3results about How to "Valid inspection" patented technology

The application discloses a kind of multiple umbrella wing unmanned aerial vehicle semi-physical simulation systems, including host computer and multiple umbrella wing unmanned aerial vehicle controllers;The host computer is connected with each umbrella wing unmanned aerial vehicle controller;The host computer includes path planning module, unmanned aerial vehicle state acquisition module and first communication module;The path planning module is used to carry out path planning to the formation of multiple umbrella wing unmanned aerial vehicle, and the expected path of each umbrella wing unmanned aerial vehicle is output;The unmanned aerial vehicle state acquisition module is used to calculate the update value of unmanned aerial vehicle state according to the control amount output by umbrella wing unmanned aerial vehicle controller and unmanned aerial vehicle dynamics model;The first communication module is used to send the expected path, the update value of unmanned aerial vehicle state to corresponding umbrella wing unmanned aerial vehicle controller;The umbrella wing unmanned aerial vehicle controller is used to calculate control amount according to expected path and unmanned aerial vehicle state.Utilize the system, multiple umbrella wing unmanned aerial vehicle cooperative air-drop system can be calibrated, and the function and performance of each equipment of system are more directly and effectively verified.

This invention relates to a rapid method for testing the sealing performance of a stepped sealing structure, comprising: assembling the stepped sealing structures together to establish a dimensional chain; converting the gasket dimension A3, which characterizes the sealing reliability of the connection parts of the stepped sealing structure, into a tightness measurement value A0 using the dimensional chain; calculating the theoretical range of the tightness measurement value A0 and determining its upper and lower limit dimensions; designing a gap rod, which includes a through end and a stop end; and measuring the tightness measurement value A0 using the gap rod. If the through end passes the measurement, it indicates that the gasket is not over-compressed; if the stop end does not pass the measurement, it indicates that the seal is properly compressed. This method can efficiently and reliably test the sealing reliability of a stepped sealing structure in a liquid rocket engine after tightening.

A PCB assembly X-ray image automatic interpretation method based on deep learning comprises the following steps that firstly, image matching is carried out on an image to be matched and an X-ray image of a qualified PCB assembly, and an image to be detected is obtained; then, identifying a target element in the to-be-detected image by using the target detection network; segmenting and extracting each target bonding pad area by using a semantic segmentation network, and filtering a target bonding pad coarse segmentation result based on a target detection result; and finally, segmenting an abnormal target again from the filtered target pad segmentation result by using a maximum between-class variance segmentation method so as to obtain an abnormal target segmentation result. And calculating size parameters corresponding to different abnormal targets, and counting abnormal conditions. According to the invention, the automatic interpretation of the X-ray image of the PCB assembly is realized, the dependence on manpower in the quality detection process of the PCB assembly is reduced, and the situation that the quality detection of the PCB assembly has problems due to artificial reasons is reduced.