Neural network control of a pneumatic robot arm

A neural map algorithm has been employed to control a five-joint pneu-matic robot arm and gripper through feedback from two video cameras. The pneumatically driven robot arm (SoftArm) employed in this inves-tigation shares essential mechanical characteristics with skeletal muscle systems. To control the position of the arm, 200 neurons formed a net-work representing the three-dimensional workspace embedded in a four-dimensional system of coordinates from the two cameras, and learned a three-dimensional set of pressures corresponding to the end effector posi-tions, as well as a set of 3×4 Jacobian matrices for interpolating between these positions. The gripper orientation was achieved through adaptation of a 1 × 4 Jacobian matrix for a fourth joint. Because of the properties of the rubber-tube actuators of the SoftArm, the position as a function of supplied pressure is nonlinear, nonseparable, and exhibits hysteresis. Nevertheless, through the neural network learning algorithm the position could be controlled to an accuracy of about one pixel (∼3 mm) after two hundred learning steps and the orientation could be controlled to two pixels after eight hundred learning steps. This was achieved through employment of a linear correction algorithm using the Jacobian matrices mentioned above. Applications of repeated corrections in each position-ing and grasping step leads to a very robust control algorithm since the Jacobians learned by the network have to satisfy the weak requirement that the Jacobian yields a reduction of the distance between gripper and target. The neural network employed in the control of the SoftArm bears close analogies to a network which successfully models visual brain maps. We conclude, therefore, from this fact and from the close analogy between the SoftArm and natural muscle systems that the successful solution of the control problem has implications for biological visuo-motor control.