In the paper, a BP network in which the factors to effect groundwater levels are set as input neurons and the groundwater level is set as output neuron, was established to forecast groundwater level.

Combining it with the adaptive linear neuron network, the multiplex control strategy takes the dynamic errors and given signals of the system as input signals to the CMAC neural network.

As the input item varying, chaotic neuron status transforms between chaos and periods, which should be used to control the interior status of chaotic neuron.