带式输送机是煤矿井下运输系统的核心设备。目前，大部分矿用带式输送机在实际生产中均以满载量对应的带速恒定运行，在轻载或空载时常处于“大马拉小车”的运行状态，不仅造成了电能耗费，而且增大了设备损耗。如何实现带式输送机的高效智能运行，是煤炭行业科技人员重点研究的课题之一。

本文依托陕西陕煤韩城矿业有限公司象山矿井主运输系统的现场数据及运行状况，围绕带式输送机煤量预测方法和调速控制策略展开了深入研究，旨在获取实时载煤量，建立运载量与带速之间的匹配关系，并采用控制算法调整运行速度，实现设备的高效稳定运行。

针对常规煤量检测方法中检测时间长、传感器安装位置受限等问题，研究了带式输送机传动系统的运行特征，提出了一种基于TensorFlow框架下全连接神经网络的煤量预测方法，采用阶段性负载电流来预测实时载煤量，并对其预测模型进行了仿真验证，结果表明，模型预测计算值与实际运载量之间的损失函数值为5.88，平均相对误差为0.41%，对于负载煤量的度量级别预测精度高。针对带式输送机在调速过程中出现的提速过慢或减速过大问题，采用负载电流变化的实时性辅助载煤量信息，结合模糊决策推理确定了速度给定策略。此外，为了保证平稳调速，设计了一种人工蜂群优化PID的带速控制算法，以超调量、响应时间和上升时间为指标进行了阶跃响应及抗扰动性仿真实验分析，结果表明，相较于传统PID控制，所提控制算法的快速性和抗扰动性能明显改善，应用在带式输送机调速控制中超调量更小、调节速度更快且稳定性更好。

在上述基础上，完成了带式输送机调速控制系统的硬件和软件设计。基于组态王开发平台设计了上位机监控界面，并在上位机软件中实现了煤量预测模型调用和人工蜂群算法寻优，将调速控制策略下载到PLC控制器中进行了现场测试。测试结果表明，本文提出的煤量预测方法和调速策略算法是可行的、有效的，实现了带式输送机的智能控制和节能运行，对于同行业科研工作有一定的借鉴意义。

The belt conveyor is the core equipment of the coal mine underground transportation system. At present, most mining belt conveyors operate at the constant speed corresponding to full load in actual production. They are often in the operation state of "a big horse pulling trolley" under light load or no load, which not only causes power consumption but also increases equipment loss. How to realize the efficient and intelligent operation of belt conveyors is one of the key research topics for scientific and technological personnel in the coal industry.

Based on the field data and operation status of the main transportation system of Xiangshan mine of Shaanxi coal Hancheng Mining Co., Ltd., a profound study on the coal volume prediction method and speed regulation control strategy of the belt conveyor is carried out in this paper, to obtain the real-time coal load, establish the matching relationship between the load and the belt speed, and use the control algorithm to adjust the operation speed to realize the efficient and stable operation of the equipment.

Because of the problems of long detection time and limited sensor installation position in conventional coal quantity detection methods, the operation characteristics of the belt conveyor transmission system are studied, and a coal quantity prediction method based on a fully connected neural network under the TensorFlow framework is proposed. The phased load current is used to predict the real-time coal load, and the prediction model is simulated and verified. The results show that the loss function value between the predicted calculated value of the model and the actual carrying capacity is 5.88, and the average relative error is 0.41%, which indicates that the prediction accuracy is high for the measurement level of coal load. Aiming at the problem of too slow speed increase or too large speed reduction in the speed regulation process of belt conveyor, the speed setting strategy is determined by using the real-time auxiliary coal load information of load current change and fuzzy decision reasoning. In addition, to ensure smooth speed regulation, an artificial bee colony optimization PID belt speed control algorithm is designed. The step response and anti disturbance simulation experiments are carried out with overshoot, response time, and rise time as indicators. The results show that compared with the traditional PID control, the rapidity and anti-disturbance performance of the proposed control algorithm are significantly improved, and the overshoot is smaller in the speed regulation control of belt conveyor faster adjustment speed, and better stability.

On the basis of the above, the hardware and software design of the belt conveyor speed control system is completed. The monitoring interface of the upper computer is designed based on the KingView development platform. The coal quantity prediction model call and artificial bee colony algorithm optimization are realized in the host computer software, and the speed regulation control strategy is downloaded to the PLC controller for a field test. The test results show that the coal quantity prediction method and speed regulation strategy algorithm proposed in this paper are feasible and effective, and the intelligent control and energy-saving operation of belt conveyors are realized, which has a certain reference significance for the scientific research work in the same industry.

[1]中电传媒能源情报研究中心. 中国能源大数据报告(2021)[J/OL]. 电力决策与舆情参考, 2021(19): 26-34.

[2]葛世荣, 刘洪涛, 刘金龙, 等. 我国煤矿生产能耗现状分析及节能思路[J]. 中国矿业大学学报, 2018, 47(01): 9-14.

[3]蒋卫良, 王兴茹, 刘冰, 等. 煤矿智能化连续运输系统关键技术研究[J]. 煤炭科学技术, 2020, 48(07): 134-142.

[4]王国法, 王虹, 任怀伟, 等. 智慧煤矿2025情景目标和发展路径[J]. 煤炭学报, 2018, 43(02): 295-305.

[5]汤家轩, 刘具, 梁跃强, 等. “十四五”时期我国煤炭工业发展思考[J]. 中国煤炭, 2021, 47(10): 6-10.

[6]申斌学, 郑忠友, 朱磊. 新时代背景下绿色矿山建设体系探索与实践[J]. 煤炭工程, 2019, 51(02): 1-5.

[7]Marasova D, Teplicka K, Boksa P, et al. Sustainable quality of conveyor belts using an integrated knowledge system to support decision-making. ACTA MONTANISTICA SLOVACA, 2021, 26(1): 117-126.

[8]Wang J, Huang Z. The radical development of intelligent mining in China[J]. Engineering Technology, 2017, 3(4): 439-444.

[9]Bing H. Selection and configuration for slag system of continuous belt conveyor in colliery inclined shaft tunnel[J]. Construction Technology, 2016, 45(21): 107-110.

[10]袁海鹏. 带式输送机多样化及发展趋势的研究[J]. 煤矿机械, 2021, 42(11): 45-47.

[11]马树焕. 井下煤炭运输带式输送机驱动控制方式的应用研究[J]. 煤矿机械, 2017, 38(06): 143-145.

[12]蒋卫良, 郗存根, 宋兴元, 等. 煤矿带式输送机关键技术发展现状与展望[J]. 智能矿山, 2020, 1(01): 98-104.

[13]张少宾, 蒋卫良, 芮丰. 矿用带式输送机输送量测量方法现状及发展趋势[J]. 工矿自动化, 2019, 45(05): 100-103.

[14]陈文钰. 电子皮带秤的误差分析与维护[J]. 计量与测试技术, 2022, 49(2): 63-65.

[15]任凤国, 刘学红, 任安祥, 等. 提高矿用X射线核子秤计量稳定性的研究[J]. 工矿自动化, 2018, 44(08): 24-27.

[16]陈湘源. 基于超声波的带式输送机多点煤流量监测系统设计[J]. 工矿自动化, 2017, 43(02): 75-78.

[17]曾飞, 吴青, 初秀民, 等. 带式输送机物料瞬时流量激光测量方法[J]. 湖南大学学报, 2015, 42(02): 40-47.

[18]关丙火. 基于激光扫描的带式输送机瞬时煤量检测方法[J]. 工矿自动化, 2018, 44(04): 20-24．

[19]代伟, 赵杰, 杨春雨, 等. 基于双目视觉深度感知的带式输送机煤量检测方法[J]. 煤炭学报, 2017, 42(S2): 547-555.

[20]李纪栋, 蒲绍宁, 翟超, 等. 基于视频识别的带式输送机煤量检测与自动调速系统[J]. 煤炭科学技术, 2017, 45(08): 212-216.

[21]陶伟忠. 基于视频的煤矿带式输送机自动调速控制系统[J]. 煤炭科学技术, 2017, 45(5): 28-33．

[22]Ristic L B, Jeftenic B I. Implementation of fuzzy control to improve energy efficiency of variable speed bulk material transportation[J].IEEE Transactions on Industrial Electronics, 2012, 59(7): 2959-2969.

[23]Mazurkiewicz D. Maintenance of belt conveyors using an expert system based on fuzzy logic[J]. Archives of Civil and Mechanical Engineering, 2015, 15(2): 412-418.

[24]He D, Pang Y, Lodewijks G, et al. Healthy speed control of belt conveyors on conveying bulk materials[J]. Powder Technology, 2018(327): 408-419.

[25]Zhang S, Xia X. Modeling and energy efficiency optimization of belt conveyors[J]. Applied Energy, 2011, 88(9): 3061-3071.

[26]任中全, 王淼. 带式输送机节能调速控制系统设计[J]. 煤炭技术, 2016, 35(05): 245-246.

[27]高赟, 徐平. 带式输送机的Fuzzy-PID调速节能分析与验证[J]. 煤炭工程, 2017, 49(02): 131-133.

[28]雷汝海, 赵强. 矿井带式输送机节能优化与智能控制系统研究[J]. 煤炭技术, 2017, 36(12): 184-186.

[29]韩东升, 杜永贵, 庞宇松, 等. 基于预见控制的带式输送机调速节能方法[J]. 工矿自动化, 2018, 44(06): 64-68.

[30]王志文, 武利生. 井下带式输送机智能调速控制系统设计与研究[J]. 煤矿机械, 2020, 41(04): 8-11.

[31]郝洪涛, 杨庭杰, 张超. 基于负载估计的带式输送机系统节能控制方法研究[J]. 煤炭科学技术, 2021, 49(07): 139-146.

[32]张振文, 宋伟刚. 带式输送机工程设计与应用[M]. 北京: 冶金工业出版社, 2015: 06.

[33]张骉, 孟文俊, 张汉中, 等. 基于有限元方法的输送带模态分析研究[J]. 煤矿机械, 2021, 42(04): 91-94.

[34]GB/T 36698-2018, 带式输送机设计计算方法[S]. 北京: 中国标准出版社, 2018.

[35]정규범. Quasi-Resonant PWM converter controlled by constant frequency[J]. Journal of Satellite, Information and Communications, 2017, 12(2): 28-32.

[36]王定龙, 王然风, 赖春林. 带式输送机双机驱动控制系统设计[J]. 工矿自动化, 2018, 44(1): 74-78.

[37]王艳萍, 周杨, 闫超. 基于改进BP神经网络的刮板输送机负载预测方法研究[J]. 矿山机械, 2015, 43(10): 17-21.

[38]王凯. 基于刮板输送机负载预测的采煤机调速技术研究[D]. 北京: 中国矿业大学, 2015.

[39]王杰. 刮板输送机负载预测方法的研究[J]. 煤炭科技, 2020, 41(01): 32-35.

[40]张东林, 赵越, 杨凡, 等. 矿山多驱带式输送机的变频传动应用[J]. 矿山机械, 2017 (9): 28-32.

[41]马鸿文. 异步电动机电磁转矩公式推导及应用分析[J]. 实验室科学, 2018, 21(05): 229-233+237.

[42]Jordan M I, Mitchell T M, et al. Machine learning: Trends, perspectives, and prospects[J]. Science, 2015, 349(6245): 255-260.

[43]Lauriola, Lavelli, Aiolli. An introduction to Deep Learning in Natural Language Processing: Models, techniques, and tools[J]. Neurocomputing, 2022(470): 443-456.

[44]张驰, 郭媛, 黎明. 人工神经网络模型发展及应用综述[J]. 计算机工程与应用, 2021, 57(11): 57-69.

[45]Ko YoungMin, Li PengHang, Sunwoo Ko. Performance improvement method of fully connected neural network using combined parametric activation functions[J]. KIPS Transactions on Software and Data Engineering, 2022, 11(1): 1-10.

[46]Hasilci Z, Bogoclu ME, Dalkilic AS, et al. Development of a prediction model using fully connected neural networks in the analysis of composite structures under bird strike[J]. Journal of Mechanical Science and Technology, 2022, 36(2): 709-722.

[47]Su Yugang, Yang Jian, Dai Xin, et al. TensorFlow neural network based load and mutual inductance identification method for magnetic coupling resonant wireless power transfer system[J]. Automation of Electric Power Systems, 2021, 18(45): 162-169.

[48]梁昱, 李彬彬, 陈志高, 等. TensorFlow中深度前馈网络优化研究及其轴承故障诊断应用[J]. 计算机应用与软件, 2019, 36(10): 175-182.

[49]蔡自兴. 智能控制原理与应用[M]. 北京: 清华大学出版社, 2014: 75.

[50]Ciucci, Davide. The theory of fuzzy sets: Basic ideas, uses and legacy 50 years after birth[J]. Mondo Digitale, 2017, 16(72): 1-16.

[51]杜岗, 马小平, 张萍. 煤矿局部通风机转速控制算法研究[J]. 工矿自动化, 2020, 46(09): 69-73+87.

[52]辛斌, 陈杰, 彭志红. 智能优化控制:概述与展望[J]. 自动化学报, 2013, 39(11): 1831-1848.

[53]张延成, 田晓丽. 人工免疫算法在带式输送机变频调速控制系统中的应用[J]. 煤矿机械, 2010, 31(09): 181-183.

[54]王卉. 基于模糊PID理论的带式输送机调速系统设计[J]. 煤矿机械, 2019, 40(09): 14-16.

[55]Tang J, Liu G, Pan QT. A review on representative swarm intelligence algorithms for solving optimization problems: Applications and trends[J]. IEEE-CAA Journal of AUTOMATICA SINICA, 2021, 8(10): 1627-1643.

[56]张少宾, 蒋卫良, 芮丰. 基于自适应模糊PID的输送机带速控制仿真[J]. 工业控制计算机, 2019, 32(06): 100-101.

[57]冯汛, 倪红军, 石健, 等. PLC的应用发展与前景[J]. 机床与液压, 2016(02): 203-206.

[58]韩相争. PLC与触摸屏、变频器、组态软件应用一本通[M]. 北京: 化学工业出版社, 2010: 2.