Oil-heater Temperature Control System Based On Neural Network

Oil-heater is applied extensively in industry production. The control result of temperature control system will directly affect quality of products and efficiency of production. so the exact precision is demanded. At present conventional PID control method is used usually by most domestic oil-heater in industries. The temperature control for oil-heater has many characteristics such as non-linearity, big inertia, great lag. It is difficult to establish its accurate mathematic model so we can't obtain the satisfied result using the conventional PID control method. Therefore, it is very valuable to design the temperature control system which has exact precision and steady operation.This paper focuses on how to apply algorithm to control the temperature and uses PID neural network which is more difficult and excessive neurons network structure. Because neural network has the characteristic and advantages which dose not rely on controlled object model, non-linearity, big inertia, great lag controlled object achieves corking control effect.Industrial Computer controling the oil-heate control system as a platform to build the control model of control system. On this basis, use of two control points principles and experimental methods.Compared the effect of traditional PID control and PID neural network control, at the certain model has lag time changing and first-order coefficient changing. These experiments indicate that the control effect of PID neural network is better than the control effect of traditional PID. Particularly in the condition of lag time and first-order coefficient changing, traditional PID will lose control function, but the PID neural network can still remain stable and obtain corking control effect. Expriments prove that the neural network PID has the advantages of no overshoot, good stability, strong self-adaptable aiming at control system and good application prospects.