Development of the En Masse Elimination Algorithm for Actuator Grouping Optimization

As large arrays of actuators become increasingly common in adaptive structures, many systems do not have enough power supplies to control each actuator individually due to design, weight, or cost constraints. This issue can be addressed by grouping multiple actuators together and powering each group with a single power supply. As this is done, it is important to determine which actuators to group together, that is, optimize the groupings for best performance. Currently, GAs and other heuristic algorithms are used to determine this grouping, where a global optimum or even acceptable solution is not always guaranteed. In order to effectively group the actuators and insure that the global optimum is found, a new method is developed in this research - the EME technique. This is an efficient algorithm that will locate the global optimum without having to examine every possible grouping combination. In this article, a detailed explanation of the EME algorithm is given and the efficacy of this method demonstrated. An example consisting of a beam with static deformation, multiple actuators, and different constraints on the number of power supplies is used to illustrate the concept. Utilizing an instrumented beam test stand, experimental investigation, and validation of the proposed method are performed.