Temperature Control and PID Parameters Optimization Based on Finite Element Model

Proportional-integral-derivative (PID) controllers have been widely used in temperature control system. It is difficult to obtain good control performances if the fixed PID parameters are used. The adaptive control schemes are complexity in practice and difficult to obtain suitable results for strong nonlinearity systems. The learning cost of neural networks and genetic algorithms are considerably large. Model-based control methods facilitate the improvement of system performance, but it often uses first-order or second-order plus dead-time models. It is not adequate for high accuracy application due to the model´s approximate. Therefore, an approach using finite element model to depict the heater unit is proposed in this paper. The finite element model is embedded in the simulation model of the thermal system. And a thermal system used in electronic packaging equipment was used as illustration. An optimal set of PID parameters is obtained based on the simulation model of the thermal system. The results show that the proposed method supplies better performance and high accuracy. The example is a generic application and the method can be used in similar situations.