The application of neural networks to excitation control of a synchronous generator is considered here. A radial basis function (RBF) network was constructed using a hybrid training algorithm, combining linear optimization of the output layer weights with singular-value decomposition, and non-linear optimization of the centres and widths using second-order gradient descent BFGS. The Jacobian of the RBF network was calculated to provide instantaneous linear models of the plant, which were then used to form linear controllers. Generalized minimum variance, Kalman, and internal model control schemes were implemented on an industry-standard VME platform linked to a network of Inmos transputers, and the performance of the neural models and neural control schemes were investigated on a 3 kVA laboratory micromachine system. Comparison was made with a self-tuning regulator, employing a generalized minimum variance strategy. The results presented illustrate that not only is it possible to successfully implement neural controllers on a generator system, but also their performance is comparable with a benchmark self-tuning controller, while avoiding the significant supervisory code needed to ensure robust operation of the self-tuning controller.