Machining parameters are essential factors affecting the machining efficiency and tool life. Tool reliability varies with the process. Tool reliability affects the life of the tool, and then impacts the processing quality and manufacturing cost. Therefore, machining parameters optimization considering tool reliability is essential and scientific. In this paper, firstly the reliability model of tool life was solved by Markov Chain Monte Carlo (MCMC) method. Then taking the average tool life as the constrain condition, a multi-objective optimization algorithm that integrates the gray correlation analysis (GRA), radial basis neural network(RBF) and particle swarm optimization(PSO) algorithm (GRA-RBF-PSO) was used to search for optimal machining parameters of blisk-tunnel processing. At last, experiments were carried out to validate optimized results. The experimental results indicated that the reliability-based optimization of machining parameters can effectively improve the tool life and as well as ensure smaller cutting force and larger material removal rate during blisk-tunnel processing.