论文部分内容阅读
针对传统的液压支架设计具有耗时、耗材、成本高等缺陷,提出一种利用ANSYS和BP神经网络相结合对掩护梁进行轻量化和寿命预测的设计方法。首先应用ANSYS软件参数化语言APDL对ZY18000-25-45型液压支架掩护梁进行建模;再应用ANSYS/OPT模块对掩护梁进行尺寸优化分析,使其在满足液压支架设计准则的前提下,对掩护梁进行轻量化设计,并对原有掩护梁模型进行改造,整体质量减少了6.9%;最后利用遗传算法改进BP神经网络对优化后的掩护梁建立神经网络模型,对掩护梁的疲劳寿命进行预测。结果表明,掩护梁平均寿命为17228次,优化后的掩护梁满足可靠性要求。
Aiming at the defect of traditional hydraulic support design, such as time consuming, consumables and high cost, a design method based on the combination of ANSYS and BP neural network is proposed to reduce the weight and life of cover beam. At first, ANSYS software parameterization language APDL is used to model the ZY18000-25-45 hydraulic support beam, and then the ANSYS / OPT module is used to optimize the size of the support beam to make it meet the design criteria of hydraulic support. Then the original quality of the beam is modified, and the overall quality is reduced by 6.9%. Finally, the neural network model is built by using improved BP neural network based on genetic algorithm and the fatigue life of the shield beam prediction. The results show that the average life span of shield beam is 17228 times, and the optimized shield beam meets the reliability requirements.