Extraction of dynamic operation strategy for standalone solar-based multi-energy systems: A method based on decision tree algorithm

Abstract

Although the operation of multi-energy system (MES) plays an important role in realizing the expected system performance, the predefined strategy commonly used in engineering with fixed operating parameters might not be an appropriate technique for optimum energy-flow management. On the other hand, the large-scale application of on-line energy scheduling is hard to be achieved due to the high computational costs. Dynamic operation strategy, in which the operating parameters vary with time, is a feasible technique to guarantee both the scheduling accuracy and calculation efficiency for MESs. However, a systematic method to generate dynamic operation strategy for MESs is non-existent in literature. In this study, the equipment capacities and operating parameters of a standalone solar-based MES are optimized to minimize the annualized system cost. Based on the optimization results, the control logic to obtain the dynamic operating parameters is extracted using the decision tree algorithm. Based upon the control logic, the dynamic operation strategy of the MES is formulated. The results demonstrate that the dynamic operation strategies could achieve up to 88.56 % effect of on-line energy scheduling in terms of operation cost savings, and greatly shorten the calculation time. The method is developed to automate the design process of dynamic operation strategies and reduce the case-specific engineering efforts.