Abstract
Health condition monitoring in wind turbine motor plays an extremely important role, as these devices are highly in demand in the energy sector, especially in renewable energy and are vulnerable to both mechanical and electrical failures, more often. As such, timely identification of internal faults in these electrical devices goes a long way in productive operations by reducing the maintenance time and costs, i.e. such internal faults, if identified at an early stage, repaired or replaced timely will aid in reliable renewable energy supply. Taking this into consideration, automated continuous monitoring of wind turbine machine is a key to making this process more effective. A web application is built in the proposed research enabling quick monitoring of faults in wind turbine motor from a remote access workstation, like a control room. An experimental setup of wind turbine motor is made and data set of stator currents from both healthy and faulty conditions as well as the power spectral density from the motors were used for condition monitoring with a web interface application. Insulation failure in stator winding is a most commonly occurring electrical failure in machines. As such in the current research stator current features from the experimental machine are used for requirement analysis under both healthy and faulty operating conditions. Among the stator insulation failure most commonly occurring stator turn-to-turn faults are taken into consideration in the current research with percentage of insulation failure varying between 25% to 75%. Fault identification is done with the help of wavelet based artificial neural network analysis at the back end and the interface displays the details in the form of dashboards, with the program mainly featuring three dashboards for the unit, stator, rotor, and components in total. Using interactive visualizations, the user will be able to obtain more in-depth knowledge about the suspected faults in the system and its components, and help to take the necessary action. i.e. whether the wind turbine motor needed to be repaired or replaced depending on the vulnerability of the fault. The application also has been experimented with handheld devices by hosting the application on local host and tunneling it over the web. Interactive visualization also includes information about the working conditions of the electrical machine, such as balanced, unbalanced, and failure conditions. Thus internal electrical fault in a wind turbine induction machine can be remotely analyzed, checked and cure can be suggested with a proper online health condition monitoring system.
