Trajectory Planning & Computation of Inverse Kinematics of SCARA using Machine Learning

In this paper an algorithm is developed for smooth trajectory with minimum jerk using Cubic-B spline and intelligent computation of inverse kinematics using different machine learning algorithms for a SCARA robot for performing pick &; place/ assembly operations. Static Obstacle is considered in the robot environment. Machine Learning Algorithms like Linear Regression (LR), K-Nearest Neighbors (KNN), and Artificial Neural Networks (ANN) are used to prevent the difficulty in computing inverse kinematics in trajectory planning. It is observed that K-Nearest Neighbor (KNN) algorithm residuals plots have better fit by comparing with linear regression and ANN. The difference between actual and predictions of KNN, gives best results as compared to LR and ANN. Therefore, KNN can be used for inverse kinematics of SCARA robot for high accuracy and fast solutions. Cubic Spline functions are used to obtain the minimum jerk for the robot path.