With the integration of Distributed Generations (DGs) into the distribution network, the traditional single-source radial network is transformed into a complex network with multiple power sources. Traditional fault location methods are difficult to adapt to the construction of complex operating conditions. In the paper, a novel fault segment localization method based on the Improved Binary Snake Optimization (IBSO) algorithm is proposed. The SPM chaotic mapping is used to generate high-quality random number sequences, which improves the randomness of individuals in the algorithm population. A dynamic mutation strategy based on genetic algorithms is also introduced. According to different search states and evolutionary stages, the mutation rate and mode are dynamically adjusted. The flexibility and accuracy of the algorithm are improved significantly. The simulation results show that the proposed method is suitable for locating single and multiple fault sections in distribution networks with distributed generation. It has better convergence, speed, and accuracy compared with the snake optimization algorithm, traditional binary particle swarm algorithm, and genetic algorithm.