Abstract : Steam humidity has a significant impact on the safety and economy of steam turbine operation. The dual-zone heating method for online monitoring of steam turbine exhaust humidity has proven effective. This method divides the heating section into an evaporation zone and a superheating zone. Each zone is heated independently, and its temperature, pressure, and heat loss are measured independently. However, due to unpredictable factors such as the location, inherent differences, and environment of each temperature and pressure sensor, the measurement data cannot fully reflect the true situation. Therefore, data fusion is necessary. This paper utilizes an improved hardware-software combined multi-sensor data fusion algorithm to process the measured information during and after the process, thus reducing the impact of unpredictable factors. This solves the problem of inaccurate humidity measurement caused by excessive uncertainties. Experiments demonstrate the feasibility and effectiveness of this method. Keywords : Multi-sensor data fusion; Steam humidity; Dual-zone heating method 1. Introduction Due to the different locations of the various temperature and pressure sensors in the humidity monitoring system, and the influence of some uncontrollable random factors in the actual environment, the measurement data of each sensor cannot fully reflect the true situation. Therefore, the exhaust steam humidity monitoring system needs to analyze, synthesize, manage, and use the data information measured at different times and spaces through data fusion, so that the measured information can truly reflect the humidity of the current superheated steam. Data fusion, also known as multi-sensor information fusion [1,2], is based on the principle of fully utilizing the information resources of multiple sensors at different times and spaces, and using computer technology to automatically analyze, synthesize, manage, and use the multi-sensor observation information obtained in time sequence under certain criteria, so as to obtain a consistent interpretation and description of the measured object, so as to complete the required decision-making and estimation tasks, and make the system achieve better performance than its individual components. The humidity of steam is measured by the dual-zone heating method [3]. After the steam passes through the evaporation zone and the superheating zone [4,6][7,8], some of the steam cannot reach the state of superheated steam. Moreover, the calculation of the heat absorption of steam in the evaporation zone and the superheating zone is inaccurate. Due to many factors, the humidity value calculated based on the temperature and pressure sensors [5] will have errors. In order to reduce the calculation error, most scholars adopt the hard measurement method, such as adding more sensors to describe the state of steam humidity more precisely. This not only increases the cost of the entire system, but also causes more uncertainties due to too much sensor information, making the accuracy of humidity calculation more uncertain. The innovation of this paper is to use the combination of hard and soft measurement. The steam state is tracked by multiple sensors such as temperature, pressure and anemometers in different directions. The least squares method is used to analyze and synthesize the data measured by multiple sensors to approximate the more realistic state of steam humidity to the greatest extent. The measured information is processed during and after the process to reduce the influence of unpredictable factors. In the experiment, good results were achieved. Full text download of "Application of Multi-Sensor Data Fusion Technology in Steam Turbine Exhaust Humidity Monitoring System"