In this study, the potential capability of 4 powerful machine learning methods including Multi-layer perceptron (MLP) neural network, adaptive boosting-support vector regression (AdaBoost-SVR), recurrent neural network (RNN) and deep belief network (DBN) was investigated to predict the flow rates of the main products of an olefin plant in industrial scale. In this regard, a large data set including 1184 actual data points was gathered during four successive years at various practical conditions. 24 different independent parameters including flow rates of different feedstock, numbers of active furnaces and coil outlet temperatures were selected as input parameters of developed models and the outputs were the flow rates of the main products i.e. pyrolysis gasoline, ethylene and propylene. The accuracy of the developed models was evaluated using different statistical techniques. Based on the obtained results, the RNN model accurately predicted the main product flow rates with average absolute percent relative error (AAPRE) and determination coefficient (R2) values of 1.94% and 0.97, 1.29% and 0.99, 0.70% and 0.99 for pyrolysis gasoline, propylene and ethylene, respectively. The relevancy factor was also calculated to evaluate the effects of different input parameters on the products flow rate (estimated by RNN model). Accordingly, the number of furnaces in service and the flow rates of some feedstock had more positive effects on the outputs. In addition, the effects of different operating conditions on the propylene/ethylene ratio as a cracking severity factor was also discussed. This research proved that the smart approaches, despite being simple and straightforward, have the ability to predict complex unit performance. So they can be easily utilized to control and optimization of different industrial scale units.
