The Application About Modified Differential Evolution Algorithm And Fractional Model In Model Predictive Control Of The Counter-current Tubular Heat Exchanger

As we know, the traditional model of counter-current tubular heat exchanger has low model precision, model mismatched when a violent change occur in the environment. To solve this problem, we promote a new model-fractional model. At the same time, we modified differential evolution algorithm to enhance its performance.Here are the main contain: modifying differential evolution algorithm and making simulation in MATLAB, building new model about the counter-current tubular heat exchanger, combining new algorithm and model in the control of the heat exchanger.1) Building a fractional model of the heat exchanger through thermal theory. Experiment results show higher precision and robust when we use a step input signal. It has a good model precision when environment changed roughly.2) Using nonlinear model predictive control as our control method, with using modified differential evolution algorithm to search the best solution online. We modify the algorithm through modifying its parameter online to get a good performance.3) At last, we build a simple platform to test and verify new model using new algorithm. We build a Weiner model by using the data which is got by the simple platform. Experiment result shows a higher precision compared with traditional model.4) We use a new control strategy that has a dynamic predictive time and input constrains. The strategy has a bang-bang control structure, when the error is bigger than a constant, we use larger predict time and input constrains, on the contrary, we use smaller predict time and input constrain. It shows a much better effect in the experiment. The system show a good performance with no output shock.