基于煤岩识别的采煤机自适应调高与调速控制策略研究

The project intends to research the adaptive control stratedies for drum height and traction speed based on coal-rock recognition for shearer on the fully mechanized coal mining face. Through some experiments, the internal correlation between the parameters of cutting audio, cutting temperature distribution field, motor current, cylinder pressure with shearer cutting state is explored. The cutting state space of shearer is constructed and operational mechanism of abnormal cutting state is presented. The sensor information denoising method based on wavelet package transformation and correlation-sample entropy is proposed and the coal-rock recognition algorithm based on fireworks algorithm and neural network is designed. Moreover, by the use of improved fruit fly optimization algorithm, the optimal models under instability conditions, such as excess regulation capacity of shear drum, over saltation of traction acceleration, operate over forbidden speed regulation section and prejudice for pushing forward the conveyer, are set up. The control models for the regulation of shear drum height and traction speed are proposed, and the adaptive control strategy is designed through integration of artificial immunity algorithm and neural networks algorithm. This project provides new theory and approach for coal-rock recognition and adaptive control for shearer, and has important theoretical significance and application prospect for realizing unmanned and less people-oriented mining on fully mechanized coal mining face.

本项目以综采工作面采煤机为对象，研究基于煤岩识别的自适应调高与调速控制策略。通过实验，探求截割声频、截割温度场、电机电流、油缸压力与采煤机截割状态之间的内在关系，构建采煤机截割状态空间，揭示其非正常截割状态的运行机理；研究基于小波包变换和相关性-样本熵的传感信号除噪方法，设计基于烟花算法与神经网络融合的煤岩识别算法，并利用改进果蝇算法，建立超出滚筒调节能力、超出牵引调节能力、禁入调速区间和不利于推溜移架等情况下非稳定工作区间的优化模型；建立采煤机滚筒高度与牵引速度的控制模型，研究基于人工免疫算法和神经网络相结合的自适应调高与调速控制策略。本项目的研究能为采煤机的煤岩识别与自适应控制提供新的理论和方法，并对实现综采工作面的“无人化”或“少人化”开采具有重要的理论意义和应用前景。

本项目以综采工作面采煤机为对象，研究基于煤岩识别的自适应调高与调速控制策略。通过搭建采煤机截割实验系统，构造典型的截割煤壁，开展了采煤机非正常截割状态的实验研究，确定了以滚筒、煤层以及岩石层三者之间的位置关系来构造采煤机截割状态空间。设计了传感信号的去噪算法和特征提取方法，研究了基于截割声音、截割温度场、截割振动的采煤机煤岩截割模式识别方法，对典型煤岩的截割模式识别精度达95%以上。以记忆截割理论为基础，综合考虑底板高度变化对记忆截割的影响，设计基于模糊理论的采煤机记忆截割路径优化算法，研究基于双坐标系的采煤机记忆截割路径平整性控制方法。提出了基于T-S云推理网络的采煤机牵引速度控制方法，建立了基于CMAC-PIλDμ控制器的采煤机液压调高模型，确定了以牵引速度控制为主、滚筒高度调节为辅的采煤机自适应控制方法，研究了基于指令优先级的采煤机自适应控制方案和流程，实现了采煤机滚筒高度与牵引速度协同调节的自适应控制。本项目的研究能为采煤机的煤岩识别与自适应控制提供新的理论和方法，并对实现综采工作面的“无人化”或“少人化”开采具有重要的理论意义和应用前景。