1. The necessity of online fault diagnosis for wind turbine units
Wind energy, as a clean and renewable energy source, is receiving increasing attention worldwide. While wind power generation is developing rapidly, problems arising during wind turbine operation are becoming increasingly prominent, especially the failure of key mechanical components leading to abnormal shutdowns for maintenance, severely reducing power generation efficiency. The inconvenience of maintaining wind turbines due to malfunctions is exacerbating application issues. For example, how to improve the utilization rate of wind turbines, how to reduce the rate of sudden accidents and downtime, and how to improve regular maintenance to ensure the safe and healthy operation of equipment are all problems. Furthermore, the increasing power of individual wind turbines is leading to rising maintenance costs after malfunctions. Therefore, in-depth research into online fault diagnosis technology for wind turbines and the development of suitable technical equipment for online fault diagnosis are crucial. This includes enabling online fault diagnosis of potentially hazardous components and issuing early warnings for impending major faults. Providing a scientific basis for reducing the rate of sudden accidents and downtime of wind turbines, lowering maintenance costs, implementing condition-based maintenance, and improving power generation efficiency and economic benefits has become one of the urgent problems facing the wind power industry.
2. Current maintenance methods for wind-powered TV groups
Currently, wind turbine maintenance in my country mainly employs two methods, relying on either manual experience or offline testing equipment. The characteristics of each are as follows:
(1) Regular maintenance refers to the periodic maintenance of wind turbine units. Regular maintenance mainly involves checking the wind turbine according to the relevant procedures for wind turbine maintenance, such as changing lubricating grease and checking for abnormal noises during wind turbine operation. Regular maintenance relies on human experience and skill, and the effectiveness varies from person to person. The disadvantage is that it is difficult to detect faults in bearings and gears that are sealed inside the machine.
(2) Fault repair, i.e., post-fault repair. Fault repair mainly refers to targeted repair of faulty components after a wind turbine unit has failed. The disadvantage of fault repair is that it cannot be effective for sudden faults between two maintenance periods, and it is also difficult to track and detect potential faults with long incubation periods. Once a potential fault develops into a major fault before the next scheduled maintenance, it will cause great losses.
3. Status of online fault diagnosis for wind turbine units
In China, the application of online fault diagnosis technology in wind turbines is still in its early stages. Currently, major wind turbine manufacturers have begun experimental work on their respective wind turbines, but mass installation of online fault diagnosis equipment is not yet underway. With the increase in the capacity of individual wind turbine units and the maturation of the insurance market, configuring online fault diagnosis equipment on megawatt-class units will become an inevitable trend.
The situation is different in the few online fault diagnosis devices currently in use.
(1) Some online monitoring technologies can provide a large amount of measurement data and rich analysis charts; however, maintenance personnel cannot obtain effective information from this data to determine the status of wind turbine units.
The method accurately locates the faulty component and the type of fault.
(2) Some online monitoring technologies. Although "fuzzy" or "learning" expert systems have been introduced, it takes time for the monitoring system to "familiarize itself" or "learn" about the monitored objects in order to establish corresponding statistical models before the monitoring system can be used. Whenever the monitored objects change, "learning" needs to be done again, making it impossible to achieve plug-and-play functionality and online automatic diagnosis.
(3) Some online monitoring technologies employ fault diagnosis expert systems, which can accurately identify the type and severity of faults and precisely locate faulty components, achieving automatic online fault diagnosis without the need for learning or training; they are ready to use immediately upon installation. This application utilizes such online monitoring technology.
4. Actual operating performance of the online fault diagnosis system for wind turbines in wind farms.
The Huitengxile Wind Farm of Inner Mongolia Huadian Huitengxile Wind Power Company has a total of 120 wind turbine units, including 30 Sinovel FL1500 wind turbines and 90 Gamesa 850 wind turbines, all of which were put into operation and generated electricity on December 1, 2007.
After more than a year of operation, the wind turbine units experienced numerous mechanical failures. To better understand the operating status of key mechanical components of the wind turbine units, Inner Mongolia Huadian Huitengxile Wind Power Company and Beijing Tangzhi Technology Development Co., Ltd. began collaborating in 2008 on the application of the JK07460 wind turbine unit online fault diagnosis system (hereinafter referred to as the JK07460 system). The core technology of the JK07460 system is "generalized resonance-resonance demodulation fault diagnosis technology," which includes vibration and impact composite sensor technology, resonance demodulation hardware information processing technology, fault impact monitoring and analysis diagnosis technology, vibration monitoring and analysis diagnosis technology, and axial movement monitoring and analysis diagnosis technology.
The JK07460 system's onboard equipment, installed within the wind turbine's nacelle, includes vibration and shock sensors, displacement sensors, the onboard main unit, a pre-processor, and related connecting cables. It collects information and performs preliminary diagnostics, transmitting the collected data wirelessly or via wired connection to the central control room's main computer. The main control room computer is equipped with fault diagnosis expert system software, which performs functions such as information collection, diagnostic reasoning, alarm output, and data management. It also controls the onboard equipment and performs onboard software maintenance and remote upgrades. One central control room main computer and fault diagnosis expert system software can diagnose, analyze, and store data from multiple onboard devices.
After more than two years of field testing, the diagnostic system withstood the harsh winter weather of Inner Mongolia, demonstrating reliable operation and identifying several fault cases. Based on the automatic alarms from the online fault diagnosis system and the data analysis reports provided by Tangzhi Technology, the Huitengxile Wind Farm carried out corresponding maintenance and repairs on some of the identified faults, promptly resolving potential safety hazards in the wind turbine units.
Case 1: Gear failure in the unit's gearbox
On February 10, 2010, wind turbine unit GL19 frequently reported signals such as high gearbox oil temperature and oil pump motor failure. Tangzhi Technology's JK07460 system analysis indicated severe wear on the gearbox gears, and that the sun gear and the secondary parallel gear on the high-speed shaft were already triggering online automatic alarms. Maintenance personnel inspected the site and found that the oil pump motor was damaged. After replacing the oil pump motor, the wind turbine unit resumed operation. March 2, 2010: Online fault detection system.
