Soft-sensing via artificial neural networks

The last decade has seen the development of many approaches for estimating those variables which are difficult to measure online in industrial process situations. In this paper one approach that can he used to provide frequent and accurate estimates of process outputs which are subject to large measurement delays is outlined. The method makes use of a neural network model. The development and application of the estimator is addressed. The results from recent industrial application studies and plant simulation studies serve to highlight the characteristics of the philosophy, and the utility of the neural network as a soft sensor, i.e. a sensor based upon software rather than hardware.<
>