AbstractThis research work proposes a modified sine cosine algorithm (mSCA) to optimize the parameters of a multistage fractional‐order PD‐PI (MSFOPD‐PI) controller for load frequency control of an isolated AC microgrid. Microgrid considered here is a cluster of renewable energy sources like photovoltaic (PV), wind turbine generators (WTG) which diverges from their optimal operating point due to change in its environment. Initially, the performance of proposed mSCA algorithms is compared with the original sine cosine algorithm (CSA) well as other techniques like Crow search algorithm (CSA), Artificial bee colony (ABC), Cuckoo Search (CS), Gravitational search algorithm (GSA), Dragonfly (DA) and Genetic algorithm (GA) techniques using benchmark test functions reported in the literature. In the next stage, proposed mSCA is applied to optimize MS FOPD‐PI controller parameters under different fluctuating scenarios of source and load. To demonstrate the supremacy of the proposed MS FOPD‐PI controller the performances are compared with conventional PID, PI and I controllers. A sensitivity analysis is made by varying the system parameters to justify the controller's robustness.