岩样内部的缺陷,由于其隐蔽性,往往是导致岩石力学试验结果离散的一个重要且不易控制的影响因素,在系列对比试验分析中,有时可能会掩盖真实的试验规律,甚至得到完全相反的结论。在大量试验数据分析的基础上,建立起纵波波速、回弹值与岩样抗压强度之间的综合关系,提出了抗压强度修正的经验公式和人工神经网络模型。在系列对比试验分析中,根据试样的初始纵波波速、回弹值等物理、力学参数对每个试样的实测抗压强度值进行修正,相当于把每个试样的初始强度修正到同一个标准,然后再进行比较分析,这样可以更好地把握试验规律。实践表明,所提出的修正方法能较好地识辨、衡量岩样之间的差异,在严格选样的基础上,可以较好地修正岩样的抗压强度,在系列对比试验中值得借鉴。 Due to its invisibility, the internal defects of rock samples are often an important and difficult to control factor that leads to the discrepancy of the results of rock mechanics test. In the comparative analysis of series of tests, the real test law may sometimes be masked, and even the opposite in conclusion. Based on a large number of experimental data, a comprehensive relationship between longitudinal wave velocity, rebound value and compressive strength of rock samples is established. Empirical formula and artificial neural network model of compressive strength correction are proposed. In the series of comparative tests, the measured compressive strength of each sample is corrected according to the physical and mechanical parameters such as the initial longitudinal wave velocity and the rebound value of the sample, which is equivalent to correcting the initial intensity of each sample to the same A standard, and then a comparative analysis, so you can better grasp the test law. Practice shows that the proposed correction method can better identify and measure the differences between rock samples. Based on the strict sample selection, the compressive strength of rock samples can be corrected well, which is worth learning from the series of comparative experiments .