Decentralised wireless networked model predictive control design for wind turbines

Abstract

An innovative Decentralised Wireless Networked Model Predictive Control (DWNMPC) is presented to regulate wind turbines speed and compensate the effect of communication constraints such as dropouts. A decentralised control system and an estimation algorithm have been developed as follows. The decentralised structure decomposes the wind farm into n turbines each with its local controller. A coordinated strategy where controllers share the turbine’s status among other controllers is implemented to adjust the power generated by each turbine. A decentralised Kalman Filter (KF), based on the state-space model, is available for each subsystem to estimate the states locally. Then, the local control performance is optimised using the state estimation while considering input constraints. Experiments using the TrueTime network simulator and a 5 MW variablespeed pitch regulated wind turbine for below rated wind speed model are provided and the results demonstrate the effectiveness of the proposed DWNMPC approach in compensating for high percentages of dropouts while providing good performance and robustness.