The invention provides a network loss equivalence load model-based power flow calculation initial value setting method, which comprises the steps of carrying out network loss equivalence load model-based DC power flow iteration calculation on a power grid to obtain a DC power flow voltage phase angle phasor theta* of a power grid node; by employing the theta* as an initial value of a node voltage phase angle, calculating a voltage amplitude estimation value phasor V* and a phase angle estimation value phasor theta* of the node through a quick decoupling method; and evaluating the voltage amplitude estimation value phasor V* and the phase angle estimation value phasor theta*, calculated through the quick decoupling method, of the node to obtain a newton method power flow initial value. The network loss equivalence load model-based power flow calculation initial value setting method is conducive to improvement of the power flow convergence and has the advantages that DC power flow calculation is free of a convergence problem and calculation through the quick decoupling method is relatively simple.