Research And Application Of Energy Efficiency Analysis For Ethylene Units Based On DEA And PCA

In order to achieve the emission reduction targets developed by the national twelfth five-year plan, more and more large enterprises increase the intensity of energy conservation. Among them, the ethylene industry, as a major energy-consuming enterprise, has increasing complexity in its production equipment, and its production process data is presented to be high-dimensional. Therefore, seeking an effective method for accurate analysis of the ethylene industry energy efficiency attracts more and more attention from researchers.Data envelopment analysis (DEA), as an optimization non-parametric method of efficiency analysis that does not rely on the data dimension, has been successfully applied to the energy efficiency analysis of multi-parameter input multiple indicators output units in the field of economic. In the energy efficiency analysis of the ethylene production process, with the increasing size of the device, more and more factors affect the energy efficiency of the ethylene plant. However, too much input and output indicators and targets and improper indicators will result in a low resolution DEA method, which thereby affects the effectiveness of DEA in the energy efficiency analysis. In order to solve this problem, we process the input and output indicators of ethylene production process by using Principal component analysis (PCA) method. The dimension of data can be reduced by PCA. Additionally, the redundant information, the linear relation ship and the noise that are bad factors between data can be deleted. Thus, in this paper, we proposed principal component analysis based data envelopment analysis (PCA-DEA) model to analyze the ethylene plant energy efficiency more effectively.Finally, the proposed PCA-DEA model was applied to analyze the energy efficiency of the actual ethylene production unit. To illustrate the effectiveness of the PCA-DEA model, we compare the results of PCA-DEA with those of DEA with simple selection key indicators and traditional DEA. The comparison results demonstrate the effectiveness of the PCA-DEA model, which broadened the idea of improving the accuracy of the ethylene energy efficiency analysis and provided a more effective method.