The interconnected power system with renewable energy sources is an intricate nonlinear system, which frequently brings to light the serious problem of the system frequency and tie-line power fluctuations due to deficient damping under severe and dynamically changing loading conditions. Primarily, the power system generation-demand equilibrium point amendments following a contingency and in this case, it is stiffer to recapture a tolerable equilibrium point via conventional control practices. To over-come this problem, advanced control techniques and fast acting energy storage systems (ESS) are requisite. The ESS such as capacitive energy storage (CES) units have tremendous capability in preserving the generation-demand balance and perpetuating the power grid frequency by effective damping of the power-frequency oscillations caused due to the sudden and variable load disturbances in power system. Hence, the impact of CES units in automatic generation control (AGC) of interconnected power system is analysed and contrasted critically in this paper. Motivated by the fact that fuzzy control techniques display superior performance under volatile operating conditions in contrast to conventional control strategies, this paper also proposes a new design of intelligent multi-stage fuzzy assisted PID with filter-(1 + PI) i.e., FPIDF-(1 + PI) controller to enhance the conduct of AGC of power system. Initially, a two-area photovoltaic-reheat thermal system is considered and the parameters of FPIDF-(1+ PI) controller are optimized utilising imperialist competition algorithm. The ascendancy of the proposed controller is substantiated by comparing the outcomes with PI/FPI/FPIDF controller based on various existing optimization techniques. It is observed that CES units installed in each control area sustain the area controller to restore the area frequency and tie-line power deviations adequately and hastily following a step load disturbance in an area. To exhibit the potency and scalability of CES and the proposed controller over other prevalent control methods, the study is also extended to a multi-unit multi-source hydro-thermal power system. Finally, robustness of the proposed controller with/without CES is validated under large changes in the system parameters and random load demands. Hence, the proposed approach asserts better and vigorous results to supply reliable and high-quality electric power to the end user. (C) 2018 Elsevier Ltd. All rights reserved.