Wind speed sensor calibration in thermal power plant using Bayesian inference

Abstract

Using natural draft dry air cooling systems in the power plant cycle is one of the proposed solutions for less water consumption. But the wind blowing will cause decreasement of cooling system performance in the power plants that work with the Rankin cycle. Therefore, it is important to know the right amount of wind speed to make the right decision to prevent reducing generating power or provide the right solution to improve the performance of the power plant cooling system. There are many methods of calibration of sensors in the world. But using optimization techniques or stochastic methods that do not require physical facilities and additional costs is almost a new approach. Therefore, in this study, wind sensor was calibrated using Bayesian inference method. Bayesian inference is a statistical method which updates the probability of a hypothesis as more evidence or information is available. Data obtained from Shazand power plant. Finally, after calibration, it is perceived that the error between observations and system models has decreased about 95%.