Improving short-term wind power prediction using hybrid improved cuckoo search arithmetic - Support vector regression machine

Wind power is fluctuant and intermittent, which has an impact on the power grid. Large-scale wind power is a serious threat to the stability and security of the power system. Accurate prediction of wind power output is significant to the safety of the power system. This study proposes a hybrid improved cuckoo search arithmetic (HICS) to optimize the hyper-parameters of support vector regression machine (SVR), which is used to predict short-term wind power output (HICS-SVR). This study uses chaotic sequences to promote the initial population. This study introduces dynamic decreasing step factor, dynamic discovery probability, dynamic inertia weight preference random walk and particle swarm arithmetic communication strategy to improve the arithmetic effect. The model has been verified with the French wind farm data set, and wind energy data has been randomly selected to form the training set and the test set of the algorithm. The regression fitting degree of the HICS-SVR is obtained under the condition of 100 iterations, with an average of 0.87 and an optimal value of 0.98. The average absolute value of the relative percentage error average is up to 7.71% and the best value is 7.12%. The HICS-SVR improves prediction precision and stability of output results, effectively. (c) 2020 Elsevier Ltd. All rights reserved.

Automated summary

This study introduces a hybrid improved cuckoo search arithmetic (HICS) for optimizing the hyper-parameters of the support vector regression machine (SVR) to predict short-term wind power output (HICS-SVR), which improves prediction precision and stability of output results effectively.