Single-stage multi-objective optimization for robust QFT controller and prefilter synthesis in process control systems
摘要
Engineering problems in the real world are often multi-objective, requiring multiple performance specifications and design constraints to be met for optimal performance. The mathematical model used in the controller synthesis approximates the actual system and is susceptible to measurement noise, disturbances, and parametric uncertainties during operation. To maintain the integrity and safety of processes, control systems must offer robust performance even under the influence of these uncertainties. The proposed work presents a single-stage procedure for the design of quantitative feedback theory (QFT) controllers using a multiple-objective optimization using the genetic algorithm. The proposed methodology is valid for both single-input single-output (SISO) and multi-input multi-output (MIMO) systems. The proposed work also explores the visualization of n-dimensional Pareto fronts using level diagrams to help choose an ideal solution. The approach is demonstrated through its application to the robust control of two process control systems: (a) liquid level control in coupled tanks (SISO) and (b) the level and temperature control in a three-stage evaporator system (MIMO). A comparison has been made with the existing techniques, and it can be observed that the proposed work offers an improvement of 28.4% in rise time, 61.5% and 29.7% in settling time, a 99.92% reduction in overshoot percentage, and 5.6% & 82.43% improvement in gain margin and phase margins when compared to already existing controllers for liquid level control systems. Moreover, in the case of level and temperature control in a three-stage evaporator MIMO system, an improvement of 41.3% can be observed in settling time