Table 1 Comparison of limitations between existing methods and the proposed method.

Distributed secondary control based on dynamic diffusion algorithm (DDA)(reference⁷)

The convergence speed is fast and the stability is good.

The controller saturation, network delay and nonlinear characteristics are not fully considered, which may lead to model mismatch.

Fuel cell based on passivity control (PBC) (reference⁸)

The structure is simple and has certain adaptability to parameter uncertainty.

Depending on the accurate parameters of load and inductance parasitic resistance, parameter drift can easily lead to control failure.

Fuzzy incremental model reference adaptive PID control (reference⁹)

Strong adaptability and flexible response to random input and load transients.

Fuzzy rules need to be designed manually. If there is insufficient running data, incomplete rule coverage will affect the adaptive ability.

Distributed event triggered secondary control (reference¹⁰)

Reduce the communication burden and support system level control.

The design of the stateless estimator depends on the real-time communication of adjacent nodes, and the communication failure is easy to lead to local control failure.

Node voltage ranking regulation algorithm (reference¹¹)

Reduce battery charging and discharging times and extend battery life.

The fixed threshold cannot cope with the rapidly fluctuating distributed load, and the response lags behind.

Battery energy storage system based on modular multilevel converter (MMC) (reference¹²)

Smooth the grid connected power of the wind farm and adjust the PCC voltage.

PI controller has poor adaptability to nonlinear load, which may lead to insufficient reactive power compensation or voltage oscillation.

Reactive self response voltage control method considering reactive power regulation characteristics of converter (reference¹³)

Effectively deal with the voltage out of limit problem of high proportion distributed photovoltaic.

In the area of weak current network or low short-circuit capacity, reactive power regulation may trigger voltage reverse overrun.

Method in this paper

Fast response to voltage fluctuation and control the overvoltage within the threshold; Multi objective optimization takes into account both regulation accuracy and energy consumption.

Power mutation and harmonic interference of multiple photovoltaic access points in complex power grid may affect the accuracy of data synchronization.