Research Of Model Predictive Control Method In Nuclear Reactor Core Temperature Control System

In the actual industrial processes, many objects have the characteristics which have constraints and are nonlinear when in a wide range change conditions, are difficult to implement effective control when using the traditional control methods. With the rapid development of the nuclear power industry and people's concern on nuclear safety, increasing high demands on the nuclear power plant safety control and effective control.By analysis the operation principle of the nuclear power plant, this paper established the reactor related system mechanism models, combined with the operating data of 1000MW Three Mile Island PWR, then gives the lumped parameter nonlinear control model of reactor core temperature. Finally, core temperature control problem is discussed the control rod movement constraints in detail. Through the theoretical study of model predictive control, temperature control problem for the reactor core is designed used the constraint optimization predictive control strategy which based on state space model, and were simulated in the different conditions. Introduces the nonlinear system model of reactor core fuzzy based on mechanism, build the global model in large condition, the simulation results shows the global model is more closed to the actual situation and convenient controlled when compare to the signal model. Then, the paper study the weighting multi-model predictive control method, adopt the MMPC to the system, simulation results showed that this MMPC asked no special restrictions model for the structure change, can independent calculated each weight of controller outputs online, does not require online dynamic identification controlled object, has a very good control performance.In summary, this paper researched nuclear core temperature use the model predictive control method in the single operating situation, then established an modeling method of unclear core temperature control system in variable conditions, and studied the predicted control system of the nuclear core temperature control system under varying operating conditions temperature, has a certain significance to optimize the reactor system.