With the development of information technology and the scale of aquaculture, the traditional breeding mode will gradually be replaced by intelligent aquaculture. This paper puts forward an intelligent control system about dissolved oxygen in culture of Eriocheir sinensis based on Zigbee, and focuses on the study of the crab pond aeration control algorithm, aiming at the characteristics of nonlinear and dissolved oxygen changes large lag, uncertainty, big inertia, the PID control strategy is improved by fuzzy RBF neural network, and through the Matlab, RBF fuzzy neural network controller are simulated and compared with conventional PID controller and fuzzy PID controller,in the same time, fuzzy PID controller and the application of experimental comparison of fuzzy RBF neural network controller in a Chinese mitten crab breeding base. It prove that the fuzzy RBF controller can meet the control of Chinese mitten crab breeding oxygen requirements.