DECISION-MAKING MODEL FOR CONTROLLING A COLLABORATIVE ROBOT-MANIPULATOR BASED ON THE SENSOR FUSION METHOD AND CNN APPROACH TO RULE FORMATION

The article discusses an approach to building a decision-making model for controlling a collaborative robot-manipulator, based on the integration of the Sensor Fusion method and convolutional neural networks (CNN). The proposed model provides the combination of multi-channel sensory information, its normalization and transformation into a feature vector for further training of the control system. The implementation of CNN allows you to automatically detect relevant input data patterns and form adaptive rules for decision-making in a dynamic environment. The simulation results showed high efficiency of the proposed approach in conditions of variable sensory signals and limited training sample. The model can be used in autonomous control systems for collaborative robots interacting with humans in a production environment.