A new design method for a generalized predictive control (GPC) based on a parametrization of two-degree-of-freedom integral controllers has been proposed. The objective is to guarantee the stability of the control system without the dependence on the design parameters and to achieve the low sensitivity against the perturbation of the plant and the disturbance. The design procedure consists of two steps. First, we design a basic integral controller for a nominal plant using the LQG (Linear Quadratic Gaussian) method and parametrize a class of two-degree-of-freedom stabilizing controllers. Next, we tune the feedforward controller to incorporate the GPC method into our control structure. A numerical example is presented to show the effectiveness of the proposed method by comparing with the conventional GPC.
In this study, first, a statistical evaluation theory for EM interference noise is newly introduced through some detailed concretization of a multi-dimensional Hankel transform type characteristic function. As a result, the probability expressions on the EM interference noise are explicitly derived in two cases when only the random noise component exists and when a signal component is contained in the EM interference field. The validity of the proposed expressions are experimentally confirmed especially through consideration on the equivalent relationship between EM and acoustical noises as one of the same wave motion type environmental problems.
This study describes a new method of global motion compensation for highly efficient video coding. We propose a method to extract pan-tilt vector quantity of image sequence by using camera position information. But as global motion compensation, the synthetic picture is distorted because by optical aberration, and this distortion stand out at joint of picture. In this paper, we propose a picture compensation processing makes use of linear approximation function for image restoration. As a result of computer simulation, we reduced the distortion within 1 pixel and confirmed improvement of picture quality.
Demands for higher image quality of color display monitors are increasing. In particular, color blurring caused by misconvergence of three electron beams is a major factor to he evaluated. The electron beams are deflected by magnetic field of a deflection yoke. Therefore, coils of the deflection yoke should be wound precisely, but it is not easily implemented due to many errors of winding wire, winding machine and forming control. From another point of view, we have proposed selective coupling of coils other than random coupling. Thus for selecting similar coils and coupling them to compensate misconvergence, fast measuring and a method for coupling are required. Misconvergence estimation by calculation of beam trajectories takes a long time because of volumes of measurement data. To overcome these difficulties, we newly developed a method which is derived through multiple regression analysis between magnetic flux density distribution of a few intersection and miseonvergence. Furthermore, we have developed a new coupling method based on efficient coil categorization. In this paper, the new methods and the comparison between experiments and theoretical approaches are presented.
For the control of incoming and outgoing persons through the gates of security rooms in offices, a method of identifying individuals by using an image of the name plate which is usually clipped on each breast of workers is proposed. Before the identification, the calibration of the image data is carried out against the three dimensional directions of name plates at the time the images were obtained. This process is carried out by transferring the two dimensional image to the image which is estimated that it would be obtained in the case of setting the camera opposite to the name plate accurately. The identification of the characters on the name plates is accomplished by using a neural network. In this paper, the authors describe the results of the experiments which were performed to demonstrate the effectiveness of the proposed method.
In this paper, a design of the model following linear control system with time delays and disturbances is discussed. The design of the control system is constructed using the easy algebraic algorithm of matrices whose elements are polynomials of the two kinds of operators. Boundedness of the inner states for the control system is given and the utility of this control design is guaranteed. It is confirmed on basis of a numerical example that the output signal of the control system asymptotically follows the reference model signal in the case of the existence of disturbances.
Almost all ancient documents were handwritten with cursive styles using writing brush, as result, successive strokes that represent characters are generally connected and individual character patterns are considerably deformed. Character segmentation for documents plays an important role in character recognition and in building image database using such documents. In order to segment each character on a document, we identify the points at which the pressure of the writing brush changes infinitesimally between individual characters. We propose a method for effective segment positioning by automatic binary level control and interactive character segmentation. Result using the above method for character segmentation indicates that the proposed method has an 84.5% ratio of segmentation, which is about 27.2% higher as compared with the result when the method is not used. This paper describes the algorithm for estimating effective segmentation position and result based on this algorithm are evaluated.
This paper proposes a method for generating a map of a large scaled environment by observing scenes both along routes and at intersections. Panoramic representation is used for describing route scenes, and the number of routes connecting at an intersection is assumed to he known. A robot can obtain a closed loop by taking the same turn (leftmost for example) at each intersections when it moves along routes. By memorizing routes at intersections, the robot can select unmove routes for finding new closed loops. Further, by fusing new closed loops with found ones, the robot can build the map of environments. The effectiveness and robustness of our method are shown by our experiment in a real-world environment.
Large vocabulary, isolated word recognition requires a large amount of training data proportional to the vocabulary size to characterize each individual word model. A subword-unit-based approach is a more viable aternative than the word-based approach to overcome the problem of the training data size, since different words can share common segments in their representations in the former approach. This paper deals with a word recognition system where the subword-unit-based approach is employed. Subword boundaries within a word are detected by finding peaks of a sequence of power spectral distance or a sequence of delta cepstra associated with the word and the resulting sequence of subword units is deciphered into the original word by means of concatenated hidden Markov models of isolated words. This system attains average word recognition rates over 98%.
This paper discusses the disturbance characteristics for the optimal deadbeat control systems. First, the design method of the deadbeat controller which use the softening filter of the integralaction or pre-compensator is presented. Then it is proved that the control system proposed here gives the deadbeat response not only for the step change of the desired value but also the disturbance. Finally, numerical examples are presented to illustrate the effectiveness of the disturbance.
In dimension inspection of painted automotive body, the conventional method using slit light cannot detect the shape of edges of adjacent body parts. Thus the conventional method is not able to measure accurately the gap of between body parts. A new vision sensor based on simultaneous imaging from two directions has been developed. The sensor has a imaging unit. The imaging unit is consisted of a slit light source, a TV-camera, two triangular mirrors and four plane mirrors. The two edge of facing body parts are observed respectively from upper left and upper right directions by the imaging unit. Thus the TV-camera of the imaging unit is able to take simultaneously two edge images. At this time, in the TV-camera image, the overlapping of the two edge images occurs. We analyzed the characteristic of the overlapping of the images and showed the design index to get the clear edge images. The sensor used the imaging unit measures the gap and flushness with accuracy of within ±0.1mm without the influence of the shape of body edge.
Reliability prediction of electric parts by using neural network is described. A mean time to failure (MTTF) in reliability test for carbon film registor is adopted and predicted in this paper. In general, engineer's experienced decision making plays an important role in reliability problems. Therefore, intelligent information processing techniques are required in the field. A neural network is an attractive technique in many fields of reliability problems because the neural network can learn complicated relationships among the factors. This paper describes some experimental results of reliability prediction using a neural network and comparative study on the accuracy between conventional auto-regressive model and the neural network. By using the time series of the mean time to failure in reliability test, we attempt to find an appropriate model for predicting the value of mean time to failure in twenty hours in advance, in order to establish a convinient method for reliability estimation. We first discuss some essential issues to be considered in reliability problems and AI techniques. It is shown that the neural prediction model gives a better prediction results compared with the statistical model. Furthermore, the robustness of each model is indicated by applying to noise confused data of mean time to failure.
In this paper, development of a CAD of control systems which realizes an integrated environment for simulation and real-time implementation is introduced. It enables us to do not only analysis of control systems and design of controllers but also simulation and real-time implementation with a digital controller in the same software environment. By utilizing this software, the control engineer is able to repeat the procedure of design of control systems efficiently to achieve the best performance in much shorter time. It is also shown that a real-time control program is generated from a simulation program with a slight modification by making two programs have the similar structure.
In this paper, we propose a coarse grain process generation method suitable for the workstation cluster environment. In our method, we adapt the functional parallelizing strategy, in which a function written in a given program is treated as a unit of process generation. Namely, our process generation algorithm decides whether each function is executed as an independent distributed process or as a normal function call. For this decision, the dependency value is calculated to evaluate the parallel executability between callee/caller functions. Our process generation method can generate the processes with coarser granularity and lower communication frequency. The workstation cluster environment consists of several autonomous workstations connected through the local-area-network, and has no parallel-processing-specific facilities. Thus, the overheads of process invocation and communication among processes are very large in comparison with the parallel computer environment which other methods deal with. Therefore, considering the property of our processes described above, our process generation method is suitable for the distributed computer environment in term with the process granularity and the communication frequency.
The ADSL technology is standardized in US by ANSI to enable digital transmission speed of 1.5Mbps-6Mbps over existing copper phone access line. The ADSL technology extracts almost full capability, up to the theoretical limitation imposed by Shannon's Theory, of the access line. The limitation is solely determined by loss and cross talk noise of the access line. The ADSL technology is not so sensitive to bridged taps or AM radio wave, which are commonly seen hazards of phone access lines. But, in Japan, phone access lines are different from those in US. In Japanese phone cables, 4 wires are twisted together, and pulp as well as polyethylene is used as insulating material. As a result the access lines in Japan suffers more loss and crosstalk. Moreover, the ISDN modulation method in Japan generates a lot more nose to ADSL than that in US. The ANSI ADSL technology is first evaluated with access line environment in Japan. Then, three approaches; separation of ISDN and ADSL access line, ADSL power increase, ISDN filter capability enhancement are proposed and evaluated. It can be concluded that the ADSL technology is applicable also in Japan.
The dynamic tunneling algorithm (DTA) is an effective method for global optimization problem. It is based on the idea that the global minimum is included in the local minima. And it is consisted of two dynamical systems: the optimization system by which a local minimum from an initial point is found, and the tunneling system by which a new point in a lower valley is searched for. In theory, the global minimum should be found by computing the trajectory of each system by turns. However, according to our numerical experiments, the tunneling system in DTA has numerical instability. In consequence, it is next to impossible to tune its parameters appropriately, and it is few cases that the global minimum is found. In this paper, a multi-trajectory DTA is proposed. In this algorithm, an interaction among each trajectory of the tunneling system is introduced. As a result, it becomes possible to search for the global minimum efficiently, and in addition, as the number of trajectories is increased, the global minimum is found more frequently. Several numerical experiments show the ability of the proposed algorithm.
Workflow is the flow of business processes. Management of the workflow can shorten the time that is needed for group work. Therefore, workflow management systems, which perform definitions, execution, and management of business processes, have been widely implemented. However, only computing management does not enable to achieve a sufficient effects such as “shortening turn-around time”. In order to manage tasks more smoothly, the improvement and reconstruction of workflow are necessary. To accomplish this purpose, it is necessary to evaluate the workflow. However, it is impossible to put a modified workflow into practice and gain numerical data. Therefore, numerical data must be estimated by simulation. This paper proposes a workflow modeling method using the log data of a real work flow monitoring system. The parameters of a model were determined through the stochastic analysis of the log data and interviews to clerks. The simulated results of the estimated workflow model coincided with the observed data. Several types of sensitive analysis on critical parameters were clarified.
We have proposed the client server system(CSS) design support system which consists of following four parts, a requirement analysis part, a design planning part, a design analysis-explanation part and a learning part. In a design planning part, it is difficlut to decide rules for plan design, because correspondences between required functions and designed plans are complex. In this part, a plan is designed by modifying a case. In this paper, to present of compatibility and correspondence of functions to hardware and software, formats of data are proposed. CSS configuration cases are described as tree structures, in which it is represented that a function is realized by a set of components. Parts data, which show combinations of components for realization of functions, supplements the requested functions that are not included in a retrieved case. Compatibility knowledge indicates the sets of components which work well in a whole CSS. With these data, the retrieved case can be compared with clients' requirement and can be modified.
This paper discribes an attempt on a construction way of human friendly man-machine interface. At first, we do a simple experiment to find out the characteristic of human verbal communication. From the experimental results, we get some rules in case in human verbal communication. We construct the man-machine interface which is based on these rules. Through teaching process, we examine our verbal communication interface comparing with conventional interfaces. From this comparison, we recognize that the verbal communication interface is valid to construct the user-friendly man-machine interface.
This paper proposes a novel estimation method of dead time and relative degree, in which wavelet transformation (WT) is applied to the step response of linear system with dead time. By wavelet analysis, it is found that the dead time is detected from the co-form point of WT iso-phase lines and the relative degree is caluculated from scailing principle of WT. An example is given with simple numerical simulations on a 3rd order linear system.