0 Introduction To smoothly carry out condition-based maintenance work in hydropower plants, an accurate and comprehensive assessment of the condition of the turbine generator units and their auxiliary equipment is essential. The turbine governor is a crucial auxiliary device in the turbine generator unit; its condition directly affects the power supply quality of the generator and impacts the safe, stable, and economical operation of the turbine generator unit. An effective, comprehensive, and objective method for evaluating equipment condition is to conduct online monitoring of the equipment's condition performance parameters and compare the monitored operating parameters with relevant instruments and regulations to assess its condition performance. Since large and medium-sized turbine governors are microcomputer-based, highly intelligent regulating devices with diverse hardware structures that are trending towards integration and modularization, resulting in increasingly compact external structures, it is challenging to comprehensively assess the condition performance of the turbine governor by monitoring the performance parameters of its various components online during field operation. Meanwhile, because the turbine governor is part of a complex system involving water, machinery, and electricity, the regulated and controlled system will inevitably affect the governor's performance to varying degrees. Therefore, simply monitoring certain state parameters of the turbine governor, such as speed (frequency), guide vane opening, blade angle, and power, is insufficient for a comprehensive and objective assessment of its performance. It must be considered within the context of the entire regulated and controlled system. Thus, to verify the performance of the turbine governor during operation, a series of tests are typically conducted, including static characteristic tests and dynamic characteristic tests such as start-up and shutdown, no-load oscillation, no-load disturbance, sudden load increase/decrease, and load shedding. Actual turbine governor tests are generally divided into dry-state tests and wet-state tests. However, using simulation technology, wet-state tests can be performed in a dry state, allowing for early prediction of equipment maintenance quality and early detection of equipment defects, providing a reliable basis for hydropower plant maintenance. [b]1. Principles and Methods of Governor Simulation[/b] There are two modeling methods for governor simulation: digital simulation and experimental simulation. Digital simulation involves artificially constructing a mathematical model of the unit. The model is input into a computer, and then this computer with measurement and control functions is connected to the governor. The model in the computer replaces the real unit, and the simulation test is carried out using software. Due to technical reasons, there are certain differences between the artificially constructed model and the actual unit model. The accuracy of the simulation test is not very high, and it is difficult to use state analysis software to determine whether its regulation quality has changed. At the same time, since artificial model construction requires human intervention, it is difficult to automate the simulation test. Experimental simulation is very different from digital simulation in terms of constructing mathematical models and experimental methods. Its modeling principle is: if the unit being simulated has previously undergone a dynamic characteristic test (such as no-load disturbance, load shedding, etc.), and the test was very successful, then the data recorded during the test already implicitly contains all the characteristic parameters of the unit (including the comprehensive characteristic curve of the turbine, the Ta of the generator, and the TW of the water intake system, etc.). Under the same operating conditions, if the previous test is repeated, as long as the control characteristics of the speed governor (and related mechanisms) remain unchanged, their transient process curves should be identical. Based on this principle, the previous test file can be saved as a template file. This template file can be retrieved for each simulation, and a realistic model can be constructed using mathematical methods to perform a full-scale simulation test on the speed governor. The curves from this test are compared with those from the template file, and analysis software is used to determine whether the test was successful and to assess whether and by what extent the regulating characteristics of the speed governor (and related mechanisms) have changed. The use of computer-automated model construction makes automated simulation testing possible.　**2. Accuracy Analysis and Verification Methods** Real-machine simulation is based on measured sampling data, constructing a realistic model using certain mathematical methods. Its error mainly comes from two aspects: the accuracy of the sampling signal and the calculation accuracy of the mathematical model. The accuracy of the sampling signal is generally between 0.2% and 0.1%, while the calculation accuracy of the computer can be controlled at a very high level, generally above one ten-thousandth. Therefore, the overall accuracy can be better than 0.2%. Because the accuracy of real-machine simulation is very high, it provides a high degree of reliability for the equipment status analysis of the governor. The method for verifying the accuracy of real-machine simulation is relatively simple: perform a real dynamic characteristic test (such as a load shedding test) on the unit, or consult relevant historical data; after shutdown, perform the same real-machine simulation test in a waterless state, and the curves of the two tests should completely overlap. **3. Hydro-turbine Governor Status Monitoring System Based on Real-Machine Simulation Tests** The governor status monitoring consists of two main parts: hardware and software. The hardware includes: a computer, a high-speed data acquisition system, a high-precision digital frequency converter, a high-precision digital frequency measurement device, a stator current detection module, and corresponding sensors. The software includes: all test software for the speed governor, real-machine simulation modeling software, real-machine simulation software, condition analysis software, online monitoring software, communication and plotting display software package, database, and trend analysis software. If the speed governor's condition monitoring is integrated into a unit's condition monitoring system or computer monitoring system, some hardware (such as the computer and high-speed data acquisition system) can be directly referenced from the monitoring system's computer without additional configuration; only dedicated hardware such as digital frequency converters, digital frequency measurement devices, stator current detection devices, and sensors needs to be configured. The test software package, real-machine simulation modeling software, real-machine simulation software, condition analysis software, and online monitoring software can be placed on the lower-level computer, while the communication and plotting display software package, database, and trend analysis software can be placed on the upper-level computer. During unit operation, the online monitoring software is used to perform simple monitoring and analysis of the speed governor's operating status. During unit overhaul or initial commissioning, the test software package can be used to complete the static and all dynamic characteristic tests of the governor, and the test file with good regulation characteristics can be selected as a template file and stored in the template library. Afterwards, whenever the unit undergoes minor repairs or is shut down without water, a real-machine simulation test can be performed. A scheduling program is designed to arrange the time and sequence of the simulation tests, which are all performed automatically. The simulated test files and analysis results are automatically transmitted to the host computer and stored in the database, and can also be displayed and printed. The trend analysis software on the host computer retrieves the simulation test files from the database over the years for trend analysis to determine whether the governor needs a major overhaul, thus realizing condition-based maintenance of the governor. [b]4 Two Important Significances of Governor Real-Machine Simulation[/b] a. Realizing governor condition analysis and condition monitoring. b. Load shedding simulation can replace actual load shedding. After a unit overhaul, a load shedding test is often performed to check the regulation and control characteristics of the governor and related mechanisms. Load shedding tests not only require application to dispatch, but the load shedding of large units also has a significant impact on the system and the unit, affecting its lifespan. Load shedding simulation tests can completely simulate real load shedding, thus replacing it. This ensures maintenance quality, improves efficiency, and extends the unit's service life.　[b]5 Application Prospects[/b] Currently, hydropower plants in our province use a variety of speed governors, but most have been upgraded to microprocessor-based speed governors. This provides a good application background for using simulation technology to implement real-machine simulation of speed governors. Real-machine simulation technology can be used to understand the overhaul quality of the speed governor before water filling and to promptly identify potential problems and hidden dangers, providing strong technical support for the entire overhaul. With the further development of condition-based maintenance, conducting real-machine simulation tests of speed governors in hydropower plants has become imperative. Through comparisons of simulation tests conducted at the Zhexi Hydropower Station and actual machine transition processes, the results of speed governor real-machine simulation tests are convincing. As the understanding of condition-based maintenance of speed governors deepens, the use of speed governor real-machine simulation test technology for condition monitoring and performance evaluation can be better applied.