This paper proposes a passive replication mechanism which establishes complete location and replication independence of program modules. The proposed method is based on an object model and a reliable broadcast mechanism, where each object has an automatic checkpointing mechanism within itself. The automatic checkpointing mechanism frees users from the need to designate checkpoints in application objects.
By the proposed mechanism, replication degree and execution mode of each object can be changed dynamically in accordance with the required reliability and responsibility, and also passive replication and active replication methods can be mixed. In addition, because checkpointing and failure detection are carried out by passive modules, extra operations for maintaining reliability can be removed from active modules.
The proposed method is applied to a train traffic control system; its effectiveness and performance characteristics are also described.

抄録全体を表示

Operation control based on the idea of antirepeat used in a system where the human operator and machine work in the same space is defined as antirepeat control. This study uses multiple-valued logic to evaluate the idea of antirepeat and clarifies the logic structure of antirepeat control. Antirepeat control is expressed by such an interlock that safety resulting from the action of the operator is taken as a signal to permit the movable part of the machine to operate. Two methods are available to assure safety in antirepeat control. One method allows the movable part of the machine to operate based on the proof that the operator has returned from the hazardous space to a safety place. The other allows the movable part of the machine to operate based on the proof that the operator is not present in the hazardous space. These two methods constitute different systems from a safety control point of view and are represented here by concrete operation control circuits. The control circuits proposed here have fail-safe output property.

抄録全体を表示

A fail-safe logic circuit must switch the output signal to the logical value “0” when the operation circuit fails. Transformer has such favorable characteristics for fail-safe logic circuits that only the magnetic flux change is transmitted and that a superposition of magnetic flux is realized. So we applied those characteristics to a fail-safe logic operator. A new fail-safe logic operator using an insulated planar transformer was developed for a fail-safe majority operation and a fail-safe interlocking. In this paper, the principle, structure, fabrication and fundamental characteristics of the new fail-safe logic operator are described.

抄録全体を表示

Our daily life is supported by powerful information networks which are composed of computers and communication equipments. If they do not work correctly or completely, our daily life would be surely confused. Under some circumstances safety could not be kept. Uninterruptible Power Supply (UPS) systems are commonly equipped to supply stable power to the critical loads, for example, large computer center telephone and telecomnication center, aviation control system and so on. Several kinds of UPS systems are now used in many important fields as the above mentioned. But they almost do not have been evaluated numerically based on system reliability.
This paper evaluates MTBF of parallel redundant UPS systems, and reports MTBF of 2 sets of parallel redundant UPS system is lower than that of 3 sets of parallel redundant UPS system under some circumstances. Continuous power supply reliability of the UPS system with bypass and uninterruptible transfer switch is much higher than any other UPS system without bypass circuit and uninterruptible transfer switch. And then the last some highly reliable parallel operation technology are shown.

抄録全体を表示

This paper deals with an estimation method of parameter variations of dynamic systems based on the disturbance observer. In this method, the disturbance observer is designed to estimate the parameter variations as disturbances. A necessary and sufficient design condition for the existence of disturbance observers is shown from the view point of the observability of extended systems. This condition is used to determine the states to be measured by sensors for corresponding parameter variations. The simple correlation method extracts the parameter variations from the estimated disturbances which include external disturbances and noise.
The proposed method is applied to the estimation of the inner pressure, i. e. the stiffness, variations of tires in the active suspension system. Several numerical and experimental results are shown to confirm the effectiveness of the method.

抄録全体を表示

This paper describes a construction method of the system that detects failure of lamps and indicates those locations in an airfield and reports the result of its field tests. This system uses lighting power lines as a transmission circuits. Because a transmitter is connected serially with each lamp as a failure detector and its power source is supplied by current for burning the lamp, the power souce of the transmitter is shut down when the filament of the lamp is burned out. Normal state of a lamp is reported periodically by signal current transmitted through SCR switching power source and its output line. We show that the methods of this lamp failure detection and signal transmission can realize a highly reliabe system. The signal current level is established lower than harmonics of noise level generated by SCR switching not to give bad influences on radio equipments.

抄録全体を表示

In the vector control system, exact estimation of the rotor flux is very important for accurate control. We have proposed the construction and the design method of the sliding observer for robust flux estimation with IM parameter deviations. In this paper, the design method of the sliding observer based on the proposed one is shown more in detail and the configuration of the vector control system with the observer is shown. To verify feasibility of the design method, the experimental results are reported.

抄録全体を表示

AC current waveform controlled self commutated voltage source power converters have various advantages in order to connect a DC power source to electric power systems, e. g., they do not increase short circuit current of electric power systems, and harmonic current is easily controllable.
In this paper, new design theory of AC current waveform control method for single phase self commutated voltage source power converters termed the current error tracking mode AC current waveform control method is proposed and its performance is theoretically evaluated.
The first attractive feature of this design theory is that it can treat an arbitrary current waveform as a target of the objective function of AC current. The second is that the error between actual AC current value and the objective function value can be controlled to be less than an arbitrary value which is set up beforehand. The third is that all control components get digitized. This design theory can also be applied widely to common single phase self commutated voltage source power converters of AC current waveform control type with capacitors at the DC main circuits instead of a DC power source.

抄録全体を表示

In this paper, we proposed a new method for reducing harmonics involved in output voltages of double connected voltage source inverter. By adding four auxiliary switching elements and an interphase transformer with secondary winding to the usual 12-step inverter, the proposed inverter can decrease the harmonics in output voltages no less than the conventional 36-step inverter. The basic theories for this circuit are discussed by analysing the output voltage waveforms, and the optimum parameters are derived.
In addition, this paper presents a static var compensator (SVC) system which employs the proposed inverter. The distortion factor of reactive currents and the capacities of system components for the proposed SVC are compared with those for conventional 12-step inverter. As a result, a capacity of interphase transformer is only 5% for the rating. Therefore, total capacities of this SVC system can be reduced, including output transformers and AC filters, by about 40% compared with the conventional one.

抄録全体を表示

In the field of robot manipulator control, one of the important issues is robustness of the servo-system. The robustness of a servo-system applied to each joint of a manipulator enables non-interferences among other joints. Several papers have been published concerning this issue^(1)_??_(6).
Particularly, the authors have proposed such a robust-servo-system, based on a combination of disturbance observer and sliding mode control^(1)_??_(3). This control law has an advantage in the robustness with the less chattering even at the relatively low sampling frequency. The introduction of the disturbance observer improves the robustness of conventional sliding mode control.
On the other hand, two-degree-of freedom control is derived from stability and achievable transfer characteristics of a linear system⁽⁷⁾. As a result, it includes equivalently various linear control systems such as a disturbance observer^{(4) (8)}. The disturbance observer is very robust servo-system against a disturbance and a fluctuation of plant parameters. And the effectiveness has confirmed in some applications.
In this paper, from the view point of disturbance reduction, a sliding mode control with disturbance observer is compared with two-degree-of-freedom control which includes an equivalent system to the disturbance observer. The comparison of these control schemes makes clear that the introduction of non-linear control (sliding mode) improves the disturbance suppression characteristics of a linear system with a disturbance observer or two-degree-of-freedom control system.
In the simulations and experiments, both control methods were applied to DC servo motor positioning control, and it was proved that the proposed non-linear control has an advantage over the two-degree-of-freedom control in the feature of disturbance reduction.

抄録全体を表示

The slip frequency type vector control and the flux feedback type vector control are typical control methods of induction motors. We show that two methods are some except for a frame of reference. The flux feedback type vector control using the rectangular coordinate uses vertical and horizontal components of the rotor flux. The slip frequency type vector control using the polar coordinate users absolute value and angle of the rotor flux. Usually, the flux feedback vector control dose not use the rotor flux model but uses the rotor flux observer for flux estimation. On the other hand, the slip frequency type control uses the rotor flux model. Therefore, we propose the slip frequency type vector control using the rotor flux observer. The proposed method is equivalent to the flux feedback type vector control using the rotor flux observer. Then it is more robust than the conventional slip frequency type vector control.

抄録全体を表示

Linear pulse motor (LPM) can be directly driven by the digital signals. However, a high accuracy positioning system using LPM needs fine position information for feedback signal. There are the optical displacement sensors, the magnetic displacement sensors and so on that detect the position of LPM mover. But no sensors satisfy all demands, size, simplification and severe environment. So we developed the linear sensor using meander coils. A meander coil is studied for miniature size inductor, micro-transformer and so on. Since the linear sensor using meander coils detects the magnetic pole tooth of LPM stator as a scale, the position detecting system will be made of miniature size. Furthermore, the system can be expected the high speed response, because the exciting current frequency of meander coils is megahertz order.
In this paper we describe the following contents.
(1) The principle operation and the construction of developed linear sensor using meander coils.
(2) The output voltage characteristics of the linear sensor using meander coils, and optimum exciting frequency of coils.
(3) Measuring displacement errors of the linear sensor using meander coils, and analyzing the errors.

抄録全体を表示

This paper describes the computational method to obtain load characteristics of permanent magnet synchronous motors including space harmonics. A time-stepping finite element technique including the rotor movement is proposed, where both terminal voltage and internal angle are given as the known values. In simulating the rotor movement, the suitable rotation step to take account of the effects of space harmonics, is determined. An agreement between calculated and measured results of the load characteristics in an experimental motor is good. As a result, it is clear that the computational method proposed is reasonable.

抄録全体を表示