The self-learning CMAC control of the oil temperature in hydraulic system

In view of the large time delay of the temperature control system of hydraulic system, a self-learning cerebellar model articulation controller (CMAC) algorithm is proposed to control the cooling system. The temperature control precision exceeds the Grade A of ISO standard (50±1°C) and reaches the level of 50±0.5°C. The cooling system is composed of digital and proportional controllers and in hybrid form. This can increase the control precision. Considering the speciality of the cooling system, an inverse dynamics model different from the one commonly used is established. The computing speed of the self-learning CMAC is fast and thus can be used online. Different learning rules are used for different error bands to meet the requirement on speed of response and steady state precision. The experimental result shows that the proposed self-learning CMAC control can be effectively used in the temperature control of hydraulic systems