Mobile Robot Transition Dynamics Dataset for LSTM-RCM Driven Delay-Compensated Modeling

This dataset contains sequential transition data collected from a mobile robot platform for training machine learning models that predict future states based on current control inputs and previous states. Each sample is structured as a (previous state, control input, next state) triplet, supporting supervised learning of system dynamics.It is specifically intended for learning predictive models used in time-delay control systems. The dataset was used to train Long Short-Term Memory (LSTM) networks and Sparse Variational Gaussian Processes (SVGP) within an LSTM-RCM Driven Smith Predictor framework, which compensates for network-induced latency in mobile robot path tracking.