RESUMEN
In the process of marketization, the lack of redundancy evaluation of the MSW incineration treatment capacity leads to the regional imbalance of treatment capacity and waste of resources. Therefore, this study aimed to develop a spatial-temporal redundancy evaluation method for the MSW incineration treatment capacity based on the accurate MSW generation prediction using artificial intelligence. To achieve this aim, this study first proposed and finalized a prediction model of the provincial MSW generation by applying the artificial neuron network (ANN) technology and using the statistical data of Jiangsu Province of China from 1990 to 2020. In the finalized model, the input variables consist of three demographic variables, three social variables, and five economic variables; the model structure that includes four hidden layers and 16 neurons in each hidden layer performed best with a coefficient of determination (R2) of 0.995 on the training samples and an R2 of 0.974 on the test set, respectively. Using the finalized model and statistical data of all provinces in China, this study proposed a redundancy evaluation method for the MSW incineration treatment capacity and evaluated the spatial and temporal redundancy status of China. The results first confirm the effectiveness of the proposed method to model and quantify the redundancy problem. Second, according to the evaluation results, even if no new treatment plant will be built before 2025, 10 of China's 31 provinces still have redundancy problems, indicating the severity of this problem. This study first contributes to the body of knowledge by modeling the redundancy problem of the MSW incineration treatment capacity. Moreover, this study provides a tool to quantify temporal and spatial redundancy using advanced technology and publicly available data. Furthermore, the results can help waste-related authorities and organizations make optimal strategies and actions to better match MSW treatment capacity and MSW generation volume.
