You can automate machine initialization for a generator-load system in MATLAB instead of manually reading Pm and Vf each time. In power-system simulations, you can try to programmatically run the load-flow solver and extract steady-state machine states before simulation or whenever your control mode requires re-initialization.
Automated Initialization Workflow:
Set your model with a "powergui" block configured for "load-flow" analysis, then call the "load-flow" solver from MATLAB code. This computes the steady-state operating point of all machines, including mechanical power, excitation voltage, rotor angle, and currents. You can retrieve these values from the returned structure and automatically assign them back to generator blocks using "set_param", allowing your controller to start from consistent initial conditions without manual intervention. This method ensures numerical consistency and is the standard practice for initializing parallel generators with droop and AVR control.
Trigger Initialization During Simulation Mode Changes:
If initialization must occur dynamically when your controller mode switches, you can script the process so that whenever the mode flag changes, MATLAB reruns the load flow and updates block parameters. This can be done by pausing simulation and updating parameters programmatically, by switching variant subsystems tied to control modes, or by triggering an external MATLAB call from a function block. The key idea is that initialization becomes a callable routine rather than a one-time manual step.
Minimal One-Click Initialization Script:
A compact script can be used to perform the entire initialization automatically:
model = 'parallel_gen_system';
load_system(model)
LF = power_loadflow(model);
assignin(
'base','Pm_init',[LF.machines.Pm])
assignin(
'base','Vf_init',[LF.machines.Efd])
Running this script computes steady-state conditions and places the required Pm and Vf values directly into workspace variables that your generator blocks or controllers can read.
Each generator’s steady-state values can be accessed from LF.machines(k), which is a structure array containing fields such as Pm, Efd, and delta. These can then be written back into generator blocks using set_param.
set_param([model '/Gen1'],'Pm0',num2str(LF.machines(1).Pm))
Using this automated method is more robust than manual initialization because the nonlinear machine equations are solved consistently each time the system configuration or operating mode changes.
For more information, refer to the following documentation:
