Multiobject Approach for Total Separation Network

Authors

Basma. D. Aly, S.A. Aly, M.S. Awad
Department of Refining and Petrochemical Engineering, Faculty of Petroleum and Mining Engineering, Suez University.

Abstract

A multiobject approach for synthesis the sharp separation network is developed in this paper .This approach consists of two parts, first part forthe selection of the best sequence of separation network (without heat integration) and second partsynthesis of separation network (with heat integration). The synthesis algorithm is based on application of three expert rules qualification of the estimated mass load, difference in boiling point and relative volatility. Two fuzzy analogical gates are employed (symmetric and asymmetric). The symmetric gate (AND gate) inputs are the normalized estimated mass load with the normalized relative volatility. The asymmetric gate (INVOKE gate) inputs are the output of AND gate and the normalized boiling point difference. The proposed approach is simple to implement and can be done by manual computation. Several illustrative examples using existing fuzzy analogical gate are given to demonstrate the effectiveness of the proposed method.