在安防系统中利用RFID定位技术可以获取大量目标轨迹数据。通过数据挖掘技术对采集的目标轨迹数据进行分析，可以挖掘出目标行为模式、找出安防系统防御薄弱点、增强安防系统内部防控。目前，针对安防系统中目标轨迹挖掘的相关研究还处于初步阶段。因此，提出模糊语义轨迹频繁模式挖掘方法和基于蚁群算法的高效用语义轨迹模式挖掘方法。将这两种方法应用于某一体化安防系统的RFID定位子系统中，并设计开发基于RFID定位技术的目标轨迹挖掘平台。主要包括以下研究内容：

针对以精确停留时间定义的语义轨迹频繁模式挖掘方法在实际应用场景中局限性较大的问题，提出模糊语义轨迹频繁模式挖掘方法。首先定义了模糊停留时间隶属度函数，从而将目标在停留点的停留时间模糊化，得到模糊语义轨迹。然后提出模糊语义轨迹频繁模式挖掘算法FST-FPM(Fuzzy Semantic Trajectory Frequent Pattern Mining)。分别在公开数据集GeoLife和某一体化安防系统的RFID定位数据集上对FST-FPM算法进行了实验验证。实验结果表明：FST-FPM算法在GeoLife数据集和RFID定位数据集上可以挖掘出模糊语义轨迹频繁模式，并且运行时间比PrefixSpan算法、PrefixSpan-x算法和LFFT2算法减少了10%以上。

针对现有频繁模式挖掘方法未考虑停留点价值差异的问题，提出基于蚁群算法的高效用语义轨迹模式挖掘方法。首先综合停留点兴趣度、目标在停留点的停留时间以及目标语义轨迹的支持度这三个参数定义了目标语义轨迹效用值的概念。在此基础上提出高效用语义轨迹模式挖掘算法HUPM-IACO(High Utility Semantic Trajectory Pattern Mining Improved Ant Colony Optimization)。该算法采用精英蚂蚁策略改进了蚂蚁种群的迭代方式，并通过轮盘赌选择法改进了蚂蚁选择下一节点的策略，运用无效用编码向量剪枝策略来提高算法的执行效率。分别在chess、mushroom、foodmart和retail四个公开数据集和某一体化安防系统的RFID定位数据集上对HUPM-IACO算法进行了实验验证。实验结果表明：HUPM-IACO算法在发现高效用语义轨迹模式的数量上比HUPE-GARM算法、HUIM-BPSO算法和HUIM-ACS算法增加了10%以上，在运行时间上比HUPE-GARM算法、HUIM-BPSO算法和HUIM-ACS算法减少了7%以上。

设计基于RFID定位技术的目标轨迹挖掘平台。将上述FST-FPM算法和HUPM-IACO算法应用于某安防系统中的RFID定位子系统，通过对目标轨迹进行模糊语义轨迹频繁模式挖掘和高效用语义轨迹模式挖掘来分析目标的行为规律。该平台采用C#设计开发，实现了安防系统对目标轨迹的模式分析。验证了本文提出的FST-FPM算法与HUPM-IACO算法在某安防系统中的有效性。

In the security system, a large amount of target trajectory data can be obtained by using RFID positioning technology. By analyzing the collected target trajectory data through data mining technology, we can mine target behavior patterns, identify weaknesses of the security system, and enhance the internal prevention and control of the security system. At present, the research on target trajectory mining in security systems is still in the preliminary stage. Therefore, a fuzzy semantic trajectory frequent pattern mining method and a high utility semantic trajectory pattern mining method based on ant colony algorithm are proposed. The two methods are applied to the RFID positioning subsystem of an integrated security system, and a target trajectory mining platform based on RFID positioning technology is designed and developed. It mainly includes the following research contents:

(1) Aiming at the limitation of semantic trace frequent pattern mining method defined by precise residence time in practical application scenarios, a fuzzy semantic trace frequent pattern mining method is proposed. Firstly, the membership function of fuzzy stay time is defined, so the target's stay time at the stay point is fuzzy, and the fuzzy semantic trajectory is obtained. Then, a fuzzy semantic trajectory frequent pattern mining algorithm FST-FPM (fuzzy semantic trajectory frequent pattern mining) is proposed. The FST-FPM algorithm is tested on the Geolife public data set and RFID positioning data set of an integrated security system. Experimental results show that the FST-FPM algorithm can mine frequent patterns of fuzzy semantic trajectories on the Geolife dataset and RFID positioning dataset, and its running time is faster than the PrefixSpan algorithm, Prefixspan-x algorithm and LFFT2 algorithm is reduced by more than 10%.

(2) Aiming at the problem that the existing frequent pattern mining methods do not consider the difference of dwell point value, a high utility semantic trajectory pattern mining method based on ant colony algorithm is proposed. Firstly, the concept of utility value of target semantic trajectory is defined by integrating three parameters: interest degree of stay point, stay time of target at stay point and support degree of target semantic trajectory. On this basis, an high utility semantic trajectory pattern mining algorithm HUPM-IACO is proposed. The algorithm adopts the elite ant strategy to improve the iterative mode of ant population, improves the strategy of ant selecting the next node through roulette selection method, and uses the low efficiency coding vector pruning strategy to improve the execution efficiency of the algorithm. The HUPM-IACO algorithm is experimentally verified on four public data sets: chess, mushroom,foodmart and retail, as well as the self collected RFID positioning data set. The experimental results show that HUPM-IACO algorithm increases the number of high utility semantic trajectory patterns more than 10% compared with HUPM-IACO algorithm, HUIM-BPSO algorithm and HUIM-ACS algorithm, and reduces the running time more than 7% compared with HUPM-IACO algorithm, HUIM-BPSO algorithm and HUIM-ACS algorithm.

(3) A target trajectory mining platform based on RFID positioning technology is designed. The above FST-FPM algorithm and HUPM-IACO algorithm are applied to the RFID positioning subsystem of a security system to analyze the behavior law of the target through frequent pattern mining and high utility semantic trajectory pattern mining. The platform adopts C# design and development, and realizes the mode analysis of target trajectory by security system. The effectiveness of FST-FPM algorithm and HUPM-IACO algorithm in a security system is verified.

[1]Ahmed H M, Essa H S. Survey of intelligent surveillance system for monitoring international border security[J]. Materials Today: Proceedings, 2021, 54(2): 1098-1104.

[2]Bagchi T, Mahapatra A, Yadav D, et al. Intelligent security system based on face recognition and IoT[J]. Materials Today: Proceedings, 2022, 62(4): 2133-2137.

[3]Verma R K, Singh P, Panigrahi C R, et al. ISS: intelligent security system using facial recognition[M]//Progress in Advanced Computing and Intelligent Engineering. Springer, Singapore, 2021: 96-101.

[4]陈博. 基于物联网的实物保护系统关键技术研究[D]. 陕西: 西安科技大学, 2021.

[5]Zhu C, Zhao S, Xia Y, et al. An improved three-point localization method based on RSS for transceiver separation RFID systems[J]. Measurement, 2022, 187(04): 110283.

[6]Hayward S J, Earps J, Sharpe R, et al. A novel inertial positioning update method, using passive RFID tags, for indoor asset localisation[J]. CIRP Journal of Manufacturing Science and Technology, 2021, 35(06): 968-982.

[7]Hwang S, VanDeMark C, Dhatt N, et al. Segmenting human trajectory data by movement states while addressing signal loss and signal noise[J]. International Journal of Geographical Information Science, 2018, 32(7): 1391-1412.

[8]Renso C, Bogorny V, Tserpes K, et al. Multiple-aspect analysis of semantic trajectories (MASTER)[J]. International Journal of Geographical Information Science, 2021, 35(4): 763-766.

[9]Alvares L O, Bogorny V, Kuijpers B, et al. A model for enriching trajectories with semantic geographical information[C]//Proceedings of the 15th annual ACM international symposium on Advances in geographic information systems. 2007: 1-8.

[10]Gao C, Zhao Y, Wu R, et al. Semantic trajectory compression via multi-resolution synchronization-based clustering[J]. Knowledge-Based Systems, 2019, 174(5): 177-193.

[11]Cao Y, Xue F, Chi Y, et al. Effective spatio-temporal semantic trajectory generation for similar pattern group identification[J]. International Journal of Machine Learning and Cybernetics, 2020, 11(2): 287-300.

[12]Zhang W, Wang X, Huang Z. A system of mining semantic trajectory patterns from GPS data of real users[J]. Symmetry, 2019, 11(7): 889.

[13]Zhang D, Lee K, Lee I. Semantic periodic pattern mining from spatio-temporal trajectories[J]. Information Sciences, 2019, 502(04): 164-189.

[14]Ahmed U, Srivastava G, Djenouri Y, et al. Knowledge graph based trajectory outlier detection in sustainable smart cities[J]. Sustainable Cities and Society, 2022, 78(6): 103580.

[15]Li J, Su J. Semantic Trajectory Frequent Pattern Mining Model: The Definitions and Theorems[J]. International Journal on Semantic Web and Information Systems (IJSWIS), 2022, 18(1): 1-20.

[16]Zheng Y. Trajectory data mining: an overview[J]. ACM Transactions on Intelligent Systems and Technology (TIST), 2015, 6(3): 1-41.

[17]Agrawal R, Srikant R. Fast algorithms for mining association rules[C]//Proc. 20th int. conf. very large data bases, VLDB. 1994, 1215: 487-499.

[18]祝孔涛, 李兴建, 王乐. 高效用项集挖掘算法[J]. 计算机工程与设计, 2013, 34(12): 4220-4225.

[19]Tsoukatos I I, Gunopulos D. Efficient mining of spatiotemporal patterns[C]//International Symposium on Spatial and Temporal Databases. Springer, Berlin, Heidelberg, 2001: 425-442.

[20]Xie H. Research and case analysis of apriori algorithm based on mining frequent item-sets[J]. Open Journal of Social Sciences, 2021, 9(04): 458.

[21]Wang H. Correlation analysis of College students’ achievement based on improved Apriori algorithm[C]//Journal of Physics: Conference Series. IOP Publishing, 2021: 012020.

[22]Kusumo H, Marlina D, Novita M, et al. Analysis of transaction patterns at drug store with Apriori Algorithm[C]//Journal of Physics: Conference Series. IOP Publishing, 2021: 012070.

[23]Bagui S, Devulapalli K, Coffey J. A heuristic approach for load balancing the FP-growth algorithm on MapReduce[J]. Array, 2020, 7(4): 100035.

[24]Ji W T, Guo Q J, Lei Y R. HM-AprioriAll Algorithm Improvement Based on Hadoop Environment[J]. Computer and Information Science, 2017, 719(2): 163-174.

[25]Chen J, Zhu L, Guo Y, et al. A mining approach for component abnormal information based on monitor log[J]. International Journal of Simulation and Process Modelling, 2016, 11(5): 353-362.

[26]Campbell E A, Qian T, Miller J M, et al. Identification of temporal condition patterns associated with pediatric obesity incidence using sequence mining and big data[J]. International Journal of Obesity, 2020, 44(8): 1753-1765.

[27]Wang Q, Davis D N, Ren J. Mining frequent biological sequences based on bitmap without candidate sequence generation[J]. Computers in biology and medicine, 2016, 69(2): 152-157.

[28]胡庆顺. 基于改进的位图存储及运算技术的spam算法优化研究与设计[D]. 山东： 齐鲁工业大学, 2019.

[29]Han J W, Pei J, Yan X F. From sequential pattern mining to structured pattern mining: a pattern-growth approach[J]. Journal of Computer Science and Technology, 2004, 19(3): 257-279.

[30]Cai G, Lee K, Lee I. Mining mobility patterns from geotagged photos through semantic trajectory clustering[J]. Cybernetics and Systems, 2018, 49(4): 234-256.

[31]吴瑕, 唐祖锴, 祝园园, 等. 近似到达时间约束下的语义轨迹频繁模式挖掘[J]. 软件学报, 2018, 29(10): 3184-3204.

[32]Davashi R. UP-tree & UP-Mine: A fast method based on upper bound for frequent pattern mining from uncertain data[J]. Engineering Applications of Artificial Intelligence, 2021, 106(4): 104477.

[33]刘春, 周燕, 李鑫. 挖掘语义轨迹频繁模式及拼车应用研究[J]. 计算机工程与应用, 2019, 55(15): 96-103.

[34]薛飞, 单征, 闫丽景, 等. 基于数据挖掘的多轨迹特征检测技术[J]. 计算机科学, 2016, 43(05): 91-95.

[35]Wu T Y, Lin J C W, Yun U, et al. An efficient algorithm for fuzzy frequent itemset mining[J]. Journal of Intelligent & Fuzzy Systems, 2020, 38(5): 5787-5797.

[36]兰志辉. 基于轨迹数据的语义频繁模式挖掘算法研究[D]. 陕西: 西北大学, 2020.

[37]Lin J C W, Zhang Y, Fournier-Viger P, et al. Efficiently updating the discovered multiple fuzzy frequent itemsets with transaction insertion[J]. International Journal of Fuzzy Systems, 2018, 20(8): 2440-2457.

[38]韩萌, 丁剑. 数据流频繁模式挖掘综述[J]. 计算机应用, 2019, 39(03): 719-727.

[39]汤小春, 樊雪枫, 周佳文, 等. 基于数据流的大图中频繁模式挖掘算法研究[J]. 计算机学报, 2020, 43(07): 1293-1311.

[40]Liu Y, Liao W, Choudhary A. A two-phase algorithm for fast discovery of high utility itemsets[C]//Pacific-Asia Conference on Knowledge Discovery and Data Mining. Springer, Berlin, Heidelberg, 2005: 689-695.

[41]Liu Y, Liao W, Choudhary A. A fast high utility itemsets mining algorithm[C]//Proceedings of the 1st international workshop on Utility-based data mining. 2005: 90-99.

[42]Ryang H, Yun U. Top-k high utility pattern mining with effective threshold raising strategies[J]. Knowledge-Based Systems, 2015, 76(1): 109-126.

[43]Ryang H, Yun U. High utility pattern mining over data streams with sliding window technique[J]. Expert Systems with Applications, 2016, 57(2): 214-231.

[44]Kim D, Yun U. Mining high utility itemsets based on the time decaying model[J]. Intelligent Data Analysis, 2016, 20(5): 1157-1180.

[45]Kim D, Yun U. Efficient algorithm for mining high average-utility itemsets in incremental transaction databases[J]. Applied Intelligence, 2017, 47(1): 114-131.

[46]Liu M, Qu J. Mining high utility itemsets without candidate generation[C]//Proceedings of the 21st ACM international conference on Information and knowledge management. 2012: 55-64.

[47]Deng Z H. An efficient structure for fast mining high utility itemsets[J]. Applied Intelligence, 2018, 48(9): 3161-3177.

[48]Lee S, Park J S. Top-k high utility itemset mining based on utility-list structures[C]//2016 International conference on big data and smart computing (BigComp). IEEE, 2016: 101-108.

[49]Duong Q H, Liao B, Fournier-Viger P, et al. An efficient algorithm for mining the top-k high utility itemsets, using novel threshold raising and pruning strategies[J]. Knowledge-Based Systems, 2016, 104(4): 106-122.

[50]Dam T L, Li K, Fournier-Viger P, et al. An efficient algorithm for mining top-k on-shelf high utility itemsets[J]. Knowledge and Information Systems, 2017, 52(3): 621-655.

[51]Dam T L, Li K, Fournier-Viger P, et al. CLS-Miner: efficient and effective closed high-utility itemset mining[J]. Frontiers of Computer Science, 2019, 13(2): 357-381.

[52]Kannimuthu S, Premalatha K. Discovery of high utility itemsets using genetic algorithm with ranked mutation[J]. Applied Artificial Intelligence, 2014, 28(04): 337-359.

[53]Lin J C W, Yang L, Fournier-Viger P, et al. A binary PSO approach to mine high-utility itemsets[J]. Soft Computing, 2017, 21(17): 5103-5121.

[54]Wu J M T, Zhan J, Lin J C W. An ACO-based approach to mine high-utility itemsets[J]. Knowledge-Based Systems, 2017, 116(05): 102-113.

[55]郭世明, 高宏. 基于滑动窗口挖掘数据流高效用项集的有效算法[J]. 哈尔滨工程大学学报, 2018, 39(04): 721-729.

[56]穆晓芳, 邓红霞, 郭虎升, 等. 基于快速高效用项集挖掘的大规模消息流预测算法研究与应用[J]. 计算机应用与软件, 2019, 36(11): 243-249.

[57]程浩东, 韩萌, 张妮, 等. 基于滑动窗口模型的数据流闭合高效用项集挖掘[J]. 计算机研究与发展, 2021, 58(11): 2500-2514.

[58]孙蕊, 韩萌, 张春砚, 等. 含负项top-k高效用项集挖掘算法[J]. 计算机应用, 2021, 41(08): 2386-2395.

[59]Zheng Y, Zhang L, Xie X, et al. Mining interesting locations and travel sequences from GPS trajectories[C]//Proceedings of the 18th international conference on World wide web. 2009: 791-800.