Search

• This paper focuses on daily natural gas consumption forecasting. • A novel hybrid model is developed for daily natural gas consumption forecasting. • The most recent datasets of four cities in three climate zones are collected. • An improved singular spectrum analysis is first proposed. In daily natural gas consumption forecasting, the accuracy of forecasting models is vulnerably affected by the noise data in historical time series. Singular spectrum analysis (SSA) is often introduced into hybrid models for denoising. However, as a deterministic-based algorithm, SSA does not give good results when a time series is contaminated with a high noise level. Considering this fact, this paper proposes an improved SSA (ISSA) that modifies the determination method of subseries selection in the reconstruction stage of SSA. Combining ISSA with long short-term memory (LSTM), a novel hybrid model, ISSA-LSTM, is thus developed. Additionally, for validating the robustness and superiority of ISSA-LSTM, the historical datasets of four representative cities located in three climate zones are collected as the training and testing datasets, and a comparison of ISSA-LSTM with five advanced models is performed. The results reveal that SSA would generate negative values when time series close to zero and the contribution of SSA in improving the forecasting accuracy of LSTM is insignificant. In contrast, ISSA avoids generating negative values and reduces the mean absolute range normalized error (MARNE) of LSTM by a range of 0.86–11.86%. Among the models, ISSA-LSTM achieves the best performance and its MARNEs for London (temperate zone), Melbourne (subtropical zone), Karditsa (subtropical zone), and Hong Kong (tropical zone) are 4.68%, 5.72%, 5.76%, and 14.10%, respectively. The MARNE of the tropical city is higher than that of others, which is caused by the complex natural gas consumption pattern of itself. [ABSTRACT FROM AUTHOR]

Compared with artificial intelligence black-box models, statistical white-box models have less application and lower accuracy in forecasting daily natural gas consumption that contains high dimensional and large samples. Parallel model architecture (PMA) is a forecasting strategy that improves the accuracy of forecasting models. However, due to the large numbers of non-stationarity subseries generated by PMA in daily natural gas consumption forecasting, the forecasting problem becomes more difficult. This paper proposes a weighted parallel model architecture (WPMA) strategy that reduces the numbers and the non-stationarity of subseries by introducing k-means clustering and weighting the forecasts of subseries for out-of-sample forecasting. By combining WPMA with principal component analysis (PCA) and multiple linear regression (MLR), a white-box hybrid model is generated called PCA-WPMA-MLR. Principal component analysis is a dimension-reduction algorithm that is used to extract the components from input variables, and MLR is a white-box forecaster. Additionally, the historical datasets of four representative cities distributed in three climate zones are collected in case studies. The results show that the PCA-WPMA-MLR model provides comparable forecasting performance with the deep learning model. WPMA outperforms PMA in improving forecasting accuracy, and it reduces the mean absolute percentage error of MLR by 39.07% in the Melbourne case. • A forecasting strategy, namely weighted parallel model architecture, is proposed. • A novel white-box hybrid model is generated. • The novel model provides comparable forecasting performance with LSTM. [ABSTRACT FROM AUTHOR]