Parallel robots with closed-loop kinematic chains have been used in industrial applications. They exhibit superior precision, stiffness, and operating speed than serial robots owing to their mechanical architecture. However, parallel robots demonstrate static singularity unique to parallel mechanisms, which further impede their application. Static singularity is caused by passive joints included in the mechanism. In this study, a 3-leg 6-DOF spatial parallel robot with actuation redundancy and its statics are analyzed. The analysis indicates that static singularity does not occur if actuation redundancy is properly introduced. This analysis is based on the decomposition of force transmission in the robot. Experimental results show that precise trajectory tracking and stable contact are achieved by virtue of the actuation redundancy. This study opens new avenues for the application of parallel robots.
Unmanned aerial vehicles (UAVs) are widely used in many fields, including agriculture and industry. Touchdown of a UAV without tipping over is a crucial but challenging issue owing to disturbances and uncertainties in the landing phase. In particular, when a breakdown occurs in a UAV system and the UAV free falls, sensors can be destroyed or the integrity of the UAV can be compromised. Therefore, developing an emergency landing system that can suppress rebound after free falling and preserve the integrity of UAVs is necessary. This paper proposes an adaptive shock response mechanism as a safe and robust emergency landing system for UAVs. A spring-damper system combined with a plastic deformation part serves as this emergency landing system to absorb and mitigate the impact during the landing phase to avoid tipping over of a UAV by reducing the rebound height. A release system that unlocks the plastic deformation part when the landing height is sufficiently high is proposed. Numerical simulations are conducted to evaluate the performance of the proposed emergency landing system, which is compared with those of two other mechanisms. The results reveal that the proposed method can deliver satisfactory rebound-reducing performance and high robustness against variations in the UAV weight and falling height. Additionally, the effectiveness of the proposed mechanism is experimentally validated using an equivalent model.
This study examines the training policies and environmental robustness of a neural network used in velocity estimation for a tracked vehicle with slippage. In the proposed method, the velocity is estimated by a neural network whose input is an estimated disturbance to the driving axle that includes slippage information. First, we experimentally clarify the proposed method's scope of applicability and effectiveness under different environmental conditions in training and estimation. Subsequently, we experimentally confirm that the estimated disturbance is robust to environmental changes and complementary to environmental information. Finally, the neural network trained on a flat surface is validated in combination with gravity compensation for acceleration to apply it to driving on a slope.
This paper proposes a hybrid angle/force control method for two-degree-of-freedom (2-DOF) magnetic screw motor. This 2-DOF motor realizes rotational and linear motions by itself. Although the considered plant model of the 2-DOF motor is complex, it can be decomposed into rotational and linear motions based on modal information. The proposed method independently controls the rotational motion angle and linear motion force based on the modal information. Simulations and experiments were conducted in this study, and the results showed the effectiveness of the proposed method.
This paper proposes a machine learning-based method to identify human hand motion using a 3D capacitive proximity sensor based on multiple sensing electrodes, which was developed in our previous studies. Although the sensor can detect nearby objects, determining their position and motion directly from the nonlinear outputs of the sensor is difficult. This study proposes a random forest method to identify the direction of movement of a human hand passing above the 3D proximity sensor unit. The time-series data obtained by combining the outputs of three channels are classified into four directions: upward, downward, rightward, and leftward. Experimental evaluation reveals that the proposed method achieves over 95% classification accuracy in all four directions.
With a rapidly aging population, declining birthrate, and decreasing number of skilled workers, automation of construction machinery is expected. On construction sites, automated earth moving work by a bulldozer requires the measurement of the soil pile. However, measuring entire pile data using sensors mounted on the bulldozer is difficult, since the back side of the soil pile becomes blind spot. To solve this problem, it is required to generate the path of the pile spread only using the image of the part visible from the bulldozer. This paper proposes a method to generate digital evaluation model (DEM) of the soil pile from occluded measurement pile data using a convolutional neural network. Since deep learning-based methods require a large amount of training data, we generated pile data and captured images via simulations. Considering the sensing device, three image patterns and their estimation accuracy were evaluated. By using the trained network model, construction path optimization for earth moving tasks was performed using the estimated DEM data. The results of DEM estimation and filling performance of soil moving tasks are also shown.
One of the challenges facing modern large-scale wind turbines stems from the shaft which has a finite stiffness, leading to shaft torsional oscillation which creates fatigue fractures and overload in the shaft system causing unexpected downtime of the wind turbines (WTs). In this study, the shaft torsional oscillation is investigated using a two-mass model of the shaft system. The permanent magnet synchronous generator is operated as a suppressor to mitigate the shaft torsional oscillation by implementing an H∞ controller in the speed control loop, thereby enabling the closed-loop control system to reduce the effect of shaft torsional oscillation. Another challenge lies in inability to directly measure the shaft torsional torque. Thus, the shaft torsional torque estimate is obtained by the H∞ observer. Both observer and controller are designed using the linear matrix inequality method. The simulation results were obtained using Matlab/Simulink@ to analyze the effectiveness of the proposed control under highly oscillated wind. Based on the results, the shaft torsional motion was mitigated by the suggested control technique as compared to the standard control method. The estimate result proved that the H∞ observer was effective under various conditions that affect the performance of the observer.
A wireless power transfer (WPT) system with parity-time symmetry has a critical magnetic coupling coefficient (kmc). A small kmc results in a highly robust, long-distance WPT. We found that the series-series/parallel topology is a good choice to obtain a small kmc. Additionally, a kmc value of 0.04 was achieved by a combination of a solenoid coil of 46µH, low frequency range of 47-56kHz, and buck converter. We investigated the optimal shape of the solenoid coil using an electromagnetic field simulator, and a critical transmission distance (dc) of 77mm was achieved by small transmitter and receiver coils (dimensions: 100mm × 20mm × 10mm). We succeeded in extending dc by 1.8 times over the results from our previous work. dc of 77mm is 2.8 times as large as the cube root of the coil volume.
This paper proposes a voltage modulation method that maintains the voltage-integral error vector within the desired tolerance specification for low-sampling-frequency motor drive systems. A conventional modulator outputs an instantaneous voltage vector per sampling period and requires a high sampling frequency. In contrast, the proposed method analytically solves the voltage vector during the sampling period. Thus, the proposed method is adaptable to the low-sampling-frequency systems while maintaining a high time resolution, and it contributes to increased versatility. The proposed method is tested through experiments on a V/f control system with sampling frequencies up to five times lower than those of the conventional method. This paper confirms that the proposed method maintains the voltage-integral error vector within the tolerance specification, regardless of the sampling frequency. Moreover, by designing an evaluation function to reduce the switching frequency, the proposed method achieves the same performance as the conventional method driven under a high sampling frequency.
In this paper, we investigated the cause of the increased wind noise by the active noise cancellation (ANC) headphones under windy conditions. Initially, we conducted experiments comparing both feedforward and feedback types of ANC under windy conditions. Subsequently, we formulated a hypothesis for the cause of amplified wind noise in ANC headphones as follows: The feedforward type ANC may introduce additional noise into the user's eardrum when its microphone, used for feedforward control, captures low acoustic power sounds. To validate this hypothesis, we conducted experiments using reproduced wind noise sounds played using a speaker (high acoustic power) or an earphone (low acoustic power) with feedforward type ANC headphones. Our findings provide robust confirmation that the experimental results align with and support our proposed hypothesis.
In this letter, we discuss control experiment classes and consider how we realize the same experiment together simultaneously in remote locations that are far apart from one another. This class operation approach may emphasize the importance of COVID-19 precautions or the rationalization of lab classes, considering an aging society with a low birth rate. We propose a system configuration that safely realizes high-education classes using a hierarchical optimal control method. In this letter, we present, for the first time, Internet-based experimental results of the simultaneous stabilization of inverted pendulums, including communication delays.
In this paper, abnormal modes during operation are organized to evaluate the reliability of circuit breakers driven by linear motors. Next, an algorithm that automatically estimates the acceleration time from the load current of the linear motor and detects abnormalities in the circuit breaker from the acceleration time is examined. We report on the results of detecting abnormalities in mechanical components in an operation test of the circuit breaker with the proposed algorithm applied.