HYDROTHERMAL STOCHASTIC OPTIMAL POWER DISPATCH - A NEWTONS BASED APPROACH

A technique that deals with the power system hydro-thermal optimal power dispatch problem accounting for the effects of uncertainty in predicting active power demand is presented. The method relates the probability distribution of the uncertain factors in the system active power generation, bus voltages, and phase angles to the probability distribution of the errors resulting from the forecast of the active power demand. It is assumed that random errors in power demand are normally distributed with zero mean μΔPD = 0 and some variance σΔPD2 ≠ 0. The equality constraints considered are the power flow equations in the polar co-oridinates, using the nodal admittance matrix formulation. The iterative solution of the reformulated problem uses Newton's method and Powell's penalty function approach. Test results for the IEEE hypothesised 14-bus, and 118-bus hydro-thermal power systems are presented.