基于风速估算和改进复合算法的风机MPPT控制OA

A maximum power point tracking(MPPT)algorithm for wind turbines is proposed for wind speed estimation.The wind speed estimation is based on a backpropagation(BP)neural network.The proposed algorithm is to reduce the dependence on wind speed measurement devices and enhance the power generation efficiency of wind power systems.In order to overcome the drawback that the initial weights and thresholds of the traditional BP neural networks are assigned randomly,an improved parrot optimization algorithm(POA)is employed to optimize the BP neural network,thereby improving the accuracy of wind speed estimation.Subsequently,a hybrid MPPT algorithm combining the tip speed ratio(TSR)method and the hill-climbing search(HCS)method is adopted to ensure both rapid response and stability of the MPPT.Since an inappropriate setting of TSR may lead to tracking failure,an adaptive TSR mechanism is introduced into the algorithm.The experimental results indicate that the wind turbine can stably and accurately track the maximum power point and reduce power fluctuations even without prior knowledge of the precise optimal TSR,which verifies the effectiveness and stability of this method for maximum power tracking.