A triple-tail cell possesses a wide linear input voltage range in spite of its simple configuration. The tripletail cell requires a voltage divider. In practical implementation, the maintenance of the intrinsic linearity necessitates a resistive voltage divider that has low resistances. This makes the input resistance of the tripletail cell low. This paper presents a method to realize high input resistances of the triple-tail cell. An active voltage divider is employed. The output resistances of the active voltage divider are lowered by current feedback. Thus, although the operating current is reduced, the output resistance is sufficiently low. This enables the voltage divider to have high input resistance. A third-order gyrator-C filter composed of the triple-tail cell is analyzed through SPICE simulation. Further, the characteristic variations of the triple-tail cell and the voltage divider due to transistor mismatches are analyzed. It is shown that the characteristic variations of the voltage divider affect strongly the transconductance of the triple-tail cell. The results of Monte Carlo simulation show that the proposed circuit is feasible.
To improve color reproducibility in computer monitor display and television receiver, an analog gamma correction circuit with not approximate but accurately exponent conversion characteristic has been developed. It consists of a logarithm converter, a variable gain amplifier and an exponent converter. Owing to making the most effective use of characteristics in differential bipolar transistor circuits, the gamma correction circuit has a small size configuration and high speed characteristics in signal processing. Its exponent gamma value is easily changeable from less than 0.1 to more than 3.1, and its frequency width is obtained 125 MHz on bread board. Consequently, it has expected to become possible that white balance adjustments for the displays are precise enough to use them in color matching applications, and color reproductions in displays of many devices are stable independently of adjustments by users too.
We designed a bit-block circuit which becomes a part of a 1 V operational, 20 Ms/s, 10-bit CMOS pipe-lined current-mode A/D converter (ADC). Full current-mode circuits and CMOS 0.6 μm process with low-threshold voltage (low-Vth) transistors are utilized throughout the design. All the node voltages in the current mirrorbased circuits are fixed; as a result, a very accurate current transfer is possible since the influence of the input current dependency is eliminated. The differential configuration is used throughout the circuit so that the switch feed-through error and even numbered harmonic distortions are suppressed. The control signal for the switches is voltage boosted to make their on-resistance low. The simulation resulted in less than 0.1% linearity and 63 dB of a spurious-free dynamic range at the output of the sample-and-hold circuit with 1 MHz full-scale input current. 0.05 μA and 0.64 μA of the error current at the output were also observed at a clock of 20 MHz, when 1 LSB of 0.4 μA current change is input and ±100 μA current change at an equivalent rate of 5 MHz is input to this bit-block, respectively.
The possibility of realizing a 1.5 V and 30 Ms/s video-rate CMOS sample-and-hold circuit with 9- to 10-bit accuracy has been studied. The current-mode circuit technique has been adopted to enable the sample-and-hold circuit operate in 1.5 V of supply voltage. The sample-and-hold uses current mirrors whose node voltages are completely fixed in order to avoid the influence of the gm and the output conductance variation of a transistor when signal current changes. Sample switches are placed differentially in between current mirror transistors, which eliminates the undesirable clock feed-through from analog switches. The designed circuit was actually fabricated by using standard CMOS 0.6 μm process and its characteristic was evaluated. The signal-to-noise ratio of 57 dB and 54 dB at a clock speed of 20 MHz and 30 MHz, respectively, were measured from the 1.5 V supply voltage with the power dissipation of 2.3 mW.
This paper describes highly efficient on-chip high-voltage multipliers using charge pump circuits. The proposed charge pump circuits use bootstrapped gate transfer switches to avoid the threshold voltage drop in conventional Dickson charge pump circuits and enables them to generate a given voltage with a smaller number of pumping stages, which results in higher efficiency. The SPICE simulation results show that the proposed circuits have high pumping gain, are suitable for low-voltage operation, and have ample current drive capability.
This paper proposes second-generation current conveyors (CCII's) for low supply voltage operation. The power consumption of the proposed circuits is much lower than that of a conventional circuit. A method for reducing the effect caused by channel-length modulation is also proposed. Simulation confirms several advantages of the proposed circuits.
A switched-current circuit is one of the typical current-mode signal processing circuits, which is suitable for low-power supply voltage operation because of small voltage swings. This paper proposes a switched-current sample-and-hold circuit with low-power consumption. The power consumption of the proposed circuit is much lower than those of conventional sample-and-hold circuits. The maximum input and output currents can be determined not by the magnitude of bias currents but by aspect ratios of MOSFET's used in current mirrors. SPICE simulations of the sample-and-hold circuit show quite good agreement with the theory.
An OTA is one of useful analog circuit blocks. In order to obtain a linear characteristic of OTA's transconductance, two MOSFET's operated in the linear region are used as input devices. The linearity of the OTA is achieved by fixing a drain-to-source voltage of the MOSFET's. A Current Transfer Voltage Regulated circuit (CTVR) is utilized to fix the drain-to-source voltage. An OTA using the CTVR realizes linear characteristic. The conventional CTVR has a problem of voltage error between the Gate-to-Sourse voltage of nMOSFET and pMOSFET In this paper, Low-Voltage OTA's are proposed. The proposed circuits use the CTVR without influence of voltage setting error. The proposed circuits are simulated by PSpice to confirm their performance.
In this paper, an application of genetic algorithm for generation of jumping motion pattern of a hopping robot is described. The parameters of a central pattern generator are regarded as genes and tuned by the genetic algorithm so that the robot car, jump to the reference height continuously with the minimum drive current. To realize online tuning of the parameters, new genetic operations such as a few individuals, rapid estimation, instant selection and intentional mutation are introduced. Experimental and simulation results demonstrate that the proposed scheme is effective to generate the suitable jumping motion pattern of the hopping robot.
To expand a dynamic range of brightness is generally effective to enhance a contrast of a gray image. On the other hand, contrast enhancement that is based on the evaluation of spatial features in an objective image such as strength of edges is also useful to generate a visual contrastive result. This paper describes experiments to evaluate effective spatial features for contrast enhancement using the method to determine a relation between input and output gray levels that adapt to an input image by the genetic algorithm. Experimental results show that increasing the sum of edge strengths was an important condition for generating a visual contrastive image.
Probabilistic Universal Learning Networks are proposed, where a calculation method of the propagation of stochastic signals through Universal Learning Networks is provided. Probabilistic Universal Learning Networks also provide a gradient learning method to optimize parameters in Universal Learning Networks by minimizing the value of the stochastic-based evaluation function. From simulations, it has been shown that identification of a nonlinear dynamic system can be realized without overfitting by using Probabilistic Universal Learning Networks.
A novel routing protocol for route reconstruction to realize high throughput on Point to Multi-Point (P-MP) communication is investigated. Route reconstruction is operated when a route becomes invalid in routing protocol. In order to transmit data packets quickly, it is important to shorten the time required for route reconstruction and to reduce broadcast packet (ROUTE DISCOVERY packet: RD) traffic needed to discover new route. In P-MP communication, the traffic of RD packet delays the transmission of data packets for other routes, so it is necessary for the routing protocol to provide high data packet throughput. A novel routing protocol to transmit RD packets to source terminals directly and to introduce the hop-limit technique is proposed in this paper. The results of route reconstruction processes are transmitted to source terminals directly, so the proposed protocol can realize rapid route reconstruction. The hop-limit technique provides to limit the route discovery area and it is possible to reduce extra traffic of RD packets. Throughput of data packets and RD packet traffic are evaluated by computer simulation. The results confirm that the proposed protocol is suitable for P-MP communication in wireless ad-hoc networks.
In sequential controllers of plants, ladder diagram is one of major programming languages. However, it is difficult to understand their control sequence and to maintain them because ladder diagram does not clearly represent the sequence of plants' behaviors. Sequential Function Chart (SFC), which is based on Petri net, recently spreads as a powerful graphical description of sequential logic. SFC can represent clearly the sequence of control programs. Therefore, the method to translate ladder diagram into SFC is required. In this paper, control sequence extraction method is proposed. In this method, a tentative SFC is created by searching all actually possible states by using transitions in the ladder diagram. However, the tentative SFC description can be complicated because of existance of non-order control, which does not depend on fixed sequence, for example manual control. Therefore, sequential flows of plants' behaviors in the tentative SFC are extracted according to the difference of transfer frequency between order control and non-order control, and the check of both-way control included by non-order control. Through an application of the proposed method to a practical plant model, it is confirmed that the adequate sequence which represents the original plants' behaviors is extracted from ladder diagram.
This paper proposes an efficient method based on neural networks to automatically determine the quality of pellet components. By using two images (front and side views, referred to below as CIFV and CISV respectively) captured through the microscope using a CCD camera, it is possible to assess the quality of pellet component. Here we define the captured image in front view as CIFV and the captured image in side view as CISV. In addition, Two corresponding template images are defined here as TIFV for front view and TISV for side view, respectively. These template images were created based on the design dimension. Prior to carrying out the quality evaluation, the captured images were processed to eliminate noise and the images were binarized. Then, the datum matching plane and the datum machining line for the captured images were calculated. Next, using a coordinate transformation related to the positions of the datum plane and the datum line, the test hole in the captured image can be moved to the center part of this image in order to match the template image. After that, the values of certain feature parameters are calculated based on the errors and on the differences between the captured images and template images. Finally, the values of parameter features serve as inputs of the neural networks for determining the quality of the component. The experimental results confirm the effectiveness and efficiency of this procedure for inspecting the quality of pellet components.
We observed the coseismic electromagnetic phenomena just at the occurrence of the earthquake in Sendai area. This earthquake's epicenter was very close, less than 10km from our observation site. Our site accidentally located on the fault of which underground tip slid at the focus. In addition to variations of the ground acceleration signals and the earth potential difference ones, we observed the clear variations of the atmospheric electricity using the system with 10ms sampling time. In its aftershock we can identify the very small coseismic variations of atmospheric electricity removing the other unnecessary signals. The model how this variation appears is proposed. More events are needed to make this clear and the observation must be continued.
The new re-order system aims to prevent running out of stock, to limit replenishment quantities and to keep reasonable stock level for multi-items. In the Lot-size re-order system and the Interval re-order system, replenishment intervals or quantities are determined by the each system after replenishment items are decided. When both re-order systems run for multi-items, total replenishment quantities usually over or under the limited capacity. The problem of the limited capacity was solved by the production scheduling or supply lead-time delay. We propose Optimal Ordering Method (OOM) to be controlled by the margin stock ratio under the limited capacity. Margin stock ratio is defined by the prediction of the out-of-stock situation that is calculated by the available stock of inventory and necessary demand quantities of supply lead-time period. After each item is arranged by the margin stock ratio, replenishment items are determined under the limited capacity. This new method has been applied to a window frame manufacturing process successfully.
Satellite communication systems suitable for applications in distributed cooperative work are proposed and evaluated. We have been proposing a new method of multimedia data communication, in which voice and video signals are exchanged via allocated SCPC channels, and burst data via code-division multiplexing (CDM). For burst data access, demand assignment and random access are compared. This burst data is transmitted in the same frequency band as voice and video data. The evaluation clarified that it is possible to support a considerable number of concurrent synchronous cooperation while conserving valuable bandwidth resources and alleviating the costs of expensive transponder usage.
This paper proposed the method of objectively evaluating arousal level through facial skin temperature. There are two types in arousal level, one is based on circadian rhythm, and the other is temporary depending on quantities of sensations. We tried to evaluate the latter. We applied local fractal dimension analysis to the thermogram. We examined the relationship between the thermogram and a rhythm amplitude of EEG, reflects arousal level. As a result, the local fractal dimension analysis of distribution of nasal temperature was valid for the system evaluating arousal level.
A fault diagnosis scheme for nonlinear time series is employed for parts and tool breakage diagnosis problems. The fault is first detected from regression lines plotted for the raw time series. Finite Impulse Response (FIR) network is then used to estimate the unknown system for the normal condition data and to filter the abnormal condition data. To confirm and analyze the fault the regression lines are again plotted for the predicted normal and filtered abnormal conditions' data. The results obtained from the above application of fault diagnosis scheme are discussed.
So far, neurons used in Chaos Neural Network (CNN) have only sigmoid function as an input output function of it. This is the reason that the neuron model should be similar to a real biological neuron. However, in case that we make a neuron model as a model of a group of the real biological neurons, the result neuron model has a non-monotonous function in general. In this paper we construct a CNN with the neuron which has the Radial Basis Function as the non-monotonous function. We call this network the RBF model of CNN. In order to evaluate the RBF model of CNN, we applied this network to memory search problem. As a_??_result, it is clarified that in case that the stored patterns have weak correlation each other, the Sigmoid model of CNN is superior to the RBF model of CNN as to memory search speed, but in case that the stored patterns have strong correlation each other, the RBF model of CNN is superior to the Sigmoid model of CNN.
A new structure of touch-type corrosion sensor for extraction of rust pattern is presented. This sensor consists of three electrodes and three coils for recognition of touch pattern with rust part. Three impedance rates between the terminals are obtained, and the unbalance of these rates reveals when the electrodes touch rust part. Therefore the existence of iron rust in the place where the sensor touched is estimated. As the result, this sensor is found to be useful as a detecting device for recognition of initial rust state.
In a network for an information and control systems, real time communication of large volume data and transmitting multimedia information such as video and voice has been required. In response to realize these requirements, introduction of an ATM (Asynchronous Transfer Mode)-LAN have been investigated. This paper describes an experimental loop-type ATM-LAN for control use utilizing an onboard distributed architecture. Realization method of a loop control mechanism (abortion of loop round-trip cells and loop-back control) using ATM switch function to miniaturizing a hardware and transmission delay of cyclic data communication caused cascade connection of an ATM switch is obtained.
The purpose of this paper is to theoretically identify that there exist stationary distributions in the system with the limited transaction lost. In order to examine this type of system, we formulate and analyze the input/output process as the counting process based on point process. Accordingly, we propose the average transit time theoretically in stationary. We presents the simulation and calculation results in order to evaluate the average processing time for this system.
This paper proposes a new self-organizing system for feature extraction. Human-beings can easily and unconsciously extract the features to distinguish objects in ambiguous and noisy environment. The extracted features are self-organized through learning, and the neurons which represent similar features are said to be assembled in a columnar network in a human brain. To realize these characteristics, the proposed system consists of three hierarchical neural networks. They are an associative memory layer, a middle layer and a symbol layer. The units in each layer connect to the units in other layers. The associative memory finds out the features of the input pattern. In this paper, the feature is defined as a combination of parts from which the system recalls the memorized patterns. The units in the middle layer represent the roughly extracted features. In the middle layer, a new method of weight duplication is proposed when a new unit is required. The duplication is to copy all the connections of the most frequently activated unit to those of the new unit. The weight duplication is effective to avoid the dead unit problem and to extract variable features of the input patterns without degradation of the system performance. The weights between the symbol layer and the associative memories express the detailed features, which are useful to distinguish the similar input patterns. The rough features and complementalry detailed features are extracted simultaneously to distinguish the input patterns without designers' prior knowledge. Simulations are carried out to show the feasibility of the proposed system.
Recognition of a working environment is critical for a mobile robot to confirm its possible intelligence. Therefore it is necessary to equip a recognition system with some sensor which can get environmental information. As an effective sensor, a CCD camera is generally thought to be useful for all kinds of mobile robots. However, it is thought to be hard to use the CCD camera for visual feedback which requires the system to be robust for lighting condition varieties and noise objects. This research presents a corridor recognition method using gray-scale image and Genetic Algorithm (GA) without any image information conversion which gives the recognition process to be robust for lighting condition and noises. The performance is evaluated using real corridor images.
Traditionally, tasks for a robotic system are specified by a desired timed trajectory in the manipulator's workspace, which the manipulator is required to track at every instant of time. However, there are many tasks in which the desired motions are specified by the state of system rather than time, such as contour following tasks. For such tasks, Passive Velocity Field Control (PVFC) has been proposed and the geometry of the controlled systems was analyzed. In this paper, a method to apply a decentralized PVFC to a cooperative multiple mobile robotic system whose sub-system is under non-holonomic constraints and which conveys a common rigid object in a horizontal plain is proposed, and the effectiveness of the control method is demonstrated by experiments.
Noninvasive pulse wave analysis is useful for evaluation of vascular load and vascular aging. The radial pulse waveforms provide more precise information concerning blood pressure changes than systolic and diastolic pressures only. For this purpose, we performed drug administration studies (study 1) and epidemiological studies (study 2) to evaluate the possibility of clinical application of the ΔBPp of the radial pulse waveform. We call this is “Locality of the waveform”. In the study 1, radial pressure pulse was measured with the noninvasive stroke volume for 6 adults with normal physical findings. The ΔBPp was defined as the amplitude from the systolic peak to that of the dicrotic notch in the pulse waveform. Although radial pressure decrease is only from 118/69 to 117/60 after 20mg sublingual nifedipine, the ΔBPp increase from 31 mmHg to 43 mmHg. From this fact, the ΔBPp may be useful for noninvasive evaluation of the effects of vasoactive agents. In the study 2, The ΔBPp was measured in 80 adults with normal physical findings. The ΔBPp showed the tendency of decreasing with age. The ΔBPp (y) decreased with age (x) (r=-0.41, P<0.001, y=-0.62 x+48.8) as aging index.
Recently it is desired to design integrated active filters with low supply-voltages. This letter proposes a design of bipolar transconductance circuits with 3 V supply-voltage. This letter also proposes a synthesis of gm-C biquad circuits using these transconductance circuits. The proposed gm-C biquad circuits are suitable for video-band active filters. Two design examples by the leapfrog and the node-voltage simulation method are given. Spice simulations of the designed filters show the effectiveness of the proposed design.
To compress moving pictures, we must consider the motion compensation and the shape adaptive coding. In this paper, we propose a novel method to compress moving picture of lip by Cubic Curve Model. It is not necessary of the information to describe the content shape by this method. Experimental results show the high compression ratio and the good quality of reconstructed images.
This paper presents a design of a neuron-MOSFET current mirror circuit. This current mirror is built with neuron-MOSFETs and a differential amplifier. The influence of the channel-length-modulation and initial charge of floating gate are minimized by adopting feedback of differential signal. The performance of the proposed circuit is predicted by theoretical analysis and PSPICE simulation.
We propose an efficient construction algorithm of the wavelet network which is proposed by one of the authors in 1994(1). The proposed algorithm is based on an adoption of only significant wavelet bases, and then the proposed algorithm can make the network size very small and the learning speed very fast comparing with the conventional one. The validity and the effectiveness of the proposed algorithm are verified by applying it to approximation of one-dimensional functions and classification of a two-dimensional spiral data.
Recently, many researches on two-degree-of-freedom optimal type-1 servo system have been done. However, in case of the digital control, though either of two integrators, without the delay or with the delay in the input-output relation, are used, there are no papers about the different of feedback property for these integrators. In this paper, we discuss on the feedback property for difference of integrators.
The influence of process parameter such as kinetic energy of the ion and sample temperatures was investigated in detail in order to discuss the dry etching mechanism of polymer. Extremely smooth polymeric optical waveguide could be fabricated by chemically assisted Nitrogen ion beam etching. From these experimental results, it was found that the heating temperature was the critical parameter for polymer dry etching.
Several sufficient conditions are derived for quadratic matrix polynomial systems to be overdamped based on an established characterization of the property. One of the conditions is shown to be less conservative than a previous sufficiency criterion, whereas another one to complement it. It is also pointed out that a pair of the coefficient matrices of the systems is interchangeable so long as the overdamping property is concerned. This fact may possibly double the line-up of sufficiency criteria for overdampedness, thus enriching alternatives to such testable conditions.
We theoretically analyzed the opotical force acting on an elliptically shaped object in order to numerically obtain the magnitude of the applied torque an ddirection of rotation. An elliptically shaped object could be rotated and trapped on condition that the major axis of the object and the propagating direction of the laser beam were coincident with each other.
Speed-up techniques for backpropagation (BP) by removing singular points are simple and powerful, and therefore they are practical. We studied the techniques to secure generalization ability by computational experiments. We introduced “useful region” for gains of the speed-up techniques, where generalization ability is equivalent to that of the standard BP. Two kinds of experiments were conducted to test the speed-up techniques. One experiment applied the speed-up techniques to modify weights between hidden and output nodes (Case 1). The other applied the techniques to all weights of neural networks (Case 2). The results of the experiments found no useful region in Case 2. In Case 1, however, experiments found useful region when gains of the speed-up techniques were small.