As its sampling frequency is independent of the scale of the system and non-linear performance, Monte-Carlo simulation is widely applied to reliability evaluation of a complex large transmission system because of its accurate result. However, it takes a relatively long period of time to obtain a higher accurate reliability index. In order to solve this problem, the combination of importance sampling method and discrete Latin hypercube sampling method is able to reduce the variance of the sample space and uniformly sample it simultaneously, which improves the convergence of Monte-Carlo simulation. This paper points out that, the combination of sensitivity analysis and linear programming in the process of overload correction is able to guarantee the solution optimality as well as improve the solving speed. Finally, the algorithm is presented for IEEE RTS79, IEEE RTS96 and a certain system to obtain the reliability index, the result proves its practicability. This work is supported by National Natural Science Foundation of China (No. 51337005) and Fundamental and Prospective Science and Technology Project of State Grid Corporation of China (No. XT71-14-002).