Event-Triggered Online Scheduling for Industrial-Integrated Energy System

Sound scheduling and allocation of multienergy media is of paramount significance for reducing energy consumption and improving operation efficiency in an industrial-integrated energy system of a SIP, and the online optimization solution of various scheduling events can be regarded as the prerequisite for such challenging tasks. Thus, a novel granular-driven extended particle swarm optimization with metacognitive component in terms of what-to-learn, how-to-learn, and when-to-learn, termed as MC-GEPSO, is proposed in this study, which were realized by the observation, learning and selection of operation mode characteristics described by granularity. Furthermore, an adaptive interval type-2 fuzzy system based on autonomous fuzzy rule learning mechanism is reported for achieving the selection of granularity described by equipment operation performance, and the multiple-objective optimizations of online scheduling can be solved by GEPSO. The performance of the proposed MC-GEPSO is experimentally validated by a number of industry study cases, where the proposed approach outperforms manual scheduling in aspect of operation cost, operation efficiency, and carbon emissions.

I. Introduction

As The critical sector to achieve the goals of emission peak and carbon neutrality, a series of highly diversified approaches were developed for tradeoff between energy consumption and production demand by improving operation efficiency and reducing carbon emissions in a steel industrial park (SIP) [1]. The industrial integrated energy system (IIES) of SIP generally involves the production, transformation, and consumption of cold, heat, power, gas, etc. [2]. However, the synergistic effect of these multiple energy media (MEM) might be neglected by an online operator when the scheduling plan was formulated due to their complicated coupling property [3]. Facing the foreseeable pressure of carbon emission reduction, those ubiquitous scheduling events (SEs) with respect to demand event (DE), response event (RE), and demand response event (DRE) incurred by multienergy system require real-time optimization solution for supporting the decision-making of online operator, and therefore pose new challenges for event-triggered online scheduling. The abbreviations and their definitions are listed in Table I.