Backbreak (BB) is a detrimental outcome of blasting activities in mineral extraction processes within mines. It involves the development of fractures and cracks at considerable distances behind the last row of blast pits, leading to reduced bench safety and increased operational costs. Given the multitude of factors influencing BB, various techniques have been developed to predict and optimize its occurrence. This particular study focused on analyzing 48 blasts in the tailings section of Gol Gohar Mine No. 1 to forecast BB using the whale optimization algorithm (WOA), multiverse optimizer (MVO), sine cosine algorithm (SCA), ant lion optimizer (ALO), and multivariate linear regression (MLR). Comparative analysis of the four BB prediction models revealed that the MVO algorithm yielded the most favorable outcomes, with the train data exhibiting parameter values of 0.9901, 0.2161, 0.1127, 98.8472, and 0.0180 for R², RMSE, MSE, VAF, and MAPE, respectively, while the test data displayed values of 0.6357, 1.4955, 1.2003, 63.5472, and 0.1951 for the same parameters. In addition, the analysis specifically emphasized the substantial influence of spacing, burden, and GSI as the primary determinants of the backbreak phenomenon. In stark contrast, however, powder factor, delay time, and joint condition are identified as having negligible effects on backbreak.

回裂 (BB) 是矿山矿物开采过程中爆破活动的有害结果。它涉及在最后一排爆破坑后面相当远的距离处产生裂缝和裂缝，导致工作台安全性降低并增加运营成本。鉴于影响 BB 的因素众多，人们开发了各种技术来预测和优化 BB 的发生。这项特殊研究重点分析了 Gol Gohar 1 号矿尾矿段的 48 次爆炸，使用鲸鱼优化算法 (WOA)、多元宇宙优化器 (MVO)、正弦余弦算法 (SCA)、蚁狮优化器 (ALO) 来预测 BB和多元线性回归 (MLR)。对四种 BB 预测模型的比较分析表明，MVO 算法产生了最有利的结果，训练数据的 R ²、 RMSE、MSE、VAF 和 MAPE 参数值为 0.9901、0.2161、0.1127、98.8472 和 0.0180，相同参数的测试数据分别显示为 0.6357、1.4955、1.2003、63.5472 和 0.1951。此外，分析还特别强调了间距、负担和 GSI 的实质性影响，作为背折现象的主要决定因素。然而，与此形成鲜明对比的是，粉末因素、延迟时间和接头状况对回缩的影响可以忽略不计。