ИСТИНА |
Войти в систему Регистрация |
|
ИПМех РАН |
||
In the talk we consider a regularization scheme based on Newton-Raphson method for minimizing squared norm of the error of the load flow equation system. A solution to the system does not always exist for a given nodal load profile. If solution does not exist Newton-Raphson method applied to load flow equations will diverge. At the same time, since there is no a simple apriori criteria of solution existence, the divergence of the method does not necessarily imply the absence of the solution. Instead of finding a root to the system of equations one might reformulate the problem to minimizing the norm of error of load flow equations. However, due to non-convexity of the goal function the convergence of the method might be still problematic. The other practical issue is that once the convergence fails a certain guidance is required as to identification or localization of the source of the problem in the energy system that caused the convergence failure. We propose a regularized Newton-Raphson method and study the character of the stationary points of the method and profile of the nodal loads from the perspective of possibility of problem spot localization and providing load corrections that bring the system to the feasible region