Detecting inverse responses in chemical reaction processes with fuzzy qualitative simulation

This paper presents a significant application of a fuzzy qualitative simulation algorithm to the detection of inverse responses in chemical reaction systems. Recognising the impracticality of the existing algorithm in handling real-world problems, an alternative method for representing system models is proposed. With the modified method the algorithm is able to generate the possible system behaviours from a given model, ensuring that if it does not find an inverse response then an inverse response cannot be produced by the system. Due to qualitative ambiguity, however spurious inverse responses may be generated. Fortunately, these are in general only, generated near to the limits of the the numerical presence of inverse responses. Experimental results are compared with those obtained by numerical simulations and future improvements pointed out.