This paper proposes a power control method for distributed photovoltaic （PV） grid-connected systems in distribution networks based on an improved aquila optimizer. This method uses internet of things （IoT） technology to collect real-time key data from distributed PV grid connection and distribution networks

including current

voltage

weather conditions

and load status. A generalized predictive control method is adopted to construct a power prediction control model by integrating these key data. The method fully considers the dynamic characteristics of PV systems and grid fluctuations

and derives future output predictions of the PV grid-connected system by employing the Diophantine equation

thereby achieving effective prediction of voltage fluctuations and flicker over a future period. Based on the prediction results and actual grid requirements

a rolling optimization strategy is designed to further optimize power control in response to the high dynamics and uncertainty of the power grid. An improved Aquila optimizer is introduced to optimize the regulation parameters of the inverter according to the prediction results and actual grid demands. By adjusting these parameters

the method reduces output power fluctuations

subsequently alleviates voltage fluctuations and flicker

and achieves accurate prediction and control of distributed PV grid-connected power. Experimental results show that

compared with the proactive control method

the proposed method exhibits superior performance in power control of distributed PV grid connection in distribution networks and is suitable for practical application scenarios.