A self-validating inferential sensor for emission monitoring

Proposes a self-validating inferential sensor approach based on principal component analysis (PCA). The input sensors are validated using a fault identification and reconstruction approach proposed in Dunia et al. (1996). A principal component model is built for the input sensor validation. The validated principal components are used to predict output variables using linear regression or neural networks. If a sensor fails, the sensor is identified and reconstructed with the best estimate from the PCA model. The principal components are also reconstructed accordingly for prediction. The self-validating soft sensor approach is applied to air emission monitoring.