The inversion of dynamic systems

A development of Berman's method is described for computing the input of a known system given its output. The method is not restricted to linear systems and allows the state equations to be used in the normal manner, only the output equation or some derivative of it needing to be used in an inverse sense. Differentiation of the measured output is usually required, but this or some similar operation is unavoidable in this type of problem. The method can accommodate non-zero initial conditions and allows one to obtain confidence limits for the computed input.