Studies On Hybrid Algorithm Of LCA-based And Its Application To Chemical Process System

A chemical process system synthesis problem is a typical example of a mixed-integer programming (MIP) problem. As the research scale of the process system increases, the synthesis problem will also become increasingly complex, and solving the problem will become more difficult. The MIP problem solving method has become a widely researched topic. Therefore, this article attempts to put forward an effective hybrid algorithm for solving chemical process system problems.This paper proposed a hybrid algorithm based on the Linear-up Competition Algorithm (LCA). It is LCA mixed with deterministic algorithm. Two strategies for solving the two-tier nesting algorithm are used; in the outer layer, LCA optimize integer variables and in the inner layer, simplex method or sequential quadratic programming algorithm (SQP) optimize continuous variables. This paper studied the combination mechanism, implementation guidelines and problem solving steps of the hybrid algorithm, and through solving the certification functions tested the effectiveness of the algorithm.The hybrid algorithm was used for the solution of MIP problems in the chemical process system, for the following three aspects:(1) liquor blending problem: a new liquor blending mixed-integer linear programming (MILP) model was established and using the hybrid algorithm proposed in this paper solved the model abtaining the minimum liquor blending costs, maximum utilization of base liquor storage space and the lowest operating cost;(2) Optimization scheduling problem of a multi-cycle boiler steam system operation: taking lowest operating costs and transit expenses as objective function; established a mixed-integer nonlinear programming (MINLP) model for the problem and solved the model using the hybrid algorithm. The optimization result obtained was closer to that quoted in literature.(3) Long distance hot oil pipeline operation optimization problem: taking lowest fuel costs and power expenses as objective function; established a MINLP model for the problem and solved the model using the hybrid algorithm. The optimization result obtained was slightly better than that quoted in literature.The hybrid algorithm based on the Linear-up Competition Algorithm proposed in this paper for solving MILP and MINLP problems in chemical process system achieved good results, showing the effectiveness of the hybrid algorithm. The algorithm can be used as a new algorithm for solving MIP problems.