地质灾害的监测和预警是防灾减灾的重要前提。GNSS（Global Navigation Satellite System）因具有全天候、测站间无需通视、提供实时三维位置信息等优势，被广泛应用于监测工作。RTK（Real-time Kinematic）技术因具有出色的实时性和监测精度而广泛用于地质灾害监测预警中。但在复杂的地质灾害环境中，由于大气条件活跃，同时受山地、植被遮挡等因素影响，GNSS信号极易受多路径效应干扰，特别是新出现观测卫星时，卫星信号较弱，容易出现粗差，形变序列会出现监测不连续、异常值多等问题，难以获取可靠的监测结果。针对以上在地质灾害监测过程中遇到的问题，本文提出一种抗差部分模糊度固定的RTK方法，该方法利用抗差技术和部分模糊度固定策略，能够同时提升模糊度固定率和RTK定位精度，此外，利用第五代大气再分析数据集（ECMEF Reanalysis v5，ERA5）进一步削弱大高差背景下对流层延迟对RTK定位高程方向的影响，保证地质灾害监测的定位精度、鲁棒性和可靠性。本文的主要研究内容如下：

（1）针对GNSS不同系统不同频率的监测精度不同，基于信噪比、多路径效应开展数据质量分析，并对不同频率、不同系统进行组合，获取地质灾害背景下精度最高的RTK监测结果。受限于RTK定位精度随基线长度增加而下降的特点，导致只有短基线条件下能达到mm级精度。因此，基于地质灾害监测数据，从定位精度、模糊度固定率、卫星可见数、精度衰减因子（Dilution of precision，DOP）进行分析，得到不同基线长度下精度最优组合。结果表明，单系统下BDS（B1+B3）定位精度优于GPS（L1+L2）和BDS（B1+B2），组合系统GPS（L1+L2）+ BDS（B1+B3）可以进一步提高定位精度，弥补RTK技术在地灾监测过程中受环境影响导致精度不足。

（2）针对多路径误差干扰RTK定位精度及固定率，采用不同抗差方法和部分模糊度固定方法改善RTK定位结果，并生成抗差部分模糊度固定算法。对伪距单点定位进行抗差处理，采用IGGIII等价权函数和零速约束的卡尔曼滤波提高伪距单点定位精度。进一步对相对定位中卡尔曼滤波分先验后验进行二次抗差处理，分别采用IGGIII、IGGI、Huber、Hample、Andrew等价权函数进行抗差，并对同一等价权函数选取不同阈值进行对比，结果表明：Huber等价权函数对精度提升最为明显，随着阈值增加，抗差后RTK定位精度提升，模糊度固定率提升。采用基于卫星高度角的部分模糊度固定算法和基于成功率和Ratio为指标的部分模糊度固定算法进一步提高RTK定位模糊度固定率，并与全模糊度固定算法进行对比。结果表明，2种部分模糊度算法固定率均在95%以上，基于成功率和Ratio为指标的部分模糊度固定算法在短基线拥有更好地精度提升效果，基于卫星高度角的部分模糊度固定算法不受基线长度影响，在10km以上基线的定位结果仍能保证精度。最后将抗差和部分模糊度相结合，形成抗差部分模糊度固定算法，基线长11.8km时，采用该算法精度较抗差方法在E、N、U三个方向上分别提高了19.6%，14.1%，8.6%，较部分模糊度方法在E、N、U三个方向上分别提高了34.9%，29.3%，49.7%，该方法模糊度固定率较传统RTK方法从21.8%提升到了100%。

（3）采用ERA5大气再分析资料计算RTK对流层延迟，并与抗差部分模糊度固定算法结合，生成一种ERA5约束的抗差部分模糊度固定RTK算法。结果表明，ERA5约束后RTK对流层湿延迟在高差为140m、189m、229m分别提升了82.2%，90.5%，94.8%，高程方向精度分别提升了2.3%，5.8%，10.7%。将ERA5约束的抗差部分模糊度固定RTK方法应用在地质灾害监测上，结果表明，该方法可以有效提升定位精度，当基线为6km，高差为140m时，传统RTK方法8小时解算水平和高程方向能达到mm级精度，本文方法6小时就可以达到mm级精度，提高了RTK在地质灾害监测领域的精度和可用性。

The monitoring and early warning of geological disasters is an important prerequisite for disaster prevention and mitigation. GNSS (Global Navigation Satellite System) is widely used in monitoring work because of its advantages of all-weather, no inter-station visibility, and real-time three-dimensional position information. RTK (Real-time Kinematic) technology is widely used in geological disaster monitoring and early warning because of its excellent real-time performance and monitoring accuracy. However, in the complex geological disaster environment, due to the active atmospheric conditions and the influence of factors such as mountains and vegetation occlusion, GNSS signals are easily interfered by multipath effects. Especially when new observation satellites appear, satellite signals are weak and prone to gross errors. Deformation sequences will have problems such as discontinuous monitoring and many outliers, and it is difficult to obtain reliable monitoring results. In view of the above problems encountered in the process of geological disaster monitoring, this paper proposes a robust partial ambiguity fixing RTK method. This method uses robust technology and partial ambiguity fixing strategy to improve the ambiguity fixing rate and RTK positioning accuracy at the same time. In addition, the fifth generation of atmospheric reanalysis data set (ECMEF Reanalysis v5, ERA5) is used to further weaken the influence of tropospheric delay on the elevation direction of RTK positioning under the background of large height difference, and ensure the positioning accuracy, robustness and reliability of geological disaster monitoring. The main research contents of this paper are as follows :

(1) Based on the signal-to-noise ratio and multipath effect, the data quality analysis is carried out, and different frequencies and different systems are combined to obtain the RTK monitoring results with the highest accuracy under the background of geological disasters. Limited by the fact that RTK positioning accuracy decreases with the increase of baseline length, mm-level accuracy can be achieved only under short baseline conditions. Therefore, based on the monitoring data of geological disasters, the optimal combination of accuracy under different baseline lengths is obtained by analyzing the positioning accuracy, ambiguity fixed rate, satellite visibility and dilution of precision (DOP). The results show that the positioning accuracy of BDS (B1 + B3) under single system is better than that of GPS (L1 + L2) and BDS (B1 + B2). The combined system GPS (L1 + L2) + BDS (B1 + B3) can further improve the positioning accuracy and make up for the lack of accuracy caused by the environmental impact of RTK technology in the process of ground disaster monitoring.

(2) Aiming at the multipath error interference RTK positioning accuracy and fixed rate, different robust methods and partial ambiguity fixing methods are used to improve the RTK positioning results, and a robust partial ambiguity fixing algorithm is generated. The robust processing of pseudo-range single point positioning is carried out, and the IGGIII equivalent weight function and zero-speed constraint Kalman filter are used to improve the accuracy of pseudo-range single point positioning. Further, the Kalman filter in the relative positioning is divided into a priori posterior for secondary robust processing. The IGGIII, IGGI, Huber, Hample, and Andrew equivalent weight functions are used for robustness, and the same equivalent weight function is selected. Different thresholds are compared. The results show that the Huber equivalent weight function has the most obvious improvement in accuracy. As the threshold increases, the RTK positioning accuracy after robustness is improved, and the ambiguity fixed rate is improved. The partial ambiguity fixing algorithm based on satellite elevation angle and the partial ambiguity fixing algorithm based on success rate and Ratio are used to further improve the ambiguity fixing rate of RTK positioning, and compared with the full ambiguity fixing algorithm. The results show that the fixed rates of the two partial ambiguity algorithms are all above 95 %. The partial ambiguity fixing algorithm based on success rate and Ratio has better accuracy improvement effect on short baselines. The partial ambiguity fixing algorithm based on satellite elevation angle is not affected by baseline length, and the positioning results of baselines above 10 km can still ensure accuracy. Finally, the robust and partial ambiguity are combined to form a robust partial ambiguity fixing algorithm. When the baseline length is 11.8 km, the accuracy of the algorithm is 19.6 %, 14.1 %, and 8.6 % higher than that of the robust method in the E, N, and U directions, respectively. Compared with the partial ambiguity method, it is increased by 34.9 %, 29.3 %, and 49.7 % in the E, N, and U directions, respectively. The ambiguity fixing rate of this method is improved from 21.8 % to 100 % compared with the traditional RTK method.

(3) The tropospheric delay of RTK is calculated by ERA5 atmospheric reanalysis data, and combined with the robust partial ambiguity resolution algorithm to generate an ERA5 constrained robust partial ambiguity resolution RTK algorithm. The results show that the tropospheric wet delay of RTK after ERA5 constraint is increased by 82.2 %, 90.5 % and 94.8 % respectively at 140 m, 189 m and 229 m, and the accuracy in elevation direction is increased by 2.3 %, 5.8 % and 10.7 % respectively. The robust partial ambiguity fixed RTK method constrained by ERA5 is applied to geological disaster monitoring. The results show that this method can effectively improve the positioning accuracy. When the baseline is 6km and the height difference is 140m, the traditional RTK method can achieve mm-level accuracy in the horizontal and elevation directions in 8 hours. The method in this paper can achieve mm-level accuracy in 6 hours, which improves the accuracy and availability of RTK in the field of geological disaster monitoring.

[1] 殷跃平. 中国地质灾害减灾战略初步研究[J]. 中国地质灾害与防治学报, 2004, 15(2): 1-8.

[2] 谭行, 蒋健, 陈炜, 等. 分布式 GNSS 地质灾害实时监测系统设计[J]. 测绘地理信息, 2020, 45(6): 118-123.

[3] 国家统计局. 中国统计年鉴[M]. 北京: 中国统计出版社, 2023.

[4] 王利, 许豪, 舒宝. 滑坡灾害 GNSS 监测技术研究进展与展望[J]. 导航定位与授时, 2023, 10(1): 12-26.

[5] 朱真, 江思义, 刘小明, 等. 基于广播 RTK 边缘计算的北斗高精度地质灾害监测系统及应用分析[J]. 水文地质工程地质, 2021, 48(5): 176-183.

[6] 王慧敏, 罗忠行, 肖映城, 等. 基于 GNSS 技术的高速公路边坡自动化监测系统[J]. 中国地质灾害与防治学报, 2020, 31(6): 60-68.

[7] 王霞迎. BDS/GPS 基线解算算法研究[D]. 北京: 中国测绘科学研究院, 2014.

[8] Piras M , Roggero M , Fantino M .Monitoring of land deformation using GNSS[J]. Advances in Geosciences, 2009, 13(4): 120-142.

[9] 王利, 张勤, 管建安, 等. 基于 GPS 技术的滑坡动态变形监测试验结果与分析[J]. 武汉大学学报 (信息科学版), 2011, 36(4): 422-426.

[10] Zhang B C , Odijk D .A method for processing GNSS data from regional reference networks to enable single-frequency PPP-RTK[J].Chinese Journal of Geophysics, 2015, 58(7):2306-2319.

[11] Rawat M S , Joshi V , Rawat B S ,et al.Landslide movement monitoring using GPS technology: A case study of Bakthang landslide, Gangtok, East Sikkim, India[J].Journal of Development and Agricultural Economics, 2011, 3:194-200.

[12] 王利, 张勤, 黄观文, 等. GPS PPP 技术用于滑坡监测的试验与结果分析[J]. 岩土力学, 2014, 35(7): 2118-2124.

[13] 张小红,李星星,李盼. GNSS精密单点定位技术及应用进展 [J]. 测绘学报, 2017, 46 (10): 1399-1407.

[14] Aponte J , Meng X , Hill C ,et al.Quality assessment of a network-based RTK GPS service in the UK[J].Journal of Applied Geodesy, 2009, 3(1):25-34.

[15] Du Y , Huang G , Zhang Q ,et al.Asynchronous RTK Method for Detecting the Stability of the Reference Station in GNSS Deformation Monitoring[J].Sensors, 2020, 20(5):1320.

[16] 李征航，黄劲松．GPS测量与数据处理[M]． 武汉: 武汉大学出版社, 2009：15-168.

[17] 张勤, 白正伟, 黄观文等.GNSS滑坡监测预警技术进展[J]. 测绘学报, 2022, 51(10): 1985-2000.

[18] Miklius A, Iwatsubo E Y, Denlinger R P, et al. GPS measurements on the island of Hawaii in 1992[R]. Reston: US Geological Survey, 1994

[19] Ashkenazi V, Bingley R, Dodson A, et al. GPS monitoring of vertical land movements in the UK[C]//Proceedings of the 11th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GPS 1998). 1998: 99-107.

[20] Lovse J W, Teskey W F, Lachapelle G, et al. Dynamic Deformation Monitoring of Tall Structure Using GPS Technology[J]. Journal of Surveying Engineering,1995,121(1):35-40.

[21] Brunner F K, Hartinger H, Richter B. Continuous monitoring of landslides using GPS: A progress report[J]. Geophysical Aspects of Mass Movements, edited by: Bauer, SJ and Weber, F., Austrian Academy of Sciences, Vienna, 2000: 51-60.

[22] Gili J A, Corominas J, Rius J. Using Global Positioning System techniques in landslide monitoring[J]. Engineering Geology, 2000, 55 (3):167-192.

[23] Malet J P , Maquaire O , Calais E .The use of Global Positioning System techniques for the continuous monitoring of landslides: application to the Super-Sauze earthflow (Alpes-de-Haute-Provence, France)[J].Geomorphology, 2002, 43(1-2):33-54.

[24] Baldi P., Casula G., Cenni N. GPS-based monitoring of land subsidence in the Po Plain (NorthemItaly)[J]. Earth and Planetary Science Letters, 2009, 288:204-212.

[25] 陆业海, 吴定洪. GPS 空间测量新技术在岩崩滑坡中的应用[J]. 大坝观测与土工测试, 1995, 19(6): 33-36.

[26] 李劲峰. GPS 应用于监测岩崩滑坡[J]. 长江流域资源与环境, 1996, 5(3): 284-288.

[27] 张勤.GPS监测滑坡形变的基准研究[J]. 地球科学与环境学报, 2001, 23(4): 69-71.

[28] 徐绍铨, 程温鸣, 黄学斌, 等. GPS 用于三峡库区滑坡监测的研究[J]. 水利学报, 2003, 1(1): 155-159.

[29] 王利, 张勤, 李寻昌, 等. GPS RTK 技术用于滑坡动态实时变形监测的研究[J]. 工程地质学报, 2011, 19(2): 193-198.

[30] 张伟斌, 金永新, 袁本银. 基于 BeiDou/GPS 双系统的三峡滑坡监测预警系统[J]. 测绘通报, 2013, (2): 120-121.

[31] 邵伯兴,彭家意,润扬长江公路大桥GPS+北斗变形监测系统应用探索[J].现代交通技术,2014, 11(4):32-34.

[32] 王利, 张勤, 范丽红,等.北斗/GPS融合静态相对定位用于高精度地面沉降监测的试验与结果分析[J]. 工程地质学报, 2015, 23(1): 119-125.

[33] 张飞跃, 汪剑, 等.北斗导航定位技术在溪洛渡水电站库区边坡监测中的应用[J]. 大坝与安全, 2016, (2): 1-5.

[34] 黄观文, 黄观武, 杜源, 等. 一种基于北斗云的低成本滑坡实时监测系统[J]. 工程地质学报, 2018, 26(4): 1008-1016.

[35] 李振. 滑坡环境下 BDS 三频 PPP 精度分析[J].测绘标准化,2021,37(3):22-25.

[36] Zhang Q, Chen M, Wu J, et al. Feasibility Verification of Virtual Reference Station Technology in Geological Hazard Monitoring[J]. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 2021, 43: 235-240.

[37] 杨元喜.抗差估计理论及其应用[M]. 八一出版社, 1993.

[38] 杨元喜.相关观测抗差估计[C]//抗差估计论文集. 北京: 测绘出版社, 1992: 14-21.

[39] 刘经南, 姚宜斌, 施闯. 基于等价方差-协方差阵的稳健最小二乘估计理论研究[J]. 测绘科学, 2000, 25(3): 1-6.

[40] Huber P J. A robust version of the probability ratio test[J]. The Annals of Mathematical Statistics, 1965: 1753-1758.

[41] Huber P J. Robust Estimation of a Location Parameter[M]//Breakthroughs in Statistics. Springer New York, 1992:73-101.

[42] Tukey J W. Study of Robustness by Simulation:Particularly Improvement by Adjustment and Combination[J]. Robustness in Statistics, 1979:75-102.

[43] Huber P J. The 1972 wald lecture robust statistics: A review[J]. The Annals of Mathematical Statistics, 1972, 43(4): 1041-1067.

[44] Hampel F R. Contributions to the theory of robust estimation[M]. University of California, Berkeley, 1968.

[45] Krarup T, Kubik K. The Danish method: Experience and philosophy[C]//Deut. Geodact. Komm. Seminar on Math. Models of Geodetic Photogrammetric Point Determination with Regard to Outliers and Systematic Errors, 1983: 131-134.

[46] 周江文. 经典误差理论与抗差估计[J]. 测绘学报, 1989, 18(2): 115-120.

[47] 周江文, 王跃进．抗差拟合推估[C]//抗差估计论文集，北京，测绘出版社，1992:41-50.

[48] 杨元喜. 测量平差模型的抗差最小二乘解及影响函数[J]. 解放军测绘学院学报, 1994, (2): 77-82.

[49] 欧吉坤. 一种三步抗差方案的设计[J]. 测绘学报, 1996, 25(3): 173-179.

[50] Koch KR and Yang Y 1998, Robust Kalman filter for rank deficient obsservation model,Journal of Geodesy,72(8),436-441.

[51] 高晓, 戴吾蛟, 潘林. 抗差 EKF 算法及其在 GPS/BDS 动态相对定位中的应用[J]. 大地测量与地球动力学, 2014, 34(4): 140-144.

[52] 蔡保杰,邵雷. 三段判别域与最小二乘拟合的抗差滤波算法[J]. 系统工程与电子技术, 2021, 43(5): 1346-1353.

[53] Remondi B W. Using the Global Positioning System (GPS) phase observable for relative geodesy: Modeling, processing, and results (Satellite, Timing)[M]. The University of Texas at Austin, 1984.

[54] Teunissen P J G. A new method for fast carrier phase ambiguity estimation[C]//Proceedings of 1994 IEEE Position, Location and Navigation Symposium-PLANS'94. IEEE, 1994: 562-573.

[55] Teunissen P J G. The distribution of the GPS baseline in case of integer least-squares ambiguity estimation[J]. Artif Satell, 1998, 33(2): 65-75.

[56] Teunnissen P J G. The least-square ambiguity decorrelation adjustment: a method for fast GPS integer ambiguity estimation[J]. J Geodesy, 1995, 70(1): 65-82.

[57] EULER H-J, SCHAFFRIN B. On a measure for the discernibility between different ambiguity solutions in the static-kinematic GPS-mode [C]//Proceedings of the Kinematic Systems in Geodesy, Surveying, and Remote Sensing. Canada, 1990: 285-295.

[58] E Frei,G Beutler. Rapid Static Positioning Based on the Fast Ambiguity Resolution Approach FARA:Theory and First Results[J].Manuscripta Geodaetica，1990,15(4):325-356.

[59] Wang J,Stewart M P , Tsakiri M. A discrimination test procedure for ambiguity resolution on-the-fly[J]. Journal of Geodesy, 1998, 72(11):644-653.

[60] Verhagen S, Teunissen P J G. New global navigation satellite system ambiguity resolution method compared to existing approaches[J]. Journal of guidance, control, and dynamics, 2006, 29(4): 981-991.

[61] Verhagen S, Teunissen P J G, van der Marel H, et al. GNSS ambiguity resolution: which subset to fix[C]//IGNSS symposium. 2011: 15-17.

[62] Teunissen P J G, Joosten P, Tiberius C. Geometry-free ambiguity success rates in case of partial fixing[C]//Proceedings of the 1999 national technical meeting of the institute of navigation. 1999: 201-207.

[63] Hsu,Li-Ta.Analysis and modeling GPS NLOS effect in highly urbanized area[J].GPS Solutions, 2018, 22(1):1-12.

[64] Odijk D , Teunissen P J G .ADOP in closed form for a hierarchy of multi-frequency single-baseline GNSS models[J].Journal of Geodesy, 2008, 82(8):473-492.

[65] Zhang S , Zhao L , Li X ,et al.A sequential and partial ambiguity resolution strategy for improving the initialization performance of medium-baseline relative positioning[J].Earth Planets & Space, 2016, 68(1):29.

[66] Henkel P, Günther C. Partial integer decorrelation: optimum trade-off between variance reduction and bias amplification[J]. Journal of geodesy, 2010, 84: 51-63.

[67] Verhagen S , Teunissen P J G .The ratio test for future GNSS ambiguity resolution[J].GPS Solutions, 2013,17(4):535-548.

[68] 曾琪. BDS/GPS 组合动态相对定位及其质量控制方法研究[D]. 武汉: 武汉大学, 2017.

[69] Brack A. Reliable GPS+BDS RTK positioning with partial ambiguity resolution[J]. GPS Solutions, 2017, 21:1-10.

[70] Li G, Geng J, Guo J, et al. GPS+ Galileo tightly combined RTK positioning for medium-to-long baselines based on partial ambiguity resolution[J]. The Journal of Global Positioning Systems, 2018, 16: 1-10.

[71] 邵晓东, 陈源军, 史小雨等.分步组合部分模糊度固定法在中短基线中的应用[J].测绘科学,2018,43(11):15-19.

[72] 刘伟, 李长庚, 满小三.基于部分模糊度解算的BDS/GPS组合RTK算法[J].测绘工程, 2018, 27(9): 11-15.

[73] Liu X, Zhang S, Zhang Q, et al. A fast satellite selection algorithm with floating high cut-off elevation angle based on ADOP for instantaneous multi-GNSS single-frequency relative positioning[J]. Advances in Space Research, 2019, 63(3): 1234-1252.

[74] 陈文涛, 程鹏飞, 徐彦田, 赵硕, 刘扬. 部分模糊度固定的GPS/BDS中长基线RTK算法研究[J].测绘科学，2020,45(11):28-32.

[75] 耿江辉, 常华, 郭将, 栗广才, 魏娜.面向城市复杂环境的3种多频多系统GNSS单点高精度定位方法及性能分析[J]. 测绘学报, 2020, 49(1): 1-13.

[76] 高乾.基于BDS/GPS 融合定位的部分模糊度解算方法研究[D]. 徐州: 中国矿业大学,2021.

[77] Li Z , Xu G , Guo J ,et al.A sequential ambiguity selection strategy for partial ambiguity resolution during RTK positioning in urban areas[J].GPS Solutions, 2022, 26(3):1-16.

[78] 柴海山, 陈克杰, 魏国光, 等. 北斗三号与超高频 GNSS 同震形变监测: 以 2021 年青海玛多 Mw 7.4 地震为例[J]. 武汉大学学报 (信息科学版), 2022, 47(6): 946-954.

[79] 李建勋,刘明娴,苑泽鹏. 基于RTK技术的低成本、高精度单频接收机性能分析 [J]. 内江科技, 2020, 41 (3): 16-17.

[80] 叶长青,范存国. GPS跨河水准测量技术探讨 [J]. 科技信息, 2012, (3): 92-93.

[81] 刘瑶. 应用于低轨导航增强卫星系统的实时高精度定位算法研究[D]. 南京: 南京航空航天大学, 2018.

[82] 李国伟, 郭金运, 原永东等. GPS测站多路径效应建模研究 [J]. 测绘科学, 2013, 38 (3): 7-9+21.

[83] 马驰. RTK数据处理及接收机测量数据质量影响研究[D]. 长沙: 国防科学技术大学, 2014.

[84] 成伟. GPS精密定位误差及其相位多路径误差分析研究[D]. 西安: 长安大学, 2003.

[85] Kamatham Y , Sarma A D , Kumar A ,et al.Spectral Analysis and Mitigation of GPS Multipath Error Using Digital Filtering for Static Applications[J].IETE Journal of Research, 2013, 59(2):156.

[86] 李成钢. 网络 GPS/VRS 系统高精度差分改正信息生成与发布研究[D]. 成都: 西南交通大学, 2007.

[87] 胡楠楠, 章红平, 李团, 等. 城市动态环境下 GNSS RTK 部分模糊度固定算法性能分析[J]. 大地测量与地球动力学, 2018, 38(3): 263-267.

[88] Chang X W , Yang X , Zhou T .MLAMBDA: a modified LAMBDA method for integer least-squares estimation[J].Journal of Geodesy, 2005, 79(9):552-565.

[89] 姚宜斌, 张良, 张琦, 等. 面向大高差 RTK 的对流层延迟改正模型及实时差分服务构建[J]. 武汉大学学报 (信息科学版), 2023, 48(7): 1019-1028.

[90] 夏赛. 基于ERA5和GNSS数据的区域对流层延迟建模与应用[D]. 南京: 南京信息工程大学, 2024.