Multi-communication protocol intelligent sensing method

Abstract

The invention discloses a multi-communication protocol intelligent sensing method, which can be used for power distribution and utilization Internet of Things edge calculation, and comprises the following steps: constructing a lightweight deep neural network model for power distribution and utilization Internet of Things edge calculation; training the lightweight deep neural network model throughprotocol frame data of various different communication protocol types to obtain a pre-training model; deploying the pre-training model into a multi-protocol intelligent sensing system of a power distribution and utilization terminal; and the multi-protocol intelligent sensing system performs sensing reasoning on the received protocol frame data according to the pre-training model so as to identifythe protocol type of the received protocol frame data. The multi-communication-protocol intelligent sensing method can be applied to edge calculation of the power distribution and utilization Internet of Things, various application protocols can be rapidly sensed on the edge side, and interconnection, intercommunication and plug-and-play of access equipment of a power distribution and utilizationterminal are promoted.

Description

technical field [0001] The invention relates to the field of edge computing, in particular to a multi-communication protocol intelligent perception method. Background technique [0002] With the rapid development of Internet of Things technology and artificial intelligence technology and the successful application of cloud computing services to promote the new computing paradigm edge computing, the traditional power grid increasingly needs to be deeply integrated with the new generation of information technology to realize the unification of smart grid services. Edge computing focuses on the analysis of real-time and short-period data. It is close to the device side and better supports real-time analysis and intelligent processing of local services. It has the characteristics of distributed and low latency, higher efficiency, and ease of traffic pressure. Cloud computing is also more efficient and secure. While deep learning continues to penetrate into the smart grid, explo...

AI Technical Summary

Problems solved by technology

[0003] In the Internet of Things for power distribution and consumption, the construction scale of terminal communication access networks in different regions is quite different, the various business requirements of the communication access layer have not been unified, and the standardization of application protocols is not enough and there are many types, which affects the efficient and two-way operation of intelligent power distribution and consumption services. transmission

When the edge terminal of the Internet of Things for power distribution and consumption is connected to multiple heterogeneous devices, it faces a variety of different application protocol identifications, such as the access of different devices such as electricity, water, gas, heat meter equipment, and charging pile equipment. The rapid identification of different types of application protocols is still a major obstacle to interconnection and plug-and-play

Images