In this paper, a master-slave synchronization scheme based on parameter identification is proposed to overcome the controller singularity problem that appears when linearization-like techniques are applied in indirect adaptive neural control, like Neural Block Control (NBC). Such a synchronization strategy requires an identifier-like recurrent neural network and an adaptive law to update the neural weights. The proposed adaptive law prevents both, specific adaptive weights zero-crossing and the 'parameter drift' phenomenon. NBC consists of two tasks; synchronizing an identifier-like recurrent neural network (slave) with the plant (master) and controlling the system based on the slave model. The effectiveness of the synchronization law is tested using NBC for controlling the angular speed and magnetic flux magnitude of an induction motor. Using a priori knowledge about the real plant, a high-order recurrent neural network is proposed as the slave system. Based on the slave neural model, a discontinuous control law is derived, which combines Block Control and Sliding Modes. NBC with the proposed synchronization strategy is tested via simulations, comparing results with a standard parameters adaptive law.