The invention discloses a robot autonomous navigation control method and system based on deep learning. The method comprises the following steps: setting a starting point and an end point of a robot in an environment map, and calculating the optimal path that the robot can drive; adopting a positioning algorithm for determining the position of the robot at the moment in the driving process, stopping movement when arriving at the end point, and otherwise continuing navigation; acquiring an environment information image in front of the robot by using a camera, inputting the environment information image into the trained convolutional neural network model, judging whether an obstacle with an unknown environment map influences the normal movement of the robot or not, if not, continuing navigation according to the optimal route, and if yes, generating a corresponding control instruction by using the network model, enabling the robot to avoid an unknown obstacle, and at the same time, replanning the optimal route, determining the position at the moment, and judging whether an end point is reached or not is judged. The system is based on a deep learning technology, can avoid various obstacles on a route planned by a global path, and is high in applicability.