Heuristic Optimization Algorithms in Robotics

Today, Robotic is an essential technology from the entertainment to the industry. Thousands of articles have been published on Robotic. There are various types of robots such as parallel robots, industrial robots, mobile robots, autonomous mobile robots, health-care robots, military robots, entertainment robots, nano robots and swarm robots. So, this variety brings a lot of problems in Robotic. Inverse kinematic for serial robots, forward kinematic for parallel robots, path planning for mobile robots and trajectory planning for industrial robots are some of the problems in Robotic studied a lot. Some of the problems are solved easily with some mathematical equations such as forward kinematic problem for serial robots and inverse kinematic problem for parallel robots. But the problems consisting of nonlinear equations and higher order terms can’t be solved exactly with the classical methods. The forward kinematics problem of the 6 degrees of freedom (DOF) 6x6 type of Stewart Platform can be given as an example to such problems. It has been shown that there are 40 distinct solutions for this problem (Raghavan, 1993). Some of the unsolved problems with classical methods are optimization problems and heuristic optimization techniques are an alternative way for the solution of such problems. The heuristic optimization techniques produce good solutions for the higher order nonlinear problems in an acceptable solution time, when the problems aren’t solved with the classical methods (Lee & El-Sharkawi, 2008). In this chapter, first in Section 2, optimization is shortly described. The introduction and some well-known heuristic algorithms including, Genetic Algorithms(GA), Simulated Annealing(SA), Particle Swarm Optimization(PSO) and Gravitational Search Algorithm(GSA) are reviewed in Section 3. In Section 4, two well-known optimization problems in Robotic are solved with GA and PSO.