Jacobian Voltage Stiffness Metric -- A Measure of Grid-Forming Capability and System Strength in IBR-Dominated Grids

What the math gives to ML

The paper proposes the Jacobian Voltage Stiffness Metric (JVSM), a frequency-domain sensitivity measure that quantifies how strongly voltage magnitude and phase respond to active- and reactive-power perturbations. Its transferable mechanism is to evaluate local input-output stiffness through a structured Jacobian rather than relying only on time-domain transients or scalar gain measures. In neural networks, the most promising use is a phase-aware Jacobian certificate for recurrent, state-space, implicit, or complex-valued models: monitor and constrain magnitude and phase sensitivity separately, with a measurable transition when the largest local gain exceeds a prescribed stiffness threshold.