Repetitively structured cascade neural network model which generates an optimal arm trajectory

The trajectory along which a human moves his arms tends to minimize the time integral of the square of the rate of change of torque. A neural network model for trajectory formation, based on the minimum torque change criterion is proposed. The neural network first acquires a forward dynamics model of a controlled object by training, and then it calculates the motor command by a relaxation computation, utilizing the learned forward dynamics model. The model can be applied to many types of ill-posed motor control problems, for example, inverse kinematics and inverse dynamics for redundant controlled objects