49results about How to "Lighten the design burden" patented technology

The invention discloses a weak supervision fine-grained image recognition method based on a visual self-attention mechanism. The method involves a student-model module, a teacher-model module and a classification-model module. The student-model and the teacher-model are combined through a Teacher-Student loop feedback mechanism based on Pairwise Approach sorting learning so as to form a self-attention region recommendation network, so that the relation between discriminative region positioning and fine-grained feature learning is enhanced, the discriminative region in the fine-grained image can still be accurately detected under the condition of lacking a target bounding box and a part marking point, and the recognition accuracy is promoted to be remarkably improved; meanwhile, a convolution layer is shared by the three modules, namely, the student-model, the teacher-model and the class-model, so that the model storage space is effectively compressed, the calculation cost is reduced, the method meets the real-time recognition task requirement, and the method is suitable for a large-scale real scene; and besides, a dynamic weight distribution mechanism is adopted in multi-task jointlearning to reduce the amount of artificially set hyper-parameters and enhance the robustness of the model, finally, the whole model is trained and learned in an end-to-end single-stage mode, and thenetwork optimization difficulty is reduced.

The invention discloses a method for compensating finite interference time of a remote operation system based on a terminal slide model. The method comprises the steps of respectively selecting a master robot and a slave robot so as to form a remote operation system, respectively measuring system parameters of the master robot and the slave robot, measuring power-off position information of the master robot and the slave robot online, obtaining joint speed information of the master robot and the slave robot by virtue of a robust precise differentiator, designing a finite time interference observer based on a terminal slide model, feeding back to the controller design of the master robot and the slave robot, and acquiring a parameter value range of the interference observer by virtue of a Lyapunov equation. By virtue of the method, the problems of low observation speed and limited capacity of an existing observer are solved, the influences caused by parameter indeterminacy and unknown interferences to the remote operation system are overcome, and the anti-jamming performance of the system is improved; and the controller design is simplified, and the application in engineering practice is benefited.

The invention relates to a geostationary orbit spacecraft electrical propulsion transfer track control method. The method comprises the following steps: determining initial conditions, a carrier rocket transmitting a geostationary orbit spacecraft to a first initial orbit, and determining a Kepler radical of the first initial orbit; establishing a first initial orbit kinetic equation by means of the Kepler radical, reducing the orbit inclination angle of the first initial orbit through a thrust azimuth alpha and performing rounding to obtain a second initial orbit, and obtaining transfer time tf1 and a propellant consumption amount mfuel1; fixing a thrust accelerated speed in a second preset plane, transferring the second initial orbit through a thrust azimuth beta to a geostationary orbit, and obtaining transfer time tf2 and a propellant consumption amount mfuel2; calculating total time tf=tf1+tf2 of a geostationary orbit transfer process and a propellant consumption amount of the transfer process; and by taking minimization of the geostationary orbit total time tf as a design object, optimizing the thrust azimuth alpha at a first initial orbit phase to obtain an optimal geostationary orbit transferring scheme.

The invention particularly relates to a design method for combining files with figures at virtual terminals in an intelligent substation. The design method comprises the steps that a digital substation is divided into a plurality of function areas according to functions by taking the digital substation as a main body; the types of IED equipment required by the function areas are determined according to the function requirements of the function areas; the required virtual terminals are determined according to the logical relationship among IED equipment; the corresponding input virtual terminals and output virtual terminals are connected to form a virtual terminal connecting figure according to the logical connecting relationship among the IED equipment; logical associations of the IED equipment are expressed through the physical relations of the IED equipment, wherein the virtual terminals are described through virtual terminal names and virtual terminal numbers; a virtual terminal connecting chart of the IED equipment is established according to the virtual terminals connected with the IED equipment. Virtual terminal visual display is provided for construction units, owner units, operating maintenance units and equipment integrators, the work efficiency is improved, and the risk that the virtual terminals are wrongly configured is reduced.

The invention belongs to the technical field of hypersonic gliding aircraft reentry trajectory planning guidance, and discloses an intelligent robust reentry guidance method and system for a hypersonic aircraft, and the method comprises the steps: carrying out the modeling based on an uncertainty system; determining the uncertainty and the type of the uncertainty in the gliding reentry process of the hypersonic aircraft, and determining the distribution form and the distribution interval of uncertainty parameters; carrying out uncertainty quantitative analysis, and establishing a robust dynamic trajectory optimization model containing random system statistical moment characteristics; establishing a robust trajectory optimization numerical value sample set; designing a deep neural network model architecture, carrying out model training, and verifying the effectiveness of the model; and loading a deep neural network model, and carrying out intelligent real-time updating on a robust trajectory planning guidance instruction. According to the method and system, the robust trajectory design time can be shortened, the active defense capability of a guidance instruction to complex uncertainty is enhanced, and the design burden of an aircraft guidance control system is reduced.

The EEPROM cell includes a writing unit having a flash cell Metal Oxide Semiconductor (MOS) for receiving from outside a gate selection signal via a gate and a drain selection signal via a drain, and writing one bit data, and a high-voltage MOS whose source is connected to a source of the flash cell MOS to have a symmetrical structure and for receiving the gate selection signal via a gate, and a sensing unit having a first sensing MOS whose source is connected to a power supply voltage, gate to a drain of a second sensing MOS and drain to the drain of the flash cell MOS, and the second sensing MOS whose source is connected to the power supply voltage, gate to the drain of the first sensing MOS and drain to the drain of the high-voltage MOS.

The invention discloses a light field display device and method, and the device comprises an image source which generates light representing virtual information; a mixing module which enables the light of the virtual information and external environment light to pass through the mixing module and a transmitting type spatial modulator to enter human eyes; the transmitting type spatial modulator which carries out the modulation of the light of the virtual information and external environment light according to a control signal of a control module; and the control module which synchronously controls the transmitting type spatial modulator and the image source, enables different independent units of the transmitting type spatial modulator to be opened or closed in an ordered manner, and enables the image source to display corresponding images in an ordered manner. The method comprises the step: quickly opening or closing different independent units in an ordered manner through the transmitting type spatial modulator, and building a simulation light field for to-be-represented virtual information. When a user see the to-be-represented virtual information through the simulation light field, there is no convergence adjustment conflicts, thereby avoiding the fatigue of human eyes and other physiological discomfort.

The invention aims at solving a problem that higher requirements on steering power easiness, safety and so on cannot be guaranteed only by adopting the traditional hydraulic power steering system in the existing hydraulic auxiliary driving heavy commercial vehicle, discloses an auxiliary differential power steering control method for a hydraulic hub motor and belongs to an automobile control system. The auxiliary differential power steering control method for the hydraulic hub motor comprises the steps of adjusting the opening value of a hydraulic electromagnetic valve through feedforward table look-up control and feedback PID correction control; controlling left and right steering wheel hydraulic motors to generate a torque difference so as to realize an active differential power steeringfunction; and providing a vehicle steering power together with the traditional hydraulic power steering system. According to the auxiliary differential power steering control method for the hydraulichub motor; the vehicle steering power characteristic is more optimized through active adjusting of the hydraulic motors; the easiness and safety under the situation that the traditional hydraulic power steering system is invalid are guaranteed; and meanwhile, vehicle energy is also effectively distributed and utilized so that a certain energy-saving effect is realized.

Enhanced memory circuits are described that maintain coherency between concurrent memory reads and writes in a pipelined memory architecture. The described memory circuits can maintain data coherency regardless of the amount of pipelining applied to the memory inputs and/or outputs. Moreover, these memory circuits may be implemented as dedicated hard circuits in a field programmable gate array (FPGA) or other programmable logic device (PLD), and can be supplemented with user-configurable logic to achieve coherency in a variety of applications.

The invention discloses a shielding structure design method for a staggered winding transformer of a flyback power supply. According to the method, the design burden of a filter is not increased, simulation and calculation are adopted in the whole design process and the additional material cost is not increased. In addition, the weight and the volume of a staggered wound transformer device are reduced; shielding layers are relatively light and relatively thin, so that the weight or the volume of the transformer or a whole power supply device is not increased; common-mode noise is suppressed toa certain extent through designing appropriate double copper foil shielding layers in the transformer, the design burden of a common-mode filter is reduced and the weight and the volume of the filtercan be appropriately reduced, thereby reducing the weight and the volume of the whole device. Meanwhile, the suppression effect on the common-mode noise is very significant; the common-mode noise ata load side of a flyback circuit is suppressed by using the double copper foil shielding layers of the staggered wound transformer designed by the method; and the suppression effect, on the common-mode noise, of the double copper foil shielding layers is obviously better than that of a traditional shielding layer.

This invention discloses an automatic plotting device for frames including a data pick-up module picking up data in the preset frame search information, a data process module computing said data to get the manufactured sizes of the frames, a data output module outputting the computed and pick-up data to generate a blueprint, which expresses the same kind of frames on a same drawing to reduce the design load and speed up the design.

The invention discloses a communication method and communication circuit between a downhole instrument and a ground control system. The downhole instrument and well equipment are connected through a main bus, the downhole instrument and the well equipment adopt a Manchester II coding format to communicate; the downhole instrument receives a command or data meeting the Manchester II coding format on the main bus through a polar isolation transformer, a signal processing circuit directly reads the command or data, the parallel data processed by the signal processing circuit is converted into thecommand or data meeting the Manchester II coding format through a coder/decoder so as to be sent to a signal driving signal; and the command or data is sent to the ground control system by a drive isolation transformer through the main bus. The use of an address latch, a shift register, a trigger and a large number of discrete devices in the prior art is omitted, the power consumption and the heat dissipation of the communication circuit are reduced, the work burden and the hardware cost expenditure of a constant temperature system are effectively reduced, and the method has the features of being reliable in communication and stable in transmission.

The invention discloses a flexible self-adaptive device of a cold chain logistics pipeline unit load device. A cold chain unit load device (3) runs in a cold chain transmission pipeline (2), a cold chain cargo compartment is arranged in the cold chain unit load device (3), the cold chain cargo is loaded in the cold chain cargo compartment, and flexible self-adaptive devices (1) are mounted in thecold chain unit load device (3) in the back and forth direction; and each flexible self-adaptive device (1) comprises a plurality of sets of multi-linkage mechanisms distributed in the circumferentialdirection of the cold chain transmission pipeline (2), an elastic mechanism and rollers (106), and the elastic mechanism at least drives the multi-linkage mechanisms to make the rollers (106) closelybe attached to the inner pipe wall of the cold chain transmission pipe (2). The flexible self-adaptive device breaks through the limitation of the fluid pipeline, realizes the pipeline transmission of the cold chain cargoes, further realizes the flexible self-adaption of the cold chain unit load device in the pipeline transportation process, and improves the transportation stability, safety and reliability.

The invention discloses a circuit and a method for achieving high-precision temperature measurement based on a 12-bit ADC. The problem that the precision and the cost cannot be simultaneously considered in the prior art is solved. The circuit comprises a Wheatstone bridge module, a differential module and a sampling module which are connected in sequence, the sampling module comprises an analog-to-digital converter ADC, the analog-to-digital converter ADC has 12 bits, and a temperature value can be obtained after data sampled by the analog-to-digital converter ADC is converted. According to the circuit and the method, cost does not need to be increased, design burdens of software and hardware of products are reduced, calculation and implementation are easy, cost is reduced while high precision is achieved, and batch production is easy.

The invention discloses a multi-scale image translation method based on semi-supervised learning, and the method comprises the steps: enabling a multi-scale discrimination generative adversarial network to complete all image translation tasks at the same time through a unified frame, and carrying out the discrimination of a plurality of scales of a generated image at the same time, and preventingmany unreasonable forged objects from locally appearing in the generated image; combining a generative adversarial network and dual learning to train a model by utilizing non-paired data, improving the semi-supervised image translation effect, meanwhile, effectively utilizing non-paired image information to improve the performance of the model, and reducing the requirement of model training for paired images. According to the multi-scale image translation method based on semi-supervised learning, the defect that a supervised image translation algorithm needs a large amount of training data iseffectively overcome, model convergence can be accelerated, and themodel performance is improved. The method has obvious effect and is suitable for wide popularization.