Data preprocessing of fan condition monitoring based on physical structure of gearbox

Condition monitoring of wind turbine can understand the health status of wind turbine in real time and reduce the maintenance cost of wind farm, which has become an indispensable means of modern wind power operation and maintenance. Fan condition monitoring is a method to achieve real-time monitoring of fan drive chain health through scientific data processing.

Vibration analysis is a low-cost and reliable monitoring technology, which has the advantages of accurate fault location, fast diagnosis speed and high diagnosis accuracy. It is the most widely used method in the field of gearbox fault diagnosis. For doubly fed wind turbine, condition monitoring system (CMS) based on vibration signal It is to install the vibration sensor to the fixed position such as the bearing seat and the gear box shell, collect the vibration signals of each part of the transmission chain, and use the data analysis methods such as spectrum analysis to diagnose the type, position and damage degree of the fault, so as to realize the health detection of the gear box and the bearing in the transmission chain.

Most large wind farms in China need to install dozens to hundreds of wind turbines, which are distributed in a wide area, which brings some difficulties to the centralized control of wind turbines and the collection of wind farm production data. In the general CMS processing flow, all the centralized control data and production data will be transmitted to the wind farm control center in real time through the wind farm LAN, which brings great load pressure to the network. And when the network is congested or disconnected, the data needs to be stored in the collection device, which brings great challenges to the storage space. However, in order to achieve comprehensive component detectability and accurate detection accuracy, the data storage capacity and analysis resolution should be considered comprehensively. Therefore, reasonable compression of data while retaining the necessary effective information, improving the speed of data access and calculation speed are also important indicators of fan condition monitoring system.