Predicting Wind Turbine Scheduling Maintenance Using Artificial Neural Network for Preventing Blade Breakage: Case Study Baron Techno-Park

Fredi Prima Sakti (Universitas Teknologi Yogyakarta, UTY)
Haidar Rahman (Ridwan Budi Prasetyo)
Ikrima Alfi (Universitas Teknologi Yogyakarta, UTY)
Ridwan Budi Prasetyo (Badan Pengkajian dan Penerapan Teknologi, BPPT & Universitas Teknologi Yogyakarta, UTY)

International Journal of Engineering, Technology and Natural Sciences (IJETS)
Vol 3 No 1 (2021): International Journal of Engineering, Technology and Natural Sciences

One of the main issue that Baron Techno-Park (hybrid power plant) is facing are the practices of finding a suitable maintenance strategy. Operation and maintenance (O&M) of wind turbines are heavily affected by weather condition, particularly wind conditions. Blade failures, such as blade breakages, can lead to catastrophic consequences. The causes of blade breakages in Baron Techno-Park is due to unpredictable high wind speed from different directions. A technique that this research propose to implement a maintenance strategy in order to create an efficient O&M and also prevent the breakage of the wind turbine blades, is by using the Artificial Neural Network (ANN). ANN performance is satisfactory with the wind speed error of 30.25 % and wind direction error of 13.74 %. Also, R2 has a highest prediction of 0.998. Analyzing the survival wind speed of 60 m/s which is specify in the wind turbine specification. Analyzing the prediction results. It is safe to say that during the month of July 2021, it is not necessary for a maintenance schedule.