Feedforward-feedback control method based on self-calibrated pump model for chilled water pump of air conditioning systems

 During the operation of air conditioning chilled water systems, the frequency of a chilled water pump is typically adjusted with different working conditions. Conventional local controllers are mainly based on PID algorithm, which adjusts the pump frequency based on the feedback error between controlled variable (pressure difference/temperature difference) and the set point. The control quality of PID algorithm greatly relies on the tuning accuracy of PID parameters. To tackle the problem, proposes a feedforward-feedback control method based on self-calibrated pump model for chilled water pump of air conditioning systems. Using similarity and self-calibrated pump model, the proposed method can quickly calculate the target frequency when the controlled variable deviates from the set point. Validated by the simulation with measured data of an underground railway station air conditioning system as inputs, the control accuracy of the proposed method is close to the accuracy of self-tuned PID controller based on single neuron. Besides, the proposed method costs less manual labor to tune the PID parameters and reduces the dependence of control quality on parameter tuning.