Generator Maintenance Scheduling And Its Improvement Based On Credibility Theory

With the deepening of marketization reform of electric power industry, the traditional operation and management mode of power system has been challenged ceaselessly. It needs to improve to adapt the new changes in the situation. The unit maintenance scheduling is closely related to the economic benefit for the power companies. Meanwhile, it has profoundly affected reliability of the power system. Therefor, it needs to arrange the unit maintenance scheduling properly for keeping balancing between power companies and the stability of power system. There is fuzzy uncertainty and stochastic uncertainty in the unit maintenance scheduling optimization problem. Previous studies concerns more on stochastic uncertainty. But fuzzy uncertainty is a factor that can’t be ignored. Especially the new energy include wind power is introduced into power system. Then, it needs to consider fuzzy uncertainty and stochastic uncertainty to optimize the unit maintenance scheduling.The unit maintenance scheduling is a mixed-integer problem with fuzzy uncertainty and stochastic uncertainty. This article builds the model for the fuzzy-random uncertainty of the unit maintenance scheduling based on Credibility Theory. The objective function is maken sure to achieve economic optimum. NSGA-II algorithm is introduced into this problem for speeding up the solving process. This article proposes an approximate simplification algorithm to work out the stochastic expected value of the irregular fuzzy number and speed up the solving speed. It takes IEEE-RBTS and IEEE RTS-96 system as example to prove that this algorithm is feasible.Otherwise, the uncertainty of wind turbine itself will influence reliability of power system. Previous studies pay less attention on fuzzy uncertainty. So, this article build the model to research the unit maintenance scheduling with wind turbine based on Credibility Theory and test in the IEEE-RTS system. It proves that the fuzzy uncertainty and stochastic uncertainty of wind turbine should be considered in the unit maintenance scheduling, which can have better result of economy and reliability and arrange the unit maintenance scheduling more efficiently.Based on the algorithm proposed in this article, a reliability evaluation software is developed with serial and parallel operation mode under QT. This software is used in Guangdong power grid with simulating the large scale wind farm offshore, which proves that the method in this article is practicability.