This paper uses the classic reactive optimization model and takes the minimum of net loss as the objective function. According to nonlinear complementary theory, a NCP function is constructed. In the method, the inequality constraints in the Karush-Kuhn-Tucker (KKT) conditions are transformed to be equivalent smooth nonlinear equations, and solved by Newton method. Discrete variables are handled in the same way, namely, constructing a discrete NCP function equivalent to discrete variables constraints and conducting an iterative calculation through Newton method. Finally, the classic IEEE calculation results show that the proposed algorithm has a good convergence speed like traditional algorithm. It reduces net loss effectively and has large convergence region.