Traditional single positioning system (such as GPS PPP) can not deal with some complex environment (such as urban canyon). In some complex environments, the distribution of satellite is poor and the number of available satellites is relatively small. The accuracy, availability and reliability of satellite navigation and positioning are directly related to the number of satellites tracked. Therefore, in the harsh environment such as urban canyon, the positioning accuracy has to be improved greatly. Currently the Beidou navigation satellite system (BDS) has been developed, and can solve these problems combined with other GNSS. Based on this, this paper implements GPS/BDS dynamic precise point positioning using Square Root Information Filtering method as parameter estimation method, evaluates the dynamic positioning performance in complex environment, and compares it with a single GPS or Beidou system, including static simulation of dynamic (different satellite cut-off angle), short distance walking gradually from open ground to indoor, vehicle urban environment (under trees, high-rise block road and relatively open road section) .The results show that with the increase of the satellite cut-off angle, the GPS/BDS positioning advantage is more obvious. When the satellite cut-off angle is 40°, it is difficult for a single GPS or Beidou system to maintain high precision positioning, while the GPS/BDS combined system can still maintain the positioning accuracy within 0.2m; In the walking PPP experiment, the GPS / BDS combined system has better positioning stability than GPS and Beidou single system, and the positioning result does not fluctuate much even when it approaches the indoor quickly; In the vehicle PPP experiment, the 3D position error of GPS/BDS combined system is improved by 31.8% and 42.6% compared with a single GPS or Beidou system in relatively open road section, and 68.4% and 73.6% respectively in the section with a signal occlusion. Compared with a single GPS or Beidou system, the GPS/BDS combined system can improve the positioning accuracy greatly when the vehicle is driving on the road with the signal occlusion than on the road without occlusion, which indicates that the combined system PPP is very suitable for solving the problem of poor positioning accuracy of the single system PPP when the vehicle is driving in the area with the signal occlusion.