explain in details the principle of this new method and use our new developed ternary logic optical processor as the equipment to realize it. By experiments the correctness and feasibility of this method are verified. At last the advantages are discussed and the future study work is planned.

As a general purpose system in optical computing, optical vector-matrix multiplier (OVMM) can be applied in a variety of fields. In 2007 a novel ternary logic optical processor has been developed in our lab. We combine light intensity and polarization to denote ternary notation and use liquid crystal array to complete the state transformation among ternary values. Here we describe a new method to realize binary vector-matrix multiplication using this optical processor and the advantages of this method are discussed in comparison with the traditional one.

Anaphora occurs throughout discourse or dialogue. Their high frequencies make anaphora resolution one key problem in discourse processing which attract attention of increasing researchers. In the paper, according to features of Chinese personal pronoun we present an approach which is based on corpus. It adopts maximum entropy method. Then we discovered characteristic of pronoun anaphora referring to inhuman is different form that referring to people. So the paper put forward a new approach about processing the Chinese personal pronoun anaphora. The approach divided the anaphora resolution system into two parts, the PARS subsystem and the IARS subsystem. They will separately process the personal pronoun anaphora referring to people and that referring to inhuman. The paper described the design and the realization of the system, tests the new system in the Chinese Tree Bank and evaluates the arithmetic in the round. The experiment demonstrates that the method achieves the desired result.

The model for precise description and measurement of user profile was studied through adopting the technology of knowledge discovery from database. The news browsing manner of users was analyzed, while news similarity was computed based on users' reading style, then the model for user profile was generated by the mining algorithm. The experiments showed that this model worked very well in describing user profile and promoted efficiency of personalized news service system.

The implication is the emphasis and difficulty in the study of logics. From the point of view of interval structure, the operations of the universal logics including interval-intersection, interval-union, interval-complement and interval-implication are redefined based on the radical models. Regularity and monotonicity of interval-implication are proved; interval-intersection and interval-implication are proved to be an adjoint pair. In the paper, operation models of the interval-implication are firstly given in the special points of h. It is very important to the further study on flexibility of the interval-logics.

We call any interval [a,b] generalized interval. In the before work, the operation models of universal logic and neurons were limited in interval [0,1]. Authors mainly studied the universal logic operation models and neurons based on generalized interval [a,b]. Using the NTS norms theories, this paper established the universal logic operation models on generalized interval [a,b]. And the paper gave a new uniform neuron model of several universal logic operators on generalized interval [a,b]. It made the reasoning of universal logic neuron more flexible. It proved that the operation models and neurons were sound. The work offered important theories and models for universal logic in correlation reasoning of neural systems. It is a new way for neuron reasoning from the points of generalized NTS norms and generalized interval.

In these years, universal logic has developed greatly. As a kind of concrete universal logic, authentic logic is used to analyze the security of cryptographic protocols. This paper firstly puts forward relational conceptions and gives the structure's expression of object-oriented universal logic. Then authors analyze the logic structure's expression of authentic logic with the example: BAN logic. Using the conceptions of general correlation and general self-correlation, authors discuss the applicability and correlation of authentic logic. This work can contribute to perfecting authentic logic in artificial intelligence reasoning, and offering a new expression way of universal logic.

With the fast development of modern sciences, opinions that people think about the natural system are basically changing, that is to say, turning from ideal simplification to objective complexity. Universal logic puts forward the conceptions of general correlation and general self-correlation, which is used to describe the flexibility of correlation of objective things exactly, and establish the architecture of universal logic. This paper analyzes the principles of universal logic and the correlation of complex system, defines the conceptions of "whole correlation" and "local correlation", and sets up the reasoning model of complex system based on universal logic, moreover, illuminates the applied steps with an example on control system of welding process. The paper states: this reasoning model strongly integrates the natural intuition and operating feasibility, and it gives the implementation procedure of universal logic in engineering applied fields, especially the flexible control for complex system firstly.