改进混洗蛙跳算法在油井措施优化中的应用

Application of Oil Well Measures Optimization Base on Improved Shuffled Flog Leaping Algorithm

针对油田传统人为安排措施工作计算量大、耗时多且经济效益不高的缺点,本文建立以经济效益最大化为目标,以年度增注目标、增产目标、含水目标及递减目标为约束条件建立优化模型,并提出一种改进混合蛙跳算法求解该模型。该算法通过反向学习机制初始化种群,再通过正态云算子求解全局最优个体和子群最优个体周围的更优值,最后利用混沌理论对个别个体进行变异来跳出局部最优解。通过经典函数极值优化对比,改进算法的性能优于PSO算法和SFLA算法,实际数据测试表明,优选出的措施优化方案取得了很好的实际应用效果。

Aiming at the defect that artificial arrangements for oilfield traditional measures work is computationally inten-sive, time-consuming and the absence of a high economic profit, a optimization model is established.The target of the model is maximizing the economic benefits.The model is used to increase injection of annual targets, increase produc-tion,hydrous and lapse rate as constraint conditions,and is solved by a improved shuffled flog leaping algorithm based on cloud model theory. The population is initialized through reverse learning mechanism, the better value around the global best individual and subgroup optimal individual in SFLAis solved by the normal cloud particle operator in this al-gorithm.Finally, the individual were mutated to jump out of local optimal solution by using the theory of chaos. Through comparison of classical function extremal optimization, the performance of improved algorithm is better than PSO algorithm and SFLA algorithm,The actual test data show that optimal optimization measures obtain very good ef-fect in practical application.