Identification of linear systems by poisson moment functionals in the presence of noise

THE parameters of several increasingly complex linear systems are identified in the presence of additive noise, using the method of Poisson Moment Functionals (PMF). Two formulations are examined. The first identifies the first order state equations of the structure from measurements of generalized displacements, velocities, and forces. The second identifies the second order equations of motion from measurements of the generalized displacements and forces. The initial conditions of the system are arbitrary and are identified. Comparisons are made with a second-order auto regressive moving average (ARMA) method.