All issues
Volume 39, Issue 3
Displaying 1-15 of 15 articles from this issue
- |<
- <
- 1
- >
- >|
-
Optimization of the Cellular Neural Networks by Designing an Output FunctionZhong ZHANG, Michihiro NAMBA, Hiroaki KAWABATA, Akihiro KANAGAWA2003 Volume 39 Issue 3 Pages 209-217
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIt is well known that the cellular neural network (CNN) is very effective as an associative memory medium. And the saturation (output) function plays an important role in CNN, because it affects the operation, the stable equilibrium points and the performance of CNN. However, to the best of our knowledge a systematic design procedure for the output function is not available in the literature. In this paper, we present a simple, yet, effective design method for the two and three-output functions. To demonstrate the effectiveness of the output functions, we tested CNN on synthesized images. In addition, we applied CNN to recongnize Chinese characteres and diagnose liver diseases, and obtained very encouraging results.View full abstractDownload PDF (4171K) -
Noriyuki FURUICHI, Tadashi HACHIGA, Koichi HISHIDA, Masaya KUMADA2003 Volume 39 Issue 3 Pages 218-224
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA novel method to measure two-component velocity profile simultaneously was developed by using the combined system of a CCD area image sensor (CAIS) and a multi-point LDV. The benefits of this method are shown by pursuing the compact and convenient over the system. Moreover, the results of some measurements such as the flow field behind the cylinder shows the reliability of the measurement for turbulence field. The basics techniques of this method are established and the applicability for the measurement of flow field is shown.View full abstractDownload PDF (3022K) -
Masanori HARIYAMA, Toshiki TAKEUCHI, Michitaka KAMEYAMA2003 Volume 39 Issue 3 Pages 225-233
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSHigh-speed acquisition of reliable three-dimensional (3-D) information plays an essential role in real-world intelligent systems such as intelligent robots and highly-safe intelligent vehicles. Stereo vision is a well-known method for 3-D instrumentation. A major issue of stereo vision is to establish reliable correspondence between images. This paper presents a reliable stereo-matching algorithm based on SAD (Sum of Absolute Differences) computation. Reliable correspondence can be established by selecting a desirable window size of the SAD computation based on the uniqueness of a minimum of the SAD graph. Moreover, to increase reliability in matching, edge information is efficiently used for selecting the window size. A pixel-serial and window-parallel architecture is also presented to achieve 100% utilization of processing elements. The performance of the VLSI processor is evaluated to be 1500 times higher than that of a general-purpose processor.View full abstractDownload PDF (5506K) -
Teturo ITAMI2003 Volume 39 Issue 3 Pages 234-243
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA practical algorithm utilizing the theorem obtained in an eigenvalue analysis of nonlinear control is proposed. This theorem on the feedback optimization of nonlinear control systems by only one eigenfunction is enforced by an analytic continuation of the designer's constant HR. Using this method of HR=iHR, typical Affine systems with 1-input and 2-states and with nonlinearities both in state equations of van-der-Pol type and in nonquadratic cost functions are optimized by feedback input. Accordingly, this report presents an algorithm to apply the eigenvalue theorem on the feedback optimization of nonlinear control systems.View full abstractDownload PDF (1682K) -
Takashi KUNITAKE, Yoshito OHTA2003 Volume 39 Issue 3 Pages 244-252
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, an l1-optimal control problem with frequency and time domain constraints on a closed-loop response is considered. The problem is formulated as an infinite dimensional linear programming problem, and the primal and dual approach is exploited. This sometimes results in non-zero gap between the primal and dual costs. This paper proposes a new method to approximate the primal and dual problems without the duality gap. As a result, the method yields a sub-optimal solution with arbitrary accuracy.View full abstractDownload PDF (1883K) -
Makoto YOSHIDA, Takashi KAWATO, Toshinori FUJITA, Kenji KAWASHIMA, Tos ...2003 Volume 39 Issue 3 Pages 253-258
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSCity gas is one of the most important necessities of daily city life and social infrastructures. City gas is delivered to every user through a pipeline network. The gas pressure in the pipeline is regulated by gas regulator. In the pressure control system, characteristics of gas pipeline is as important as characteristics of regulator. There are many reports about the transfer function model of the fluid pipeline. But suitable model about the gas transmission pipeline is not known. In this paper, a new model considering the heat transfer between pipe wall and gas and temperature change of gas is proposed. To evaluate this model, frequency response tests are used. As a result, the proposed model shows a better agreement when compared with the experimental results than conventional models. The results show the effectiveness of the model.View full abstractDownload PDF (1416K) -
Keisuke OZAWA, Yoshiyuki NODA, Takanori YAMAZAKI, Kazuyuki KAMIMURA, S ...2003 Volume 39 Issue 3 Pages 259-265
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSMany studies have been devoted to development of tuning methods based on different principles in complexity, flexibility, and in the amount of process knowledge use. There is, however, a need for simple, easy-to-use, intuitive methods that require little information and that give moderate performance. Optimization is one of the powerful tools for tuning of controllers because the method is conceptually simple. This paper describes a tuning method for PID controller using optimization subject to constraints on derivatives of control input. Its control performance for a first-order lag plus a deadtime system is shown as an example of the commonly approximated controlled plants in process industries. Graphs by which the optimal PID parameters can be obtained are presented as functions of a normalized deadtime of a plant. To evaluate the control performance, the responses to load disturbance and reference input changes are presented compared to those by the partial model matching method. Moreover, it is found that the PID controller tuned by this optimization technique is robust and useful in simulations where a traditional PID controller would be used.View full abstractDownload PDF (1724K) -
Kazuaki YAMADA, Kazuhiro OHKURA, Kanji UEDA2003 Volume 39 Issue 3 Pages 266-275
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWe describe a distributed approach to controlling autonomous arm robots. The robots need to acquire cooperative behaviors in order to smoothly lift an object. Each arm robot has its own reinforcement learning unit for decision-making. In investigating this task, we are primarily interested in the question of how to design a reinforcement learning control system for a multi-agent system. An applied reinforcement learning algorithm uses Bayesian discrimination method to segment continuous state and action spaces simultaneously, thereby generating of a set of effective rules. The proposed approach is examined empirically with two real arm robots. The basic dynamics of the reinforcement learning process are also analyzed.View full abstractDownload PDF (4594K) -
Masatoshi HARIGAE, Hiroshi TOMITA, Takeshi NISHIZAWA2003 Volume 39 Issue 3 Pages 276-285
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSNAL and NASDA develop High Speed Flight Demonstrators (HSFDs) to examine an automatic take-off and landing technology (Phase 1) and to measure the transonic aerodynamic characteristics of a reusable space plane for getting the reference data of a CFD (Computed Fluid Dynamics) technology (Phase 2). For HSFD navigation, we develop a GPS aided inertial navigation system (GAINS) whose distinctive feature is using the carrier-phase DGPS (CDGPS)/INS hybrid navigation technology. This paper describes the design of GAINS. We show its accuracy, integrity and continuity performances that are analyzed by numerical simulations using the covariance analysis technique. Ground and flight tests are also carried out to confirm the performance of the GAINS flight model. Both results prove that GAINS meets the requirements of the navigation performance under severe flight conditions of HSFD.View full abstractDownload PDF (4475K) -
Hiroshi KAWAKAMI, Eiji IRIE, Tadataka KONISHI, Osamu KATAI2003 Volume 39 Issue 3 Pages 286-294
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSA method of artifact fusion is proposed based on the interactions among physical causalities. Imaging some relevant design examples is beneficial for designing artifacts. It involves however some risks to fall into a trap the so-call “mental inatea”. To bring out a special feature of each design example, designers should not be caught up by its structural features e.g. shapes, materials etc. In other words, in the “structural space”, sophisticated solutions are not near to the examples. For supporting functional design without falling into mental inatea, we will propose a method to propose a design candidate which is not focused on structures but on physical causalities. Structural features do not play a main role in this method, they are only referred as the conditions on the validity of each “physical causal relation”. Design examples in the “case-base” of our method are encoded into forms what we call “Functional Diagrams” and “Physical Causal Networks (PCNs)”, which provide purely causal aspects of the design example, and hence they can be processed in various ways. Based on these types of codings, we can develop various methods for supporting functional design processes. Merging PCNs of different design examples is one of the above mentioned method, which can be interpreted as an “artifact fusion” in the level of physical causalities. This process results in a novel artifact which cannot be obtained by merely structural combinations or modifications.View full abstractDownload PDF (4120K) -
Daisuke KATO, Manabu KOTANI2003 Volume 39 Issue 3 Pages 295-301
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSWe propose a method of feature extraction to improve the performance of pattern recognition technique. The extracted features are defined as polynomial expressions, which are composed of the original input information. These polynomial expressions are searched by a coevolutionary genetic programming. We introduce a new fitness function based on competition between individuals. Experiments were performed for real-world datasets with k nearest neighbor classification rule. From experimental results, we have confirmed that the proposed method could maintain the diversity of populations and improve discrimination accuracy on most datasets.View full abstractDownload PDF (1254K) -
Ryoji KATAYANAGI2003 Volume 39 Issue 3 Pages 302-304
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSThis paper proposes a design method which is based on “-90° phase locus” for the two-input system associated with the optimal regulator. This leads to a simple technique, which permits a designer to achieve the objective to get a system with large phase margin.View full abstractDownload PDF (396K) -
Toshinori NAWATA, Hitoshi TAKATA2003 Volume 39 Issue 3 Pages 305-306
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper a nonlinear feedback control, called an augmented automatic choosing control (AACC), for nonlinear systems with constrained input is presented. Parameters of the control are suboptimally selected by minimizing the Hamiltonian with the aid of a genetic algorithm.View full abstractDownload PDF (287K) -
Sadaaki KUNIMATSU, Takao FUJII, Taro TSUJINO2003 Volume 39 Issue 3 Pages 307-309
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper we consider quadratic stabilization of ILQ servo systems with generalized reference inputs. We first solve the inverse problem of quadratic stabilization in the ILQ servo system with an uncertain plant. We then derive a new sufficient condition for quadratically stabilizing those systems subject to perturbations of both state and input matrices. Finally we propose a renewed robust ILQ design method.View full abstractDownload PDF (449K) -
Minoru ITO, Masanori SUGISAKA2003 Volume 39 Issue 3 Pages 310-312
Published: March 31, 2003
Released on J-STAGE: March 27, 2009
JOURNAL FREE ACCESSIn this paper, we propose an “elite correlation selection operator (ecs)” as a new selection operator for survival. The proposed method selects individuals with a fitness and a genotype similarity among populations. The performance of the proposed method is examined with several standard test functions. We compare the performance of the proposed method with conventional methods. The experimental results show good performance.View full abstractDownload PDF (1128K)
- |<
- <
- 1
- >
- >|