Research on latin hypercube random simulation of optional operation of complex hydropower stations

The hydrological forecast error inevitably leads to the uncertainty of reservoir inflow, so the optimal operation of reservoirs group is also uncertain. This paper presents a method which combines the random simulation with the optimal operation of reservoirs group. The random simulation uses the method of Latin hypercube sampling to improve Monte Carlo method, which reduces the sampling frequency and improves the sampling accuracy compared with traditional sampling method. Ant colony algorithm is found application for solving the optimal operation problem of reservoirs group. According to the method presented we can obtain the scheduling scheme of reservoirs group under a given risk rate. At the same time the method changes the decision-making of stochastic operation into that of determinable operation. The study case of giant hydropower stations in Hubei Province shows that the method is practical.