Abstract: Current heater anomaly diagnosis methods predominantly focus on single-signal processing or model fitting, with insufficient consideration of coupling relationships between high-pressure heater terminal differences and multi-system parameters. This results in limited diagnostic accuracy under complex operating conditions. To address this, this paper proposes a multi-parameter correlation-based diagnostic method for high-pressure heater terminal differences. The study first analyzes the mapping relationships between terminal differences at the upper and lower ends of high-pressure heaters and characteristic parameters such as saturation temperature under extraction steam pressure, inlet/outlet feedwater temperatures, and condensate temperature. A multi-parameter correlation matrix is then constructed to identify correlation patterns and determine thresholds for extraction steam pressure, condensate level, feedwater flow rate, and terminal differences. Finally, based on correlation analysis results, indicators such as extraction steam valve flow capacity deviation coefficient and condensate cooling section heat exchange efficiency are calculated according to parameter anomaly priority levels, enabling fault localization and terminal difference anomaly diagnosis. Experimental results demonstrate that this method maintains comprehensive diagnostic errors below 5% across seven typical operating conditions, significantly outperforming comparative methods with higher diagnostic accuracy.