Adaptive Estimation-Based Control for Uncertain Processes with Low Order Dynamics

This paper investigates modeling and control of many high-order processes that have typical applications in industry via capturing their dominant dynamics and reformulating them as low order models with some parameters uncertainties. Many control applications are candidates for the proposed techniques, including industrial processes (e.g. level, flow, temperature, etc.) and some automotive applications (e.g. active suspension). The shortcomings and disadvantages of classical PID controllers, with static gains, are highlighted to motivate the need for the proposed designs and to use them as the ground for comparison, as well. Lyapunov-based techniques are used to design the first controller in order to implement an additional estimation loop that captures the dominant dynamics of the process to be controlled. The key feature in the design is arriving at the best parameter update law that guarantees both stability and satisfactory transient performance. The gradient-based optimization method is used to design the second controller that relies on filtration techniques, in addition to implementing another conventional PI control loop. Similarities and differences between the proposed controllers arc explored, while investigating their performance in the existence of saturation, uncertainties, noise, and variable time delays. A simulated fifth-order process is used to verify the effectiveness of the proposed techniques, while suggesting some feasible real-time improvements and extensions.