Pollutant concentration profile reconstruction using digital soft sensors for biodegradation and exposure assessment in the presence of model uncertainty.

A new approach to the problem of environmental hazard assessment and monitoring for pollutant biodegradation reaction systems in the presence of uncertainty is proposed using soft sensor-based pollutant concentration dynamic profile reconstruction techniques. In particular, a robust reduced-order soft sensor is proposed that can be digitally implemented in the presence of inherent complexity and the inevitable model uncertainty. The proposed method explicitly incorporates all the available information associated with a process model characterized by varying degrees of uncertainty, as well as available sensor measurements of certain physicochemical quantities. Based on the above information, a reduced-order soft sensor is designed enabling the reliable reconstruction of pollutant concentration profiles in complex biodegradation systems that can not be always achieved due to physical and/or technical limitations associated with current sensor technology. The option of using the aforementioned approach to compute toxic load and persistence indexes on the basis of the reconstructed concentration profiles is also pursued. Finally, the performance of the proposed method is evaluated in two illustrative environmental hazard assessment case studies.