基于复合蛙跳算法的火山岩最优化测井解释方法

Logging interpretation optimization of volcanic rocks using the complex frog-leaping algorithm

火山岩岩性复杂,不同岩性的矿物种类差异大,因此火山岩储层的矿物含量计算难度大。最优化测井解释方法是一种有效的储层参数计算手段,其中的关键在于建立准确的测井解释体积模型以及选取合适的最优化方法。根据火山岩主要矿物特征建立火山岩体积解释模型,将石英、长石、铁镁矿物作为骨架矿物,凝灰质作为填充物,并首次引入蛙跳算法(SFLA)计算火山岩储层参数。鉴于SFLA在寻优后期搜索速度变慢,精度不高的缺陷,以及复合形方法(CM)具有极强的局部搜索能力,但优化结果受初始点的影响较大,易陷入局部极值的特点,提出了一种将蛙跳算法的进化思想与复合形方法极强的搜索能力相结合的复合蛙跳算法(CFLA)。首先按照种群中顶点的目标函数值大小进行排序、分组,然后在每个子种群中利用复合形方法进行局部进化,最后将所有子种群中的顶点进行混合以保证全局信息的交流。该方法应用于实际资料处理的结果表明,相较于蛙跳算法,复合蛙跳算法的计算精度明显改善,它同时克服了蛙跳算法在寻优后期搜索速度变慢的缺陷,计算效率提高约1倍。

Volcanic rocks exhibit complex lithology and mineral compositions,which bring challenges to the mineral content calculation of volcanic reservoirs.The optimal logging interpretation method is effective for the inversion of reservoir parameters,and the key is to establish an accurate logging volume interpretation model and select an appropriate optimization method.In this study,the volume interpretation model was established with quartz,feldspar,mafic mineral as frame minerals,tuff as filled mineral.The shuffled frog-leaping algorithm(SFLA)was introduced to calculate the mineral contents of volcanic reservoirs.However,its search speed slows down and the accuracy is not as high in the late stage of optimization.The complex method(CM)has a strong local search ability.However,its optimization results are greatly affected by the initial points,and it is susceptible to falling into the local extremum.Therefore,we propose a complex frog-leaping algorithm(CFLA),which combines the evolution idea of the SFLA with the strong searching ability of the CM.First,we ranked and grouped the vertices according to the value of the objective function of the population.We then used the CM to locally evolve in each subpopulation,and finally mixed the vertices of all subpopulations to ensure the exchange of global information.The processing results of the actual data showed that the calculation accuracy of the CFLA was improved,and the calculation efficiency of the CFLA was double that of the SFLA.