MODELING AND OPTIMIZATION OF FAST-ACTING ELECTROMAGNETIC ACTUATORS

Fast-acting electromagnetic actuator designs are optimized numerically by combining minimization techniques with both lumped-parameter and distributed-parameter models. A lumped-parameter (equivalent-circuit) model that accurately simulates dynamic actuator response is developed and experimentally verified. This model accounts for magnetic saturation, flux leakage, eddy currents in solid and laminated regions, and the dynamic coupling of actuator state variables. Two-dimensional distributed-parameter (finite-element) models are developed and used to predict linear and nonlinear steady-state actuator performance and to estimate the airgap and leakage reluctances in a hybrid, lumped/distributed actuator model. An improved constrained optimization technique is then presented which requires substantially fewer function evaluations to minimize or maximize an unknown function than do most other optimization techniques. The performance of this new technique is compared to the performance of a wide range of existing methods using nine published test functions. Finally, the new optimization technique is coupled with the hybrid actuator model to optimize dynamic actuator performance. Three examples of constrained actuator optimization are given where up to eight design parameters are optimized. One example problem is to minimize the variation in the response time of a fuel-injection control valve caused by assembly tolerances. Two other examples deal with minimizing the response time of a high-speed gas sampling valve.