抄録
This paper concerns the development of a new design and tuning method for use with PID controllers that are commonly used to heating, ventilating and air-conditioning (HVAC) fields. The robust PID controller is designed for temperature control of a single-zone environmental space. Although the dynamics of environmental space are described by higher-order transfer functions, most HVAC plants are approximated by first-order lag plus deadtime systems. Its control performance is examined for this commonly approximated controlled plant. Since most HVAC plants are complex with nonlinearity, distributed parameters and multivariables, a single set of PID gains does not necessarily yield a satisfactory control performance. For this reason, the PID controller must be designed as a robust control system considering model uncertainty caused by changes in characteristics of the plant. The PID gains obtained by solving a two-disk type mixed sensitivity problem can be modified by contrast to those tuned by the traditional Ziegler-Nichols rule. The results, which are surprisingly simple, are given as linear functions of ratio of deadtime to time-constant for robustness. The numerical simulations and the experiments on a commercial sized test plant for air-conditioning suggest that the robust PID controller proposed in this paper is effective enough to practical applications.
