Probabilistic System Identification Of Pelton Turbine Using Bayesian Least Squares

This paper presents a system identification approach for a Pelton turbine using Bayesian Least Squares (BLS) to estimate its transfer function. Pelton turbines, widely used in high-head hydroelectric power plants, exhibit dynamic behavior that can be difficult to model due to noise, nonlinearities, and limited measurements. Traditional least squares methods often fail to handle uncertainty effectively, leading to less reliable models. Bayesian Least Squares offers a probabilistic framework that incorporates prior knowledge and provides a distribution over the model parameters, rather than just point estimates. This enhances robustness and enables quantification of uncertainty in the estimated transfer function. In our study, a parametric transfer function structure was assumed for the Pelton turbine. The BLS method was applied to simulated input-output data to estimate the system’s dynamics. Results demonstrate that BLS yields accurate parameter estimates and captures model uncertainty effectively. Overall, the proposed approach shows promise for identifying reliable models of Pelton turbines, especially in scenarios where data is uncertain or limited, making it suitable for practical engineering applications.

کلیدواژه ها:

Bayesian Least Squares ، System Identification ، Pelton turbine ، Estimation