Clutch friction work and vehicle jerk were synthesized to evaluate the shift process performance of dry dual clutch transmission (DCT). The entire shift process was analyzed and divided into 4 sections by the key points which were found based on the synthesized criterion. A control algorithm including proper control methods in every respective section was designed to achieve optimization of the entire process. Especially, the micro-slip control method was applied in the torque swap section and the discrete dynamic programming method was applied in the speed regulation section. At last, the algorithm was applied in the experimental vehicle and the experiment result showed the effectiveness in improving shift performance of DCT. Compared with the traditional control algorithm, the results prove that the new algorithm can decrease vehicle jerk during the entire shift process without producing more clutch friction work.