A Hybrid Forecasting Model Based on EMD-GASVM-RBFNN for Power Grid Investment Demand

Table 3

Summary of prediction methods and application literature based on intelligent algorithms.

Author	Forecasting object	Forecasting model	Characteristics
Pedro A. González et al. [13]	Energy consumption in buildings	Feedback Artificial Neural Network	It has a powerful function approximation ability and can fit the function expression of unknown system. It is an effective method to deal with complex nonlinear systems. However, it is difficult to fully identify and extract the internal characteristics of complex nonlinear and non-stationary time series.
Ping-Feng Pai et al. [14]	Stock price	ARIMA-SVM
ZOU Zheng-da et al. [15]	Short-Term Load Forecasting	Recurrent Neural Network-Ant Colony Optimization Algorithm
Ping-Feng Pai et al. [16]	Electricity load	SVM-GA
Thanasis G. Barbounis et al. [17]	Long-Term Wind Speed and Power Forecasting	Local Recurrent Neural Network Models
Nima Amjady [18]	Electricity Prices	Fuzzy Neural Network
Ping-Feng Pai et al. [19]	Rainfall Forecasting	Recurrent Support Vector Regression
J.P.S. Catalão et al. [20]	Electricity prices	Neural network approach
Nicholas I. Sapankevych et al. [21]	Time series prediction	SVM
Dongxiao Niu et al. [22]	Power load	SVM and ant colony optimization Algorithm
LI Jin et al. [23]	Mid-long Term Load Forecasting	Simulated Annealing and SVM Algorithm
Hong-ze Li et al. [24]	Power load	Generalized regression neural network with fruit fly optimization algorithm
Wei-Chiang Hong [25]	Traffic Flow Forecasting	SVR with Chaotic Immune Algorithm
Geng, J et al. [26]	Load Forecasting	SVR