I am using navmesh graph consisting of triangles. To find the shortest path from start to goal point, I have taken into consideration A* algorithm. Though I want to perform a rough preliminary kind of search to find the path from start to goal, it becomes really cumbersome needing a lot of calculation.
While calculating fitness(f) from goal(g) and heuristic(h), most tutorials tell you to calculate 'h' by the difference in distance between current position and goal and then find the mesh/triangle which gives you least value of 'f'. For a path like a semi-circle with start and goal at the opposite edges of diameter, least 'f' will be towards the direction of goal which will lead to a dead end. Hope you understood what I`m saying. So, least value of 'f' is not always the right one. Consequently, you probably need to calculate each triangle connected to every other and check until it leads to goal.
Is there a better way to calculate heuristic? Is there a better pathfinding algorithm for navmesh?
