Customer-owned demand side management techniques are being developed to maintain stability and through optimal regulation of power and frequency. Inverter-driven Thermostatically Controlled Loads (i-TCLs) are gaining attention as flexible demand-side resources for frequency regulation. This paper develops a coordinated control strategy by incorporating thermo-electrical model of the i-TCLs in the frequency loop of the power grids. A controller is developed which will regulate the speed of the i-TCL’s compressor to attain optimal frequency control of a single area reheat steam turbine-based power grid, considering the nonlinearities such as GDB and GRC. Several different nature-inspired metaheuristic algorithms such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Grey Wolf Optimization (GWO), Ant Colony Optimization (ACO), Teacher-Learner Based Optimization (TLBO), and Coronavirus Disease Optimization Algorithm (COVIDOA) are used to tune the controller. The effectiveness of each algorithm is demonstrated using performance parameters of settling time, peak time frequency deviation, and the Rate of Change of Frequency (RoCoF) following a step load disturbance. Simulation results demonstrate that the COVIDOA controller exhibited the best dynamic performance, achieving the lowest overshoot (M
p
= 1.1950%), shortest transient time (T
trn
= 101.3136 s), fastest settling time (T
st
= 191.5702 s), and least nadir frequency (f
n
= −0.02070 Hz), indicating a rapid and stable frequency recovery. As the involvement of i-TCLs for frequency regulation increases, the power adjustment managed by i-TCLs rises, while the demand for power adjustment from the generator decreases. The developed framework shall enable the modern power grid operators to optimally choose the frequency regulator for different levels of i-TCL penetration and load perturbations.
