论文部分内容阅读
综放回采巷道支护设计大多依赖于设计者的经验,采用工程类比法。由于其影响因素很多,且定量因素与定性因素并存,有相当一部分综放回采巷道的支护形式与围岩(煤)的变形特性不相适应,巷道支护的参数选择不合理,造成巷道支架折损严重,维护困难,严重影响煤矿的生产和安全。为此,建立了综放回采巷道支护设计二级神经网络模型,并以实际支护方案为样本,对模型进行学习训练,形成智能推理网络。实际应用表明:该方法简便,支护决策结果与实际相吻合,实用价值较高。
The fully mechanized caving mining roadway support design mostly depends on the designer’s experience, using engineering analogy. Because of its many influencing factors and the coexistence of quantitative and qualitative factors, a considerable part of the fully mechanized top-coal caving roadway is not compatible with the deformation characteristics of surrounding rock (coal), the parameters of roadway support are not selected properly, and the roadway Serious damage to the brackets, maintenance difficulties, seriously affecting the production and safety of coal mines. To this end, a secondary neural network model of supporting design of fully mechanized sub-level caving roadways has been set up. Taking the actual supporting scheme as a sample, the model is learned and trained to form an intelligent inference network. The practical application shows that the method is simple, the support decision-making result is in accordance with the actual situation, and the practical value is higher.