Abstract: About convergence rate and solution precision are not high in high-dimensional Nonlinear Optimization Problem (NOP), an improved Artificial Bee Colony (ABC) optimization algorithm is proposed in this paper.Firstly, the orthogonal experimental design algorithm was used to generate initial population and discover a new food source for the scout;Secondly, employed bees uses Gaussian Distribution Estimate Algorithm (GDEA) to search, according to fitness value, onlooker bees select one employed bees and search new nectar source in an self-adaptive differential search algorithm.At last this algorithm is tested on 4 standard benchmark functions, and the experimental results show this algorithm has some advantages in convergence velocity, solution precision, and stabilization.