Reflection.md

Reflection

In this section, I am describing the code model for generating paths.

Structure

A single header file named Vehicle.h was created for our solution. The solution is designed in such a way that the Car is an object which will interact with the simulator.

Initialization

The car is initialized in line 59 of main.cpp,

 Vehicle ego(1, 0, 0, 0, 0, 0, 0);

It is initializing the Vehicle object named ego with lane=1 , Cartesian co-ordinates (0, 0), frenet coordinates (0, 0) , yaw=0. As defined in line 18 of Vehicle.h.

The map waypoints are saved using set_map_waypoints method in line 60

Setting the parameters per iteration

In Line 110-114 the parameters of the car at each iteration along with the sensor fusion data from the simulator are set using the set_* methods.

Lane Cost

The way this solution was approached was that the car measures cost of each lane at each iteration. The cost is defined as

if lane is the same lane
    if the distance from the next car is less than 30 meters
        cost = - abs(distance)
    else
        cost = 0
else
    if the distance from behind and next cars is less than 15 meters
        cost = -abs(distance)
    else cost = 0

The method name is get_lane_cost() which is called in Line 135 of main.cpp . The method is defined in Line 129-148 of vehicle.h for all of the lanes. Cost per lane is calculated in Line 150-187 of vehicle.h. The costs of each lane is saved in the map member of the Vehicle object named lane_cost

Choosing the appropriate behavior

In Line 136-158 of main.cpp the behavior planning is done. The pseudocode is described below

If the cost of the current lane is negative
    if  the cost of the left lane is not negative
        change lane left
    else if the cost of the right lane is not negative
        change lane right
    else if all of the lanes' costs are negative
        slow down
    else
        speed up