Numerical method for the solution of the singular riccati matrix
M. B. Subrahmanyam · Optimal Control Applications and Methods · 1989
Abstract A computational scheme is presented for the solution of the Riccati matrix associated with time‐varying quadratic regulation. It is shown that the local truncation error of the algorithm is at most O(Δt3) where Δt is the interval of computation. The main contribution of this short communication is to demonstrate that the scheme yields the solution even in the face of singularities. Higher‐order formulae are useful when the sensitivity of the Riccati matrix to error is large.