4results about How to "Resolve synchronization" patented technology

This invention relates to the field of protocol conversion and transmission technology, specifically an FPGA-based AFDX to Aurora protocol transmission and conversion system. The system includes: an AXI-AFDX module responsible for parsing and encapsulating standard AFDX frame signals, and separating AFDX control signals from data signals; an AFDX over Aurora module that achieves cross-clock domain synchronization by storing control and data signals separately in dual FIFOs, extracting frame groups in an orderly manner based on virtual link static priority scheduling and same-priority polling mechanisms, and encapsulating them into IP frame format with extended fields to ensure that critical AFDX control information is not lost during Aurora transmission; and an Aurora protocol module that performs bidirectional conversion between IP frame signals and Aurora protocol high-speed serial signals, adapting to external GT high-speed interfaces. This invention achieves reliable transmission to high-speed serial links.

The present application belongs to the technical field of artificial intelligence chip design, and particularly relates to a LogSoftmax function hardware implementation system and method thereof. The system comprises a control module, a data preprocessing module connected with the control module, a configuration data generation module connected with the control module, a LUT storage module connected with the control module and the configuration data generation module, and an operation module connected with the control module, the data preprocessing module and the LUT storage module. The data preprocessing module comprises a maximum value calculation circuit, a data parallel processing circuit and a precision conversion circuit. The configuration data generation module comprises a control flow analysis module, a data flow analysis module, a control information generation module, an exponential FIFO and a logarithmic FIFO. The LUT storage module is used for storing LUT tables of exponential values and logarithmic values. The method can significantly reduce storage resource consumption and improve computing efficiency.

The utility model provides a kind of shoveling tooth machine, it is related to shoveling tooth machine technical field.It includes base, portal frame, shoveling tooth unit, first single motor, first single screw and first sliding unit, portal frame is connected with base, first single motor and first single screw are centrally connected on portal frame, first single motor is driven connection with first single screw, shoveling tooth unit is transmission connection with first single screw and is movably connected with portal frame by first sliding unit.By using first single motor and first single screw cooperation, the synchronization problem of double motor and double screw in control process can be solved, the shoveling precision is greatly improved, and because single motor and single screw design are used, the shoveling tooth machine is more miniature, smaller in size, the start-stop position of shoveling tooth unit processing can be accurately controlled, the quality and performance of product are improved, the requirement of high standard is satisfied, and production efficiency is improved.

This invention relates to the field of data processing technology and discloses a real-time data anomaly detection method integrating time series analysis. The method includes: acquiring a high-frequency target time series signal and a low-frequency operating condition feature vector; generating a synchronization operating condition vector using mask dimensionality reduction and zero-order preservation mechanisms; generating a steady-state and transient composite hash key through hash mapping and comparison; reconstructing the high-frequency signal and constructing a high-dimensional state trajectory matrix; determining the routing query key based on the hash key; extracting benchmark features from the dictionary when a match is found; performing regularization on the sample covariance matrix using a noise floor parameter; and projecting the sample onto the tangent space to calculate anomaly metrics for anomaly determination; and calculating incremental benchmark features and updating the dictionary when a match is not found and the physical continuity and manifold stability conditions are met. This invention solves the multi-rate data synchronization problem, reduces hardware noise interference and calculation errors in direct ranging in non-Euclidean geometric space, and enables online learning of new operating conditions without requiring system downtime for retraining.