A neurogenetic model of muscle EMG to torque

The major problems of neural network applications include optimizing architecture selection and network parameter selection. In this study, a neurogenetic model of muscle EMG to torque for a motor control task utilizes genetic algorithms to evolve the optimal neural network genetic algorithms (GA) which are highly powerful and robust search mechanisms that solve optimization problems. The GA evolve the best population of neural network solutions by optimizing the inputs, architecture, connection weights and biases of the neural network. Comparison of the neurogenetic solution with a backpropagation neural network solution from previous work is presented in the analysis.