To reasonably plan a transmission network with high-speed railway load and wind and solar power stations in complex mountainous areas, this paper proposes a stochastic planning method for the network considering the uncertainty of that load and wind and solar power. First, a Laplace mixed model combined with a binomial distribution is used to model the intermittent and shock high-speed loads. Second, for the problem that the DC random power flow cannot take into account the voltage distribution, and the AC random power flow model is relatively complex, this paper proposes a semi-invariant random power flow calculation method based on a decoupling linear power flow calculation. From this, a transmission network planning model considering voltage deviation is constructed. Then, from the characteristics of high dimension of decision variables and complex constraints in the transmission network planning and problem solution, a genetic algorithm is improved by adaptively and dynamically adjusting the probability of crossover and mutation in the evolution process. Finally, a simulation study of the power grid along a railway in a high-altitude mountainous area is carried out to verify the correctness and effectiveness of the model and the solution algorithm.