In order to solve the problem that most of the current construction plan optimization methods ignore the impact of the randomness of the construction progress indicators of each key node on the construction plan optimization, and at the same time, it is difficult for the attribute weight determination method to achieve the overall optimization of high-dimensional and multi-level characteristic problems. This paper proposes a stochastic and intelligent optimization method for time-series multi-attribute construction schemes of high arch dams based on the prospect theory (PT) and the sparrow search algorithm (SSA). According to the construction characteristics of high arch dams, a time-series multi-attribute stochastic intelligent optimization model is established based on PT, and a dynamic reference point setting method is built based on the stage development characteristics. The optimal attribute weight and stage weight search model is established based on SSA, and the fitness function is constructed with the idea of maximizing differences to realize the model solving. The results of engineering application show that the proposed method is consistent with the optimization results of the prospect stochastic dominance-CRITIC method (PT-CRITIC) and the stochastic dominance method, and the proposed method has more advantages in scheme discrimination than the other two methods.