Abstract: Given the ineluctable uncertainties of renewable generation and load demand, it becomes challenging to achieve optimal energy management for integrated energy microgrid clusters under acceptable performance. To overcome this challenge, a novel two-layer energy optimization management framework is proposed based on tube-based model predictive control for integrated energy microgrid clusters with a peer-to-peer energy transaction mechanism, which significantly enhances the enforceability of optimal energy management strategies under the influence of uncertainty factors. Furthermore, to overcome the conservatism common in the application of conventional tube model predictive control technology, a flexible tube model predictive control technology is proposed, which enhances the strategy with a higher degree of elasticity to obtain a trade-off between the operational robustness and cost reduction under different convincing levels for renewable prediction, by adaptively adjusting the shape of the “tube” according to the rolling optimization time domain internal source and load prediction accuracy. Case studies on typically integrated energy microgrid clusters demonstrate that the proposed method can ensure that the optimal energy optimization management strategy has impressive economic benefits, renewable energy utilization efficiency, and robustness in confronting high system uncertainties.