Solving the Forward Kinematics Problem in Parallel Manipulators Using Neural Network

Abstract

A parallel manipulator is a closed kinematic structure with the necessary rigidity to provide a high payload to self-weight ratio suitable for many applications in manufacturing, flight simulation systems, and medical robotics. Because of its closed structure, the kinematic control of such a mechanism is difficult. The inverse kinematics problem for such manipulators has a mathematical solution but, the forward kinematics problem (FKP) is mathematically intractable. This paper presents a Stewart platform with asymmetric payload and proposes a neural-network-based strategy that solves the problem of FKP to a desire level of accuracy, and can apply the solution for any other desire trajectory. The Neural-network concepts with backpropagation learning is implemented. The inverse kinematic is used to obtain the required force for desire trajectory. Then, neural network is taken into account to take actuator forces as inputs and train the system in order to reach the desire path. It is concluded that, neural network results are acceptable and can be applied for any desire path.