9results about How to "Guaranteed correspondence" patented technology

This invention discloses a drilling device for processing and producing wrought iron railings, relating to the field of wrought iron railing processing technology. It includes a support leg, a drilling mechanism, and a fourth transmission rail. The drilling mechanism is located on the top of the support leg, and the fourth transmission rail is installed on the top of the support leg. The drilling mechanism includes multiple positioning frames that slide on the top of the fourth transmission rail. Each positioning frame has a railing placement groove on its top to ensure the holes are at the same horizontal position. After drilling, the railings are directly welded and fixed, ensuring that the holes of the connecting pieces align with the connecting posts during subsequent bolt connections. This improves installation quality and avoids enlarged holes and misalignment of the railings after installation. This process ensures that the holes of the connecting pieces and connecting posts correspond perfectly, improving hole accuracy. Direct welding after drilling reduces process steps, increases processing efficiency, and ensures overall structural dimensional stability.

This invention discloses a training method and apparatus for a 3D meandering river recognition model, and a method for identifying meandering rivers. The method includes: acquiring several interpretation results and corresponding seismic data volumes in a target area; initializing a preset label data volume based on the interpretation results and the seismic data volume; masking the label data volume to update it; inputting each seismic data point in the seismic data volume as sample data into a pre-constructed image segmentation model for training; calculating a loss function value based on the output of the image segmentation model and the updated label data volume; adjusting the parameters of the image segmentation model inversely based on the loss function value; iteratively training until the loss function reaches a preset convergence condition, stopping the training, and obtaining a trained 3D meandering river recognition model. This model enables the identification of meandering rivers in a target area using sparse labels, improving the efficiency of seismic data interpretation.

The application provides a Zhuang language audio extraction method, including obtaining a first text set and Zhuang language audio through video processing, the first text set including a plurality of first texts A and corresponding text time axes a, step two, inputting a video, extracting an audio file, performing Fourier transform on the audio file to obtain an audio spectrum, setting a waveform amplitude threshold, filtering noise, the noise being audio with a waveform amplitude lower than the threshold, obtaining target audio B and a target audio time axis b, step three, aligning the text time axes a and the target audio time axes b, and adjusting the length of the text time axes a to be consistent with the length of the target audio time axes b; the scheme first aligns the audio and the text in the time domain, and then aligns them in the frequency domain based on the alignment in the time domain, so that the result is more accurate, and the audio and the text are aligned in the frequency domain and the time domain, so that the text and the audio information completely correspond.

The utility model discloses a kind of non-contact corneal perception instruments, comprising: base;Mandibular support, mandibular support is connected in base one side;Main body, main body is movably connected in the upper of base, it is equipped with plane driving device between main body and base, plane driving device is used to drive main body relative base horizontal movement, and main body forms detection surface corresponding mandibular support one side;Air nozzle and gas supply device, air nozzle is set corresponding detection surface, and gas supply device is used to air supply to air nozzle;Focusing device, focusing device includes photographic assembly and two symmetrical light sources arranged in the two sides of air nozzle, two sides light source is used to irradiate eye to form reflection point in eye, photographic assembly is used to obtain eye image by shooting, focusing action can be completed by the setting of focusing device, to accurately obtain patient eye center point, to ensure that air nozzle and cornea position corresponding.

The invention provides a pressure release valve switch control system and method, and relates to the technical field of oil injection control, and the system comprises a pressure obtaining unit which is used for obtaining the real-time pressure of a pressure release valve; the condition generation unit is used for generating a pressure relief triggering condition according to the real-time pressure, the valve opening threshold pressure and the delay time corresponding to the opening and closing process of the pressure relief valve; the trigger detection unit is used for generating a pressure relief control signal according to the pressure relief trigger condition; the oil injection prediction unit is used for predicting the oil injection duration corresponding to the current oil injection period according to the initial flow characteristic coefficient and the corresponding historical opening and closing pressure of the pressure release valve in the historical oil injection period; and the oil injection control unit is used for controlling oil injection of the pressure release valve at regular time according to the pressure release control signal and the oil injection duration and then closing the valve. According to the system and method provided by the invention, the correspondence between the valve pressure of the switching valve and the threshold pressure of the switching valve can be ensured, and the accuracy of switching valve control based on the threshold pressure of the switching valve is improved.

A scheduling and training method, system, device, and medium for a large language model are disclosed, relating to the field of artificial intelligence. The method includes the following steps: obtaining a vertical question; selecting at least one sub-vertical model from a comprehensive vertical model based on the vertical question, wherein the comprehensive vertical model contains multiple sub-vertical models, each of which is a lightweight model; inputting the vertical question into the selected sub-vertical model to obtain multiple first-answer results output by the sub-vertical model; and organizing and concatenating the first-answer results to obtain a second-answer result corresponding to the vertical question, thus completing the answer to the vertical question. By adopting the technical solution provided in this application, the scheduling of multiple lightweight sub-vertical models is achieved, thereby selecting a suitable sub-vertical model to answer the vertical question, effectively balancing the computational power requirements and representation capabilities of the question-answering model.

The utility model discloses a press-down mechanism of a test fixture and an earphone test tool. The press-down mechanism of the test fixture comprises a base, a cover plate, a guide assembly and a second connecting assembly. The cover plate can cover the base and comprises a base cover part, a floating cover, a pressure head part, a first elastic piece and a second elastic piece, the first elastic piece is arranged between the pressure head part and the floating cover, and the pressure head part can float in the first direction relative to the floating cover; a second elastic piece is arranged between the floating cover and the base cover part, and the floating cover can float in the first direction relative to the base cover part; the guide assembly comprises a guide column and a guide hole which can be matched in an inserted mode. The floating cover is detachably connected with the base cover part through a second connecting assembly; by setting the floating of the pressing head part and the floating of the floating cover, the downward pressing process of the pressing head part is divided into the movement of the floating cover and the movement of the pressing head part, errors in the horizontal direction are corrected, and the jig positioning precision is improved.

This invention relates to a method, system, and electronic device for synchronizing vehicle-mounted electronic maps. The method includes: acquiring multiple vehicle-mounted electronic maps organized by route; storing the multiple vehicle-mounted electronic maps in a vehicle-mounted host; the vehicle-mounted host selecting the corresponding vehicle-mounted electronic map based on the locomotive's location; the vehicle-mounted host and the dispatch center exchanging data wirelessly, and notifying and downloading the vehicle-mounted electronic maps; and triggering the synchronization of the vehicle-mounted electronic maps based on the locomotive's scenario. Specifically, selecting the corresponding vehicle-mounted electronic map based on the locomotive's location ensures the correspondence between the locomotive's location and the vehicle-mounted electronic map. Notifying and downloading the vehicle-mounted electronic maps allows them to be handled by different servers, avoiding disruption to normal locomotive control and dispatching operations. Furthermore, triggering the synchronization of the vehicle-mounted electronic maps based on the locomotive's scenario enables synchronization to be triggered in multiple application environments.

The present application relates to the technical field of unmanned aerial vehicle, especially to a photovoltaic power station unmanned aerial vehicle dual-camera intelligent cruise detection device, comprising a multi-source data acquisition module, which synchronously acquires photovoltaic panel images through an infrared thermal imaging camera and a visible light camera, and uses a laser range finder to acquire the vertical distance from the photovoltaic panel, wherein the visible light camera is equipped with a stepping motor and a rotatable circular polarizer; a dynamic parallax compensation module, which calculates a dynamic parallax compensation angle according to the vertical distance and the camera installation interval, drives the stepping motor to adjust the angle, and makes the optical centers of the two cameras intersect; and a spatial pose analysis module, which extracts the current unmanned aerial vehicle attitude data, position data, time data and the dynamic parallax compensation angle when detecting a suspected hot spot. In the present application, the vertical distance is acquired in real time by using a laser range finder, and the dynamic parallax compensation angle is calculated in combination with the camera installation interval, thereby effectively eliminating the dynamic parallax caused by the physical baseline.