The invention belongs to the field of industrial robots, and particularly relates to an online
robot collision detection method and
system based on a
momentum model. The method comprises the steps that firstly, structure parameters are set, and a dynamics model is established and converted into a
robot state space equation; secondly, the angle measuring value and the torque measuring value of eachjoint of a
robot at the time t are monitored, and t is larger than or equal to 0; thirdly, the angle measuring value and the joint torque measuring value, obtained in the second step, of each joint are substituted into the robot
state space equation in the first step, and the external torque
estimation value of each joint of the robot at the time t is obtained; and fourthly, the external torque
estimation value of each joint and the corresponding external torque collision threshold value are compared, if the external torque
estimation value is larger than or equal to the external torque collision threshold value, it is judged that collisions happen, and if the external torque estimation value is smaller than the external torque collision threshold value, it is judged that collisions do not happen. According to the online robot
collision detection method and
system based on the
momentum model, the collision detecting
operand and the calculating time consumption are reduced, and the collision detecting precision is improved.