With the technological innovation, public transit has truly entered the era of big data. Bus operators have been possible to monitor the system conditions and obtain the real-time transit demand. This situation inspires us to rethink the transit demand prediction and real-time control problems. This project proposes a proactive real-time control method based on data-driven transit demand prediction with multi-source traffic data. A proactive control strategy is to predict the possible disturbance in the future by monitoring and inferring the system operation, and takes measures in advance to prevent the disturbance from disrupting the service regularity. Firstly, the further service reliability is assessed based on the evolution of the latest service reliabilities, to justify whether to conduct control actions. Secondly, if a control action is required, predict the transit demand and the number of alighting passengers. Thirdly, according to the predicted results, the bus dispatching time is optimized by minimizing passenger waiting time. A calculation process is introduced to solve the problem and the effectiveness of the proposed method is evaluated with the data of a real transit route.