煤矿电网电容电流超限会导致接地点产生不稳定电弧，进而引发弧光过电压和漏电故障，严重危害矿井电气设备和人身安全，因此，《煤矿安全规程》对煤矿电网电容电流有严格的限值要求。煤矿井下是典型的易燃易爆生产场所，必须采用防爆设备供电，不允许接入测试仪器；在煤矿电网设计阶段，系统接地方式和消弧线圈容量配置也需要通过计算电容电流才能确定。此外，智慧矿山建设与智能化技术的快速发展也对煤矿电网电容电流计算提出了智能化要求，在此背景下，研究煤矿电网电容电流计算方法对保障矿井安全供电具有重要意义。

论文首先分析了煤矿电网电容电流的产生机理，采用Sobol敏感性分析方法研究影响煤矿电网电容电流的关键参数以及各参数之间的相互作用，明确需采用智能计算模型来反映各参数的相互作用，以及对煤矿电网电容电流的非线性影响关系，在此基础上，设计适用于煤矿电网的电容电流计算方案。

针对现有煤矿电网电容电流计算方法误差较大的问题，提出一种基于支持向量机的电容电流计算方法。根据Sobol敏感性分析结果，选取解释度高且方便获取的参数作为输入特征量；分析不同类型矿用电缆电容电流参数的相似性与差异性，采用模型迁移思想扩充和增强电容电流数据样本，解决了电容电流数据样本不完备的问题。在此基础上，采用基于稀疏技术的支持向量机建立小容量电容电流数据样本的智能计算模型，综合考虑系统运行方式和电气设备对电容电流的影响，实现对煤矿电网电容电流的准确计算。

针对支持向量机模型在对煤矿电网电容电流计算时会出现计算结果不稳定的问题，采用一种混合策略改进的鲸鱼算法对模型超参数进行优化。由于鲸鱼算法存在收敛速度慢和易陷入局部最优等问题，引入Tent混沌映射使初始解在空间分布更均匀，并结合随机性学习策略增强迭代过程的种群多样性，通过自适应改变权重提高算法的全局寻优和局部搜索能力，进一步提升支持向量机模型对电容电流的计算精度与稳定性。

采用矿用电缆电容电流测试数据和多家煤矿电网电容电流实测数据验证本文方法的计算精度，相较于现有电容电流计算方法，本文方法能够实现对煤矿电网电容电流的准确计算，其平均相对误差控制在2.26%以下。

The over-limit capacitive current of coal mine power grid will lead to unstable arc at the grounding point, which will lead to arc overvoltage and leakage fault, which will seriously endanger the electrical equipment and personal safety of the mine. Therefore, the Coal Mine Safety Regulations have strict limit requirements for the capacitive current of coal mine power grid. The underground coal mine is a typical inflammable and explosive production place. The explosion-proof equipment must be used for power supply, and the access to test instruments is not allowed. In the design stage of coal mine power grid, the system grounding mode and arc suppression coil capacity configuration also need to be determined by calculating the capacitance current. In addition, the rapid development of intelligent mine construction and intelligent technology also puts forward intelligent requirements for the calculation of capacitive current in coal mine power grid. Under this background, it is of great significance to study the calculation method of capacitive current in coal mine power grid to ensure the safe power supply of coal mine.

Firstly, the generation mechanism of capacitive current in coal mine power grid is analyzed. The Sobol sensitivity analysis method is used to study the key parameters affecting the capacitive current of coal mine power grid and the interaction between the parameters. It is clear that the intelligent calculation model should be used to reflect the interaction of the parameters and the nonlinear influence relationship on the capacitive current of coal mine power grid. On this basis, the calculation scheme of capacitive current suitable for coal mine power grid is designed.

Aiming at the problem of large error in the existing calculation method of capacitance current in coal mine power grid, a calculation method of capacitance current based on support vector machine is proposed. According to the results of Sobol sensitivity analysis, the parameters with high interpretation and convenient acquisition are selected as input feature quantities. The similarity and difference of capacitance current parameters of different types of mine cables are analyzed. The model migration idea is used to expand and enhance the capacitance current data samples, which solves the problem of incomplete capacitance current data samples. On this basis, the intelligent calculation model of small capacity capacitive current data samples is established by using support vector machine based on sparse technology. Considering the influence of system operation mode and electrical equipment on capacitive current, the accurate calculation of capacitive current in coal mine power grid is realized.

Aiming at the problem that the support vector machine model will have unstable calculation results when calculating the capacitive current of the coal mine power grid, a hybrid strategy improved whale algorithm is used to optimize the model hyperparameters. Because the whale algorithm has the problems of slow convergence speed and easy to fall into local optimum, the Tent chaotic map is introduced to make the initial solution more evenly distributed in space, and the random learning strategy is combined to enhance the population diversity of the iterative process. The global optimization and local search ability of the algorithm are improved by adaptively changing the weight, and the calculation accuracy and stability of the support vector machine model for capacitive current are further improved.

The calculation accuracy of this method is verified by the test data of mine cable capacitance current and the measured data of capacitance current of many coal mine power grids. Compared with the existing capacitance current calculation method, this method can realize the accurate calculation of capacitance current of coal mine power grid, and its average relative error is controlled below 2.26 %.

[1]唐珏, 王俊. “双碳”目标下煤炭发展及对策建议[J]. 中国矿业, 2023, 32(09): 22-31.

[2]袁亮. 我国煤炭工业高质量发展面临的挑战与对策[J]. 中国煤炭, 2020, 46(01): 6-12.

[3]杨仁树, 李成孝, 陈骏, 等. 我国煤矿岩巷爆破掘进发展历程与新技术研究进展[J]. 煤炭科学技术, 2023, 51(01): 224-241.

[4]刘建华, 唐润恒, 刘晓丹, 等. 煤矿智能供电系统单相接地网络保护[J]. 煤矿安全, 2018, 49(01): 115-117.

[5]KALYUZHNY A. Analysis of temporary overvoltages during open-phase faults in distribution networks with resonant grounding[J]. IEEE Transactions on Power Delivery, 2015, 30(1): 420-427.

[6]中华人民共和国应急管理部. 煤矿安全规程[M]. 北京:应急管理出版社, 2022: 153.

[7]中国煤炭建设协会. 矿山电力设计标准: GB 50070-2020[S]. 北京: 中国计划出版社, 2020: 5-6.

[8]韦明杰, 张恒旭, 石访, 等. 基于谐波能量和波形畸变的配电网弧光接地故障辨识[J]. 电力系统自动化, 2019, 43(16): 148-154.

[9]COSTA F B, SOUZA B A, BRITO N S D, et al. Real-Time detection of transients induced by high-impedance faults based on the boundary wavelet transform[J]. IEEE Transactions on Industry Applications, 2015, 51(6): 5312-5323.

[10]KHAN W A, BI T, JIA K. A review of single phase adaptive auto-reclosing schemes for EHV transmission lines[J]. Protection and Control of Modern Power Systems, 2019, 4(3): 1-10.

[11]吴丽芳, 白浩, 欧阳健娜, 等. 基于接地变压器绕组分级调节的配电网对地电容精确测量方法[J]. 电力科学与技术学报, 2022, 37(05): 109-114.

[12]全国高电压实验技术标准化分技术委员会. 中性点不接地系统电容电流测试规程: DL/T 308—2012[S]. 北京: 中国电力出版社, 2012: 5-6.

[13]薛永端, 李广, 徐丙垠. 利用熄弧后暂态信息测量谐振接地系统的对地电容[J]. 电工技术学报, 2020, 35(07): 1521-1528.

[14]杨磊, 曾祥君, 喻锟, 等. 新型谐振接地系统接地故障全补偿方法[J]. 电力自动化设备, 2018, 38(11): 57-62.

[15]刘宝稳, 王晨雨, 曾祥君, 等. 三相分布参数不对称配电线路接地故障检测与消弧技术综述[J]. 高电压技术, 2023, 49(09): 3684-3695.

[16]马亚琦, 蔡杰, 林俊超, 等. 基于改进最大位移电压法测量谐振接地系统电容电流[J]. 电力电容器与无功补偿, 2017, 38(05): 126-129+135.

[17]张帆, 薛永端, 徐丙垠. 基于信号注入的谐振接地系统故障点残流测算方法[J]. 电力系统保护与控制, 2023, 51(13): 132-138.

[18]WU H, SUN Y, LIU X, et al. The Signal Injection Method with High-Frequency for Distribution Network Capacitive Current Measurement[J]. Journal of Electrical Engineering & Technology, 2021, 16: 867-872.

[19]李忠奎, 吴文臻, 张子良, 等.煤矿井下机电设备保护和动作特性试验装置设计研究[J]. 煤炭工程, 2023, 55(04): 187-192.

[20]ZENG X, YU K, WANG Y, et al. A novel single phase grounding fault protection scheme without threshold setting for neutral ineffectively earthed power systems[J]. CSEE Journal of Power and Energy Systems, 2016, 2(3): 73-81.

[21]刘扬, 刘建功, 王毅颖, 等. 煤矿坚强智能电网建设理论与技术探讨[J]. 煤炭学报, 2020, 45(06): 2296-2307.

[22]王国法, 刘峰, 庞义辉, 等. 煤矿智能化——煤炭工业高质量发展的核心技术支撑[J]. 煤炭学报, 2019, 44(02): 349-357.

[23]周婷, 解绍锋. 电气化铁路新型电缆供电方案[J]. 电力自动化设备, 2018, 38(07): 189-195+206.

[24]电力工业部电力规划设计总院. 电力系统设计手册[M]. 北京: 中国电力出版社, 2018: 190-191.

[25]王炳. 铁路10kV电力系统电容电流计算探讨[J]. 铁道标准设计, 2010, (S1): 153-155.

[26]刘群杰, 樊艳芳, 侯俊杰, 等. 基于开关控制的风电场集电线路复合中性点接地方式研究[J]. 太阳能学报, 2019, 40(09): 2424-2432.

[27]许京遥. 中性点不接地系统电容电流数据管理系统的设计与实现[D]. 成都: 电子科技大学, 2021.

[28]雷志鹏, 李蔚, 何庆辉, 等. 基于介电响应法的矿用电缆绝缘性能评估方法研究[J]. 工矿自动化, 2023, 49(01): 46-55.

[29]王彦文, 张旭然, 高彦, 等. 三芯矿用电缆线芯温度计算及故障预警方法[J]. 煤炭学报, 2023, 48(03): 1439-1448.

[30]苏继锋. 配电网中性点接地方式研究[J]. 电力系统保护与控制, 2013, 41(08): 141-148.

[31]李建忠. 煤矿电网安全分析研究[D]. 徐州: 中国矿业大学, 2021.

[32]吴君, 朱瑞卿. 煤矿井下高压电网中性点接地方式与漏电保护方案研究[J]. 矿山机械, 2014, 42(08): 120-122.

[33]陈景涛. 中压电缆系统接地电容电流计算方法[J]. 电力勘测设计, 2015, (02): 41-43.

[34]代国勇. 铁路10kV电力系统的电容电流计算分析[J]. 科技创新与生产力, 2017, (02): 92-94.

[35]张红涛, 王星, 蔡文龙, 等. 煤矿10kV供电系统电容电流计算方法[J]. 西安科技大学学报, 2016, 36(02): 282-287.

[36]赵陆尧, 卢睿, 李全皎, 等. 基于实数编码遗传算法的海底电缆电容电流实用计算方法[J]. 科技与创新, 2023, (06): 108-113.

[37]耿蒲龙. 矿井电缆交流杂散电流产生机理及分布规律研究[D]. 太原: 太原理工大学, 2018.

[38]WU W, YANG J, LI L, et al. Residual life prediction of mining cables based on RBF neural network[C]//Journal of Physics: Conference Series. IOP Publishing, 2021, 1754(1): 012194.

[39]徐政, 李斯迅, 金砚秋. 海底电缆谐波频段电气参数计算[J]. 电力系统自动化, 2021, 45(21): 76-84..

[40]姚家祎. 高压电动机磁平衡差动保护[J]. 建筑电气, 2017, 36(03): 29-32.

[41]咸日常, 高鸿鹏, 孙晓维, 等. 基于雷电耐受能力的配电变压器选型策略[J]. 科学技术与工程, 2021, 21(24): 10270-10274.

[42]佟永焕. 动车组高压电气设备绝缘故障机理研究及在线监测系统设计[D]. 北京: 北京交通大学, 2023.

[43]王光辉, 彭勇, 代伟, 等. 基于灵敏度分析与增强捕食-食饵优化的重介质选煤过程动态模型[J]. 煤炭学报, 2021, 46(09): 2813-2823.

[44]HASAN K N, PREECE R, MILANOVIC J V. Priority ranking of critical uncertainties affecting small-disturbance stability using sensitivity analysis techniques[J]. IEEE Transactions on Power Systems, 2017, 32(4): 2629-2639.

[45]YE K T, ZHAO J B, HUANG C, et al. A data-driven global sensitivity analysis framework for three-phase distribution system with PVs[J]. IEEE Transactions on Power Systems, 2021, 36(5): 4809-4819.

[46]李景禄. 实用配电网技术[M]. 北京: 中国水利水电出版社, 2006: 142-143.

[47]《煤矿用电缆》编辑组. 煤矿用电缆[M]. 北京: 煤炭工业出版社, 2012: 100-101.

[48]GUO L, LEI Y, XING S, et al. Deep convolutional transfer learning network: A new method for intelligent fault diagnosis of machines with unlabeleddata[J]. IEEE Transactions on Industrial Electronics, 2019, 66(9): 7316-7325..

[49]兰健, 郭庆来, 周艳真, 等. 基于生成对抗网络和模型迁移的电力系统典型运行方式样本生成[J]. 中国电机工程学报, 2022, 42(08): 2889-2900.

[50]WANG J Z, ZHOU Q P, JIANG H P, et al.Short-term wind speed forecasting using support vector regression optimized by cuckoo optimization algorithm[J]. Mathematical Problems in Engineering, 2015, 2015(1): 1-13.

[51]YIN Z Y, HOU J. Recent advances on SVM based fault diagnosis and process monitoring in complicated industrial processes[J]. Neurocomputing, 2016, 174: 643-650.

[52]LV G Y, CHENG H Z, ZHAI H B, et al. Fault diagnosis of power transformer based on multi-layer SVM classifier[J]. Electric Power Systems Research, 2005, 75(1): 9-15.

[53]REN L K, LV W M, JIANG S W, et al. Fault diagnosis using a joint model based on sparse representation and SVM[J]. IEEE Transactions on Instrumentation and Measurement, 2016, 65(10): 2313-2320.

[54]POLI R, KENNEDY J, BLACKWELL T. Particle swarm optimization[J]. Swarm Intelligence, 2007, 1(1): 35-57.

[55]ETATRABY A I A, WAHBA R R T. Features selectionusing parametric and non- parametric methods: tag snps selection using GA-SVM and GA-KNN[J]. Advances and Applications in Statistics, 2015, 45(2): 105.

[56]肖白, 邢世亨, 王茂春, 等. 基于改进KDE法和GA-SVM的多风电场聚合后输出功率长期波动特性预测方法[J]. 电力自动化设备, 2022, 42(02): 77-84.

[57]MIRJALILI S, LEWIS A. The whale optimization algorithm[J]. Advances in Engineering Software, 2016, 95(6): 51-67.

[58]张君. 基于WOA-FOPID算法的钻锚机器人机械臂运动控制研究[J]. 煤炭科学技术, 2022, 50(06): 292-302.

[59]秋兴国, 王瑞知, 张卫国, 等. 基于混合策略改进的鲸鱼优化算法[J]. 计算机工程与应用, 2022, 58(01): 70-78.

[60]ZHANG J, WANG J S. Improved whale optimization algorithm based on nonlinear adaptive weight and golden sine operator[J]. IEEE Access, 2020, 8: 77013-77048.

[61]徐航, 张达敏, 王依柔, 等. 混合策略改进鲸鱼优化算法[J]. 计算机工程与设计, 2020, 41(12): 3397-3404.