Hybrid training approach for artificial neural networks using genetic algorithms for rate of reaction estimation: Application to industrial methanol oxidation to formaldehyde on silver catalyst

摘要: A novel reactor simulator for the methanol oxidation to formaldehyde on silver catalyst was presented in this paper, including an original kinetic model based artificial neural networks. The network training performed using genetic algorithms associated with standard back-propagation, order improve efficacy, eliminating effect of random initial weights estimates. Experimental data (rates reaction) were obtained from process (conversion and selectivity), a back-calculation procedure through simplified deterministic implemented simulator. Process are widely available at industrial plants or literature proposed approach improves response time train cases where rigorous experimental work cannot be conducted due resource limitations. containing trained successfully validated data, especially operating conditions, models system have failed. here, as well net consist powerful tool plant engineers optimize timely economical fashion.