Human-like model-free control for heat exchangers with a new form of actuator bandwidth limitation

The design of control schemes for heat exchanger systems in large-scale industrial scenarios is challenged by complex dynamics and safety requirements. This paper proposes a human-like model-free control (HLMFC) scheme. The scheme is independent of model information, learning decision-making rules during manual control processes of heat exchangers from experienced operators. Exploiting the thermal equilibrium characteristic of the heat exchangers, a control efficiency factor and control interval iteration mechanism are introduced, which gradually compresses the control input to the desired value. The proposed control strategy eliminates the need for dynamic analysis, suppresses instabilities in the heat transfer process, and allows flexible adjustment of the actuator execution frequency. The bounded stability of the scheme is rigorously proven. Experimental results demonstrate that the proposed strategy achieves control objectives while flexibly allocating actuator frequencies, effectively reducing temperature fluctuations during operation.