Mitigation of windfarm power fluctuation by adaptive linear neuron-based power tracking method with flexible learning rate

Most wind turbine generators installed in large windfarms are of variable speed types operating at the maximum power point tracking mode to generate the maximum amount of power. Owing to this fact and regarding the random nature of the windspeed, the output power of the windfarm fluctuates. Fluctuating power is a serious problem for high capacity power plants and should be smoothed. As an effective factor on the required battery energy storage system (BESS) capacity value, tracking is the most important part performed by a coordinated control system in the power smoothing process. An ADAptive LInear NEuron (ADALINE)-based power tracking method with a flexible learning rate is proposed in this study. Furthermore, a particle swarm optimisation-based calculation of the learning rate is presented for optimising the proposed tracking method which reduces the required BESS capacity and the investment cost. Moreover, a charging/discharging algorithm for the BESS units is proposed which decreases the number of required BESS units and increases their useful life by reducing the switching activity as well. To evaluate the performance of the proposed coordinated control approach, the real output data of a 99 MW windfarm are tested. The simulation results verify the effectiveness of the proposed approach.