Research On Temperature Model Building And State Estimation For Batch Reactor

Batch reactors are widely used in fine chemicals, biochemicals, pharmaceuticals and food industries for the production of small amounts of products with high added value. In today's fierce market, it is an inevitable trend to ensure consistent high quality products and to improve the performance of batch operations. In the Batch polymerization reaction, the temperature control for reactor impacts directly on the quality and quantity, so the temperature control is very important. In order to optimize system operation and to achieve the best control performance, an accurate temperature model of batch reactor is indispensable.First, this paper analyzed the characteristics of batch reactor system and the polymerization process. Second, determine the temperature modeling methods according to temperature characteristics of the batch polymerization reactor and the difficulty of modeling. According to heat balance, heat transfer and cooling principle, the batch reactor temperature mechanism models of warm-up phase, temperature reaction stage and the cooling recovery phase were established basing on the analysis of the principles of heat transfer. The warm-up phase established a relational model about the reactor temperature and the heating power of heating rods using distributed modeling, the exact values of the reactor heat transfer coefficient were obtained according to thermal path analysis in heat transfer; a relationship model between the reactor temperature and the cooling water flow was established in the temperature reaction stage, the transfer function parameters were estimated by using the step response of constant temperature reaction stage, multiple integral to the output through the composite simpson algorithm and Least squares; the cooling phase established the model using heat transfer, cooling principle and calculate the heat loss accurately. These models were validated through experimental methods. The simulation results show that the model could describe the measured data with satisfied accuracy. Finally, Strong tracking filter algorithm was using in constant phase model for state estimation, according to the dynamic model. The algorithm can track the actual state of the system.