Quality-Cost Trade-Off on Constructing Logical Views for Vehicular Cyber-Physical Systems: a Deep Reinforcement Learning Approach