煤矿巷道锚护网络结构多目标进化优化设计方法

发布时间：2018-06-29 19:11

本文选题：锚护网络 + 支护质量　；参考：《中国矿业大学》2017年硕士论文

【摘要】：巷道支护作为保障煤矿安全生产的重要手段,一直受到煤矿企业和研究机构的广泛关注。在诸多支护方式中,锚杆支护是一种主动支护,具有结构简单、施工方便、便于对围岩主动加固等优点,已在实际的煤矿生产中得到了广泛应用;相应的,由锚杆和锚索构成的锚护网络成为提高巷道围岩稳定性的有效方法。锚护网络结构的合理性决定了支护巷道的稳定性,在锚护网络中,锚杆数量越多,巷道的安全性越高,但是,这也提高了支护成本,增加了支护时间,从而降低了采掘效率。如果锚护网络包含的锚杆少,那么,将难以维持巷道围岩的稳定性,使得巷道冒顶、片帮。此时,需要多次锚杆补打,从而增加了支护成本,降低了支护效率。可见,在巷道断面形状及其地质条件已知时,选取合适的支护形式和锚护网络结构,是保证巷道支护安全、经济和高效的关键问题。首先,针对现有锚护网络结构设计方法的不足,提出了锚护网络结构多目标进化优化设计方法。该方法构建了兼顾支护质量、支护成本和支护时间等三个目标的锚护网络结构多目标数学模型,模型中的支护质量通过构建支护质量代理模型进行描述。进而,采用多目标优化算法求解数学模型,得到一组最优锚护网络结构方案集。面向某一实际巷道,基于所提方法获取最优锚护网络结构方案集。实验结果表明最优锚护网络结构方案集能够兼顾支护质量和支护成本等多个优化指标,与已有的优化后的设计方案相比,有了更多选择。另外,开发了以该方法为理论基础的锚护网络结构多目标优化设计系统,为将锚护网络结构多目标进化优化设计方法应用于实际工程中提供了切实可行的途径。其次,考虑到决策者期望获得满足偏好的某一锚护网络结构,且偏好难以预先获得,提出交互式偏好下的锚护网络结构多目标进化设计方法。该方法在进化初始阶段,基于语义描述的偏好,以目标空间的最差点划分先验偏好区域,保证了偏好区域划分的合理性;在进化过程中,基于从候选解中所选的满意解,动态迁移和缩放偏好区域,以精准定位决策者的偏好;在决策者偏好取向发生变化时,给出相应的检测方法和应对策略,以及时跟踪决策者偏好的变化。面向某一实际巷道,实验验证了采用所提方法设计锚护网络结构时,不但缩小了搜索空间,而且提高了设计效率。最后,考虑到锚护网络结构多目标数学模型中的支护质量代理模型精度不高的不足,提出了交互式偏好下的支护质量代理模型更新策略。首先,将融入交互过程的多目标进化算法与支护质量代理模型的更新并行化,在缩短设计周期的同时,减轻了决策者的疲劳程度;然后,在进化算法求解的过程中,根据优势进化方向给出支护质量代理模型更新条件和新增样本点选取准则,以提高支护质量代理模型在偏好区域内的精度,针对已有的静态支护质量代理模型,采用本方法进行更新,实验结果表明,基于更新后的支护质量代理模型获得的最优锚护网络结构可靠性更高。
[Abstract]:As an important means to ensure the safe production of coal mine, roadway support has been widely concerned by coal mine enterprises and research institutions. In many supporting methods, bolt support is a kind of active support, which has the advantages of simple structure, convenient construction and easy to reinforce the surrounding rock. It has been widely used in the actual coal mine production; The anchoring network made up of anchor and anchor has become an effective method to improve the stability of the roadway surrounding rock. The rationality of the structure of the anchor network determines the stability of the supporting roadway. In the anchor network, the more the bolt number and the safety of the roadway, the higher the cost of the support, the increase of the support time and the reduction of the excavation. Efficiency. If the anchor network contains fewer bolts, it will be difficult to maintain the stability of the surrounding rock of the roadway, make the roadway roof and help. At this time, many bolts need to be supplementing, thus increasing the cost of support and reducing the support efficiency. It is obvious that when the shape of the tunnel section and its geological conditions are known, the appropriate support form and the anchor network are selected. Structure is the key problem to ensure the safety of roadway support, economy and high efficiency. Firstly, in view of the shortage of the existing design method of the existing anchor network structure, a multi-objective evolutionary optimization design method of the anchor network structure is proposed. The method constructs the multi target number of the anchor network structure with three targets, such as the support quality, the support cost and the support time. The support quality in the model is described by the construction of the support quality agent model. Then, the multi-objective optimization algorithm is used to solve the mathematical model, and a set of optimal anchor network structure schemes are obtained. The optimal anchor network structure scheme is obtained based on the proposed approach, and the experimental results show the optimal anchor net. The complex structure scheme set can take account of the multiple optimization indexes such as the support quality and the support cost. Compared with the existing optimized design schemes, there are more choices. In addition, the multi-objective optimization design system of the anchor network structure based on this method is developed, which is applied to the multi-objective evolutionary optimization design method of the anchor network structure. Practical ways are provided in practical engineering. Secondly, considering that the decision-makers expect to obtain a certain anchor network structure which satisfies the preference, and the preference is difficult to obtain in advance, the multi objective evolutionary design method of the anchor network structure under the interactive preference is proposed. This method is based on the preference of semantic description in the initial stage of evolution, and the target space is in the target space. The difference division of the prior preference region ensures the rationality of the partition of the preference region; in the process of evolution, based on the satisfactory solutions selected from the candidate solutions, the dynamic migration and scaling of preference regions are used to accurately locate the preference of the decision-makers; the corresponding detection methods and coping strategies are given when the preference orientation of the decision-makers is changed, and the time follows. In a practical tunnel, the experiment verifies that the design of the anchor network structure by using the proposed method not only reduces the search space, but also improves the design efficiency. Finally, considering the insufficient precision of the support quality model in the multi-objective mathematical model of the anchor network structure, the interactive deviation is proposed. First, the multi-objective evolutionary algorithm and the update of the support quality agent model are merged into the interactive process. At the same time, the fatigue degree of the decision-makers is reduced while the design cycle is shortened. Then, the support quality agent is given in the evolutionary direction of the evolutionary algorithm in accordance with the dominant evolutionary direction. The model updating condition and the new sample point selection criteria are used to improve the precision of the support quality agent model in the preference region. This method is used to update the existing static support quality agent model. The experimental results show that the optimal anchor network structure based on the updated support quality agent model is more reliable.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TD353

【参考文献】