TY  - JOUR
PY  - 2019//
TI  - In-depth research on qualitative simulation of coal miners' group safety behaviors
JO  - Safety science
A1  - Cao, Qinggui
A1  - Yu, Kai
A1  - Zhou, Lujie
A1  - Wang, Linlin
A1  - Li, Chunai
SP  - 210
EP  - 232
VL  - 113
IS  - 
N2  - Gaining in-depth and systematical knowledge of qualitative simulation technology and the related change rules is of great significance to controlling group unsafe behaviors of coal miners. In this study, the group safety behavior model of coal miners was established based on self-organized and hetero-organized behaviors. Using the qualitative simulation technology and the software platform of group safety behaviors based on the QSIM's (Qualitative Simulation) filtering optimized by GA-BP (Genetic Algorithm - Back Propagation Neural Network Algorithm), group safety behaviors of coal miners are qualitatively simulated comprehensively and systematically. By comparing and analyzing the simulation results using 15 different schemes, the evolutional rules of group safety behaviors are concluded. <br><br>RESULTS show that, under any group initial states, if the group is disturbed by a number of external environmental factors and no control strategies are taken, the level of group safety behavior deteriorates. In addition, the management and control strategies should be implemented in a targeted manner when the level of group safety behavior drops. Finally, the research results are applied for guiding the safety management of a coal mine enterprise, which can reduce the incidence of unsafe behaviors. Accordingly, it can be concluded that desirable results can be achieved after the adoption of reasonable control strategies. Additionally, the proposed qualitative simulation technology based on the optimized QSIM algorithm not only improves the simulation efficiency and reliability of group safety behaviors, but also can provide references for relevant research and applications in the field.<p />  <p>Language: en</p>
LA  - en
SN  - 0925-7535
UR  - http://dx.doi.org/10.1016/j.ssci.2018.11.012
ID  - ref1
ER  -