Energy Efficiency Analysis and Optimization of Industrial Processes Based on a Novel Data Reconciliation

Abstract

The energy management and energy efficiency optimization are of particularly significant for promoting the sustainable development of industrial processes. However, industrial raw data with uncertain, relevant and inaccuracy characteristics influence the reliability and accuracy of energy efficiency analysis and optimization modeling. Therefore, an energy efficiency analysis and optimization method based on a novel data reconciliation (DR) integrating Gaussian mixture model (GMM) and mutual information (MI) is put forward. First, the material flow information with multiple data characteristics corresponding operation modes is divided through the GMM. Moreover, the novel data reconciliation model integrated with critical variable and mutual information is established considering time-scale redundancy information in different modes, then the comprehensive data reconciliation result is evaluated by the hypothesis testing. Furthermore, the reconciled data is applied to analyze the exergy balance and built the energy efficiency optimization model with multi-objective for a case study of industrial evaporation process. Finally, simulation case and industrial application case are used to analyze and discuss, and the results show that the validity and applicability of the proposed approach are illustrated in energy saving potential which is about 14.81%.