Mechanical failures caused by ultra-low frequency vibrations are common in large rotating machinery of large new energy generator sets, such as the rotors of super-large wind turbines and the rotating impellers of pumped storage power stations. The main causes of ultra-low frequency faults include: complex operating conditions of the generator set, strong randomness and non-periodicity of external disturbances, and harsh installation environments. In 2008, a single low-frequency fault at Huizhou Pumped Storage Power Station resulted in significant losses of over one hundred million yuan. In 2009, low-frequency vibration of the turbine at Russia's Sayano-Shushenskaya Power Station caused excessive swinging amplitude of the impeller, leading to mechanical failure that resulted in 75 casualties and direct losses of 13 billion US dollars. In 2010, Dongfang Electric Wind Power Co., Ltd. suffered losses of one billion yuan due to fan vibration failure. The loss from major failures of large generator sets is significant, and reliable operation must have advanced diagnostic technology. With the continuous development of the new energy industry and increasing installed capacity, Ultra-low frequency vibration monitoring and fault diagnosis have become a technical bottleneck restricting the utilization of new energy. The main causes of ultra-low frequency vibration are the non-periodicity of rotating machinery operation and the uncertainty of load. Ultra-low frequency faults have the following characteristics: vibration frequency below 1 Hz, signals are extremely weak and drowned out by high-frequency noise, making signal measurement an internationally recognized technical challenge. After more than a decade of joint research and development involving production, academia, and research institutions, the project has focused on vibration monitoring of large new energy generating units as a starting point. It has conducted in-depth research on detection technologies based on contact and non-contact weak signals, developed a complete set of technology for the status monitoring and diagnosis of large new energy generating units, and created a series of new technical products for fault monitoring of generating units. These efforts have effectively promoted the application of ultra-low frequency fault diagnosis technology in China's new energy power field and provided effective technical support for the safe operation of large power equipment. Innovation achievements are as follows: 1. Expanded new principles, methods, and technologies of low-frequency vibration testing technology: Targeting the operating characteristics of generator sets under non-rated load conditions, revealed the derivative mechanisms of low-frequency vibration frequency and faults, proposed the dual magnetic circuit structure design principle for vibration test benches, established a model to study the impact of test frequency on sensors made of conical piezoelectric ceramics, and invented a new type of sensor structure for low-frequency vibrations, providing a strong technical foundation for nonlinear filtering of weak signals. 2. Developed high stability and long-term effective non-periodic random disturbance vibration measurement technology: Addressing the issue of large non-periodic random disturbances in wind turbines and their all-weather operation, established a sensor stability model with long-term effectiveness and low drift; invented a composite environmental test system, developed high reliability and low maintenance cost methods and technologies for sensing large gaps using eddy currents. Achieved non-periodic weak signal perception and temperature compensation technology under strong noise backgrounds, invented a new type of rare inert alloy material preparation technology, successfully developed new sensor structures and processes, and formed long-term low drift sensor design standards. 3. Systematically developed and applied online monitoring and diagnostic complete sets of technologies. Based on vibration monitoring technology, through the development of sensors for new energy generator units in 12 series and 50 varieties, calibration instruments, established a system-level fault diagnosis expert system based on wavelet analysis of vibration signals, suitable for different units; achieved vibration monitoring and application of large new energy generator units. The achievements have been applied in the domestic wind power industry represented by Goldwind Technology, achieving good results. They have also been successively applied in China's Three Gorges Group Company, Gezhouba Group Company, East China Tianhuangping Pumped Storage Power Station, More than a hundred enterprises and units, including East China Power Grid Company, have promoted the application of low-frequency monitoring and fault diagnosis technology. This technology has been widely applied across the domestic pumped storage industry. It has also been promoted in the wind energy sector, accounting for 60%. The achievements have resulted in 31 patent applications, with 16 granted and 2 software copyright registrations. A total of 38 papers have been published in core domestic and international journals, including 18 that are indexed by SCI/EI; 10 graduate students have been trained. After third-party testing and verification, the technical indicators of the achievements exceed those of similar products domestically and internationally, reaching an international advanced level. In the past three years, the achievements have generated direct economic benefits of 475 million yuan and tax revenue of 203 million yuan.