Fouling detection in a heat exchanger: A polynomial fuzzy observer approach

• Detection of fouling in heat exchangers.
• Development of a fuzzy polynomial observer.
• Comparisons are performed with neural network, Kalman filter and two other fuzzy approaches.

The paper presents new results on state estimation of nonlinear systems based on a fuzzy polynomial Takagi–Sugeno representation. Using sum of squares tools and some matrix properties, conditions to compute the gains of the observers can be obtained easily. Then, the results are applied on the detection of the fouling occurring in a counter flow heat exchanger, which involves many polynomial terms. Simulations and a comparison with previous approaches show the interest of the approach.