In order to improve the convergence accuracy and speed of path optimization by ant colony algorithm, a grid map based on effective turning point and an ant colony algorithm based on the shortest distance and minimum number of steps path (shortest-minimum path) were proposed to search the shortest path from the beginning to the end. In the standard path planning of ant colony algorithm, the searching method of ants is finite neighbor with finite direction. A searching method with infinite neighbor and infinite directions was proposed, which can take a shortcut to any grid point that can be directly connected. The concept of effective turning point was put forward to reduce the single-step searching amount: firstly, based on the relationship between shortest path and obstacles, two windows were used to scan the whole map to find all the inflection points, and then a through-through tree was established from the end point to the starting point according to the hierarchical relationship, and only the points within the through-through tree were reserved as the effective inflection points. The concept of shortest-minimum path was proposed, which was used to replace Euclidean distance as heuristic value to improve the accuracy and reliability of heuristic value. At the same time, the shortest-minimum distance from the beginning to the end was used to guide the updating of pheromone and improve the quality of ant colony algorithm iteration. Finally, in the grid map environment with different scales and different proportions of obstacles, the rationality of using the shortest-minimum path distance to replace the Euclidean distance was proved through experiments, and it was verified that the method presented can search the path with shorter distance and fewer steps with faster convergence speed while reducing the calculation cost.