External disturbance identification of a quadruped robot with parallel–serial leg structure

External disturbances can greatly influence the walking and working performance of legged robots. Identification of external disturbances is the basis of legged robots’ ability to adapt to a given environment. However, there are no appropriate sensors to measure the location, direction, and magnitude of external disturbances. In this paper, an indirect method is presented to identify external disturbance forces and torques for a high-payload quadruped robot. To improve carrying capacity, the robot leg is designed as a parallel–serial mechanism. This leg design ensures that there are only mechanical parts on the lower half of the robot. To increase anti-radiation capability, no sensors are installed on the robot feet. The robot is driven by twelve hydraulic actuators. A dynamic model has been proposed to identify external disturbances using hydraulic actuation forces and robot inertia. The obtained disturbances are expressed as a 6-DOF force screw, including forces and torques. The calculation process is simple because the equations can be explicitly solved without differential calculus. Simulations were performed to test the effectiveness of the identification model. A pulling experiment was carried out to evaluate the identification error. Finally, as an application, the proposed model was used to recognize changes in additional weight during crawling gait walking.