Modeling forward kinematics with neural networks allows for efficient handling of nonlinear relationships and realistic error correction in time-critical applications by relying on accurate training data. This paper presents a Multi-Layer Feed-Forward Neural Network (MLFFNN) to solve the forward kinematics of a 3-DOF robot. The proposed MLFFNN consists of 50 hidden neurons and was trained using 628319 samples to find only the position (x, y, z) of the end-effector. Data were generated by MATLAB, assuming an incremental motion of joints. The joint variables ( , , and ) are the inputs of the NN, which outputs the positions of the end effector (x, y, z) calculated using the Denavit-Hartenberg (DH) method. The results demonstrate that the proposed MLFFNN has high performance and is efficient for solving the forward kinematics, with a Mean Squared Error (MSE) between the desired and estimated position of 4.3881×10-11. This performance clearly demonstrates that, despite the large size of the dataset, it can be effectively mastered with only a small number of neurons. The simplicity of the network allows it to learn a compact and efficient representation of the data. This improves the reliability of using the proposed network for similar applications in other robotic systems.
