Comparison of numerical methods and metaheuristic optimization algorithms for estimating parameters for wind energy potential assessment in low wind regions

Abstract

Recently, with energy crises and environmental problems becoming increasingly obvious, the utilization of wind power has become a big concern. Meanwhile, the inconsistent relationship between China's economy and wind energy potential distribution has caused inevitable difficulties in transportation of wind power and even in grid integration. Therefore, the establishment of electrical power system integrated with local-used low-speed wind power has got considerable attention.Weibull, Rayleigh, Gamma and Lognormal probability distributions are evaluated. Then three numerical methods (NMs) - method of moment (MM), maximum likelihood estimation (MLE), and least squares method (LSM), are applied to get parameter estimation in the these distributions. Additionally, another three comparison metaheuristic optimization algorithms (MOAs), including bat algorithm (BA), cuckoo search algorithm (CS) and particle swarm optimization (PSO) are employed as comparison methods to tune the optimal parameters.Experimental results conclude that in this case MOAs perform better than NMs. Moreover, BA-Weibull, CS-Weibull, and PSO-Weibull with only a slight difference outperform all of the other distributions. Specifically, BA-Weibull and PSO-Weibull are only slightly superior to CS-Weibull. The average wind power density, the effective wind power density, the available factor and the capacity factor of wind turbine are considered as key determinant factors in assessing the low-speed wind energy potential, which are directly influenced by the parameters in Weibull model. Moreover, the wind potential assessment in the low-speed wind areas can provide an essential technique support for further investment and development, even for further wind farm construction and economy evaluation. Consequently, accurate parameter estimation is of great importance in low speed wind energy resource assessment.