Streaming-Data-Driven Multi-Objective Optimal Operation for Ammonia-based Integrated Energy Chemical Systems: An Event-Triggered Scheme

As an important component in the industry sector, the integrated energy chemical system is inescapably responsible for carbon reduction in addition to economic and reliability performance. We propose a novel streaming-data-driven multi-objective optimal operation for ammonia-based integrated energy chemical systems (AIECS). In the day ahead scheduling, a detailed mathematical model for AIECS is presented, upon which a multi-objective optimization considering cost, CO
2
emission and reliability is established. For the intra-day operation, an event-triggered mechanism is firstly developed in the real-time operation of AIECS, and it enables the AIECS to be risk-aware just in time. By monitoring the difference between the predicted state of charge and the capacity limitations of battery, the AIECS identifies whether it is in hazardous regime. Then, the intelligent switch between certainty equivalence model predictive control and online-learning based distributionally robust model predictive control is proposed, such that their respective advantages are leveraged. Case studies demonstrate that the proposed approach guarantees the system safety, and brings good revenue.