Dynamic Q–V Coordination Strategy for Voltage Stability and Reactive Power Sharing in Hybrid Hydro–PV–Hydrogen–EVCS Microgrids

Hybrid microgrids combining hydro generation, photovoltaic (PV) plants, hydrogen storage, and electric vehicle charging stations (EVCSs) lose voltage stability under large disturbances because each converter regulates reactive power from local measurements and the units cannot observe one another. This article proposes a dynamic Q–V coordination strategy (DQVCS). The controller takes the voltage deviation at the point of common coupling, the terminal voltage of each unit, and the measured reactive output and capability limit of each unit as inputs, and returns a reactive power reference and an excitation correction for every controllable device. Two mechanisms distinguish it from a fixed Q–V droop. The local Q–V slope is rescaled online by the depth of the system voltage error, so a unit responds sharply during a deep sag and gently near nominal voltage. The aggregate reactive requirement is then distributed through participation factors proportional to the reactive headroom remaining in each unit, so support is shared by capability rather than by electrical proximity, which prevents the units nearest a disturbance from saturating while others stay idle. Only one scalar is broadcast, which avoids the latency of centralized dispatch. The strategy is evaluated by time‐domain simulation on a hybrid hydro–PV–hydrogen–EVCS microgrid and benchmarked against a conventional Iq‐based voltage stabilizer (IQVS), a fixed Q–V droop, an adaptive droop, a centralized reactive power dispatch, and a model predictive voltage controller across nine disturbance scenarios, giving 54 cases. Under a single‐line‐to‐ground fault at the point of common coupling, the maximum voltage deviation falls from 0.1218 to 0.0973 p.u. and the recovery time from 0.225 to 0.149 s. Under a simultaneous fast‐charging surge at two EVCS buses, the deviation falls by one‐half against the adaptive droop, which differs from the proposed law only by the absence of the allocation rule, and a reactive sharing index defined on the normalized reactive output of the controllable units falls from 0.4501 to 0.1977. Across the nine scenarios, the maximum frequency deviation differs between the six strategies by less than 2 × 10 ⁻⁴ p.u. in five cases, so the benefit of coordinated Q–V action lies in voltage regulation and reactive power sharing rather than in frequency support.