Short-Time Wind Speed Forecast Using Artificial Learning-Based Algorithms

Algorithm 1

ConvLSTM training.

	Input: the wind speed time series data
	Output: forecasting performance indices
(1)	The wind speed time series data are measured every 5 minutes, being averaged two times for 30 minutes and 1 hour, respectively.
(2)	The wind datasets are split into Training, Validation, and Test sets.
(3)	Initiate the multi-lags-one-step (MLOS) arrays for Training, Validation, and Test sets.
(4)	Define MLOS range as {1 : 10} to optimize the number of needed lags.
(5)	loop 1:
	Split the first set based on MLOS range
	Initiate and Extract set features with CNN layer
	Pass the output to a defined LSTM layer
	Select the first range of the number of hidden neurons
	Generate prediction results performance indices
	Count time to execute and produce prediction results
	Save and compare results with previous ones
	loop 2:
	Select next MLOS range
	If MLOS range = maximum range, then goto loop 3 and initialize MLOS range
	goto loop 1
	loop 3:
	Select new number of hidden neurons
	If number of hidden neurons range = maximum range, then goto loop 4 and initialize number of hidden neurons range
	goto loop 1
(6)	loop 4:
	Select new dataset from the sets {5 min, 30 min, and 1 hr}
	goto loop 1
	If set range = maximum range, then:
	Generate the performance indices of all tested sets.
	Select the best results metrics