This paper describes the method to quickly estimate the vocal area function by using a digital signal processor (DSP). The method of estimation on the basis of a reflection coefficient derived by Levinson-Durbin algorithm is the best one. However, it is necessary to remove (i) glottis characteristic and (ii) radiation characteristic from the voice wave that is observed by a microphone, to estimate precisely the vocal area function that corresponds to a vocal of human. We paid attention to an adaptive inverse filter in this pre-processing. However, a conventional adaptive inverse filter is not suitable to DSP in terms of process and accuracy. This paper proposes the method that the effect of an adaptive inverse filter is given to autocorrelation directly. The proposed method reduces a calculation volume to 70% in comparison with the conventional method, and increases accuracy by about 5_??_10 times.
A high-speed design pattern generating method and its hardware for pattern inspection of printed circuit board (PCB) photomasks are described. PCB photomask patterns consist of some ten kinds of aperture patterns and straight and arc lines by scanning the aperture. Remarking that these patterns are categorized into aperture stop parts and its transfer parts, we developed a new method which generates aperture stop parts by referring stored pattern elements in memories and the transfer parts by high-speed approximation of straight and arc lines. 64 digital signal processors (DSPs) calculate start and end addresses of each pattern in each scanning line in parallel, but in preceding to each raster scanning of a linear image sensor for photomask pattern detection. Input/output buffers of the DSPs are owned commonly and equal services are realized by patrolling to detect requests from each DSP. Pattern generating speed as fast as 0.1μs/pixel has been achieved. 220 pattern elements can be generated in a scanning line in maximum. Scale of the hardware of this parallel, asynchronous pattern generator is about a half of conventional synchronous ones.
In order to control an electric servo motor precisely, the state variables of the motor should be known exactly. For that purpose, position sensor, velocity sensor and current sensor are mainly used. However sufficient control performance is not always obtained when the fluctuations of the parameters such as torque constant and inertia moment are excited, or when a large step-like load is imposed. In particular, these problems are serious in the robot controller or the numerical controller in the machine tools. To overcome these problems, an advanced control method which also utilizes the information of the acceleration has been proposed recently. Since the practical angular accelerometer is not available, the differentiation of the speed signal is carried out to obtain the acceleration signal. This method does not give the precise acceleration signal to the controller due to the noise problem in the high frequency area. More effective method should be developed. In this paper, a method which estimates an accurate acceleration by using both a position sensor and an AC tachogenerator is proposed. In the proposed method, the higher harmonic component except the fundamental component in the AC tachogenerator outputs is cancelled out by the proposed method based on the Fourier analysis. Because of a considerable amount of the comprehensive calculation, the digital signal processor (DSP) will be applied to the real applications. Several experimental results based on the proposed method with DSP are also shown.
The rotor bars and end-rings of a squirrel cage induction motor are forced to break when it is driven under the overload duty cycle as frequently accelerating and braking operations, although it has essentially mechanical ruggedness. In this paper we present a model of squirrel cage induction motor with the above rotor faults, show a method of analysis, and discuss the relations between the various faults conditions and the amplitude of abnormal frequency components of the currents and torque. The results are summarized as follows. (1) The amplitude of abnormal frequency component increases with the number of broken bars, while it decreases with increasing the number of broken end-rings, in the case of less than five consecutive fault positions. (2) The amplitude of abnormal frequency component becomes maximum when the distance between two broken bars is 180°k(k=l, 2, …, 2P-1: P; No. of pole pairs) in electrical angle, whereas it becomes minimum at 90°×(2k-1) (k=1, 2, …, 2P) in angle. (3) The influence of end-ring faults is maximum at 180°×(2k-1) (k=1, 2, …, 2P) in electrical angle. It is assured that the measured results obtained by the artificially faulted test machine have a good agreement with the analytical results.
The linear induction motor (LIM) is structurally divided into the flat and the tubular-type. In this paper, we study one kind of tubular linear induction motor which has the primary configuration composed of four pieces of flat-type core. We call this type of motor “the square tubular LIM”. The square tubular LIM has various merits of the cylindrical tubular LIM, in addition to those of the usual flat-type LIM. There are two remarkable features of the square tubular LIM that the flat-type LIM doesn't have. The first feature is that it has no edge effect, the second is the normal force in secondary conductor is cancelled by the that of the opposite side of the secondary conductor. Comparing with the cylindrical tubular LIM, moreover, the square tubular LIM is easy to manufacture and maintain it. The tubular LIM also has such a feature that its primary winding is a kind of ring winding, so we can independently supply the primary currents in each slot and hence easily obtain the various winding schemes. In this paper, we study the features of the square tubular LIM. After analyzing our experimental model numerically, the motor is driven by the independently controlled slot currents, resulting expected performance.
The authors developed a compact system for finding intruders, by using a vision sensor with a fiber grating optical device (FG vision sensor). When this FG vision sensor was installed at the ceiling and a laser beam was projected onto the FG, a bright spots pattern was projected on the floor. The authors succeeded to realize a pilot system of good performance for detecting the person's volume, position and direction in a short time, by calculating the three-dimensional position of the translated spots. This paper describes the method of finding intruders using the FG and the configuration of the developed system. And we discuss the results of the experiment which we performed to verify the effectiveness of the proposed method.
Recently, the voltage fed type PWM converters have been investigated from ideal and inherent properties as DC power source. But, the current control system generally used may be somewhat complicate due to the DC voltage detection. In this paper, we propose the simple control system of the PWM converter by means of power factor control in which the DC voltage detection is not always needed. In this system, the DC voltage is given by the line voltage and the modulating index of the PWM converter. The principle of the PWM control system is shown at first. From the simplified circuit model and its phasor diagrams, some basic relations of the PWM converter system are obtained and some fundamental characteristics are shown. From these investigations, it is seen that the phase control region is restricted to a certain angle, which depends on the setting power factor, the modulating index of the PWM and the load. Under the operating condition of unity power factor, it is given by ±π/4 for both the converting and inverting operations. The system can be operated over the all control range of the modulating index by restricting the angle to this critical value for the angle over this control range. The ideal operating characteristics as DC power source can be verified by some experimental results such as the operating waveforms of line current, the steady state characteristics and the transient responses between both operations of the converter and inverter. The features of the system are in the simple construction of the control circuit and some inherent and excellent characteristics as a voltage fed type DC power source. This control system may be applicable to any types of PWM converter used now.
This paper presents steady-state performance evaluations of PWM inverter for 3φ induction motor using generalized and optimum trapezoidal-wave signal and wide-band variable carrier frequency, as compared with PWM inverter characteristics of introducing sinewave signal and wide-band carrier frequency. PWM output voltage with lower harmonic contents and larger effective AC value is obtainable under the optimum trapezoidal-wave signal PWM scheme. The optimun trapezoidal signal pattern is characterized through introducing slope factor L of the generalized trapezoidal-wave signal and modulation factor M. This is theoretically determined as L=0.2π from the conditions of the realization of minimized lower harmonic components and maximized utilization factor of DC voltage to PWM voltage. In particular, the induction motor variable-speed drives using the optimum trapezoidal-signalbased PWM inverter are more effective because the stator current distortion and torque pulsation are considerably reduced in the range of lower carrier frequency in the comparison with those in the case of using sinewave signal. The experimental results are illustrated in breadboard.
A number of configurations for interior permanent-magnet type synchronous motors have been developed in recent years. This paper describes a form of motor having large output and high efficiency. Use can be made of samarium cobalt magnets, and the volume of the magnets used in the rotor is small compared with that of the conventional permanent-magnet synchronous motors. The characteristics of the motor are predicted by using the finite-element method. The agreement between predicted and measured results is good. Lastly, the characteristics of the motor using Neodymium-iron-boron magnets instead of samarium cobalt magnets are predicted.