Pulse jet fabric filters (PJFF) have become an attractive option of particulate collection utilities due to the feature that they can meet the stringent particulate emission limits regardless of variation in the operating conditions. The dynamics of the filter has complex nonlinear characteristics as reported in [1]. In this paper, the framework of local model networks (LMN) is employed to approximate the process dynamics of the pulse jet fabric filter that is subject to multiple operating regimes. To do so, an extended self-organizing map is employed to partition the PJFF's operating range and construct the LMN automatically. Simulation results illustrate the proposed approach and a comparison with the conventional approach is made.
展开▼