It is necessary to describe the mechanical characteristics of materials by appropriate mathematical models as a basic technique for the manipulation of their corresponding virtual objects. In this study, virtual objects are assumed to consist of combinations of tetra-polyhedron viscoelastic elements. Their dynamic characteristics is given by the calculation of their discrete-time models. Then, a sheet-like virtual object with given physical properties is theoretically cut by scissors. The resistance force to the grips of the scissors on their cutting of synthesized material is confirmed similar to the actual resistance force on the cutting of its corresponding actual material.
An inverse system for a given time-invariant system has been used as a compensator. However, in control systems where traditional inverse system is applied, the adverse effects of sensor noise may give seriously influence to the systems because the inverse system generally has a characteristic of high gain in high frequency area. Therefore, in these control systems other compensators except the inverse system must be constructed with complex characteristics. In this paper, a design method of a new inverse model with cut-off filter for Model Feedback Control System (MFCS) is proposed. Inverse model designed by the proposed method is an approximate inverse model for a given model. The approximation can be estimated by norm criterion for the difference between the model and the biproper transfer function used for construction of the inverse model. In order to demonstrate the usefulness of the proposed inverse model, this inverse model is applied to the MFCS. By theoretical and numerical analysis it can be shown that the proposed inverse model can reduce sufficiently the effects of sensor noise in the MFCS.
Assuming that the input vector field is not affected by unknown constant parameters, the main topic of this paper is the design of robust adaptive output tracking for uncertain nonlinear systems which are dependent on both unknown constant parameters entering linearly and unstructured uncertainties, via indirect feedback linearization (or estimation-based feedback linearization) by state feedback. For this control purpose, in a combined local-domain of state variables and parameter estimates, a nonlinear control scheme with robust adaptive law is derived indirectly. It is then shown that the nonlinear controller implies the robustness and adaptive output tracking of the closed-loop control system. This result is also confirmed with a computer experiment for uncertain single-link rigid robot.
A new method of feature generation and association based on attention is proposed in this paper. Attention to features of patterns plays an important role in recognition by human beings. A lot of models have been proposed to realize the attention in computational models. But the attention functions are determined by designers a priori and they are fixed. In the proposed model, the attention functions are organized automatically. They are realized by using the features obtained through learning. The proposed model is a 2-layered hierarchical network which consists of an associative memory layer and a feature layer. The connections between the layers represent the features of the input patterns. The connections are generated automatically and their weights are updated through the modified Hebbian learning rule. The features are extracted through a bottom-up procedure when the pattern is presented to the input layer. The connections corresponding to the most excited neuron in the feature layer are used as attention points in a top-down procedure. Through these procedures, the system can associate appropriate patterns.
The optical flow provides an impotant information on analyzing three dimensional structure and movement of moving objects. The gradient method has been developed to extract the optical flow. However, in the constraint equation of the gradient method only parallel motion is considered. Therefore, if rotate or magnification motion exist, it is difficult to extract the accurate optical flow from sequntial images correctly by using local method. The affine transform is introduced so as to expand the constraint equation of the optical flow for cope with the rotation and magnification motion. Moreover, we thought that the obtained optical flow from the moving object will have an effect on position of center point. But traditional method was not considerd moving objecter's position of center point. Therefore using this paper which is to estimate optical flow with consideration the positon of the center point is unknown. We could estimate the accurate optical flow in the case of any center point of moving object. We shown the validity of this method from experiment results.
This paper proposes a method for document validation to authorize documents on the Internet to follow the standards of descriptive rules which have been established by an industrial organization or a consortium in the electronic component industry field. This method adopts SGML (Standard Generalized Markup Language) as a standard document format and newly introduces the datadictionaries and the content validation rules in order to represent the standards of descriptive rules. The datadictionaries stipulate standard technical terms. And the content validation rules stipulate the specification of document validation. Furthermore, this paper proposes a document content validation tool, a software which validates whether SGML documents follow the standards of descriptive rules by interpreting the content validation rules. Lastly the effectiveness of the method is evaluated through the experiments in the E-CALS (Electronics CALS) project. This evaluation shows that the method can promote reliable document sharing by circulating validated documents to the content descriptive rules.
This paper presents a study of relaxation labeling of line images using Genetic Algorithms (GA). The proposed technique considers the labeling problem as two combination optimization problems and uses a GA search to obtain an optimal solution. A novel multi-GA system approach is proposed in the first optimization stage to produce a GA population for optimization by the second stage. Image partitioning for large images prior to labeling followed by reconstitution after labeling with appropriate edge processing was also implemented to achieve better optimization and shorter processing time. The proposed scheme was tested on several noiseless line drawings and achieved an average labeling success rate of over 80% for the test images used. Enlarging the same images by a factor of three while keeping the line thickness constant, led to average labeling success rates of almost 95%. Image restoration from random noise degradation was also achieved to some extent. The noise rejection capability of the scheme proved to be quite good; between 36.5% and 96.9% for different line images. The proposed labeling scheme was found to perform comparatively well compared with a popular deterministic labeling method, being superior in some cases but with the disadvantage of significantly longer processing time.
Users of conventional digital videoconferencing tools are typically limited to a single static view of their environment with no opportunities to move about or to obtain a general mental model of the communication space. The Ubiquitous Media Space (UMS) offers a viable solution to these problems, using the internet, while observing the restrictions of limited bandwidth. UMS supports background awareness and navigation with an active floor plan or another motorized camera and exploits sensors to provide additional background information regarding remote activity. Because the system is accessible through a web browser, UMS offers to make the world into a media space.
We have already developed an optical fiber trapping system and verified that optical trapping and manipulation of a micro object were easily achieved by a focused laser beam emerging from an optical fiber inserted into a sample chamber. In this paper, we described the performance of an optical fiber trap realized using a tapered hemispherically lensed optical fiber and experimentally analyzed the axial and transverse optical forces exerted on a micro-sphere to corroborate the optical trapping using a lensed optical fiber. Experimental results were as follows. (i) Transverse forceFtr, acting on a sphere was restoring force that acted to pull the micro-sphere back to the center of trap. (ii) Axial force Fax always acted to push a sphere in the direction of the beam away from the trapping fiber end. (iii) Vector sum of Ftr and Fax acting on a sphere gave a restoring force directed back to the stable point. (iv) Transverse force played a significant role in trapping a micro-sized object by means of an optical fiber.
Recently, it has been important to decrease fuel consumption of automobile engine for the sake of reducing the amount of CO2. The consumption is less, if idle speed is lower. However idling stability at low idle speed tends to be worse than the one at the high idle speed. So idle speed control must be more accurate. In this paper, we propose an engine model based feed-forward idle control system. The model based control system has obtained several better control results.
This study is to propose a method to improve the noise immunity of electronic circuits on printed circuit boards (abbr. PCBs). The noise on the PCB traces is induced from external cables which are different from the PCB traces in length. This paper experimentally and analytically clarifies that the large noise can be transferred from a noisy wire to a victim wire, even when two wires have different length and loose coupling at short-section. The reason why the large noise is transferred to PCB traces is because the base resonant frequency f0 of the shorter wire can be the same as the multiple resonant frequency N•f0 (N=2, 3, …) of the other wire. Further, in this paper, the characteristics of PCB traces are clarified, such as the resonant frequency dominating noise frequency is changed according to kinds of load and the location to which a load is connected, and then noise reduction methods are proposed. The bus circuit with non-power-consuming termination proposed by authors is verified to be effective on reducing induced noise as well as high speed bus data transmission.
Bidirectional passive optical network systems including a single trunk fiber and a bidirectional optical amplifier is proposed for the purpose of gathering ITV signals distributed on a line. A FM-SCM technique is used to discriminate the signals. The bidirectional transmission function is realized by using a WDM technique. A novel bidirectional EDFA is applied to long-haul analog signal transmission. The characteristics of gain and noise figure of the EDFA are clarified and they are found to be desirable for the bidirectional analog signal transmission. The transmission properties of the proposed system are confirmed for a 100-km SMF in a cable installed in the field. The satisfied CNR of more than 19dB for the FM modulated video channel can be attained for over the optical line loss of 60dB. Finally, stability of the system is confirmed by a continuous measurement of the SNR for two-weeks.
Designing a robot which works under unknown environment is a difficult problem. Evolutionary approach such as Genetic Algorithms (GA) is a possible solution to this problem. In this paper we propose a new improved GA incorporating gene evaluation into conventional GA and apply it to evolution of robot behavior. In the proposed method mutation probability of each gene is determined independently based on the evaluation value (fitness) of the gene. The proposed method prevents mutation of good genes which have high fitness so as to preserve scheme which is good for robot behavior, and promotes mutation of bad genes which have low fitness. We demonstrate the advantages of the proposed method by applying it to a problem of carrying a baggage by robot. Computational experiments show that the proposed method constructs a robot which has better ability of carrying a baggage comparing conventional GA.
This paper proposes a handwritten memo interface system that is basically independent of an application program. This system is managed to support works in which workers have to go out of an office with information concerning their job in a computer terminal like an inspection/sales work. They have to refer to the information in a site and they also have to impart information concerning the site situation in the site. For this purpose, a certain part of availability which these conventional pen interface systems have is lost because they can be used in limited timings and to limited memo areas. In the proposed system, image information of an application program is shown on a display as a background information for memo writing. And memo information is saved as vector information for transformation while the memo is being written. When the memo writing is finished, an image memo with background information is generated. The proposed system can be launched instantly for memo writing and can also save memo information with its background information. By applying the proposed system to electrical-supply-equipment inspection work, it is easily possible to reduce a time for inputting an inspection information to a host computer and to manage the information.
Fuzzy control behaves more robustness than conventional control that has been proved by many researches. A problem associated with the design of fuzzy control has been the size of the rule-base. As the number of system variable increases, the number of rules in a conventional complete rule set increases exponentially which will require the computer to process a huge data base, leading to memory overload and longer computational time. To make the problem manageable, cascade structure, in which the number of rules will increase linearly instead of exponentially with the number of system variables is proposed. This makes it possible to apply fuzzy rule based controllers to large scale system. On the other hand, cascade fuzzy controller is also an effective method to achieve a good performance such as robutness for a fuzzy control system. In this paper, the principle of cascade fuzzy controller is analyzed and its possibility and feasibility applying to large scale system have been discussed. Simulation results show the advantages of using the cascade structure fuzzy control to these models.
This paper deals with a segmentation method of an image composed of some kinds of textures with randomness by using a wavelet transform and neural networks. After applying a wavelet transform to strict non-stationary texture image, it is divided into a number of small arias with the same size, and the feature vectors in those arias are extracted by using two-dimensional autoregressive model and fractal dimension. The clustering of feature vectors is performed by applying the unsupervised and supervised neural networks. The feature vectors which are classified by the neural networks are mapped to the original image. In numerical examples, the validity of proposed methods is verified. This segmentation method using a wavelet transform and neural networks will be useful for strict non-stationary texture segmentation.
In this paper, we propose an algorithm to transform discrete-time systems with negative real-axis poles into continuous-time systems. Complementing as many degrees as the number of negative real-axis poles of a discrete-time system yields a continuous-time system with real coefficients. In order to calculate the system matrix Ac of a continuous-time system from the system matrix A of an identified discrete-time system, the closed form general solution of the matrix equation eX-A=0 is also utilized.
Recently, Evolutionary Robotics (ER) approach has been attracting a lot of concern in the field of robotics and artificial life. In this approach, neural networks are widely used to construct controllers for autonomous mobile robots, since they intrinsically have generalization, noise-tolerant abilities and so on. However, the followings are still open questions: 1) the gap between simulated and real environments, 2) the evolutionary and learning phase are completely separated, and 3) the conflict between stability and evolvability/adaptability. In this paper, we particularly focus on the evolvability. In order to overcome this problem, we propose a concept of dynamically rearranging function of neural networks by neuromodulators. We apply this concept to construct a gait controller of a six-legged robot by carrying out simulations.
This paper describes the layer-oriented development of object-oriented frameworks and the framework usage by introducing domain-specific CASE tools. We can verify the structure and the function of a framework by defining three processes: layer definition, layered framework development, and framework refinement. A layered framework consists of three layers: infrastructure, generic, and domain layer. Then, we can correctly and rapidly use the framework by introducing domain-specific CASE tools. The framework supplies the basic part as reusable components, and the domain-specific CASE tools automatically generate the system-specific part. We apply our approach to automatic teller machine software that is kind of an embedded system. We show the result of applying to this software, and discuss the effects and the potential problems in our approach.
We present a new pH distributed sensing system based on color imaging of pH-sensitive dye-doped hydrogel membrane. The RGB imaging data from a CCD camera were related to pH using the y value of CIE-xy chromaticity. The y-pH relationship were expressed as a first- or second-order function (calibration curve). The pH value at each pixel point was determined from the RGB imaging data using the calibration curve. The use of phenol red/ager gel membranes has been found to result in satisfactory distributed sensing of pH in the swollen gel membranes.