The authors describe a special type of dynamic neural network called the recursive neural network (RNN). The RNN is a single-input single-output nonlinear dynamical system with a nonrecursive subnet and two recursive subnets arranged in the configuration shown. The authors describe the architecture of the RNN, present a learning algorithm for the network, and provide some examples of its use.
Proceedings of the 1991 IEEE Workshop Neural Networks for Signal Processing 
Multilayer perceptrons, Neural networks, Nonlinear control systems
Abdallah, Chaouki T.; Don Hush; and B. Horne. "Recursive neural networks for signal processing and control." Proceedings of the 1991 IEEE Workshop Neural Networks for Signal Processing  (1991): 523-532. doi:10.1109/NNSP.1991.239489.