Investigation of wind resource characteristics in mountain wind farm using multiple-unit SCADA data in Chenzhou: A case study

Abstract

Wind resource characteristics in a mountain wind farm is more complicated and changeable than that in a flat wind farm due to the influence of complex mountain terrain-topography on air flow. Although there have been a large number of investigations on wind resource characteristics, the local wind resource characteristics in a mountain wind farm are still a lack. To fill this knowledge gap, high-resolution wind data are extracted for investigating from multiple-unit SCADA data, that is, wind data are collected from four selected wind turbines with SCADA system, which means wind data are from four different locations. For each location, more than thirty million sets of wind data are extracted for investigation in 2015. Then, several kinds of probability density functions (PDFs) are compared, and one-dimensional and multidimensional kernel density estimation method is selected for the investigation of the frequency distribution. Data preprocessing methods of both wind speed and wind direction are also presented. Finally, the results of this investigation reveal the specific change characteristics of wind speed and wind direction in the investigated mountain wind farm. Yearly mean wind speeds for four different locations are 3.

Multifractal analysis of the time series of daily means of wind speed in complex regions

“…Wind has been gaining an increasing attention in the context of renewable energy because it represents one important substituent to conventional fuels that should play a major role in the future energy mix [1]. In fact, wind power has the advantage to be widely produced (in the three consecutive years, from 2013 to 2015, the world total wind capacity grew up more than 10% [2]) with little environmental pollution, becoming economically competitive as a type of clean energy source capable of withstanding environmental damage and avoid future crises [3].…”

SSAE‐MLP: Stacked sparse autoencoders‐based multi‐layer perceptron for main bearing temperature prediction of large‐scale wind turbines

“…Negative numbers and null values have no significance in our study, which belong to data recording errors. Step 4:Smooth abnormal data with the mean of the previous and the following data. According to the literature and expert knowledge, it is impossible for the condition parameter variable value to increase or decrease to two times of the original value in one‐second interval 48 …”

“…According to existing SCADA‐based condition monitoring and abnormal detection literature, 12,48‐50 the data resampling method is applied by calculating the mean for each parameter variable in this study. The 5‐min mean of SCADA data is utilized to verify the proposed approach.…”

Investigation of Inverter Temperature Prediction Model in Wind Farm Based on SCADA Data

“…There have been many relevant researches based on the SCADA data, such as WT power curve calculation, operating state evaluation, identification of abnormal states and fault diagnosis, fault identification and vibration analysis, etc. [13][14][15]. For example, WT gearbox was monitored through analyzing lube oil and bearing temperature data in [4].…”