In general, the effects of modeling errors, parameter variations, and external disturbances of the controlled object degrade tracking control performance. To address these problems, we have developed a control method based on deep reinforcement learning. Accordingly, a reference signal self-organizing control system based on a deep deterministic policy gradient (DDPG) is proposed, which is an extension of an existing control system using DDPG. In a previous study, we confirmed the realization of the swing-up and stabilizing motions of an inverted pendulum using the proposed control system⁽⁸⁾. However, the addition of a new ability to the system could not be verified in that study. Thus, in this work, we aim to verify whether a new function can be added to the proposed control system. By performing a control simulation, we verified whether the proposed system can achieve robustness by using the inverted pendulum with an inertia rotor. A control simulation of the system was performed by adding noise into the system, and its control performance was investigated to confirm the robustness of the system. The simulation results indicate that the pendulum could not be inverted using an un-retrained control system. However, it was confirmed to have been inverted and stabilized by the swing-up and stabilizing control using a retrained control system. Moreover, the retrained control system could effectively function under the effect of noise with an accuracy close to that of a noise-free state. Therefore, we confirmed that the addition of robustness can be realized in the proposed control system.

Service robots working in human environments must have the ability to grasp a wide variety of unseen objects with ability to grasp a wide variety of unseen objects with appropriate grip force without knowing their properties and its response to action in a human environment. This research uses a novel Cyber Physical System framework to estimate and apply minimum grasping force for unknown objects under a task motion using only tactile motion data captured from sensor-less acceleration-based controlled common gripper. The goal is to use this novel framework for developing functions for position-controlled robots to perform real-time cognitive grasping and enable quick implementation across multiple robots. This paper uses soft sensing inputs from the earlier research to characterize unknown object properties like stiffness, mass and surface interaction and show how a data-based analytics algorithm learns a wide range of object properties and slip status along with task action dynamics features, to propose the grasping action with minimum applied force. We validated that the minimum grasping force is able to hold the object firmly during any task motion. In order to improve the learning, the paper proposes a part of the framework to use virtual simulation to gather more learning data and show results that compare well between simulated results with experimental data. We use the simulated data to train Reinforcement Learning approach and show that small variation in object width can be learnt to identify optimum gripping position.

A method for three-dimensional (3D) point cloud-based object recognition and a method that uses the recognized objects for six-degree-of-freedom simultaneous localization and mapping (SLAM) with a high accuracy are presented. For object recognition, we use a convolutional neural network to identify the meaning of each point inside an input 3D point cloud. For scan registration, we present a highly accurate hybrid method that combines the iterative closest point with particle swarm optimization (PSO) to match the recognized points to be archived. Using PSO to match the recognized object's points in each neighboring scan can help decrease incorrect correspondences and enhance the robustness of scan matching. Compared to state-of-art methods, the proposed method achieved good performance on the KITTI odometry benchmark and our SLAM experiments.

This paper presents design studies on a rotor for a high-speed homopolar type permanent magnet motor that can be applied to super chargers for automobiles, cordless stick vacuum cleaners and so forth. The stator comprises a core and a permanent magnet. The rotor is composed of only a core, resulting in enabling the removal of a retaining ring on the surface. Firstly, acceleration performance simulation using a prototype motor with a load blade was developed based on the data measured in an air compressor drive. Subsequently, the rotor shape design refinements based on three-dimensional (3D) finite element analysis were conducted to achieve the maximum torque and power required from a target application under a constraint of core stack length. The design refinement results proven by the acceleration performance of the developed simulation reveal that following the refinements to the rotor design, the homopolar type permanent magnet motor meets the requirement of the target application.

This paper describes a bonding technology for suppressing the thermal imbalance of a reverse conducting insulated gate bipolar transistor (RC-IGBT). The thermal imbalance is a difference of the heat distribution between the IGBT and free-wheeling diode (FWD) regions operating of the RC-IGBT. The heat distribution is determined by design of each active area. The different heat distribution unexpectedly increases the maximal thermal resistance and temperature at the bonded area. In this study, the temperature-uniform effects by the bonding are calculated by simulation to reduce the difference of the heat distribution. It was found that covering the FWD region with the bonding effectively decreased the thermal imbalance. The proposed structure improved the difference of the thermal resistance by more than 74% compared to the conventional structure. Additionally, the difference in temperature at the bonded area of the proposed structure was almost zero. Finally, the highest temperature of the bonded area in the proposed structure reduced by more than 13^°C compared to the conventional structure, which is 25% higher power cycle capability.

This paper presents a coupling estimation based decoupled current vector control scheme for current control of a Dual Three Phase (DTP) Surface Permanent Magnet Synchronous Motor (SPMSM) drive with two micro control units (MCUs), to solve its inherent magnetic coupling issue. Aiming for an enhanced fault tolerance, the SPMSM drive utilized is a redundant and modular system composed of two subsystems in terms of three-phase inverters, sensing, and controllers, with respect to each winding set in the DTP SPMSM. The current of each winding set along with the rotor position is separately detected and applied for control by individually assigned MCU. This comprehensive modularity prevents interference between subsystems in local fault conditions, thus improving the fault tolerance capability. However, the magnetic coupling existing between the winding sets would deteriorate the performance and even cause instability in the whole system without proper decoupling control. Previous studies adopted communication between MCUs for real time current sampling sharing demanded by decoupling control. Nevertheless, communication bandwidth becomes a limiting factor of system design. Some general communication approaches would not satisfy the current sharing requirement of decoupling control for implementation. To eliminate this undesirable limit, a novel decoupled current vector control that utilizes a current observer to estimate the relevant magnetic coupling between winding sets is proposed in this paper. As validated from the experiment results, the proposed decoupled control scheme is able to decouple the system for dynamic improvement so that modularity of the drive can be achieved without communication, leading to an eventual improvement in fault tolerance capability.

This paper proposes a power-leveling method in which a lithium-ion battery and a flywheel energy storage are installed in a photovoltaic power generation plant. The proposed method is a division method that focuses on the amplitude component of the compensating power required to level the generated power of photovoltaic power generation. Furthermore, by adding the flywheel state-of-charge automatic adjustment method to the proposed method, the coverage period of the flywheel is improved. The effectiveness of the proposed method is verified on an actual 40kW plant.

Magnetization reversal motors (MRMs) have been proposed as motors with lower power losses. MRMs comprise the same configuration as interior permanent magnet synchronous motors; however, a low coercive force core is applied to the stator. Driving an MRM requires a pulse current for magnetizing the stator core. In this paper, a bipolar pulse current magnetizer is proposed for driving MRMs Moreover, the relationship between passive components is reported in terms of voltage, current, and circuit constant. Finally, this paper presents experimental verification of the proposed magnetizer using an inductive load to simulate a single phase of an MRM.

In electric railways, rail potential rise because of return current flows in the rail. As the rail potential rises, the return current flows in both rails and the earth. Some railway systems install VLD (Voltage Limiting Devices) to prevent rail potential from exceeding the standard voltage. VLDs are designed to short-circuit between the rail and the earth if the rail potential exceeds a threshold voltage. When the VLD is operated, the rail potential decreases to almost 0V. However, it causes the stray current to increase. The stray current causes an electrical corrosion, which corrodes water pipes, gas pipes, steel bars, and other underground metals. Therefore, it is important to reduce the stray current. We devised a novel VLD concept to reduce the stray current while preventing the rail potential from exceeding the standard voltage. Furthermore, we simulated a suburban line to evaluate the reduction effect of stray current using the novel VLD. It was confirmed that the stray current of the case with the novel VLD was reduced by about 90% compared to the case with the conventional VLDs while preventing the rail potential from exceeding the standard value even momentarily.

This paper proposes a method to reduce the torque ripple vibration of an integrated power factor correction (PFC) converter by using the zero-phase inductance of an interior permanent magnet synchronous motor (IPMSM) when the motor is not running. When an integrated PFC converter is applied to electric vehicle (EV) chargers, it is necessary to reduce the torque oscillation of the motor caused by the charging current. In an integrated PFC converter using an IPMSM, the zero-phase current and magnetomotive force harmonics of the permanent magnet cause torque vibration. In the proposed control method, the generated torque is estimated using the back EMF table, and the phase current is controlled to cancel it. The proposed method can be used to suppress torque vibration irrespective of the mechanical angle of the rotor. The effectiveness of the proposed method is experimentally verified. Using the proposed method, the torque vibration can be suppressed by up to 91.3% as compared to that in the case of the conventional control method. Furthermore, a power factor of 99.9% is achieved under a load condition of 1kW.

This study proposes a closed-loop current balancing control scheme for mitigating the current imbalance during switching intervals between IGBTs connected in parallel. The proposed control technique is based on a current balancing control scheme that utilizes a gate driver circuit, in which the output resistance can be varied using a segmented gate driver technique. Moreover, it reduces the current imbalance by minimizing the delay time difference between IGBTs connected in parallel.

Gate driver ICs equipped with the closed-loop current balance controller is also proposed and fabricated using XFAB's 0.35 µm HV CMOS process. To verify the effectiveness of the closed-loop current balancing control, a multi-pulse switching test is conducted with a PCB test board comprising a simple chopper circuit with two parallel connected IGBTs rated at 600V and 90A. The test results reveal that the current balance controller determines and controls the output resistance of the gate driver ICs, thereby significantly reducing the current imbalance between parallel connected IGBTs during the switching transient.

This paper proposes a control method for an electrolytic capacitor-less dual inverter for compressors. Single-phase motor drive systems are required to minimize vibrations, maintain a high power factor, and meet the limit for harmonic current emissions. The proposed control method achieves the harmonic reduction of the dq-axis currents, a high power factor, and a low grid current THD under periodic load torque fluctuations. The effectiveness of the proposed method is demonstrated via experimental results, using an open-end winding IPMSM and an electrolytic capacitor-less dual inverter. The proposed method reduced the 100Hz harmonic components of dq-axis currents up to 98% under periodic load torque fluctuations compared with an electrolytic capacitor-less single inverter. The results of FFT analysis of the input current show that the power factor and input current THD of the proposed method are equivalent to or superior to those of conventional methods. Therefore, the proposed control method can effectively achieve vibration reduction, a high power factor, and a reduction in the input current THD under periodic load torque fluctuations.

Common-mode choke (CMC) plays an important role in suppressing the leakage current produced during the operation of pulse width modulation (PWM) inverters. However, when the common-mode path exhibits a large series stray capacitance, particularly the capacitance produced by long shielded cables, the CMC core reaches magnetic saturation. This paper presents a technique of selecting a suitable carrier frequency for the inverter to suppress the magnetic saturation of the CMC. First, the analytical approach leads to a non-saturation condition, and clarifies the maximum and minimum carrier frequency design conditions. Subsequently, the effect of the design is demonstrated by a non-linear circuit simulation and experimental tests using an actual inverter system demonstrate the effectiveness of the design technique.

This paper proposes a small and high efficiency circuit topology for series connected devices with regenerative snubber circuits. The proposed circuit topology introduces an important break-through in the trade-off between switching losses and dv/dt. Regenerative snubber circuits store most of the switching losses that occur in general circuits and regenerate at high efficiency. A 10kW-200V class prototype of the proposed circuit topology recorded a maximum efficiency of 99.49%, and a rated efficiency of 99.21%. Loss analysis of the prototype demonstrated no reduction in efficiency due to the regenerative operation by regenerative snubber circuits.