15results about How to "Guaranteed output accuracy" patented technology

This invention discloses an inference method and electronic device, relating to the field of model inference technology. The method is applied to a pre-set accelerator and includes: calculating the rate of change of data features, the coefficient of variation of the rate of change of data features, and the cumulative deviation within the target sliding window based on input data and inference output data; calculating an initial decision threshold based on the rate of change of data features, the coefficient of variation, and the cumulative deviation; determining a final decision threshold based on the initial decision threshold and the current graphics processor utilization; and performing either cache reuse or full inference based on the final decision threshold. Thus, by dynamically adjusting the decision thresholds for cache reuse and full inference, manual threshold calibration is avoided, allowing the thresholds to adapt to changes in computing resources. This reduces the computational load on the graphics processor during cache reuse and ensures output accuracy during full inference.

The utility model relates to the technical field of coating die precision control, specifically relates to an actuating mechanism and coating machine. An actuating mechanism, include: integrated seat body, displacement piece, be located integrated seat body one side, displacement sensor, displacement sensor be located integrated seat body, displacement sensor is equipped with elastic probe, elastic probe with displacement piece is connected, elastic probe with displacement piece same side setting, displacement piece is driven ground to move towards or away from integrated seat body, displacement probe follows displacement piece movement, displacement sensor senses displacement piece's movement distance. The utility model provides an actuating mechanism and coating machine to solve in traditional actuating mechanism, multiple installation error's superposition leads to the overall assembly error of screw rod and displacement sensor between big, influence the whole mechanism's assembly precision problem.

The embodiment of the present application provides a method and device for data query, the method comprising: presetting multiple query plans; obtaining real-time query conditions; wherein the real-time query conditions include query conditions for structured data and query conditions for unstructured data; for the real-time query conditions, determining the target selection rate of the pre-constructed center point in each query plan; according to the target selection rate, determining the query cost corresponding to each query plan; using the query cost, determining the target query plan from the multiple query plans, and executing the query for the real-time query condition according to the target query plan. Through the embodiment of the present application, the optimization of the fusion query is realized, and multiple query plans are generated while ensuring the output accuracy, and a query plan with good performance is selected from them, which can not only ensure the accuracy of the fusion query, but also reduce the cost of the fusion query and ensure the query response time.

The utility model belongs to lighting device technical field, concretely relates to a high accuracy light adjusting device and adjustable light lighting equipment, this high accuracy light adjusting device includes: microprocessor, interactive module, clock signal calibration module and drive module, interactive module is configured as sending luminance adjusting signal to microprocessor, microprocessor is configured as obtaining the 48MHz clock signal after calibration through clock signal calibration module, microprocessor is configured generation corresponding PWM signal to drive module to adjust the luminance of light emitting element, the utility model discloses through the selection 48MHz clock signal can ensure the output precision of drive module, and through clock signal calibration module real -time to 48MHz clock signal is adjusted, can overcome the influence of external factor to clock signal, guarantee microprocessor with high accuracy PWM signal control drive module drive light emitting element work, realize lighting equipment luminance stepless regulation.

The utility model belongs to the technical field of sensors, and discloses a micro-differential pressure sensor, which comprises a shell, a first pressure pipe, a second pressure pipe, a first pressure sensor, a second pressure sensor and a second pressure sensor, and is characterized in that the shell is provided with an inner cavity, and the bottom of the shell is provided with a first pressure pipe; the pressure chip is connected with the shell through a substrate, and the pressure chip and the substrate are arranged in the inner cavity; the upper end of the switching module is provided with a second pressure pipe, the second pressure pipe is communicated with the inner cavity, and the switching module is buckled on the pressure chip; and the flexible circuit component is connected with the pressure chip and external equipment through the switching module. According to the micro differential pressure sensor, the pressure chip is arranged on the substrate, the switching module is arranged to buckle the pressure chip, the pressure chip is wrapped, the protection degree is high, the influence of dirt on the output of the pressure chip is reduced, the pressure chip and the substrate are arranged in the inner cavity, the sealing performance of the whole structure is ensured, and the reliability of the micro differential pressure sensor is improved.

The invention relates to a multi-channel power supply calibration method and device, computer equipment, a computer readable storage medium and a computer program product. The method is applied to a controller of a calibration system of a multi-channel power supply, the system comprises a measuring instrument and a power supply test box, the power supply test box is electrically connected with a plurality of channels of a to-be-tested power supply, and the method comprises the following steps: acquiring a calibration preset value, and controlling the power supply test box to connect the to-be-tested channels and the measuring instrument; determining a calibration value corresponding to the calibration preset value according to the calibration data set corresponding to the to-be-detected channel; adjusting a circuit parameter corresponding to the to-be-measured channel according to the calibration value, and obtaining a measurement value of the measurement instrument and an electric power parameter value corresponding to the to-be-measured channel; and under the condition that a first difference value between the calibration preset value and the measurement value and a second difference value between the electric power parameter value and the measurement value are within a preset range, determining that the to-be-measured channel passes calibration. The method can simplify the process and improve the efficiency.

The present invention provides a pressure transmitter, which includes: a shell, the shell is provided with a mounting cavity; a data processing component is arranged in the mounting cavity; a data interface is arranged in the shell and electrically connected to the data processing component, the data interface is used to connect to a host computer; a pressure detection component is arranged in the mounting cavity and is located on a side of the data processing component away from the data interface, the pressure detection component is electrically connected to the data processing component; a first buffer component is arranged between the pressure detection component and the inner wall of the shell, and is located on the peripheral side of the pressure detection component, the first buffer component is respectively against the pressure detection component and the inner wall of the shell.

The embodiment of the invention provides a Buck-type switching power supply topological structure and a control method, the topological structure comprises a Buck-type switching circuit, the Buck-type switching circuit comprises at least one phase of Buck-type switching sub-circuit, the output voltage of any phase of Buck-type switching sub-circuit is connected with a first power MOS tube, the first power MOS tube is connected with a second power MOS tube and a first energy storage inductor, and the second power MOS tube is connected with a second energy storage inductor. The first energy storage inductor is connected with a first output capacitor and a first output load, and the first output capacitor is connected with the first output load in parallel; and the main control chip is used for outputting a control signal for controlling the Buck type switching circuit based on the received sampling value of the Buck type switching circuit. According to the embodiment of the invention, the conduction loss of the diode can be reduced, and the working efficiency of the Buck-type switching power supply is improved.

A circular grating axial deformation test and compensation device comprises a bearing seat to be tested installed in a base, a shaft to be tested is installed in the bearing seat to be tested, a bearing to be tested is fixedly installed between the bearing seat to be tested and the shaft to be tested via a bearing cap of the shaft to be tested and a nut of the shaft to be tested, a handle is connected to one end of the shaft to be tested, a transition shaft is connected to the other end of the shaft to be tested, the transition shaft is connected to one end of a grating shaft via an elastic shaft, a grating bearing seat is mounted on the grating shaft, a grating bearing is fixedly mounted between the grating shaft and the grating bearing seat via a grating shaft bearing cap and a grating shaft nut, a grating seat is connected to the other end of the grating shaft, a circular grating is mounted on the grating seat, a reading head seat and a reading head mounted on the end face of the grating bearing seat are arranged outside the circular grating, and 6 gaskets are evenly distributed between the base and the bearing seat to be tested. When testing the output accuracy of the circular grating, the gasket is installed between the bearing seat to be tested and the base to simulate the change in the axial direction of the shaft system to be tested due to the influence of temperature.

The present invention relates to the technical field of switching power supplies, and in particular to an error amplifier circuit and a voltage regulation system, wherein a first positive input terminal, a second positive input terminal and one end of a resistor R1 are all connected to a reference voltage port, a second reverse input terminal and one end of a resistor R2 are all connected to the other end of the resistor R1, the other end of the resistor R2 is connected to an error amplifier output port, a feedback port is connected to the first reverse input terminal through a resistor Rc, the error amplifier output port is connected to the first reverse input terminal through a capacitor Cz, the output end of the error amplifier is used to drive a field effect transistor Mn, an input end of a current source Ip is connected to an external power supply, an output end of the current source Ip is connected to a drain of the field effect transistor Mn, an input end of a current source In is connected to a source of the field effect transistor Mn, and an output end of the current source In is grounded. The present invention can achieve the effect of ensuring the output accuracy of the error amplifier circuit by using a low-cost structure.

The invention provides a microwave device large-range modeling method and system based on an adjoint neural network, and the method comprises the steps: obtaining a range of a plurality of device parameters of a microwave device, dividing the range of the device parameters into a plurality of sub-intervals with overlapping intervals, building a sub-model based on the plurality of sub-intervals of the plurality of device parameters, and carrying out the modeling of the plurality of device parameters. Each sub-model is correspondingly provided with an adjoint neural network; pre-training each sub-model based on the training data, and performing auxiliary training through the adjoint neural network corresponding to each sub-model; constructing a sigmoid function corresponding to each overlapping interval for the upper boundary and the lower boundary of each overlapping interval; determining a correction function of each sub-model based on the sub-interval of each device parameter corresponding to each sub-model; and constructing all the sub-models provided with the correction functions into a large-range model, outputting electromagnetic frequency domain characteristic values corresponding to the to-be-calculated device parameter group based on the large-range model, and outputting derivative information corresponding to the device parameter group through an adjoint neural network.

The utility model discloses a discharging device, a seasoning mechanism and kitchen equipment, relates to the technical field of kitchen equipment, and mainly aims to prevent seasoning from being bonded in the discharging device so as to ensure the discharging precision and avoid blockage. According to the main technical scheme, the discharging device is applied to the seasoning mechanism of the kitchen equipment, the seasoning mechanism comprises a first magnetic attraction part, the discharging device comprises a shell, a second magnetic attraction part is arranged on the shell, and the second magnetic attraction part is used for being matched with the first magnetic attraction part in a magnetic attraction mode.

Embodiments of the present application provide an impedance phase detection circuit, an NFC chip, a near field communication circuit and an electronic device. The impedance phase detection circuit comprises: a first comparison circuit and / or a third comparison circuit, wherein the first comparison circuit is configured to compare a signal of a first sending end with a reference signal, and output a first pulse signal; the third comparison circuit is configured to compare a signal of a second sending end with the reference signal, and output a third pulse signal; a second comparison circuit configured to compare the signal of the first sending end with the signal of the second sending end, and output a second pulse signal; and a digital processing circuit configured to sample and process the first pulse signal and / or the third pulse signal, the second pulse signal, and output a sign bit signal representing a positive or negative of an antenna impedance phase and a duty cycle signal representing a size of the antenna impedance phase. The phase detection is realized by using waveform characteristics of the NFC chip and an antenna port, without using an ADC or a phase-locked loop circuit, so that the circuit structure is simple, the power consumption is low, and the area is small.

The present application provides a superposition operation circuit and a floating voltage digital-to-analog conversion circuit, which realizes the superposition of analog quantities by using the indirect current superposition principle. Among them, the negative phase input terminal and the output terminal of the first operational amplifier are connected, so that the first operational amplifier has the characteristics of a voltage follower, and its output terminal is embedded to the voltage inputted by the positive phase input terminal, that is, the analog quantity to be superimposed; then, the voltage signal is converted into a current signal by the current generating circuit, and the current signal generates a voltage drop on the first resistor connected to the output terminal of the first operational amplifier, and the voltage drop is superimposed with the voltage signal outputted by the first operational amplifier. It can be seen that the output operating point of the first operational amplifier in the superposition operation circuit is always fixed at the analog quantity to be superimposed, thereby ensuring the output accuracy of the operational amplifier; moreover, the high-precision current superposition amount is obtained by converting the voltage signal with low absolute accuracy, rather than being directly generated by the digital-to-analog conversion circuit, therefore, the input accuracy of the current generating circuit is relaxed, and then the requirements of the current generating circuit on power consumption and area are relaxed.

The present application relates to the technical field of rock mechanics, and provides a device for accurately acquiring the geometric morphology of rock structural surfaces and a fine quantification method. The device is used to assist in positioning the undulating structural surface when optically scanning the undulating structural surface of the sample to be tested; in the acquisition device, a tray for placing the sample to be tested is rotatably mounted on a base, and the sample to be tested placed on the upper surface of the tray is clamped by a fixture mounted on the top surface of the tray, and the rotation angle between the upper surface of the tray and the base is adjusted by the tray rotating around an axis extending along a first direction connected to the base, so that the undulating structural surface of the sample to be tested can face the scanning lens of the optical scan, thereby realizing the positioning of the undulating structural surface of the optical scan, and realizing the leveling of the undulating structural surface with an inclination; finally, the positioning coordinate mark set on the fixture is used to assist in establishing a unified coordinate system for the geometric morphology point cloud data acquired by optical scanning.