Development and application of lightweight AIoT chips for peripheral applications

Artificial intelligence technology is developing rapidly, and various new technologies related to artificial intelligence are emerging in the industrial world, but along with the massive growth of data, algorithmic models tend to be complex, processing objects are heterogeneous, and the computational performance requirements are high, and the artificial intelligence chip is gradually extended to the end. The end of the end side of the artificial intelligence chip deployment scale and decentralized, the future edge of the intelligent chip demand will grow rapidly, which requires the chip needs to have the ability to adapt to various types of complex environments. At present, the electric power field is centered on the intelligent operation and inspection of the power grid, operation and control, enterprise management and electricity service and other fields to carry out independent innovation of artificial intelligence. Electricity parameter analysis for the end side, through the analysis of the characteristics of electric parameters such as current, voltage, active power, reactive power, etc., to improve the data analysis effect of non-involved load identification, anti-stealing, station identification, predictable operation and maintenance, etc., and then improve the level of refined management and green electricity consumption.