In the modern electric vehicle industry, the fast charging of lithium-ion batteries is essential. Charging at a high Crate minimizes the charging time; however, this results in degradation due to the rapid rise in the temperature and voltage. Therefore, we propose an advanced charging protocol that reflects degradation conditions by integrating multistage constant current-constant voltage and pulse protocols. The proposed protocol was efficiently evaluated using a high-fidelity lithium-ion battery model based on porous electrode theory. However, this model-based optimal protocol design has challenges regarding the expansion of the design space based on the increasing number of parameters and lack of information about the degradation conditions. A Bayesian optimization was applied to perform sample-efficient optimization without using a first-principles model and to incorporate the variability of battery cells into a stochastic prediction model. The charging protocol design guidelines identified in the nai ̈ve design space are used as prior knowledge to improve the efficiency of charging protocol design considering degradation. As a result, advanced protocols that suppress degradation and allow for minimized charging times compared to reference protocols (i.e., multistage constant current-constant voltage) are given as a Pareto front.