Abstract: The railway centralized signal monitoring system is the core infrastructure for ensuring traffic safety and improving operation and maintenance (O&M) efficiency. However, the proportion of invalid alarms in field application is relatively high, which seriously interferes with fault diagnosis and greatly increases the O&M load. Taking the CSM-KA monitoring system of the Qinghai-Tibet Railway as the research object and combining with field O&M data, this paper systematically sorts out four core causes: insufficient adaptability of alarm thresholds, unsynchronized acquisition timing of switching values and analog values, non-standard skylight maintenance management, and limitations of the alarm mechanism. An integrated optimization strategy is proposed from four dimensions: dynamic threshold optimization, acquisition mechanism transformation, whole-process control of skylight maintenance, and intelligent alarm cascade. Field application results show that the scheme can effectively reduce the proportion of invalid alarms and on-site O&M workload, and improve the efficiency of fault alarm analysis and handling.