Robust control of multi joint robotic arm visual servoing based on neural network and adaptive instruction filtering

This paper proposes a multi joint robotic arm visual servo robust control strategy based on neural networks and adaptive instruction filtering, aiming to solve dynamic uncertainty and nonlinear problems. By constructing an image plane feature point error model and deriving the image Jacobian matrix, a mapping relationship between image velocity and joint velocity was established. Using radial basis function neural network (RBFNN) to approximate the uncertainty of the dynamic model of the robotic arm online, the dynamic model error is reduced by 15% -20%. Designed an adaptive gain instruction filter, reducing filtering error by about 20% and computational complexity by 10% -15%. Experiments have shown that the tracking error (RMSE) is reduced by 15% -25%, and the tracking success rate is increased to over 90%. The anti saturation mechanism effectively ensures system stability under actuator constraints.