To enhance the monitoring of power system and ensure grid stable operation,power companies need to actively apply advanced technology to optimize network security attack detection of power monitoring system,and improve detec-tion utility.This paper focuses on the use of incremental learning to enhance detection of network security attacks in power monitoring systems by carrying out a case analysis on the utility of an attack detection model established for an actual power supply company.It may contain useful information facilitating improvement in operational quality of power monito-ring systems and accuracy of network attack detection,and is of certain significance to the security of the State Grid.