You are on page 1of 1

 Summary

RPO is a method of optimization based on minimization of error. It is Heuristic-Based technique and uses
Artificial Potential Fields (APF’s) & bacterial foraging for collision avoidance. In this paper, RPO is applied
on a mobile robot. We use different techniques for path planning of a mobile robot where a mobile robot
moves towards a target avoiding obstacles. Obstacles can be fixed or moving and technique is depends on
static or dynamic environment. For example if we knows the coordinates of target and obstacles, graphical
techniques or classical techniques are used.

In this online algorithm, given the initial position of a mobile robot in a dynamic environment i.e. (moving
obstacles), mobile robot has to reach to the destination avoiding obstacles with minimum distance possible.

Given robots current position coordinates, mobile robot generates artificial point around it with a circle of
radius C (t) where C (t) is a fraction of sensors range. From each artificial point, the robot calculates the
distance from that point to the target position. If obstacle is not in the sensor range, robot moves to the
artificial point with lowest distance error and negative error cost function. Gaussian cost function employs
to every obstacle and the target. Every obstacle is given a repellent Gaussian cost function while target is
given attractant Gaussian cost function.

Jobs= αobs exp(-µobs * (Pi(t)-Po(t))1/2)

Jtar= -αtar exp(-µtar * (Pi(t)-PG(t))1/2)

Jtotal=Jobs+Jtar

Where α and µ represents the height and width of obstacles and goal. Pi(t) is position of artificial point,
Po(t) represents current position of robot and PG(t) represents goal position.

We calculates error function of distance and total cost function, minimum value of distance error and
smallest negative value of cost function is chosen while putting every artificial point and then robot moves
toward that artificial point and then this process is done for every step size until target is reached.

RPO does not stuck in local minima while moving towards target with smallest time possible thus solving
the short comes of Artificial Potential Fields.

You might also like