Research on Edge Computing of Automatic Control System of Unattended Intelligent Manufacturing Equipment

In order to solve the problem that unattended substations are difficult to monitor in real time due to the large number of points, this paper proposes an intelligent classification edge computing algorithm for real-time monitoring of inbound personnel. The paper uses the continuous-time Markov algorithm to complete the intelligent classification of the personnel entering the station: record the name and time of the entry for the known personnel, and perform the alarm function and other prescribed actions for the stranger. Experimental results in practical applications show that adjusting the hyperparameters of the algorithm will achieve different sensitivity and recognition rates. After fine- tuning the hyperparameters, the accuracy of the algorithm reaches about 90%. The monitoring platform developed based on this algorithm has been deployed on smart terminals, relying on edge computing technology to realize automatic identification of unattended substation personnel entering the station, and achieved expected results in production practice.