Forecasting Natural Gas Consumption in the US Power Sector by a Randomly Optimized Fractional Grey System Model

Algorithm 1

The process of optimizing parameters by random search algorithm.

(1)	Input the original data sequence , need to search for the optimal parameter model , parameter optimization interval , iteration number
(2)	Divide the data set into two parts: training set and validation set
(3)	Define the objective function
(4)	Initialize the objective function judgment value as
(5)	for ; ; do
(6)	For , randomly select a set of parameter values with uniform distribution in the interval
(7)	Pass the training set into this model to train the and predict the time series node where the validation set is located to get the forecast result
(8)	Calculate of the forecast result and the verification set as
(9)	if () do
(10)
(11)	Update judgment value
(12)	end
(13)	end
(14)	return
(15)	Output optimal parameter