多参数耦合优化煤岩界面主动红外感知识别

Coal-Rock Interface Recognition Based on Active Infrared and Coupling Optimization of Multiple Parameters

为实现煤岩界面的预先感知与精准识别,在开采前为采煤机提供精准的截割轨迹,提出一种基于多影响因素耦合优化的煤岩界面主动红外感知识别方法。考虑光照时间、光照距离和光照强度多因素耦合作用对煤岩界面识别精度的影响,通过测试、采集各影响因素不同参数工况下煤岩试件的主动激励红外图像信息,利用正交实验方法确定实现煤岩界面高精度识别的多因素参数的最优组合。结合迭代优化方法在最优组合附近搜索各影响因素的最优参数,克服局部参数最优的问题,实现煤岩界面的高精度识别。由实验结果可知,该方法能够实现煤岩界面的快速、精准识别,最低识别精度达到97.96%以上,具有非常好的普适性,为实现井下智能化、无人化采煤提供了一种有效的技术手段。

To realize the pre-perception and accurate recognition of coal-rock interface and provide an accurate cutting trajectory for a shearer before mining,a coal-rock interface identification method based on active infrared and coupling optimization of multiple influencing factors was proposed.Considering the influence of the coupling effect of illumination time,distance and intensity on the accuracy of coal-rock interface identification,we tested the active excitation infrared images of coal-rock specimens with different parameters.Then,an optimal combination of multiple factor parameters for achieving high-precision identification of coal-rock interface was determined by orthogonal experimental method.Moreover,combined with the iterative optimization method,the optimal parameters of each influencing factor were searched near the optimal combination to overcome the problem of local optimal parameter and realize the high precision identification of coal-rock interface.Experimental results show that the proposed method can realize the coal-rock interface rapidly and accurately,and the minimum identification accuracy is over 97.96%,which provides fine universality,as well as an effective technical means for realizing the underground intelligent and unmanned mining.