Abstract: Disclosed in the present invention is a flexible modeling method for a timing constraint of a register. Simulation ranges of input terminal transition time, clock terminal transition time, and output load capacitance of a register are determined first, simulation is performed under each combination of input terminal transition time, clock terminal transition time, and output load capacitance to obtain a timing constraint range, then setup slack and hold slack are extracted in this constraint range with a particular interval, and then simulation is performed to obtain a clock terminal-to-output terminal delay. Finally, a mutually independent timing model of the register is established by using an artificial neural network, where the clock terminal-to-output terminal delay is modeled as a function of the input terminal transition time, the clock terminal transition time, the output load capacitance, the setup slack, the hold slack, and an output terminal state.

Abstract: The present invention discloses a calculation method of eddy current loss in magnetic materials based on magnetic-inductance. The present invention proposes a vector model of a magnetic circuit, an eddy current reaction is equivalent to a magnetic-inductance component in the magnetic circuit, and the eddy current loss can be fast calculated by the vector model of the magnetic circuit. When the frequency is high, the eddy current loss dominates an iron loss and can be estimated as an entire iron loss. The present invention proposes the vector model of the magnetic circuit based on which the calculation method of eddy current loss in magnetic materials is proposed as well. Through the proposed method the eddy current loss in magnetic materials can be directly calculated by using the magnetic-inductance and the magnetic flux in the magnetic circuit, which can provide guidance for design and performance evaluation of high-frequency electrical equipment from a brand new viewpoint.

Abstract: A device for magnetic field-assisted simulation of zero-microgravity flame synthesis of nanoparticles includes a gradient magnetic field device, a combustor and a product collection device. The gradient magnetic field device is composed of two magnetic field devices arranged face to face. The combustor is located between the two magnetic field devices. The outlet of the combustor is vertically upward. The position is below the magnetic field center of the gradient magnetic field device. The body force acting on the flame and surrounding magnetic species thereof by the gradient magnetic field device counteracts the gravitational buoyancy lift effect, so that flame synthesis is carried out under a simulated zero/microgravity flame to prepare the nanoparticles. The device is able to use a gradient magnetic field to simulate the zero/microgravity flame on the ground to synthesize the nanoparticles under special flame characteristics, with reduced flame disturbance, improved stability, and no overheated region.

Abstract: The present invention discloses a method for segmenting pedestrians in roadside images using a variable-scale multi-feature fusion convolutional network. It addresses the challenge of significant changes in pedestrian scale by using two parallel convolutional neural networks to extract the local and global features at different scales, and then fusing them to obtain a variable-scale multi-feature fusion convolutional neural network, and this network is trained using roadside pedestrian images to realize accurate pedestrian segmentation, avoiding issues with boundary fuzziness and missing segments commonly found in single-network methods.

Abstract: An electrically-driven gyroscope having a housing capable of alternate rotation and a control method thereof, which relate to the field of electrically-driven gyroscopes. The electrically-driven gyroscope includes a rotating housing, an internal rotator, a drive motor composed of an electrical stator part and a rotating part, a drive circuit, and a battery. The internal rotator is mechanically connected to the rotating part of the drive motor, and they can coaxially rotate. The electrical stator part of the motor, the drive circuit, and the battery are mutually electrically connected, and are all connected inside the rotating housing.

Abstract: A robot system for active and passive upper limb rehabilitation training based on a force feedback technology includes a robot body and an active and passive training host computer system. Active and passive rehabilitation training may be performed at degrees of freedom such as adduction/abduction and flexion/extension of left and right shoulder joints, and flexion/extension of left and right elbow joints according to a condition of a patient. In a passive rehabilitation training mode, the robot body drives the upper limb of the patient to move according to a track specified by the host computer, to gradually restore a basic motion function of the upper limb. In an active rehabilitation training mode, the patient holds the tail ends of the robot body with both hands to interact with a rehabilitation training scene, and can feel real and accurate force feedback.

Abstract: A flyback converter and an output voltage acquisition method therefor and apparatus thereof, wherein the output voltage acquisition method comprises the following steps: acquiring the reference output voltage of a flyback converter; sampling the current output voltage of the flyback converter within a reset time of each switching period among M continuous switching periods of the flyback converter, wherein M is a positive integer; and according to the reference output voltage and the current output voltage, sampling a dichotomy to successively approximate the current output voltage until the M switching periods are finished, and acquiring the output voltage of the flyback converter.

Abstract: Disclosed are a system and method for controlling an active clamp flyback (ACF) converter. The system includes: a drive module configured to control turning-on or turning-off of a main switching transistor SL and a clamp switching transistor SH; a main switching transistor voltage sampling circuit configured to sample a voltage drop between an input terminal and an output terminal of the main switching transistor SL; a first comparator connected to the main switching transistor voltage sampling circuit and configured to determine whether a sampled first sampling voltage is a positive voltage or a negative voltage; and a dead time calculation module configured to adjust, according to an output of the first comparator and a main switching transistor control signal DUTYL of a current cycle, a clamp switching transistor control signal DUTYH of next cycle outputted by the drive module.

Abstract: The present disclosure relates to a system and a method for preparing carbon nanofiber and hydrogen through continuous microwave pyrolysis. The system includes four apparatus. The melting and feeding apparatus is to heat and melt feedstocks. The microwave pyrolysis apparatus is for catalytic pyrolysis and includes a feedstock inlet, a gas outlet and a carbon outlet. The gas purification and utilization apparatus is for hydrogen purification and residual gas separation, The power generation apparatus includes a generator and a small internal combustion engine utilizing residual gas as fuel, and the generated smoke is conveyed to the melting and feeding apparatus for feedstocks melting. According to the present disclosure, a poly-generation system for co-producing high-performance carbon materials and hydrogen through plastic wastes with greatly increased energy utilization rate is formed to solve the technical problems of low product yield and high energy consumption in traditional pyrolysis.

Abstract: Disclosed in the present invention is a method for predicting a delay at multiple corners for a digital integrated circuit, which is applicable to the problem of timing signoff at multiple corners. In the aspect of feature engineering, a path delay relationship at adjacent corners is extracted by using a dilated convolutional neural network (Dilated CNN), and learning is performed by using a bi-directional long short-term memory model (Bi-directional Long Short-Term Memory, BLSTM) to obtain topology information of a path. Finally, prediction results of a path delay at a plurality of corners are obtained by using an output of a multi-gate mixture-of-experts network model (Multi-gate Mixture-of-Experts, MMoE).

Abstract: A motion constraint-aided underwater integrated navigation method employing improved Sage-Husa adaptive filtering includes establishing a Doppler log error model; constructing a state equation for an underwater integrated navigation system employing Kalman filtering; according to a relationship between a centripetal acceleration and a forward velocity of an underwater vehicle, establishing a constraint condition, and constructing a complete motion constraint model; establishing two measurement equations; and establishing a filter equation, conducting calculation by using a standard Kalman filtering algorithm when an underwater glider normally runs, and conducting time updating, measurement updating and filtering updating by using an improved Sage-Husa adaptive filtering algorithm when a measurement noise varies.

Abstract: The present invention discloses a distributed active/passive hybrid cable drive system, including a direct current motor, a magnetorheological actuator module, and a drive gear set. The direct current motor provides power for the system, the magnetorheological actuator adjusts output, a cable is driven to move by using a reel fixedly connected to an output shaft, the power is transmitted to a tail end by using the cable along a Bowden cable, and the tail end is connected to a controlled object, to implement control. The distributed active/passive hybrid cable drive system can implement controllable force output with large torque, small inertia, and high bandwidth, and has a small volume, high efficiency, and low costs compared with a pure motor drive system.

Abstract: The present disclosure discloses a distributed secure state reconstruction method based on a double-layer dynamic switching observer.

Abstract: A dual-band shared-aperture antenna array based on a dual-mode parallel waveguide is applicable to the field of wireless communications technologies. The dual-band shared-aperture antenna array includes: a waveguide-substrate integrated waveguide transition structure, a multi-stage cascaded power splitter, a transverse coupling slot, a longitudinal coupling slot, a matching through-hole, a shared-aperture dual-mode parallel waveguide, and a shared-aperture parallel slot array. The entire structure includes two dielectric substrates. A signal is input to the multi-stage cascaded power splitter through the waveguide-substrate integrated waveguide transition structure located on a bottom-layer substrate. The foregoing dual-band shared-aperture antenna array based on a dual-mode parallel waveguide can separately implement a low-band fixed-beam slot array antenna, a high-band multi-beam long slot leaky-wave antenna, or a fixed-beam slot array antenna.

Abstract: A laterally double-diffused metal oxide semiconductor device is provided, including: a drift region (3) having a first conductivity type; a first body region (10) disposed on the drift region (3) and having a second conductivity type, the first conductivity type and the second conductivity type being opposite conductivity types; a first conductivity type region (13) disposed in the first body region (10); a second body region (12) disposed in the first conductivity type region (13) and having the second conductivity type; a source region (11) disposed in the second body region (12) and having the first conductivity type; and a contact region (9) disposed in the first body region (10) and having the second conductivity type.

Abstract: A current fed high-frequency isolated matrix converter and the corresponding modulation and control schemes are provided. The converter includes a current source full-bridge converter, a high-frequency transformer, a matrix converter, and a three-phase filter. An optimized space vector modulation solution is used for controlling the converter, and by comparing magnitudes of three-phase filter capacitor voltages to determine an action sequence of space vectors, switch tubes are turned on at zero voltage. A current source full-bridge circuit adopts a commutation strategy of a secondary clamping, and by calculating a leakage inductive current commutation time, full-bridge switch tubes are turned off at zero current to achieve safe and reliable commutation, and having advantages of a low system loss, a high efficiency, and a high power density.

Abstract: A palm-supported finger rehabilitation training device comprises a mounting base, a finger rehabilitation training mechanism mounted on the mounting base, and a driving mechanism for driving the finger rehabilitation training mechanism; wherein the finger rehabilitation training mechanism comprises four independent and structurally identical combined transmission devices for finger training corresponding to a forefinger, a middle finger, a ring finger and a little finger of a human hand, respectively, and the mounting base is provided with a supporting surface capable of supporting a human palm; wherein each combined transmission device for finger training comprises an MP movable chute, a PIP fingerstall, a DIP fingerstall and a connecting rod transmission mechanism; a force sensor is provided to acquire force feedback information to determine and control force stability, and a space sensor is provided to acquire space angle information to control space positions of fingers in real time.

Abstract: It discloses a serial FFT-based low-power MFCC speech feature extraction circuit, and belongs to the technical field of calculation, reckoning or counting. The circuit is oriented toward the field of intelligence, and is adapted to a hardware circuit design by optimizing an MFCC algorithm, and a serial FFT algorithm and an approximation operation on a multiplication are fully used, thereby greatly reducing a circuit area and power. The entire circuit includes a preprocessing module, a framing and windowing module, an FFT module, a Mel filtration module, and a logarithm and DCT module. The improved FFT algorithm uses a serial pipeline manner to process data, and a time of an audio frame is effectively utilized, thereby reducing a storage area and operation frequency of the circuit under the condition of meeting an output requirement.

Abstract: A loss optimization control method for modular multilevel converters (MMCs) under fault-tolerant control is disclosed. The method includes the following steps: when a fault of a SM in a MMC occurs, bypassing the faulty SM to achieve fault-tolerant control; suppressing the fundamental circulating current using a fundamental circulating current controller; respectively calculating the loss of each SM in faulty arms and healthy arms by using loss expressions of different switching tubes in SMs of the MMC; aiming at the loss imbalance between the arms of the MMC, taking the loss of a healthy SM as the reference, adjusting the period of capacitor voltage sorting control in the faulty SMs, achieving the loss control over the working SMs in the faulty SMs, and finally achieving the loss balance of each SM in the faulty arms and the healthy arms. Compared with the conventional methods, the proposed method is easier to implement and does not increase the construction cost of MMCs.

Abstract: The present invention discloses a high-precision point positioning method and device based on a smartphone. The method of the present invention, which belongs to the technical field of satellite positioning, improves the conventional PPP uncombined positioning model, and only uses original GNSS observation values received by a smartphone to carry out high-precision positioning without GNSS reference stations.