Micro-stepping motion of ultrasonic motors satisfies biomedical applications, such as cell operation and nuclear magnetic resonance, which require a precise compact-structure non-magnetization positioning device. When the pulse number is relatively small, the stopping characteristics have a non-negligible effect on the entire stepwise process. However, few studies have been conducted to show the rule of the open-loop stepwise motion, especially the shutdown stage. In this study, the modal differences of the shutdown stage are found connected with amplitude and velocity at the turn-off instant. Changes of the length in the contact area and driving zone as well as the input currents, vibration states, output torque, and axial pressure are derived by a simulation model to further explore the rules. The speed curves and vibration results in functions of different pulse numbers are compared, and the stepwise motion can be described by a two-stage two-order transfer function. A test workbench based on the Field Programmable Gate Array is built for acquiring the speed, currents, and feedback voltages of the startup-shutdown stage accurately with the help of its excellent synchronization performances. Therefore, stator vibration, rotor velocity, and terminal displacements under different pulse numbers can be compared. Moreover, the two-stage two-order model is identified on the stepwise speed curves, and the fitness over 85% between the simulation and test verifies the model availability. Finally, with the optimization of the pulse number, the motor achieves 3.3 µrad in clockwise and counterclockwise direction.
In the UAV electro-optical pod of the two-axis four-gimbal, the characteristics of a coarse–fine composite structure and the complexity of dynamics modeling affect the entire system’s high precision control performance. The core goal of this paper is to solve the high precision control of a two-axis four-gimbal electro-optical pod through dynamic modeling and theoretical study. In response to this problem, we used finite element analysis (FEA) and stress study of the key component to design the structure. The gimbals adopt the aerospace material 7075-t3510 aluminum alloy in order to meet the requirements of an ultralight weight of less than 1 kg. According to the Euler rigid body dynamics model, the transmission path and kinematics coupling compensation matrix between the two-axis four-gimbal structures are obtained. The coarse–fine composite self-correction drive equation in the Cartesian system is derived to solve the pre-selection and check problem of the mechatronic under high-precision control. Finally, the modeling method is substituted into the disturbance observer (DOB) disturbance suppression experiment, which can monitor and compensate for the motion coupling between gimbal structures in real time. Results show that the disturbance suppression impact of the DOB method with dynamics model is increased by up to 90% compared to PID (Proportion Integration Differentiation method) and is 25% better than the traditional DOB method.
With the impact of COVID-19 on education, online education is booming, enabling learners to access various courses. However, due to the overload of courses and redundant information, it is challenging for users to quickly locate courses they are interested in when faced with a massive number of courses. To solve this problem, we propose a deep course recommendation model with multimodal feature extraction based on the Long- and Short-Term Memory network (LSTM) and Attention mechanism. The model uses course video, audio, and title and introduction for multimodal fusion. To build a complete learner portrait, user demographic information, explicit and implicit feedback data were added. We conducted extensive and exhaustive experiments based on real datasets, and the results show that the AUC obtained a score of 79.89%, which is significantly higher than similar algorithms and can provide users with more accurate recommendation results in course recommendation scenarios.
This project was supported by the Ministry of Science and Technology of China and the project number is 2015CB057503. In this project, 500 000 RMB could be used for this research and the publication of related papers.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.