A scene data fusion analysis management system and method based on multi-source heterogeneity

Abstract

The present invention discloses a scene data fusion analysis management system and method based on multi-source heterogeneity, belonging to the field of data fusion technology. The present invention analyzes and extracts features from each collected data for classification, and obtains a multi-source data set in the scene; searches for the same information and different information in the multi-source data, and constructs a common information set and a different information set; analyzes the multi-source information in the common information set, and determines a unique time and space reference system; uses the time and space reference system to mark the time and space of the data information of each multi-source information in the common information set; aligns and fuses the multi-source information in the common information set according to the mark; uses the constructed common and different database to analyze the scene data, uses the different information in the common and different database to obtain different decisions, analyzes each decision, calculates the confidence of the decision, fuses multiple decisions, and obtains a comprehensive decision for the analysis scene.

Description

Technical Field

[0001] The present invention relates to the technical field of data fusion, and in particular to a scene data fusion analysis management system and method based on multi-source heterogeneity. Background Art

[0002] Data fusion technology was first applied in the military field for radar signal processing and target tracking. The original goal was to improve the accuracy of target detection and tracking by integrating data from multiple radars or sensors; with the emergence of multi-sensor systems, data fusion technology has further developed. The focus of this period was how to effectively fuse data from different sensors to provide more reliable and comprehensive information; with the rapid development of computer technology, there are many algorithms for data fusion, for example: probabilistic statistical methods such as Bayesian reasoning and Kalman filtering have become important tools for data fusion, and the empster-shafer evidence theory provides a mathematical framework for dealing with uncertainty and imprecision, which is widely used in multi-sensor data fusion and decision support systems; and now the rapid development of machine learning and deep learning technology has brought new opportunities for data fusion. By training complex models, features can be automatically extracted and fused from multi-source data; however, when fusing and analyzing the data in the scene, on the one hand, the different time and spatial locations of the data may cause the fused data to be unusable, and different data use different reference systems when determining time and space, which makes it easy to misjudge the time and spatial location of multi-source data; and when fusing the current data, only the same information and data are fused and analyzed, and the data with differences is not considered. However, the difference data in the scene may have an important impact on the analysis results, and not considering the difference data will cause the analysis results to lose authenticity. Summary of the invention

[0003] The purpose of the present invention is to provide a scene data fusion analysis management system and method based on multi-source heterogeneity to solve the problems raised in the above background technology.

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0005] A method for fusion analysis and management of scene data based on multi-source heterogeneity, the method comprising the following steps:

[0006] S100, collecting all data in the scene, analyzing and extracting features from each collected data, classifying the data using the data features, and obtaining a multi-source data set in the scene;

[0007] Furthermore, the specific steps of classifying data using data features are:

[0008] S101, collect all data in the scene in the history, and extract all types of information contained in each data , Indicates the 1st, 2nd, 3rd, ..., nth information contained in each collected data, where n is a positive integer; the time interval for data changes in the scene in the collection history is , Indicates the 1st, 2nd, 3rd, ..., mth time intervals at which the data in the scene changes in the collected history, where m is a positive integer; calculates the average time intervals at which the data in the scene changes in the extracted history, and the formula is: In the formula, Tc represents the average time interval of data changes in the historical scene. The calculated average value is used as the period for collecting all data in the historical scene. The data in the historical scene is collected and various types of information are extracted according to the calculated period Tc;

[0009] S102. After collecting the data in the scene according to the collection cycle, extract the information in the data and obtain the total number of each type of information as P. Let the extracted p pieces of information be , Indicates the 1st, 2nd, 3rd, ..., Pth information of each extracted information, and calculates the change speed of each information when the data changes in the historical scene. The formula is: In the formula, It indicates the latter information value when calculating the difference between two adjacent information values. It represents the previous information value when calculating the difference between two adjacent information values, Tc represents the interval period between two adjacent information values, P represents the total number of extracted information, and Su represents the change speed of each type of information calculated;

[0010] The n types of information in each data are calculated, and the change rate of each type of information is obtained as , Indicates the calculated change speed of the 1st, 2nd, 3rd, ..., nth information. Compare the calculated change speeds of each information and select the maximum value among the n information change speeds. As the feature Te of the corresponding data;

[0011] S103, using the method described in S102 to calculate the characteristics of each collected data, the characteristics of all collected data are obtained as follows: , represents the features of the 1st, 2nd, 3rd, ..., uth data collected, and the data with the same features are classified as the same type of data. Finally, the data in the scene are classified as , It represents the 1st, 2nd, 3rd, ..., jth type of data classified in the scene, where j is a positive integer. The multi-source data set in the scene is constructed using the classified data.

[0012] The data is analyzed based on the changing speed of the information contained in each data in the history. The faster the information changes, the more it can reflect the characteristics of the corresponding data. Therefore, the information with the fastest changing speed is classified as the feature of the data. The data in the scene can be clearly distinguished according to different changing characteristics. When analyzing the data in the scene, data can be collected according to different changing characteristics.

[0013] S200, analyzing each type of data in the multi-source data set, finding the same information and different information in the multi-source data, using the same information to construct a common information set, and using the different information to construct a different information set;

[0014] Furthermore, the specific steps of using the same information to construct the common information set and using the different information to construct the different information set are as follows:

[0015] S201, compare the information contained in each data in the multi-source data set, and select the same type of information contained in each data in the multi-source data set as , Indicates the first, second, third, ..., g pieces of information of the same type contained in each type of data in the multi-source data set, where g is a positive integer; constructs a common information set using the collected information of the same type;

[0016] S202, compare the information contained in each data in the multi-source data set, and select different types of information contained in each data in the multi-source data set as , It represents the 1st, 2nd, 3rd, ..., hth pieces of information of different types contained in each data in the multi-source data set, where h is a positive integer; the different types of collected information are used to construct a variance information set.

[0017] S300, analyzing the multi-source information in the common information set to determine a unique time and space reference system; using the time and space reference system to mark the time and space of data information of each type of multi-source information in the common information set;

[0018] Furthermore, the specific steps of using the time and space reference system to mark the time and space of the data information of each multi-source information in the same information set are as follows:

[0019] S301, extract the time of each multi-source data generated in the scene, and obtain the generation time of each data in the multi-source data set: , Indicates the generation time of the 1st, 2nd, 3rd, ..., uth data in the extracted multi-source data set. The generation time of the extracted data is analyzed and the earliest time when the data is generated is used as the time reference system in the multi-source data set. ;

[0020] Use the generated time reference system to timestamp each data generated in the multi-source data set, and calculate the generation timestamp of each data. The formula is:

[0021] In the formula, represents the timestamp when each data in the multi-source dataset is generated, t represents the generation time of each data in the multi-source dataset, and each data in the multi-source dataset is marked with the calculated timestamp. , Indicates that the 1st, 2nd, 3rd, ..., uth data in the multi-source data set are timestamped;

[0022] S302, extract the number of information contained in each data of the multi-source data set, assuming that the number of information contained in each data is , Indicates the number of information contained in the 1st, 2nd, 3rd, ..., uth data in the multi-source data set. Compare the number of u pieces of information, select the spatial coordinates of the information in the data containing the largest number of information as the spatial reference system, number all the information in the data containing the largest number of information, set the position of the initial information to 1, and then number them in sequence to obtain the spatial reference system: , Indicates the position of the 1st, 2nd, 3rd, ..., Dth information in the data containing the most information, where D is a positive integer;

[0023] Extract the information type contained in each data in the multi-source data set and find the location of the same type of information in the spatial reference system , using the location of the same kind of information in a spatial reference system The information contained in each data is marked with a location, and all the information in each data is marked with a location in the spatial reference system. , It means marking the position of the 1st, 2nd, 3rd, ... information in each data.

[0024] When conducting a comprehensive analysis of the data in the scene, it is necessary to fuse different data, and then comprehensively consider and analyze them to determine the time and space reference systems in the multi-source data. Using the same reference system can ensure that when the information to be fused is time-space labeled, the time and space markings are accurate, avoiding different time and space marking rules caused by different reference systems, which may cause fusion errors.

[0025] S400, using the constructed database for seeking common ground while reserving differences to analyze the scenario data, using the divergent information in the database for seeking common ground while reserving differences to obtain different decisions, analyzing each decision, calculating the confidence of the decision, integrating multiple decisions, and obtaining a comprehensive decision for the analyzed scenario.

[0026] Furthermore, the specific steps to obtain the comprehensive decision of the analysis scenario are:

[0027] S401, after selecting the common information set in the multi-source data, extract the timestamp and position of each information in the common information set in the mark, and set the timestamp of each information in the extracted common information set as , represents the timestamps of the 1st, 2nd, 3rd, ..., R information in the same information set. The R timestamps are analyzed and the information with the same timestamps are grouped into a group of fused information sets. After classifying all the timestamps, the fused information set is obtained as , represents the 1st, 2nd, 3rd, ..., Yth fusionable information sets after classification, where Y is a positive integer;

[0028] S402: After determining the fusionable information sets of the multi-source data sets, extract the position of the information contained in each fusionable information set. , It is represented by the position of the 1st, 2nd, 3rd, ..., Eth information in the fusionable information set, where E is a positive integer. After extracting the position of each information in the fusionable information set, find the amount of information in each information, and select the information with the same position for fusion. The formula is:

[0029] In the formula, XR represents the fused information, Q represents the number of information with the same position, Hr represents the information with the same position, and H represents the amount of information contained in the same information at each position;

[0030] S403, fuse the information in each fusible information set, and finally obtain the fused information as follows: , Represents the 1st, 2nd, 3rd, ..., rth information after fusion; combines the fused information with the different information sets in the multi-source data, and splices the information in the different information sets in the fused information with the same timestamp to obtain the fused comprehensive data.

[0031] After the information is marked in time and space using the same reference system, the fusion method of time and space alignment can avoid the incorrect fusion of information that needs to be fused in multi-source data during fusion, resulting in inaccurate scene results in the final analysis.

[0032] S500, using the constructed database for seeking common ground while reserving differences to analyze the scenario data, using the divergent information in the database for seeking common ground while reserving differences to obtain different decisions, analyzing each decision, calculating the confidence of the decision, integrating multiple decisions, and obtaining a comprehensive decision for the analyzed scenario.

[0033] Furthermore, the specific steps to obtain the comprehensive decision of the analysis scenario are:

[0034] S501: The user uses the integrated data to analyze the scene data and obtains different initial decisions based on the different information. , It represents the 1st, 2nd, 3rd, ...hth initial decisions obtained by the user after comprehensive data analysis;

[0035] S502, calculate the h kinds of initial decisions obtained by analysis, and fuse the h kinds of initial decisions to obtain the final comprehensive decision, the formula is:

[0036] In the formula, Ju_end represents the comprehensive decision after fusion, Ju represents each initial decision obtained after analysis, A represents the number of information fused in the multi-source data corresponding to each initial decision, and h represents the number of initial decisions; after weighted processing of the initial decisions and calculating the average value, the comprehensive decision of the scene data analysis is finally obtained.

[0037] When analyzing the information in the scene, the information with differences is retained. Finally, after the initial decision is made using the difference information, the initial decision is integrated to obtain a comprehensive decision, so as to avoid the situation where the decision result obtained from the final analysis is untrue due to insufficient consideration of the difference information.

[0038] A scene data fusion analysis management system based on multi-source heterogeneity, the scene data fusion analysis management system includes a data collection module, a data analysis module, a spatiotemporal labeling module, an alignment fusion module and a decision fusion module;

[0039] The data collection module is used to collect all data in the scene in history;

[0040] The data analysis module is used to analyze the data in the scene to obtain a multi-source data set, and then judge the information type of each data in the multi-source data set to obtain a common information set and a different information set;

[0041] The spatiotemporal marking module is used to analyze the time when the multi-source data is generated and the number of information contained therein, obtain a time reference system and a space reference system, use the time reference system and the space reference system to generate a timestamp and position of each information, and mark them;

[0042] The time reference system and the space reference system are used to generate the timestamp and location of each information, and the marking is specifically as follows:

[0043] The time of each multi-source data generated in the scene is extracted to obtain the generation time of each data in the multi-source data set: , Indicates the generation time of the 1st, 2nd, 3rd, ..., uth data in the extracted multi-source data set. The generation time of the extracted data is analyzed and the earliest time when the data is generated is used as the time reference system in the multi-source data set. ;

[0044] Use the generated time reference system to timestamp each data generated in the multi-source data set, and calculate the generation timestamp of each data. The formula is: In the formula, represents the timestamp when each data in the multi-source dataset is generated, t represents the generation time of each data in the multi-source dataset, and each data in the multi-source dataset is marked with the calculated timestamp. , Indicates that the 1st, 2nd, 3rd, ..., uth data in the multi-source data set are timestamped;

[0045] Extract the number of information contained in each data of the multi-source data set, and set the number of information contained in each data to be , Indicates the number of information contained in the 1st, 2nd, 3rd, ..., uth data in the multi-source data set. Compare the number of u pieces of information, select the spatial coordinates of the information in the data containing the largest number of information as the spatial reference system, number all the information in the data containing the largest number of information, set the position of the initial information to 1, and then number them in sequence to obtain the spatial reference system: , Indicates the position of the 1st, 2nd, 3rd, ..., Dth information in the data containing the most information, where D is a positive integer;

[0046] Extract the information type contained in each data in the multi-source data set and find the location of the same type of information in the spatial reference system , using the location of the same kind of information in a spatial reference system The information contained in each data is marked with a location, and all the information in each data is marked with a location in the spatial reference system. , Indicates that the position of the 1st, 2nd, 3rd, ... information in each data is marked;

[0047] The alignment and fusion module is used to fuse the information in the common information set according to the timestamp and position, and to splice the information in the different information set in the corresponding fused information to obtain comprehensive data;

[0048] The decision fusion module is used to analyze the initial decision generated by using the comprehensive data and obtain the comprehensive decision by weighted average calculation.

[0049] The data analysis module includes a multi-source data set construction unit and an information set construction unit;

[0050] The multi-source data set construction unit is used to analyze all the data in the scene, extract features, and classify the data using the features to obtain a multi-source data set;

[0051] The information set construction unit is used to judge the types of information in the multi-source data set, and regard the same type of information as a common information set, and the different types of information as a different information set.

[0052] The spatiotemporal marking module includes a timestamp marking unit and a position marking unit;

[0053] The timestamp marking unit is used to analyze the generation time of each data in the multi-source data to obtain a time reference system, calculate the time of each information according to the time reference system to obtain a timestamp, and mark the information with the timestamp;

[0054] The position marking unit is used to analyze the number of information contained in the multi-source data, find out the spatial reference system, and use the spatial reference system to mark the information position in each data.

[0055] The alignment fusion module includes a time alignment unit and a space alignment unit;

[0056] The time alignment unit is used to judge the timestamp of each information in the same information set, regard the information with the same timestamp as a group of fusionable information, and classify the information in the same information set by using the timestamp after judgment;

[0057] The spatial alignment unit is used to determine the position of information in each fusible information set, fuse the information according to the same position, and finally obtain fused data.

[0058] Compared with the prior art, the beneficial effects achieved by the present invention are:

[0059] 1. The present invention classifies the information with the fastest change speed as the feature of the data, and can clearly distinguish the data in the scene according to different change characteristics. When analyzing the data in the scene, the data can be collected according to the different change characteristics.

[0060] 2. The present invention only retains different divergent information in multi-source data. When finally generating a decision, it uses the divergent information to obtain different initial decisions, and obtains a comprehensive decision by fusing the initial decisions, thereby avoiding the situation where the decision result obtained by the final analysis is untrue due to insufficient consideration of the difference information.

[0061] 3. After the present invention uses the same reference system to mark the information in time and space, the fusion method of time and space alignment can avoid the information in multi-source data from being incorrectly fused when fusion is performed, resulting in inaccurate scene results in the final analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0063] Figure 1 It is a module distribution diagram of a scene data fusion analysis management system based on multi-source heterogeneity of the present invention;

[0064] Figure 2 It is a schematic diagram of the steps of a multi-source heterogeneous scene data fusion analysis management method of the present invention. DETAILED DESCRIPTION

[0065] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] See also Figure 1-Figure 2 , the present invention provides a technical solution:

[0067] A method for fusion analysis and management of scene data based on multi-source heterogeneity, the method comprising the following steps:

[0068] S100, collecting all data in the scene, analyzing and extracting features from each collected data, classifying the data using the data features, and obtaining a multi-source data set in the scene;

[0069] The specific steps to classify data using data features are:

[0070] S101, collect all data in the scene in the history, and extract all types of information contained in each data , Indicates the 1st, 2nd, 3rd, ..., nth information contained in each collected data, where n is a positive integer; the time interval for data changes in the scene in the collection history is , Indicates the 1st, 2nd, 3rd, ..., mth time intervals at which the data in the scene changes in the collected history, where m is a positive integer; calculates the average time intervals at which the data in the scene changes in the extracted history, and the formula is: In the formula, Tc represents the average time interval of data changes in the historical scene. The calculated average value is used as the period for collecting all data in the historical scene. The data in the historical scene is collected and various types of information are extracted according to the calculated period Tc;

[0071] S102. After collecting the data in the scene according to the collection cycle, extract the information in the data and obtain the total number of each type of information as P. Let the extracted p pieces of information be , Indicates the 1st, 2nd, 3rd, ..., Pth information of each extracted information, and calculates the change speed of each information when the data changes in the historical scene. The formula is:

[0072] In the formula, It indicates the latter information value when calculating the difference between two adjacent information values. It represents the previous information value when calculating the difference between two adjacent information values, Tc represents the interval period between two adjacent information values, P represents the total number of extracted information, and Su represents the change speed of each type of information calculated;

[0073] The n types of information in each data are calculated, and the change rate of each type of information is obtained as , Indicates the calculated change speed of the 1st, 2nd, 3rd, ..., nth information. Compare the calculated change speeds of each information and select the maximum value among the n information change speeds. As the feature Te of the corresponding data;

[0074] S103, using the method described in S102 to calculate the characteristics of each collected data, the characteristics of all collected data are obtained as follows: , represents the features of the 1st, 2nd, 3rd, ..., uth data collected, and the data with the same features are classified as the same type of data. Finally, the data in the scene are classified as , It represents the 1st, 2nd, 3rd, ..., jth type of data classified in the scene, where j is a positive integer. The multi-source data set in the scene is constructed using the classified data.

[0075] The data is analyzed based on the changing speed of the information contained in each data in the history. The faster the information changes, the more it can reflect the characteristics of the corresponding data. Therefore, the information with the fastest changing speed is classified as the feature of the data. The data in the scene can be clearly distinguished according to different changing characteristics. When analyzing the data in the scene, data can be collected according to different changing characteristics.

[0076] S200, analyzing each type of data in the multi-source data set, finding the same information and different information in the multi-source data, using the same information to construct a common information set, and using the different information to construct a different information set;

[0077] The specific steps of using the same information to construct a common information set and using different information to construct a different information set are as follows:

[0078] S201, compare the information contained in each data in the multi-source data set, and select the same type of information contained in each data in the multi-source data set as , Indicates the first, second, third, ..., g pieces of information of the same type contained in each type of data in the multi-source data set, where g is a positive integer; constructs a common information set using the collected information of the same type;

[0079] S202, compare the information contained in each data in the multi-source data set, and select different types of information contained in each data in the multi-source data set as , It represents the 1st, 2nd, 3rd, ..., hth pieces of information of different types contained in each data in the multi-source data set, where h is a positive integer; the different types of collected information are used to construct a variance information set.

[0080] S300, analyzing the multi-source information in the common information set to determine a unique time and space reference system; using the time and space reference system to mark the time and space of data information of each type of multi-source information in the common information set;

[0081] The specific steps of using the time and space reference system to mark the time and space of each multi-source information in the same information set are as follows:

[0082] S301, extract the time of each multi-source data generated in the scene, and obtain the generation time of each data in the multi-source data set: , Indicates the generation time of the 1st, 2nd, 3rd, ..., uth data in the extracted multi-source data set. The generation time of the extracted data is analyzed and the earliest time when the data is generated is used as the time reference system in the multi-source data set. ;

[0083] Use the generated time reference system to timestamp each data generated in the multi-source data set, and calculate the generation timestamp of each data. The formula is:

[0084] In the formula, represents the timestamp when each data in the multi-source dataset is generated, t represents the generation time of each data in the multi-source dataset, and each data in the multi-source dataset is marked with the calculated timestamp. , Indicates that the 1st, 2nd, 3rd, ..., uth data in the multi-source data set are timestamped;

[0085] S302, extract the number of information contained in each data of the multi-source data set, assuming that the number of information contained in each data is , Indicates the number of information contained in the 1st, 2nd, 3rd, ..., uth data in the multi-source data set. Compare the number of u pieces of information, select the spatial coordinates of the information in the data containing the largest number of information as the spatial reference system, number all the information in the data containing the largest number of information, set the position of the initial information to 1, and then number them in sequence to obtain the spatial reference system: , Indicates the position of the 1st, 2nd, 3rd, ..., Dth information in the data containing the most information, where D is a positive integer;

[0086] Extract the information type contained in each data in the multi-source data set and find the location of the same type of information in the spatial reference system , using the location of the same kind of information in a spatial reference system The information contained in each data is marked with a location, and all the information in each data is marked with a location in the spatial reference system. , It means marking the position of the 1st, 2nd, 3rd, ... information in each data.

[0087] When conducting a comprehensive analysis of the data in the scene, it is necessary to fuse different data, and then comprehensively consider and analyze them to determine the time and space reference systems in the multi-source data. Using the same reference system can ensure that when the information to be fused is time-space labeled, the time and space markings are accurate, avoiding different time and space marking rules caused by different reference systems, which may cause fusion errors.

[0088] S400, using the constructed database for seeking common ground while reserving differences to analyze the scenario data, using the divergent information in the database for seeking common ground while reserving differences to obtain different decisions, analyzing each decision, calculating the confidence of the decision, integrating multiple decisions, and obtaining a comprehensive decision for the analyzed scenario.

[0089] The specific steps to obtain a comprehensive decision for the analysis scenario are:

[0090] S401, after selecting the common information set in the multi-source data, extract the timestamp and position of each information in the common information set in the mark, and set the timestamp of each information in the extracted common information set as , represents the timestamps of the 1st, 2nd, 3rd, ..., R information in the same information set. The R timestamps are analyzed and the information with the same timestamps are grouped into a group of fused information sets. After classifying all the timestamps, the fused information set is obtained as , represents the 1st, 2nd, 3rd, ..., Yth fusionable information sets after classification, where Y is a positive integer;

[0091] S402: After determining the fusionable information sets of the multi-source data sets, extract the position of the information contained in each fusionable information set. , It is represented by the position of the 1st, 2nd, 3rd, ..., Eth information in the fusionable information set, where E is a positive integer. After extracting the position of each information in the fusionable information set, find the amount of information in each information, and select the information with the same position for fusion. The formula is:

[0092] In the formula, XR represents the fused information, Q represents the number of information with the same position, Hr represents the information with the same position, and H represents the amount of information contained in the same information at each position;

[0093] S403, fuse the information in each fusible information set, and finally obtain the fused information as follows: , Represents the 1st, 2nd, 3rd, ..., rth information after fusion; combines the fused information with the different information sets in the multi-source data, and splices the information in the different information sets in the fused information with the same timestamp to obtain the fused comprehensive data.

[0094] After the information is marked in time and space using the same reference system, the fusion method of time and space alignment can avoid the incorrect fusion of information that needs to be fused in multi-source data during fusion, resulting in inaccurate scene results in the final analysis.

[0095] S500, using the constructed database for seeking common ground while reserving differences to analyze the scenario data, using the divergent information in the database for seeking common ground while reserving differences to obtain different decisions, analyzing each decision, calculating the confidence of the decision, integrating multiple decisions, and obtaining a comprehensive decision for the analyzed scenario.

[0096] The specific steps to obtain a comprehensive decision for the analysis scenario are:

[0097] S501: The user uses the integrated data to analyze the scene data and obtains different initial decisions based on the different information. , It represents the 1st, 2nd, 3rd, ...hth initial decisions obtained by the user after comprehensive data analysis;

[0098] S502, calculate the h kinds of initial decisions obtained by analysis, and fuse the h kinds of initial decisions to obtain the final comprehensive decision, the formula is:

[0099] In the formula, Ju_end represents the comprehensive decision after fusion, Ju represents each initial decision obtained after analysis, A represents the number of information fused in the multi-source data corresponding to each initial decision, and h represents the number of initial decisions; after weighted processing of the initial decisions and calculating the average value, the comprehensive decision of the scene data analysis is finally obtained.

[0100] When analyzing the information in the scene, the information with differences is retained. Finally, after the initial decision is made using the difference information, the initial decision is integrated to obtain a comprehensive decision, so as to avoid the situation where the decision result obtained from the final analysis is untrue due to insufficient consideration of the difference information.

[0101] A scene data fusion analysis management system based on multi-source heterogeneity, the scene data fusion analysis management system includes a data collection module, a data analysis module, a spatiotemporal labeling module, an alignment fusion module and a decision fusion module;

[0102] The data collection module is used to collect all data in the scene in history;

[0103] The data analysis module is used to analyze the data in the scene to obtain a multi-source data set, and then judge the information type of each data in the multi-source data set to obtain a common information set and a different information set;

[0104] The spatiotemporal marking module is used to analyze the time when the multi-source data is generated and the number of information contained therein, obtain a time reference system and a space reference system, use the time reference system and the space reference system to generate a timestamp and position of each information, and mark them;

[0105] The time reference system and the space reference system are used to generate the timestamp and location of each information, and the marking is specifically as follows:

[0106] The time of each multi-source data generated in the scene is extracted to obtain the generation time of each data in the multi-source data set: , Indicates the generation time of the 1st, 2nd, 3rd, ..., uth data in the extracted multi-source data set. The generation time of the extracted data is analyzed and the earliest time when the data is generated is used as the time reference system in the multi-source data set. ;

[0107] Use the generated time reference system to timestamp each data generated in the multi-source data set, and calculate the generation timestamp of each data. The formula is:

[0108] In the formula, represents the timestamp when each data in the multi-source dataset is generated, t represents the generation time of each data in the multi-source dataset, and each data in the multi-source dataset is marked with the calculated timestamp. , Indicates that the 1st, 2nd, 3rd, ..., uth data in the multi-source data set are timestamped;

[0109] Extract the number of information contained in each data of the multi-source data set, and set the number of information contained in each data to be , Indicates the number of information contained in the 1st, 2nd, 3rd, ..., uth data in the multi-source data set. Compare the number of u pieces of information, select the spatial coordinates of the information in the data containing the largest number of information as the spatial reference system, number all the information in the data containing the largest number of information, set the position of the initial information to 1, and then number them in sequence to obtain the spatial reference system: , Indicates the position of the 1st, 2nd, 3rd, ..., Dth information in the data containing the most information, where D is a positive integer;

[0110] Extract the information type contained in each data in the multi-source data set and find the location of the same type of information in the spatial reference system , using the location of the same kind of information in a spatial reference system The information contained in each data is marked with a location, and all the information in each data is marked with a location in the spatial reference system. , Indicates that the position of the 1st, 2nd, 3rd, ... information in each data is marked;

[0111] The alignment and fusion module is used to fuse the information in the common information set according to the timestamp and position, and to splice the information in the different information set in the corresponding fused information to obtain comprehensive data;

[0112] The decision fusion module is used to analyze the initial decision generated by using the comprehensive data and obtain the comprehensive decision by weighted average calculation.

[0113] The data analysis module includes a multi-source data set construction unit and an information set construction unit;

[0114] The multi-source data set construction unit is used to analyze all the data in the scene, extract features, and classify the data using the features to obtain a multi-source data set;

[0115] The information set construction unit is used to judge the types of information in the multi-source data set, and regard the same type of information as a common information set, and the different types of information as a different information set.

[0116] The spatiotemporal marking module includes a timestamp marking unit and a position marking unit;

[0117] The timestamp marking unit is used to analyze the generation time of each data in the multi-source data to obtain a time reference system, calculate the time of each information according to the time reference system to obtain a timestamp, and mark the information with the timestamp;

[0118] The position marking unit is used to analyze the number of information contained in the multi-source data, find out the spatial reference system, and use the spatial reference system to mark the information position in each data.

[0119] The alignment fusion module includes a time alignment unit and a space alignment unit;

[0120] The time alignment unit is used to judge the timestamp of each information in the same information set, regard the information with the same timestamp as a group of fusionable information, and classify the information in the same information set by using the timestamp after judgment;

[0121] The spatial alignment unit is used to determine the position of information in each fusible information set, fuse the information according to the same position, and finally obtain fused data.

[0122] Embodiment 1: Now analyze the data in the factory scene, collect all the data in the scene, and calculate the change speed of each information. The formula is:

[0123]

[0124] After calculation, the change speeds of the information in the data are 3, 5, and 9 respectively. The information corresponding to 9 is selected as the feature of the data. According to the features, the data in the scene is divided into environmental data, equipment data, and personnel data.

[0125] By judging the types of information in the three types of data, the common information set is: temperature; the different information sets are humidity, amplitude, and moving speed;

[0126] After analyzing the three types of data, it is found that the time reference is 00:00 for environmental data, the spatial reference is the equipment data, the temperature data position in the equipment data is 5, and the temperature in the equipment data and personnel data is marked with 5, and the timestamps of the three types of data are 0, 1, and 0 respectively;

[0127] Use timestamp and position to fuse temperature information. Assume that the temperature information in the three data is 25, 55, and 35 respectively. Use the formula for fusion:

[0128] The amount of information contained in the temperature information of the three data is 3, 5, and 10 respectively; the final fusion temperature is 38; the comprehensive data is constructed by combining the different information sets;

[0129] The user uses the comprehensive data to make a decision. The probability of equipment abnormality is 40%, 80%, and 60% respectively. The decision is integrated using the formula:

[0130]

[0131] The final comprehensive decision was calculated, and the probability of equipment abnormality was 76%.

[0132] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0133] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for fusion analysis and management of scene data based on multi-source heterogeneity, characterized by: The method comprises the following steps: S100, collecting all data in the scene, analyzing and extracting features from each collected data, classifying the data using the data features, and obtaining a multi-source data set in the scene; S200, analyzing each type of data in the multi-source data set, finding the same information and different information in the multi-source data, using the same information to construct a common information set, and using the different information to construct a different information set; S300, analyzing the multi-source information in the common information set to determine a unique time and space reference system; using the time and space reference system to mark the time and space of data information of each type of multi-source information in the common information set; The specific steps of using the time and space reference system to mark the time and space of each multi-source information in the same information set are as follows: S301, extract the time of each multi-source data generated in the scene, and obtain the generation time of each data in the multi-source data set: , Indicates the generation time of the 1st, 2nd, 3rd, ..., uth data in the extracted multi-source data set. The generation time of the extracted data is analyzed and the earliest time when the data is generated is used as the time reference system in the multi-source data set. ; Use the generated time reference system to timestamp each data generated in the multi-source data set, and calculate the generation timestamp of each data. The formula is: In the formula, represents the timestamp when each data in the multi-source dataset is generated, t represents the generation time of each data in the multi-source dataset, and each data in the multi-source dataset is marked with the calculated timestamp. , Indicates that the 1st, 2nd, 3rd, ..., uth data in the multi-source data set are timestamped; S302, extract the number of information contained in each data of the multi-source data set, assuming that the number of information contained in each data is , Indicates the number of information contained in the 1st, 2nd, 3rd, ..., uth data in the multi-source data set. Compare the number of u pieces of information, select the spatial coordinates of the information in the data containing the largest number of information as the spatial reference system, number all the information in the data containing the largest number of information, set the position of the initial information to 1, and then number them in sequence to obtain the spatial reference system: , Indicates the position of the 1st, 2nd, 3rd, ..., Dth information in the data containing the most information, where D is a positive integer; Extract the information type contained in each data in the multi-source data set and find the location of the same type of information in the spatial reference system , using the location of the same kind of information in a spatial reference system The information contained in each data is marked with a position, and all the information in each data is marked with a position in a spatial reference system, indicating that the first, second, third, ... information in each data is marked with a position; S400, after marking the multi-source information in the common information set, aligning and fusing the multi-source information in the common information set according to the marks; after fusing the multi-source information, using the fused multi-source information in the different information set to obtain comprehensive data of the scene; S500, using the constructed database for seeking common ground while reserving differences to analyze the scenario data, using the divergent information in the database for seeking common ground while reserving differences to obtain different decisions, analyzing each decision, calculating the confidence of the decision, integrating multiple decisions, and obtaining a comprehensive decision for the analyzed scenario.

2. The method for fusion analysis and management of scene data based on multi-source heterogeneity according to claim 1 is characterized by: The specific steps of classifying data using data features in S100 are: S101, collect all data in the scene in the history, and extract all types of information contained in each data , Indicates the 1st, 2nd, 3rd, ..., nth information contained in each collected data, where n is a positive integer; the time interval for data changes in the scene in the collection history is , Indicates the 1st, 2nd, 3rd, ..., mth time intervals at which the data in the scene changes in the collected history, where m is a positive integer; calculates the average time intervals at which the data in the scene changes in the extracted history, and the formula is: In the formula, Tc represents the average time interval of data changes in the historical scene. The calculated average value is used as the period for collecting all data in the historical scene. The data in the historical scene is collected and various types of information are extracted according to the calculated period Tc; S102. After collecting the data in the scene according to the collection cycle, extract the information in the data and obtain the total number of each type of information as P. Let the extracted p pieces of information be , Indicates the 1st, 2nd, 3rd, ..., Pth information of each extracted information, and calculates the change speed of each information when the data changes in the historical scene. The formula is: In the formula, It indicates the latter information value when calculating the difference between two adjacent information values. It represents the previous information value when calculating the difference between two adjacent information values, Tc represents the interval period between two adjacent information values, P represents the total number of extracted information, and Su represents the change speed of each type of information calculated; The n types of information in each data are calculated, and the change rate of each type of information is obtained as , Indicates the calculated change speed of the 1st, 2nd, 3rd, ..., nth information. Compare the calculated change speeds of each information and select the maximum value among the n information change speeds. As the feature Te of the corresponding data; S103, using the method described in S102 to calculate the characteristics of each collected data, the characteristics of all collected data are obtained as follows: , represents the features of the 1st, 2nd, 3rd, ..., uth data collected, and the data with the same features are classified as the same type of data. Finally, the data in the scene are classified as , It represents the 1st, 2nd, 3rd, ..., jth type of data classified in the scene, where j is a positive integer. The multi-source data set in the scene is constructed using the classified data.

3. The method for fusion analysis and management of scene data based on multi-source heterogeneity according to claim 2 is characterized in that: The specific steps of constructing the same information set using the same information and constructing the different information set using the different information in S200 are: S201, compare the information contained in each data in the multi-source data set, and select the same type of information contained in each data in the multi-source data set as , Indicates the first, second, third, ..., g pieces of information of the same type contained in each type of data in the multi-source data set, where g is a positive integer; constructs a common information set using the collected information of the same type; S202, compare the information contained in each data in the multi-source data set, and select different types of information contained in each data in the multi-source data set as , It represents the 1st, 2nd, 3rd, ..., hth pieces of information of different types contained in each data in the multi-source data set, where h is a positive integer; the different types of collected information are used to construct a variance information set.

4. The method for fusion analysis and management of scene data based on multi-source heterogeneity according to claim 1 is characterized by: The specific steps of obtaining the comprehensive data of the scene in S400 are: S401, after selecting the common information set in the multi-source data, extract the timestamp and position of each information in the common information set in the mark, and set the timestamp of each information in the extracted common information set as , represents the timestamps of the 1st, 2nd, 3rd, ..., R information in the same information set. The R timestamps are analyzed and the information with the same timestamps are grouped into a group of fused information sets. After classifying all the timestamps, the fused information set is obtained as , represents the 1st, 2nd, 3rd, ..., Yth fusionable information sets after classification, where Y is a positive integer; S402: After determining the fusionable information sets of the multi-source data sets, extract the position of the information contained in each fusionable information set. , It is represented by the position of the 1st, 2nd, 3rd, ..., Eth information in the fusionable information set, where E is a positive integer. After extracting the position of each information in the fusionable information set, find the amount of information in each information, and select the information with the same position for fusion. The formula is: In the formula, XR represents the fused information, Q represents the number of information with the same position, Hr represents the information with the same position, and H represents the amount of information contained in the same information at each position; S403, fuse the information in each fusible information set, and finally obtain the fused information as follows: , Represents the 1st, 2nd, 3rd, ..., rth information after fusion; combines the fused information with the different information sets in the multi-source data, and splices the information in the different information sets in the fused information with the same timestamp to obtain the fused comprehensive data.

5. The method for fusion analysis and management of scene data based on multi-source heterogeneity according to claim 1 is characterized by: The specific steps of obtaining the comprehensive decision of the analysis scenario in S500 are: S501: The user uses the integrated data to analyze the scene data and obtains different initial decisions based on the different information. , It represents the 1st, 2nd, 3rd, ...hth initial decisions obtained by the user after comprehensive data analysis; S502, calculate the h kinds of initial decisions obtained by analysis, and fuse the h kinds of initial decisions to obtain the final comprehensive decision, the formula is: In the formula, Ju_end represents the combined decision after fusion, Ju represents each initial decision obtained after analysis, A represents the number of information fused in the multi-source data corresponding to each initial decision, and h represents the number of initial decisions; after weighted processing of the initial decisions and calculating the average value, the comprehensive decision of the scene data analysis is finally obtained.

6. A scene data fusion analysis and management system based on multi-source heterogeneity, characterized by: The scene data fusion analysis and management system includes a data collection module, a data analysis module, a spatiotemporal labeling module, an alignment fusion module and a decision fusion module; The data collection module is used to collect all data in the scene in history; The data analysis module is used to analyze the data in the scene to obtain a multi-source data set, and then judge the information type of each data in the multi-source data set to obtain a common information set and a different information set; The spatiotemporal marking module is used to analyze the time when the multi-source data is generated and the number of information contained therein, obtain a time reference system and a space reference system, use the time reference system and the space reference system to generate a timestamp and position of each information, and mark them; The time reference system and the space reference system are used to generate the timestamp and location of each information, and the marking is specifically as follows: The time of each multi-source data generated in the scene is extracted to obtain the generation time of each data in the multi-source data set: , Indicates the generation time of the 1st, 2nd, 3rd, ..., uth data in the extracted multi-source data set. The generation time of the extracted data is analyzed and the earliest time when the data is generated is used as the time reference system in the multi-source data set. ; Use the generated time reference system to timestamp each data generated in the multi-source data set, and calculate the generation timestamp of each data. The formula is: In the formula, represents the timestamp when each data in the multi-source dataset is generated, t represents the generation time of each data in the multi-source dataset, and each data in the multi-source dataset is marked with the calculated timestamp. , Indicates that the 1st, 2nd, 3rd, ..., uth data in the multi-source data set are timestamped; Extract the number of information contained in each data of the multi-source data set, and set the number of information contained in each data to be , Indicates the number of information contained in the 1st, 2nd, 3rd, ..., uth data in the multi-source data set. Compare the number of u pieces of information, select the spatial coordinates of the information in the data containing the largest number of information as the spatial reference system, number all the information in the data containing the largest number of information, set the position of the initial information to 1, and then number them in sequence to obtain the spatial reference system: , Indicates the position of the 1st, 2nd, 3rd, ..., Dth information in the data containing the most information, where D is a positive integer; Extract the information type contained in each data in the multi-source data set and find the location of the same type of information in the spatial reference system , using the location of the same kind of information in a spatial reference system The information contained in each data is marked with a location, and all the information in each data is marked with a location in the spatial reference system. , Indicates that the position of the 1st, 2nd, 3rd, ... information in each data is marked; The alignment and fusion module is used to fuse the information in the common information set according to the timestamp and position, and to splice the information in the different information set in the corresponding fused information to obtain comprehensive data; The decision fusion module is used to analyze the initial decision generated by using the comprehensive data and obtain the comprehensive decision by weighted average calculation.

7. The multi-source heterogeneous scene data fusion analysis management system according to claim 6 is characterized by: The data analysis module includes a multi-source data set construction unit and an information set construction unit; The multi-source data set construction unit is used to analyze all the data in the scene, extract features, and classify the data using the features to obtain a multi-source data set; The information set construction unit is used to judge the types of information in the multi-source data set, and regard the same type of information as a common information set, and the different types of information as a different information set.

8. The multi-source heterogeneous scene data fusion analysis and management system according to claim 6, characterized in that: The spatiotemporal marking module includes a timestamp marking unit and a position marking unit; The timestamp marking unit is used to analyze the generation time of each data in the multi-source data to obtain a time reference system, calculate the time of each information according to the time reference system to obtain a timestamp, and mark the information with the timestamp; The position marking unit is used to analyze the number of information contained in the multi-source data, find out the spatial reference system, and use the spatial reference system to mark the information position in each data.

9. The multi-source heterogeneous scene data fusion analysis and management system according to claim 6, characterized in that: The alignment and fusion module includes a time alignment unit and a space alignment unit; The time alignment unit is used to judge the timestamp of each information in the same information set, regard the information with the same timestamp as a group of fusionable information, and classify the information in the same information set by using the timestamp after judgment; The spatial alignment unit is used to determine the position of information in each fusible information set, fuse the information according to the same position, and finally obtain fused data.

Images