The sustainable energies take increasing proportion in the power systems due to the “net-zero emission” goal, and the future trend is to make the new type power systems operate safely and stably while maintaining low carbon and economic efficiency. This paper proposes a novel smart generation system (SGS) architecture and a smart generation system consensus (SGSC) algorithm from the perspective of automatic generation control. The SGS is based on stratified sequencing method and multi-agent theory, which describes the dynamic and optimal generation dispatch of the units as a hierarchical decentralized multiple objectives programming problem. Further, the SGSC is proposed based on distributed Newton algorithm and leaderless consensus algorithm to solve the security performance degradation caused by decentralization. All the units are configured in parallel with the designed communication topology to achieve dispatch decentralization, thus ensuring low-carbon, economical and safe operation of the new type power system. The effectiveness of the proposed algorithm is verified on a multi-layer, multi-zone smart generation system model with a high proportion of sustainable energies from macro and micro aspects. Different control methods are also used for performance comparison on the two-area load frequency control model with lower regulation costs and lower carbon emissions achievement.

中文翻译：

---

智能发电系统：基于改进共识算法的分散式多智能体控制架构，用于可持续能源系统的发电指挥调度

由于“净零排放”目标，可持续能源在电力系统中所占的比例越来越大，未来的趋势是使新型电力系统安全稳定运行，同时保持低碳和经济高效。本文从自动发电控制的角度提出了一种新颖的智能发电系统（SGS）架构和智能发电系统共识（SGSC）算法。 SGS基于分层排序方法和多智能体理论，将机组的动态优化发电调度描述为分层分散多目标规划问题。进一步，基于分布式牛顿算法和无领导共识算法提出SGSC，以解决去中心化带来的安全性能下降。所有机组按照设计的通信拓扑并联配置，实现调度分散，保证新型电力系统低碳、经济、安全运行。在可持续能源比例较高的多层、多区域智能发电系统模型上，从宏观和微观两个方面验证了该算法的有效性。还采用不同的控制方法对两区域负荷频率控制模型进行了性能比较，以较低的调节成本和较低的碳排放实现。