Ray. Thank you for your responses. The "match" in my last email means one measurement in z should have a corresponding sub-matrix of the measurement Jacobian H, when compute F = H'*R_inv*(z-z_est) . But the bus power injections and the partial derivatives of it don't have the same form, the former includes the local PD and QD while the latter doesn't. May the "the partials of those terms are zero" is the reason, is it? Thanks again. Ji
At 2015-04-23 23:18:15, "Ray Zimmerman" <[email protected]> wrote: See my responses below … On Apr 23, 2015, at 5:15 AM, 陈吉 <[email protected]> wrote: When compute net injection at generator buses in the m file doSE.m line 107-110, the local PD and QD are include. %% compute net injection at generator buses gbus = gen(:, GEN_BUS); Sgbus = V(gbus) .* conj(Ybus(gbus, :) * V); Sgen = Sgbus * baseMVA + (bus(gbus, PD) + j*bus(gbus, QD)); %% inj S + local Sd While when compute partial derivatives of the complex bus power injections in dSbus_dV.m line 9, local PD and QD are unincluded. % S = diag(V) * conj(Ibus) = diag(conj(Ibus)) * V This is because the loads, PD and QD are assumed to be constant with respect to voltage (i.e. the partials of those terms are zero). Can the sub-matrix of z_est real(Sgen(idx_zPG)) and the sub-matrix of H dPG_dVa(idx_zPG, nonref) dPG_dVm(idx_zPG, nonref) match? I’m not sure what you mean by “can … match”? The former is a vector, the latter is a matrix. Best regards! Jimmy
