In the 21 st century, when all kinds of industries are developing into intelligence, the driverless technology, as a representative of the combination of IT industry and traditional automobile, has developed rapidly. The path planning of driverless cars has always been a difficulty in its development. From the perspective of simplification, the path planning problem of driverless cars can be transformed into a single target optimization problem, which is the shortest path. By using Floyd Algorithm, the path planning model of driverless car can be built on this basis. After analyzing the implementation mechanism of Floyd Algorithm, by means of matlab, the model is solved, and the shortest distance from the starting point to any intersection and the corresponding driving path are obtained. Finally, the advantages and disadvantages of the model are evaluated, and the improvement direction of the driverless car in the future is analyzed.