The chicken swarm optimization algorithm was validated using 15 complex functions, and the simulated results were compared with those of the Wolf algorithm, particle swarm optimization algorithm, fish swarm algorithm, and genetic algorithm. Using the chicken swarm optimization to search for the optimal projection direction of the projection pursuit model, a projection pursuit evaluation model was established. Using assessment of flood and drought disasters from 1990 to 2013 in Wenshan Prefecture as an example, five flood disaster evaluation indices, including disaster-affected population, and four drought disaster evaluation indices, including the disaster area of crops, were selected, and grading standards for flood and drought disaster assessment were constructed with the mean and standard deviation of the flood and drought disaster projection series. The results show that the chicken swarm optimization algorithm is robust and has high global optimization ability. Using the algorithm to select the optimal projection direction of the projection pursuit model can effectively improve the evaluation accuracy and prevent a high degree of variation from occurring in the optimal projection direction.