Tip-position Control of a Two-segment Flexible Robot based on GABP Neural Network

The actual bending shape of the rope-driven spine-like continuum mechanism deviates from the standard arc shape, and it is difficult to establish an accurate motion control model. In order to solve the problem of precise control of the end position of the two-stage flexible robot by fusing genetic algorithm and backpropagation neural network GABP, this paper studies the problem of precise control of the end position of the two-stage flexible robot. Through the joint simulation of Solidworks and ADAMS, the mapping database of the X, Y, and Z coordinates at the end of the flexible robot and the tensile length of the six soft axes were established, and the standard backpropagation BP neural network and GABP neural network models were trained by using the database, and the model parameters were optimized. Finally, the end motion trajectory of the robot was designed, and the BP and GABP neural network models were used to verify the position accuracy of the end of the flexible robot. The results show that both the standard BP and GABP models can realize the position control of the end of the flexible robot, and the position accuracy of the neural network model fused with genetic algorithm (maximum error: ∈_{{x}}=0.73 {mm},∈_{{y}}=0.78 {mm},∈_{{z}}=1.82 {mm}) in controlling the motion of the flexible robot has been significantly improved compared with the standard BP neural network model (maximum error: ∈_{{x}}=3.13 {mm},∈_{y}=1.78 {mm},∈_{{z}}=1.95 {mm}).