Field oriented control of an induction motor fed by a VSI is achieved by means of a state feedback static controller designed according the H_∞ control theory. Designing procedure is described in the paper. The performances of the control system are numerically analyzed and experimentally verified, also in order to prove the validity of the design procedure. The results confirm that good dynamic performances and high robustness to variations of system's parameter can be achieved by means of the proposed controller.

View full abstract

While automation and labor-saving functions have been incorporated into production systems in the past, human-oriented production systems that maximize human abilities, such as flexible decision-making and action-taking capabilities, are being reassessed for their effectiveness for responding to today's diversified market needs and rapidly changing product trends. Simulation is one method to evaluate production systems, but in the past it was utilized mainly for the purpose of improving productivity Generally, conventional simulation systems are not designed to take into consideration human factors, which are vital elements in human-oriented production systems.
In this paper, we propose and evaluate a simulation model that incorporates human factors such as workers' cooperation logic and workers' human error in a human-oriented production line, and make fundamental simulation experiments considering human factors in the model. It is aimed at not only improving productivity but also providing higher worker satisfaction. From the experiments, we confirmed possibility that our model can lead to the improvement of both productivity and worker satisfaction.

View full abstract

The regenerative braking method has been employed in DC-fed electric railway vehicle traction system in order to reduce energy consumption. The load's power consuming is sometimes insufficient for the regenerated power, and it causes the input voltage of traction inverter rises. In such a case, the regenerative current is regulated to prevent the over voltage. However, in applying this regulation, especially to the permanent magnet synchronous motors (PMSM) traction system, a continuous oscillation of regenerative current is sometimes observed. Unfortunately, It deteriorates passengers' riding comfortability due to torque fluctuation.
On this paper, we analyze the phenomena of the oscillation of regenerative currents and suggest an anti-oscillation regenerative current regulation method based on the analyzing results. The detail is as follows.
(1) We construct a nonlinear time-varying model, which consists of the substation, power feeding system and the main traction circuit onboard and divide the nonlinear time-varying model into four linear time-invariant systems.
(2) Through the analysis with the models, we make it clear that the continuous oscillation is caused by the traditional regenerative braking control method
(3) Based on the analysis we suggest a novel anti-oscillation regenerative braking control method, whose effectiveness is proved through experiments.

View full abstract

Surges are undesirable with respect to both the reliability and the efficiency of power devices. This study presents the mechanism of surge occurrence in an induction heating power supply. A 2kW full bridge series resonant inverter based on zero-voltage-switching (ZVS) and phase-locked loop (PLL) control is used as a power supply. In this inverter, the modified PLL does not only keep the output current and the voltage in phase but also makes the system operate in ZVS mode. When the PLL mismatches the resonant frequency in a manner of current phase lead, the inverter operates at the non-ZVS mode and the surge voltage heavily affects the switching elements to an extent mainly decided by the recovery energy of the drain-source diode in the power devices. In order to solve this problem, this paper proposes a modified PLL to be the phase error zero.

View full abstract

In recent years autonomous distributed scheduling architecture with flexibility and robustness is required in manufacturing systems to facilitate effective manufacturing operation. We proposed a self-organised scheduling methodology as a new scheduling concept based on multi-agent paradigm. Each work is defined as agent in the methodology and cooperates so as to formulate self-organised workflows. This paper studies the abilities of reinforcement learning algorithm with multi-agent paradigm to move appropriately in relation to several process machines. We analyse the effectiveness of a major reinforcement learning algorithm, profit sharing method, in terms of scheduling robustness, and show the rationality of the algorithm in the self-organised scheduling environment. Profit sharing method is proved to enhance the robustness and reliability of work agents coping with the several scheduling demands.

View full abstract

A novel compact slip-power recovery system having sinusoidal rotor currents is proposed. In the proposed system, a PWM boost rectifier is used as a substitute for a diode rectifier and a boost chopper in a conventional compact slip-power recovery system. The conventional compact system has the disadvantage that it has a rectangular rotor current, and the motor torque with large ripple, because a diode rectifier remains in the system. Also, the rotor current can not reach the current reference value near the synchronous speed, because the voltage drop by the resistance of the semiconductor devices and so on can not be neglected when the rotor voltage becomes smaller near to the synchronous speed. The use of the system proposed in this paper has solved these problems. The effectiveness of the proposed system was verified through computer simulations and experiments. As the results, the proposed system brings the sinusoidal rotor current, the small torque ripple, and wide controllable range near to synchronous speed.

View full abstract

In this paper, the design strategy of anti-windup controller (AWC) for feedback control system is proposed. The proposed method is based on Youla parameterization of linear controllers in the framework of left coprime factorization. Conventional anti-windup controllers have difficulty in optimization of parameters. The suppression effect of windup phenomena can be adjusted easily with the proposed AWC. The effectiveness of proposed method is shown with some numerical and experimental results.

View full abstract

A modified hill-climbing method for a photovoltaic generation system (PV system), using an adaptive algorithm, is proposed for the maintenance of its output at maximum power levels. The switch duty ratio of the DC-DC converter of the PV system and its output power are selected as the input and output value, respectively, and its characteristics are described by the characteristic curves (the switch duty ratio vs. output power). The characteristics of the PV system are changed due to the changes of external environmental conditions such as atmosphere, loads and so on. The changes of its characteristic curves are detected as the changes of its gradients, which are estimated as parameters of linear functions by adaptive theory. The change of the switch duty ratio is determined based on the detected characteristic changes. That is, the non-linearity of the PV system and its changing characteristics due to the changes of its external environmental conditions are regarded as deviations of the parameters of the linear functions. It is applied in some experiments under various external environmental conditions and its effectiveness confirmed. Therefore, the proposed method is considered to be generally applicable method as a maximum power controller, for which detailed information of the characteristics of the PV system, including its external environmental conditions, is not required.

View full abstract

Several orders are processed simultaneously while sharing limited resources in the job shop production floor. If an order is viewed as a project, each order can be regarded as the scheduling of a concurrent project. When many orders are handled, several specific resources tend to receive concentrated load; thus, conflicts are always generated. For this reason, scheduling optimization becomes an extremely difficult task. Furthermore, even if a schedule has been obtained, load balancing actions may not be found when a delay in the due date is generated. We solved this problem by studying critical paths contained in schedules of concurrent projects. A novel approach is proposed in this paper for planning load balancing actions by detecting and analyzing critical paths based on Linear Programming. By this approach, a scheduling manager can determine efficient load balancing actions with computer support.

View full abstract

The train service in Japan looses rapidity, comfort and robustness against delays in morning peak hours especially, and it has been a serious social problem. On the other hand, the cost to purchase land for a new track is very high. An efficient train operation is proposed in this paper. The proposed train operation has substantially high rapidity and train density identical to a quadruple-track line, with a novel train scheduling based on full usage of a triple-track line. The significance of the proposed train operation on a triple-track line has been verified based on the calculations of waiting time and boarding time of passengers. Consequently train operators can keep the rapidity, moreover, raise track capacity by operating express and local trains like knitting on triple-track lines.

View full abstract

This paper proposes a new controller design method for the multirate sampling case such that the sampling period of the feed-forward control is greater than that of the feedback control. Also this paper proposes a new plant model which includes the same feedback controller. Effectiveness of the proposed method is confirmed by the simulation.

View full abstract