Recently, the number of electronic devices handling confidential information has increased. In these devices, cryptographic circuits are applied to protect the confidential information. It has been sufficiently confirmed that the decryption of the encryption standards used in cryptographic circuits is computationally impossible. However, it was recently reported that when a theoretically safe encryption algorithm was embedded in the hardware, confidential information could be illegally specified by fault analysis attacks. Here, fault analysis attacks specify the secret keys by intentionally generating a fault during the encryption processing and by comparing the fault and normal cases. Almost all previous studies have strictly constraints related to the number of faults and the position of faults. Therefore, this study proposes a new fault analysis attack which has no constraint for the number of faults and considers the architecture. Experimental results using FPGA show the validity of the proposed attack.
Electromagnetic disturbances in vehicle-mounted FM radios are mainly caused by conducted noise currents flowing through wiring-harnesses from printed circuit boards (PCBs) with the slits of ground patterns. To suppress these kinds of noise currents from PCBs, we previously performed the FDTD simulation using eight simple two-layer PCB models in which the ground patterns were divided into two parts with different widths, which revealed that the cross-talk cannot always be reduced by widening the width of divided ground patters. In the present study, to confirm this finding, we made actual PCB samples, having the same geometry as that described above, and measured cross-talks between the two parallel signal traces with respect to the different widths of ground patterns. As a result, we confirmed that the measured results agree with the FDTD simulation, and also that the cross-talks have the smallest values at specific spaces between the divided ground patterns. In order to clarify the latter result, we calculated cross-talks from a simplified equivalent circuit model consisting of inductances, which were derived from their geometrical mean distance for the traces and ground patterns, and found that the circuit model can qualitatively explain the dependence of ground pattern width on cross-talks. The minimum reduction mechanism was found to come from the fact that a specific width of the ground pattern provides the smallest interlinking magnetic flux between the trace and the ground pattern.
A compensator with a function of estimating unmeasurable inputs based on a full-order observer has been already proposed. However, it seems very difficult to synthesize the compensator based on the minimal-order observer, because the innovation is equal to zero. In this paper, a novel synthesis method of the compensator based on the minimal-order observer is studied. The basic design idea is based on the fact that the full-order observer can be decomposed into the minimal-order observer and the output estimator.
As a great number of induction motors are used in industry, their safe and consistent operation is highly required. There seem many motors close to the end of their lifetime after a long term operation but little attention has been paid to their maintenance due to requirement of cutting cost. But early identification of a weak motor and its replacement is typically very inexpensive compared to the costs of plant down time and rushed repairs. Thus, establishment of an easy and effective fault diagnosis method has been strongly required to assure operation with high reliability. Short-circuit faults of winding is one of the most probable faults in motor drive systems. This paper proposes a novel and simple method for diagnosing short-circuit failure in stator windings of motors under load condition at site. The method involves probabilistic process for a gauging the probability of fault occurrence by considering distributions of features, i.e. magnitude and phase of current flowing into stator windings, depending on the condition of the short-circuited winding. A series of laboratory experiments and analyses of the results for motors with artificially introduced short-circuit fault proved that the proposed method is promising under load conditions.
We propose an inchworm-type microactuator which uses levitation caused by vertical vibration. The vertical vibration is generated by a piezoelectric actuator. The inchworm usually consists of thrust elements and clamp elements. The proposed inchworm-type microactuator however does not use any clamp elements. Three levitation elements are connected with two thrust elements which are stacked-type piezoelectric actuators. The levigation element consists of a mass, a piezoelectric actuator and a plate. The vertical vibration of the piezoelectric actuator lifts the levitation elements, and eliminates the friction. By controlling the phase of the horizontal and vertical piezoelectric actuators, the inchworm moves in the linear directions. The levitation height and the position of the inchworm are measured, and some experimental results show the feasibility of the proposed inchworm. The mechanism described in this paper is effective in a precise positioning system in a cleanroom.
A method of predicting the path of a pedestrian using both of the locus and the apparent posture of the pedestrian was proposed. As a result of an experiment, it was revealed that the method of using the feature of the posture based on HOG (Histograms of Oriented Gradients) together with the locus of a pedestrian is effective in the prediction of the pedestrian's path immediately before the pedestrian changes the course.
Stacked type DTMOS which enables to realize both high-speed low-power characteristics of FinFET type DTMOS and small pattern area of stacked transistor has been newly proposed. The delay time of substrate of stacked type DTMOS can be reduced to less than 10% compared with that of conventional FinFET type DTMOS by using the sidewall connection between gate and substrate. By using stacked structure of NMOS with (110) substrate on PMOS with (100) substrate high speed performance with the optimized mobility value can be realized without sacrificing the pattern area. Furthermore, the pattern area of inverter/NAND circuit, LSI for communication and DRAM buffer circuit with stacked type DTMOS has been compared with that of conventional FinFET type DTMOS. Newly proposed stacked type DTMOS is a promising candidate for realizing high performance system LSI such as the microprocessor of GHz operation.
This paper proposes a novel non-invasive and palpable measurement sensor for carotid pulse pressure. The unit consists of a pair of coil printed circuit boards, a pair of springs and a sensing plastic chip, with each spring attached between the circuit board and the chip. The distance between the boards is monitored from the displacement of the springs, and the information is converted into a voltage signal based on electromagnetic induction. First, the optimal forces externally applied to the proposed sensor were examined to allow accurate measurement of carotid pulse wave amplitude variations. It was found that the force applied when the measured maximum amplitudes of the sensor were obtained yielded the best performance. Next, carotid pulse waves were measured using the sensor with these optimal forces, and the results were compared with carotid pulse pressure values measured using a commercialized pulse wave transducer. The resulting coefficient of correlation between the two carotid waves was 0.9 or more. It was therefore concluded that the proposed sensor enables non-invasive measurement of carotid pulse waves.
The implantable cardioverter-defibrillator (ICD) is an effective therapeutic device for rescuing patients with cardiac diseases from death caused by life-threatening arrhythmias. The authors previously proposed a detection algorithm of life-threatening arrhythmias with a multiple regression model. To enhance the classification accuracy, in the present study, we have introduced an autoregressive filter and a multiple detection process into the previous detection algorithm. The experimental results showed that the proposed method could attain a high accuracy such that all ventricular fibrillation rhythms could be exactly detected. In addition, detection errors of sinus rhythms or supraventricular tachyarrhythmias provoking the ICD malfunction were reduced.
This paper is concerned with stability and guaranteed cost control of Takagi-Sugeno fuzzy bilinear systems with time-varying delay. We first consider the stability and guaranteed cost of the closed-loop systems by assuming a special form of a fuzzy controller. Delay-dependent conditions for the stability with guaranteed cost are given in terms of linear matrix inequalities (LMIs). Free-weighting matrix method is introduced to reduce the conservatism of the conditions. Based on such conditions, we propose a design method of guaranteed cost control for fuzzy bilinear systems with time-varying delays. Finally, we give a numerical example to illustrate our design procedures and to show the effectiveness of our approach.
In recent years, with increased globalization of manufacturing sites in the industrial field, the demand of minimization of TCO (Total Cost of Ownership) from system design phase to maintenance phase has risen rapidly in FA (Factory Automation) systems. The machine of the manufacturing devices which utilizes servo systems has conventionally intended to the high speed and high torque control of the servo motors, and raised the operation speed of the machine to realize higher productivity. However, reduction of the machine stop time and machine stop frequency, which occurs from the abnormalities in the environment, comes to be also considered as a very important factor for total productivity. When a power failure occurs or an excessive load is applied to the servo system, the machine has been stopped for safety in the conventional way of control. This paper proposes the position droop compensation control method to maintain the operation in such abnorrmalities of environment and to keep the productivity level by ensuring the stable speed control. The proposed method is like that, first to accumulate the position droop of the torque which cannot be transitionally generated by the motor in an abnormal environment and then return it into the position control after having enough torque margin. The effectiveness of this method is verified by implementing to control software of servo amplifiers and spindle amplifiers of machine tools.
The Amami archipelago has suffered from natural disasters by typhoons almost every year. Typhoons badly damage the electric power systems, which causes often a long-time blackout. To perform speedy restoration of electric power supply, it is important to predict the amount of damage accurately for an approaching typhoon. This paper deals with the prediction of the amount of electric power damage in the Amami archipelago using a Gaussian process (GP) model. The predicted amount of damage is given by the predictive mean of the GP and the confidence measure of the predicted amount is estimated by the predictive covariance of the GP. Simulation results based on the actual data are shown to illustrate the effectiveness of the proposed predictor.
In this present paper, we apply a stochastic differential equation(SDE) for mathematical model which describes manufacturing process from input materials to an end of manufacturing process. We utilize a risk-neutral principal in stochastic calculus based on the SDE. We extend the mathematical model to allow for objective evaluation. In order to improve a lead-time in manufacturing process, we introduce a evaluation equation which satisfy a certain compatibility constraint with lead-time. A certain compatibility constraint with lead-time means a strategy, that is, “Cut off irrelevant process in entire process, and Bring back its improved the process”. It is assumed that the probability density function of log-normal type which is a probability density function from input materials to an end of process. This present paper evaluate a certain compatibility constraint with lead-time by applying the risk neutrality and martingale measure of stochastic integral under this assumption. After its evaluation value calculated by a evaluation equation, a process control system evaluates the each sub-process, and reschedules to improve the utilization efficiency of the entire manufacturing process. Thus, a front-end-process is to cancel the request to the process, and follows the instructions on the process control systems. But, this present paper gives only simulation evaluation calculated the parameter, the paper does not discuss a compatibility between a evaluation value of calculated manufacturing lead-time compatibility constraint and a the real system. However, this present paper gives a idea to search bottleneck sub-process in the manufacturing process, and gives a idea of software tool for process synchronization.
Traffic jam has become a serious problem with exponential increase of vehicles recently. In urban area, almost all traffic congestion occurs at intersections. The intersections can be divided three types according to the number of road at the intersections: T-junction (3-way intersection), Crossroads (4-way intersections), and Multi-way intersection (more than 4). The traffic signal phases are generally more complicated at the multi-way intersections. Consequently, traffic jams occur very often. One of the ways to solve this problem is to extend roads, but it is not easy to realize in urban areas. In such cases, traffic signal control is considered as an effective way to solve the problem. In this paper, we propose a real-time traffic signal control system for the multi-way-intersection. The aim of the control is to search the optimal traffic signals to minimize traffic-jam probabilities using a new H-GA-PSO algorithm, which includes a modified PSO, an initial generation of particles, and a modified GA.
In this paper we propose a method of scene analysis based on a horse vision system. A horse vision system (HVS) consists of a pair of fisheye cameras which have a hemispherical field of view, respectively, and are laid to overlap each other partially. The characteristics of the HVS result in a representation which enables a wide omnidirectional monocular vision and a limited-field-of-view binocular vision simultaneously. We present the method of the rectification of the proposed HVS and the preliminary experimental results of scene analysis based on the HVS.
This paper proposes a novel stage lighting controller in which users can change stage lighting effects by selecting impressions through Kansei words. First, the paper describes a mathematical model that calculates output brightness value of each stage light by emotions to be perceived by audiences and by light information that includes position/orientation, color, and shape of each stage light. Second, a subjective evaluation of stage lightings was carried out in which the subjects observed computer graphics of various stage lightings on a projector screen and answered perceived emotions. Third, a stage lighting system is proposed and implemented, which automatically calculates the output brightness value of each stage light. A user of the system only has to input desired emotions and the light information of each stage light. To evaluate the proposed system, a user study was carried out in which the subjects observed stage lightings generated by the system in a real stage set. The subjects gave positive feedbacks on the usefulness of the system.
During the past several years, there has been a significant number of researches conducted in the field of Multiple Resources Scheduling Problem (MRSP). Intelligent manufacturing planning and scheduling based on meta-heuristic methods, such as Genetic Algorithms (GAs), Simulated Annealing (SA), and Particle Swarm Optimization (PSO), have become some of the common tools for finding acceptable solutions within reasonable computational time in real settings. However, limited researches were conducted at analysing the effects of interdependent relationships between each activity of group decision-making processes. Moreover for a complex and large problem, local constraints and objectives from each managerial entity, and their effects on global objectives of the problem cannot be effectively represented using a single model. In this paper, we propose a novel Cooperative Bayesian Optimization Algorithm (COBOA) to overcome the challenges mentioned afore. The COBOA approach employs the concepts of divide-and-conquer strategy and it is incorporated with an innovative co-evolutionary framework. Considerable experiments were performed, and the results confirmed that COBOA outperforms recent research results for scheduling problems in FMS.
This paper proposes a new optimizing system for stock portfolios which uses evolutionary computation techniques to derive a highly suitable combination and investment ratio of brands as well as an appropriate trading-strategy tree. Accurately predicting price trends in the stock market is a difficult task to achieve with the result that investors often suffer great losses. Because stock portfolios are thought to be a valid means of avoiding such risks in terms of financial engineering, they have the effect of reducing risk by diversifying investment into several different brands. Based on this, it was attempted to determine an optimal combination of brands that constitute a portfolio and to derive the investment ratio using a multi-objective genetic algorithm, and also to optimize a trading strategy tree using genetic programming. When a performance evaluation was carried out, the system was found to generally obtain the operative results by making it possible to obtain stable profits using a combination of low risk brands. The system was also able to realize low risk investments in all test periods.
Network security becomes an extraordinary important issue since various attacks are launched frequently with increased usage of computers and networks in many fields. Therefore, many techniques for intrusion detection have been studied in order to build a secure Internet atmosphere. One of them, that is, the rule-based classification methods usually rank the rules in order of precedence to build the classifier for different classes. When a new data comes and matches with rules, the most confident rule is usually used for the class label of the data. However, simply matching with the most confidence rule leads to lower classification accuracy. In addition, the domain knowledge should be clear and its explicit analysis is necessary in such methods. Therefore, in this paper, the average matching degree is calculated by matching data with rules, i.e., normal rules and misuse intrusion rules in a two dimensional space in order to form more realistic classification model. In this paper, a rule-based classification method using the average matching degree and distance concept has been proposed for classifying unknown network connections into normal, misuse intrusion or anomaly intrusion, where the model uses the distance between a data and rules in the average matching degree space. The benchmark data KDD Cup 1999 and NSL-KDD are used to evaluate the performance of the proposed method.
An ant system with two colonies is proposed for the combinatorial optimization problems. The proposed method is inspired by the knowledge that there are many colonies of ants in the natural world. The proposed two-colony mechanism plays an important role in enhancing the diversification while maintaining the intensification of the algorithm. At first, ants perform solution search procedure by cooperating with each others in the same colony until no better solution is found after a certain time period. Then, communication between the two colonies is performed to build new pheromone distributions for each colony, and ants start their search procedure again in each separate colony, based on the new pheromone distribution. Furthermore, in order to enhance the quality of the iteratively found solutions, a local search algorithm is also adopted. The proposed algorithm is tested by simulating the Traveling Salesman Problem (TSP). Simulation results show that the proposed method performs better than the traditional ACO algorithms.
In this paper, we propose an approach for classifying customers in retail stores into given types according to their shopping paths, each of which is a sequence of sections visited by the corresponding customer and is gathered by an RFID tag. The approach vectorizes a sequence of sections; that is, the approach splits such a sequence into tuples of sections, then sums up occurring counts of those tuples. This vectorization is based on the hypothesis that a customer's type has relation to sub-sequences of sections in his/her shopping path and a conjecture that customers' types can be attributed to co-occurrences of such sub-sequences. After vectorization, the proposed approach applies a general discrimination method to such vectors of equal length. In computational illustrations, the principal component regression is selected as a representative of general discrimination methods and is applied to shopping paths collected in an existing retail store so as to predict whether a customer purchases items much than average or not. Computational results display the effectiveness of the proposed approach as higher forecast accuracies than known works.
The long-term aim of our research is to develop near infra-red spectroscopy for non-invasive measurement of blood alcohol concentrations. The absorption spectrum of ethyl-alcohol could possibly provide the means for such an approach, but the absorption bands of other molecules, such as glucose and acetaldehyde, together with the very low in vivo ethyl-alcohol concentrations (EAC), represent significant analytical challenges. Here we present results of our in vitro investigation of ethyl-alcohol absorption spectra, and the use of multiple linear regression analysis to predict EAC in the presence of glucose and acetaldehyde. We used ethyl-alcohol and mixed solutions of ethyl-alcohol, glucose, and acetaldehyde. The latter were divided into three experimental systems as ethyl-alcohol/glucose, ethyl-alcohol/acetaldehyde, and ethyl-alcohol/glucose/acetaldehyde solutions. The range of EAC used was the same as that found in blood. The results showed good correlation between the actual EAC and the predicted EAC (control; R=0.99, mean absolute error (MAE)=0.12mg/ml: glucose system; R=0.97, MAE=0.18mg/ml: acetaldehyde system; R=0.99, MAE=0.23mg/ml: glucose & acetaldehyde system; R=0.98, MAE=0.19mg/ml). We conclude that the change of glucose and acetaldehyde concentration does not measurably affect the absorption bands of ethyl-alcohol.
A design of lumped dual-frequency matching network for frequency-dependent complex loads having two equivalent circuits is introduced in this letter. First, a design procedure of the proposed circuit for transistor 2SC5509 is reported concretely. Next, numerical examples of the proposed circuit is presented, and effectiveness of the proposed design procedure is confirmed.
This study proposes a copy-learning model Elman network (CE-net) which can recognize overlapped character strings by only learning the standard character strings. CE-net learns standard character strings by a part of the 3-layer net, copies obtained weights and biases on an unused part of the net by a copy-learning rule, and recognizes overlapped character strings. We had confirmed the effectiveness of CE-net from several experiments.
The resolution of a satellite image depends on the satellite's posture and geographical features of the captured area. In this paper, we propose a method of synthesizing arbitrary pointed ALOS/PRISM images from textures and DEMs that are acquired from a stereo pair of ALOS/PRISM images. At first, intersections of DEMs and each sight vector of the virtual satellite are calculated. Then, the simulated image of the virtual satellite is generated by collecting texture of the intersections. We have performed experiments and the results show that the simulated images are geometrically true.
With the advent of multi-core architecture, the data distribution servers have been improving dramatically in performance. However, the servers cannot exert their improved full powers because Gigabit Ethernet networks are the mainstream for more than 10 years without bandwidth progress. In this paper, the simple method for making the best use of server abilities is proposed which surpasses IEEE802.3ad in bandwidth, load balancing and cost-effectiveness.