التنبؤ باستهلاك الطاقة الشمسية الكهروضوئية في الصين باستخدام عشرة حدوديات

Abstract EN

The problem of this paper is to forecast China's consumption of PV-solar energy up to the year 2030.

To solve this problem, ten polynomial models are built with degrees from first to tenth.

Computing the model's parameters is based on the actual consumption in the period from 1989 to 2018 and applying the method of least-squares with the aid of Excel, xuru, and php software.

A comparison between the ten models is performed based on four measures; the root mean Square Residuals (RMSE), the complement of coefficient of determination (1-R²), the Bayesian Information Criterion (BIC), and the olute difference between the actual and forecasted consumption in 2019.

The best model is the fifth-degree polynomial.

This model is tested for sensitivity and significance and it passes the tests.

The model is used for forecasting the consumptions for the period from 2019 to 2030.

The forecasted consumption is 3788 TWh in the year 2030.