Optimal Process Technology Of Products Pipeline

In recent years, the construction of products pipeline has been developing rapidly inChina. With the pipes are put into operation, the optimal operation problem is to be payedclose attention. How to guarantee the safety, economic operation of the pipeline, improve theefficiency, to reduce the cost of pipeline transportation, is attracting more increasing attention.Considering the characteristics of one inlet and multi-outlets products pipeline, theoptimal operation problem is formulated as an optimization process of automatic compilationschedule. Using the minimum total of storage cost, pump cost and contamination loss as theobjective function, taking supply and demand situation, hydraulic conditions as constraints,and the products pipeline operation scheme optimization mathematical model is established.The model is a mixed integer nonlinear programming (MINLP) problem. It is difficult toobtain the global optimal solution with the traditional numerical algorithm. The GeneticAlgorithm (SGA) is introduced, together with the local improvement, new adaptive crossoveroperator and mutation operator. An improved adaptive genetic algorithm (IAGA) is formedtherefore.Through the analyses and comparison of different contamination quantity calculationmethod, the best algorithm is determined. The influences of temperature and pressure isconsidered to correct improve the accuracy of interface tracking, as well as the intermediatestation distribution and initial contamination. The maximum allowed mixing ratio betweenaviation kerosene, gas and diesel oil mixing are found through experiments.Power charge of pumps is a main part of operating cost. In order to overcome theshortcomings of the existing optimization methods, partheno-genetic algorithm (IPGA) isused to solve the module. Pump’s speeds are considered as decision variables. Chromosome iscomposed of the solution of the optimization question with integer coding method. Geneticoperators and the fitness function are designed corresponding.In order to overcome the shortcoming of the partheno genetic algorithm to deal with thechoke pressure, a new method named Two Phases Optimization Method (TPOM) isintroduced. The solution process of the original function was divided into two phases. First,determine the ideal conditions of pump operation scheme through the hydraulic grade line translation method. Then, the pump runs closest to the ideal working conditions as a secondextremist target function, which is solved by dynamic programming method. Examples showthat, partheno genetic algorithm (IPGA) and the two stages optimization method (TPOM) canbe used to complete the pumps distribution schemes optimization of products pipeline. Theformer is applicable to a wide range of pipelines kinds, but it takes longer time; the latter hasmore fast calculation speed, but lacks of the general usage.To verify the problems above, four examples, which has different throughput of theWestern Products Pipeline, Lanzhou-Chengdu-Chongqing Products Pipeline, are used. Resultsshow that, it is feasible to use the automatic scheduling as the core of the whole optimizationprocess. Optimal scheduling of complex products pipeline can be got, and pumps distributionschemes throughout the process cycle can be completed synchronously.