PROPORTIONAL-PLUS-INTEGRAL CONTROL OF EXPERIMENTAL OZONE CONCENTRATIONS IN A LARGE OPEN-TOP CHAMBER

The main limitation in using a proportional controller to regulate experimental ozone concentrations is that proportional controllers generate zero output when the actual concentration equals the target concentration. Adding an integral controller to a proportional controller can potentially increase the controller can potentially increase the precision in regulating experimental ozone concentrations because the controller generates a constant output when the actual and target ozone concentration are equal. The realized precision of a proportional-plus-integral (PI) controller in regulating an ozone treatment in a large open-top chamber was assessed by analyzing hourly ratios of the ozone concentration inside a chamber and of the ambient bulk air (C(c):C(a)) from May to September 1989. The target C(c):C(a) ratio for this study was 3.0, and the actual ratio was deemed in control if it lay between 2.7 and 3.3 (3.0+/-10%). For 89 uninterrupted days of exposure, the median ozone ratio averaged 2.99 and was in control during the exposure period. The 90 percentile range of ozone ratios for each was also in control for all but the early and late hours of exposure. Comparing the precision of ozone control for a calm day and for a breezy day showed little difference, indicating that the PI controller successfully compensated for ambient-air ingression into the chamber. Implementation of a PI controller is described as well as strategies for dealing with the potential risk of overexposure by the integral controller.