基于遗传算法和人工神经网络的巷道支护研究

Study on roadway support based on genetic algorithm and artificial neural network

针对软岩巷道围岩的复杂性和离散性特点,采用单一围岩稳定性影响因素无法进行巷道围岩准确分类,进而无法准确地确定支护设计方案的现状,采用遗传算法和人工神经网络建立了围岩稳定性分类预测模型。通过算例验证了该模型能在考虑多影响因素下准确地代表围岩稳定性影响因素与围岩类型之间的非线性关系,并预测出软岩巷道的围岩分类,从而为软岩巷道稳定性分类及控制技术提供参考依据。

Based on the complex soft rock surrounding rock condition, surrounding rock cannot be classified based on single surrounding rock stability influencing factor, genetic algorithm and artificial neural network model was established to overcome the above- mentioned disadvantage to predict the stability classification of surrounding rock. The example was used to demonstrate that the established model can represent the nonlinear relationship between stability influence factors and types for surrounding rock under considering multi- influence factors. Meanwhile, the proposed model classified surrounding rock, which can provide reference for soft rock stability classification and control of soft rock.