Inclusion of inter-temporal constraints into a distributed Newton-Raphson method

Newton-Raphson based methods are widely used for solving Optimal Power Flow (OPF) problems. Convergence can be sensitive to the starting point of the algorithm, the step size, and the condition number of the Jacobian. The inclusion of inter-temporal constraints, i.e., constraints that link successive time steps in the optimization, can in certain cases cause the Jacobian to become singular and Newton-Raphson to diverge. These cases occur when the binding inter-temporal constraints do not fulfill the Linear Independence Constraint Qualification (LICQ). In this paper, we discuss the conditions under which this happens, and analyze when singularities occur in a particular storage device model test case.