A foil type flexible pressure sensor based on piezoelectric of highly C-axis oriented nitelide aluminum (AIN) thin film deposited on an aluminum foil substrate was developed. The sensor is composed of two AIN thin films and three aluminum foils as electrodes which detect the electric charge generated by pressure acting on the AIN film. These components are alternately piled with each other. Therefore, the inner electrode is shielded from external unknown environment by the outer two electrodes. This laminated structure of electrodes contributes to avoid unexpected leakage of electric charge and induced electric charge by unknown external electric field. Furthermore, two AIN films generate double amount of electric charge in comparison with the case of single film. Experimental results using a prototype of the sensor shows excellent linearity of sensitivity. However, because of internal loss of the generated electric charge, the sensor cannot be directly applied to measurement of static pressure. We analyze the frequency response obtained experimentally, and construct transfer function of the sensor. We show that it is able to detect quasi-static pressure by using inversed transfer function with estimated parameters.

View full abstract

For the purpose of developing acoustical environment recognition system in autonomous mobile robots, a practical method for simultaneous estimation of 3D-position and velocity of a moving random acoustic source in near-field is proposed by analyses of wavelet patterns of signals detected at 4 fixed points. By analysing correlation of time-scale (time-frequency) 2D-patterns of wavelet transforms of detected signals, information on ratios of relative compression and/or extension rates of the patterns due to the Doppler effect, those of the amplitude attenuation of wave propagation, and differences of wave propagation time are derived independently to the shape of power spectrum of the source radiating random signal. As a result, it can be applied to any moving acoustic source, radiating a stationary random signal. Then, with an aid of geometric relations of detection, the source position and velocity are estimated by the least mean squared (LMS) method on the basis of the information extracted above. After the principle of the proposed estimation method and the required conditions being made clear, the effectiveness and special features of the proposed method are ascertained through the fundamental experiments.

View full abstract

The tuning problem of two-degree-of-freedom PID (2DOF-PID) controllers is studied. The focus is placed on the effects of the location where the major disturbance enters into the plant. First, analysis is made about their effects on the response curves, and then the optimal 2DOF-PID parameters are recalculated assuming generalized location of the disturbance entrance using the two-step optimization method proposed by the authors in the previous report. The results are studied from the viewpoint of the overall performance and a guideline of parameter tuning is presented.

View full abstract

In this paper, PDS cooperative control of two one-link flexible arms is discussed. Dynamics of the flexible arms are described as an infinite-dimensional distributed parameter system. PDS Control is a simple sensor output feedback control law for flexible arms which guarantees asymptotic stability of the closed-loop system without finite dimensional approximation. The applicability of the PDS control law has been restricted to flexible mechnical structures without constraint conditions. We extend application of the PDS control law to flexible structures with a geometric constraint condition. We propose PDS cooperative control law for two one-link flexible arms holding a rigid object and prove the asymptotic stability of the closed-loop system. Experimental results are shown.

View full abstract

In this paper, we consider “Wall-Bouncing” task in which we repeat the process of hitting the ball that rebounds from the wall and the floor as a typical example of dexterous batting tasks. Through the stability analysis of a discrete system, we reveal that required dexterity varies depending on the location of equilibrium points, that is, the shape of the ball trajectory. A feedback control scheme for the racket at the moment of impact is also proposed to continue “Wall-Bouncing” task under disturbances.

View full abstract

This paper presents a practically efficient method for designing an interpolator that minimizes the L²-induced norm of the error system between the interpolator and a time-delay. The method is based on the so-called cutting-plane method for nondifferentiable convex optimization. The advantage of the proposed method is that it can solve the design problems of practical size with reasonable amounts of computation. Numerical examples show the effectiveness of the proposed method in comparison with the conventional methods.

View full abstract

In this paper, we propose an incremental state-space construction method using ART neural network in order to construct appropriate state-space for reinforcement learning. The proposed method is inspired by the notion of contradiction studied by Piagget. In this method, a state-transition table which represents the learner's states and actions is recorded. Then, if the current state transition against a certain perception is in conflict with the record, a new state for such perception is generated. We introduce two kinds of contradiction:“a contradiction such that different results are caused by the same states and the same actions” and “a contradiction due to ambiguous states” Several computer simulations on pole-balancing problem and light seeking problem for autonomous mobile robots confirm us the effectiveness of the proposed state-space construction method.

View full abstract

In senior high schools collaborative learning has become a reality because an idea not obtained by an individual learning can emerge. It is well known that the effectiveness of collaborative learning depends on the combination of students with various aptitudes. Therefore it is important how to determine the pairs of students from all members in a classroom. In this paper, a method for determining the optimal combination of pairs of students is proposed when technical high school students learn to design printed circuit boards collaboratively. For this purpose, first the correctness rates of their pairwise works are predicted by using neural networks. Consequently, the contribution of the individual aptitudes for the design problem to the correctness rate becomes clear. Secondly, a method to form the pairs of students optimally is proposed by applying an improved local search method. It is found from a numerical experiment that the correctness rate of the whole class with sixty two students attains more than seventy percentage as compared with that of the best student.

View full abstract

Recently many studies have been made on automatic design of complex systems by using evolutionary optimization techniques such as Genetic Algorithm (GA), Genetic Programming (GP) and Evolutionary Programming (EP). In this paper, a new method named Genetic Network Programming (GNP) is proposed in order to develop a more effective evolutionary optimization technique. GNP is composed of plural nodes which execute simple judgement/processing and they are connected with each other. To put it another way, GNP forms network structures, while general GPs form tree structures. This paper shows the detailed description about GNP and points out the differences among GNP, GP and other conventional. graph based methods. The comparison between GNP and GP is also shown by using a virtual simulation, the tileworld.

View full abstract

For the purpose of assisting for the blind to cross a crosswalk alone in safety, this paper presents a method for detecting an endline of a crosswalk and estimating the length of the crosswalk from an input image. First, we extract a region of white lines at a crosswalk by binarizing a grey scale image. And outlines of white lines are detected by using the Hough transformation. Then, we get the direction of their three-dimensional (3-D) vectors from the image information under the perspective projection, and obtain the direction of the road surface normal on the basis of vector analysis. Finally, we estimate the distance between the blind and an endline of a crosswalk. Extensive experiments show the effectiveness of the proposed method for real images of crosswalks.

View full abstract

One of the most important technologies for electric vehicles (EV's) will be control technology for driving AC motors. It is desired for control technology of the AC motors to have the following characteristics. 1) It does not require a position/speed sensor for controlling motor drive, which is mounted on rotor shaft. 2) It is a kind of vector controls that can produce torque quickly, efficiently and/or precisely. 3) It has wide driving-range and allows to realize EV's with no variable transmission. 4) It has so good regeneration performance that it can bring EV's to a full stop by electric brake only. In order to attain all of the above characteristics simultaneously, so-called sensorless vector control schemes will be indispensable. The frequency-hybrid (FH) vector control scheme will be one of high potential candidates as such for EV's. This paper presents design, realization and field test results of a radical new sensorless-vector-controlled EV with no variable transmission based on the FH vector control scheme, and clarifies its usefulness.

View full abstract

A new method utilizing sound is proposed for a non-destructive test of pipes. Random noise is transmitted from one end of a pipe and the sound pressure is picked up by a microphone at the other end. The output signal of the microphone is pre-whitened by an inverse filter and processed to calculate the partial auto-correlation (PARCOR) coefficients; the inverse filter which compensates the effects of the sound source and the microphone can be obtained from the PARCOR coefficients which are calculated in advance from the microphone signal for a straight pipe. Since the PARLOR coefficients give the set of reflection coefficients of a nonuniform acoustic tube represented as a series of sections with uniform area and equal length, the cross-sectional area function of the pipe is determined from the calculated PARCOR coefficients. An experimental apparatus was built and model experiments were performed for several pipes with an artificial defect, including a pipe with a partial plugging, a pipe with an extension of cross-sectional area and a pipe with an orifice in the wall. The results show that the cross-sectional area functions of pipes can be estimated by the proposed method.

View full abstract

A new device is proposed which simulates an acoustic inductor using active elements. The sound pressure at a certain point is picked up by a microphone. The medium there is vibrated by an piston at such a velocity as the medium should vibrate if there were a pipe. There is no difference between two cases from the acoustical point of view; the case the medium is vibrated by a piston and the case the medium within a real pipe is vibrating. A trial device was built which employed an electret microphone for picking up the sound pressure and a piezoelectric acuator for vibrating a piston. An experiment of the measurement of cavity capacity was performed with this device. The device was mounted on a container with an empty capacity of 62×10^-6m³ to form an acoustic resonator. The capacity of the cavity was varied by filling with water and the resonant frequencies were measured. The reuslts show that the trial device, although the actual dimension of the device is 12mm in diameter and 70mm in length, acts as an acoustic pipe with a length of 1200mm. Finally, the possibility of application of the simulated inductor to the measurement of other quantities is dicussed.

View full abstract