Application of Asymmetric Fuzzy Linear Programming in EIT

There are inconsistent, uncertain and incomplete characteristics in Electrical impedance tomography (EIT) due to two natural problems of ill-posedness and ‘soft field’ effect. The traditional EIT inverse problem solving algorithms are based on deterministic objective functions and constraints which cannot efficiently represent the natural characteristics in the EIT process. Consequently, the EIT images have low spatial resolution. Inversely, the fuzzy set theory has been validated to represent imprecise data. To represent the natural characteristics in the EIT process, an asymmetric fuzzy linear programming (AFLP) model is applied to find the optimal solution for EIT imaging process. Moreover, the parameters in AFLP are alternatively optimized. Experimental results show that AFLP has better imaging effect and higher robustness than the traditional algorithms. Compared with the existing symmetrical fuzzy linear programming (SFLP) algorithm, AFLP algorithm has high resolution for both discrete and continuous objects. These results show that AFLP algorithm provides an effective solution to enhance the EIT imaging resolution.