Integrating statistical methods and nonlinear control tools for modeling, control and fault-handling in batch processes

点击翻译按钮获取中文摘要

Batch processes are commonly used in the manufacturing industry either specifically to produce low volume, high value product, or exist as inherent operational aspects of primarily continuous operation (start-ups and shutdowns being prime examples). Yet, compared to the problem of modeling, control and fault handling of continuous processes, significantly less attention has been paid to the problem of batch process control. This research will focus on the problems of utilizing the synergies between first principles modeling and empirical modeling to develop an integrated and dedicated framework for batch process control. Several practical issues such as lack of measurements, constraints, inherent process non linearity and the occurrence of sensor and actuator faults will be considered. The results of the proposed research are expected to bring significant benefits to a range of industries from specialty chemicals to pharmaceutical to petrochemicals, both in the form of savings due to efficient operation and control, but also in the form of preventing environmental accidents resulting from faults. The proposed research will result in the development of highly qualified personnel in the area of batch process control in particular, and process control in general, fulfilling an important current and future need in Canadian industry.