With the goal of improving service levels, given the known origin destination(OD) distribution of bicycle trips in a specific area, an optimization method for urban bicycle network parameters mainly referring to bicycle lane spacing and bicycle lane width was proposed. Several test scenarios were built by different combinations of key design parameters, and Cube simulation model was used to test and analyze the effects of these scenarios. Level of service indicators such as total travel time, average speed and road saturation were obtained. The recommended key design parameters can be obtained by comparing the results. Multivariate analysis of variance was used to compare the effects of two key parameters. It is found that increasing lane width could significantly improve the level of service of bicycle network. Finally, the planning and design strategy for improving bicycle travel conditions is put forward.