摘要
针对传统控制方法存在滞后性和难以实现对采煤机进行自主调高控制的问题,构建了采煤机自主调高隐式数字孪生模型(IDT),实现了采煤机的高效自主调高。提出了IDT和长短时记忆神经网络(LSTM)联合驱动的采煤机截割轨迹预测方法,首次采用蜜獾算法(HBA)优化LSTM的超参数,算例分析结果表明:蜜獾算法优化长短时记忆神经网络(HBA-LSTM)相比普通LSTM具有更高的精确度。提出了基于隐式数字孪生的采煤机自主调高控制策略,其结构包含数据采集、IDT模型和采煤机自主调高3个部分,在自主调高部分引入控制项修正了采煤机的控制参数。实例仿真表明:在IDT环境下,参数修正后的采煤机自主调高轨迹曲线更接近采煤机实际调高轨迹,最大误差仅为0.028m,有效地提高了采煤机自主调高的适应性。
Aiming at the problem that the traditional control method has hysteresis and is difficult to realize the autonomous elevation control of shearer,the implicit digital twin(IDT)model of the autonomous elevation of the shearer was constructed,and the efficient autonomous elevation of the shearer was realized.A shearer cutting trajectory prediction method driven by IDT and long-short-term memory(LSTM)neural network was proposed.The honey badger algorithm(HBA)was used to optimize the super-parameters of LSTM for the first time.The results of example analysis showed that HBA-LSTM had higher accuracy than the ordinary LSTM.An autonomous height adjustment control strategy of shearer based on implicit digital twin was proposed.Its structure included data acquisition,IDT model and autonomous height adjustment of shearer.The control item was introduced in the autonomous height adjustment part to correct the control parameters of the shearer.The case simulation showed that in the IDT environment,the self-aligning trajectory curve of the shearer after parameter correction was closer to the actual height adjustment trajectory of the shearer,and the maximum error was 0.028m,which effectively improved the adaptability of the automatic height adjustment of the shearer.
作者
蔡安江
刘俊强
刘亚东
任志刚
CAI Anjiang;LIU Junqiang;LIU Yadong;REN Zhigang(College of Mechanical and Electrical engineering,Xi’an University of Architecture and Technology,Xi’an,Shaanxi 710055,China;Yulin Smart Energy Big Data Application Joint Key Laboratory,Yulin,Shaanxi 719000,China)
出处
《矿业研究与开发》
CAS
北大核心
2022年第11期188-194,共7页
Mining Research and Development
基金
工信部物联网集成创新与融合应用项目(2018-470)
陕西省自然科学基础研究计划项目(2019JLZ-06).
关键词
采煤机
隐式数字孪生
蜜獾优化算法
截割轨迹预测
自主调高
Shearer
Implicit digital twinning
Honey badger algorithm
Cutting trajectory prediction
Autonomous elevation
