论文部分内容阅读
从研究人的不安全行为本身出发,将设计与使用、管理、行为失误这3个方面作为构建矿工不安全行为指标体系的影响因素。结合专家咨询和煤矿实际调研情况,尝试使用遗传算法优化神经网络,构建基于矿工不安全行为的煤矿安全预测评价模型。选取山东、河南等地大型煤矿的实时监测数据进行样本学习及实证分析。实例检验表明:安全预测评价模型短期的预测结果与实际情况相符合,能够提前对煤矿安全状况进行较为准确的预测。使用遗传神经网络算法实时的预测煤矿整体安全状况,通过反馈的结果反向作用于煤矿的安全管理决策,有利于为管理者提供决策支持。
Based on the study of human unsafe behavior itself, the author considers the three aspects of design and use, management and behavior errors as the factors to construct the indicator system of unsafe behavior of miners. Based on the consulting of experts and the actual investigation of coal mines, this paper attempts to optimize the neural network using genetic algorithm and build a coal mine safety prediction model based on unsafe behavior of miners. Select real-time monitoring data of large coal mines in Shandong, Henan and other places for sample learning and empirical analysis. The case study shows that the short-term prediction results of the safety prediction model are in accordance with the actual conditions, and the coal mine safety conditions can be predicted more accurately in advance. The use of genetic algorithm to predict the overall safety status of coal mines in real time, and the results of feedback can reverse the safety management decisions of coal mines and help to provide managers with decision support.