Gainswitching

Gainswitching is a control strategy that selects among a discrete set of controller gains according to the current operating condition or state of the system. It is related to but distinct from gain scheduling; while gain scheduling uses a parameterized model to compute gains as a function of operating point, gainswitching switches between fixed gains at runtime based on predefined criteria. It is applicable to nonlinear and time-varying systems.

Techniques for implementing gainswitching include threshold-based switching with hysteresis to avoid chattering, state observers to determine conditions, and supervisory logic that determines when to switch. Designers must consider dwell time, switching instants, and stability analysis, often using hybrid systems theory or multiple Lyapunov functions to guarantee stability across modes.

Applications include adaptive PID control in robotics, aerospace, automotive control, power electronics, and any scenario where system dynamics change with operating conditions. An example is a PID controller that uses a aggressive gain for fast response at high error and a conservative gain for precision near the setpoint, switching when error magnitude crosses a boundary.

Benefits include improved performance across operating regimes and robustness to parameter uncertainty; drawbacks include potential instability if switches are poorly designed, chattering, and added design complexity. Gainswitching complements other adaptive methods such as gain scheduling and model predictive control.