A rail transit data automatic monitoring system and method based on big data

Abstract

The present invention relates to the field of rail transit, and specifically to an automatic rail transit data monitoring system and method based on big data, comprising a sensor monitoring event extraction module, a unit event inspection cycle analysis module, a necessary sensor data analysis module, a verification and correction module and an optimal sensor early warning module; the sensor monitoring event extraction module is used to extract historical records of sensor monitoring events of rail vehicles on corresponding rail paths based on redundant sensor groups with different execution purposes; the unit event inspection cycle analysis module is used to construct a unit event inspection cycle for the corresponding rail path; the necessary sensor data analysis module outputs necessary sensor data of redundant sensor groups for sensor monitoring event records within a first analysis cycle; the verification and correction module verifies and corrects the necessary sensor data of each redundant sensor group; and the optimal sensor early warning module is used to determine the optimal sensor source for non-essential sensor data of each unit analysis object.

Description

Technical Field

[0001] The present invention relates to the field of rail transit technology, and in particular to a rail transit data automatic monitoring system and method based on big data. Background Art

[0002] As an important part of urban public transportation, the safety and reliability of rail transit systems are directly related to the safety of passengers and the smoothness of urban traffic. With the continuous development of technology, rail transit systems are increasingly dependent on various sensors to monitor and control the running status of trains. However, a single sensor often has the risk of failure, and once a failure occurs, it may cause the instability or even paralysis of the entire system. Therefore, the application of redundant sensor layout has become an important means to improve the safety and reliability of rail transit systems; the layout of redundant traffic sensors not only guarantees the effectiveness of monitoring in terms of quantity, but also improves the accuracy of computing power analysis to a certain extent from the software direction to achieve efficient monitoring; but there are also certain problems in the layout of redundant sensors. From the perspective of storage pressure, the layout of redundant sensors greatly increases the storage of rail transit data, and the redundant sensor group maintains a monitoring state in the entire rail path. The amount of stored data will increase exponentially, but in reality, rail transit events that need to be responded to and decided based on redundant sensor data account for a small proportion in a traffic path. Therefore, how to implement partial storage and partial screening of non-essential data generated by redundant sensor groups for each rail vehicle in each rail path to increase storage capacity and ensure effective decision-making of rail transit events is worth analyzing. Summary of the invention

[0003] The purpose of the present invention is to provide a rail transit data automatic monitoring system and method based on big data 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: a method for automatic monitoring of rail transit data based on big data, comprising the following analysis steps:

[0005] Step S100: Extract the sensor monitoring events of the rail transit monitoring system history records based on the rail vehicles on the corresponding rail paths with redundant sensor groups with different execution purposes. The sensor monitoring events refer to the events that use the redundant sensor groups to collect data for analysis and decision-making based on the characteristics of the rail environment; the redundant sensor groups refer to the data collection groups composed of multiple sensors with the same execution functions distributed at different positions; analyze the sensor monitoring events of each rail path history records, and construct the unit event inspection cycle of the corresponding rail path;

[0006] Step S200: extracting sensor monitoring events that record the same execution purpose in various track paths and contain one redundant sensor group as the first monitoring event, using the unit event inspection period corresponding to the first monitoring event as the first analysis period, and outputting necessary sensor data of the redundant sensor group recorded by the sensor monitoring event in the first analysis period;

[0007] Step S300: Filtering sensor monitoring events recorded in various track paths with the same execution purpose and containing different numbers of redundant sensor groups as second monitoring events, and verifying and correcting necessary sensor data of each redundant sensor group based on the second monitoring events;

[0008] Step S400: Mark the non-essential sensor data based on the necessary sensor data, analyze the non-essential sensor data based on the source redundant sensor group, and determine the optimal sensor source of the non-essential sensor data for each unit analysis object; when the storage space alarm is triggered, retain the non-essential sensor data recorded by the optimal sensor in the redundant sensor group and eliminate other non-essential sensors.

[0009] Furthermore, constructing a unit event inspection cycle corresponding to the orbital path includes the following steps:

[0010] Extract several sensor monitoring events of the same rail vehicle in the valid driving events recorded in each rail path history, where the valid driving event is the driving event when the number of sensor monitoring events of the rail vehicle in the corresponding rail path record is the maximum;

[0011] Construct a coordinate system, where the horizontal axis records the length of the track path, and the vertical axis records the rail vehicles at different departure times; the longer the horizontal axis, the longer the track path, and the longer the vertical axis, the closer the travel time of the rail vehicles in the historical record is to the real time;

[0012] The sensor monitoring events recorded by different rail vehicles on the same rail path are marked with point coordinates in the coordinate axis; each point coordinate stores the response time and end time of the sensor monitoring event recorded by the corresponding rail vehicle; the response time is the time when the redundant sensor group starts to collect data for decision-making and judgment based on the characteristics of the rail environment; the end time is the time when the decision-making and judgment are implemented;

[0013] The interval between the response times of adjacent sensor monitoring events marked on the same vertical coordinate is the normal monitoring period of the next recorded sensor monitoring event. The normal monitoring period and the sensor analysis period of the next recorded sensor monitoring event are used to construct the unit event inspection period of the corresponding track path; the sensor analysis period refers to the interval period from the response time to the end time of each sensor monitoring event;

[0014] All coordinate points with the same ordinate recorded in the coordinate system are traversed to generate a unit event inspection period corresponding to each rail vehicle, and a set of unit event inspection periods corresponding to the rail path is constructed by all rail vehicles.

[0015] Further, generating necessary sensor data of the redundant sensor group for recording sensor monitoring events in the first analysis period includes the following specific steps:

[0016] Step S210: Mark the first analysis cycles with the same execution purpose as a type of analysis cycle, obtain the sensor parameters recorded by the redundant sensor group corresponding to the normal monitoring cycle in the i-th first analysis cycle of each type of first analysis cycle record, and form a first parameter type set Q i , Q i ={q 1 ,q 2 ,......,q n} i , where q 1 ,q 2 ,......,q n represents the sensor parameters of the 1st, 2nd, ..., nth types in the first parameter type set, and the sensor parameters recorded by the redundant sensor group required when the sensor analysis cycle in the i-th first analysis cycle corresponds to the decision-making judgment, and constitutes the second parameter type set F i , F i ={f 1 ,f 2 ,...,f m} i ;f 1 ,f 2 ,...,f m represents sensor parameters of types 1, 2, ..., m in the second parameter type set;

[0017] Step S220: The first parameter type set Q formed in the i-th first analysis period i and the second parameter type set F i Perform intersection calculation and output the target type set w corresponding to the i-th first analysis cycle i , w i =Q i ∩F i ; Traverse all first analysis cycles of the same type of first analysis cycle records to obtain the target type set w corresponding to each first analysis cycle; calculate the first intersection c, c={w 1 ,w 2 ,...,w k}, k≥i, k represents the total number of first analysis cycle records of the same type; if c is the same as the target type set of each first analysis cycle, then the sensor parameters in the output target type set are the necessary sensor data; if c does not exist and is the same as the target type set of each first analysis cycle, then the sensor parameters in the output first intersection are the necessary sensor data;

[0018] Step S230: extract the target type set corresponding to the record when the number of sensor parameter types in the k first analysis cycles is the largest as the cache set w max , filter the cache sensor data set D, D=w max -c; store the cached sensor data in the cached sensor data set in correspondence with the track environment characteristics.

[0019] Analyzing the necessary sensor data is to effectively focus on monitoring the different rail vehicles designed with redundant sensor groups in the corresponding rail paths, and to make reasonable storage reminders for the resources required by the source sensors of the necessary sensor data, so as to ensure that when a large amount of similar sensor data is generated by setting up redundant sensor groups, the space utilization rate is improved while ensuring the effectiveness and accuracy of using sensors to obtain data for analysis and decision-making; and after determining the necessary sensor data, the cached sensor data is screened, which can effectively distinguish the differences in the degree of data emphasis, and provide a range of selection and traceability for the sensor data required for the system to identify deviations in different scenarios.

[0020] Furthermore, step S300 includes the following specific steps:

[0021] Step S310: taking the unit event inspection period corresponding to the second monitoring event as the second analysis period, and replacing the first analysis period in steps S210 to S230 with the second analysis period, and obtaining necessary sensor data and cached sensor data corresponding to the second analysis period;

[0022] Step S320: Retrieve each sensor monitoring event recorded in each second analysis cycle as a verification event, extract the sensor monitoring events corresponding to the same execution purpose and recorded in the first analysis cycle as control events, compare the necessary sensor data recorded in the verification event and the control event, and when the number of necessary sensor data types recorded in the verification event is less than or equal to the number of necessary sensor data types recorded in the control event, output the necessary sensor data recorded in the control event as the necessary sensor data of the corresponding redundant sensor group; when the number of necessary sensor data types recorded in the verification event is greater than the number of necessary sensor data types recorded in the control event, correct the union of the necessary sensor data recorded in the verification event and the necessary sensor data recorded in the control event to be the necessary sensor data of the corresponding redundant sensor group.

[0023] Because it is impossible to effectively predict the type of rail transit event that may require sensor data to implement decision-making when actually collecting data from redundant sensor groups, it is necessary to increase the diversity of data types under different rail transit scenarios when considering data to determine the necessary sensor data, thereby increasing the intelligence and comprehensiveness of rail transit data automatic monitoring. Comprehensive improvement is achieved through verification. Further, step S400 includes:

[0024] Step S410: marking the sensor data other than the necessary sensor data and the cached sensor data as non-essential sensor data for each rail vehicle on the same rail path;

[0025] Step S420: Obtain the sensor recognition event of the unit analysis object history record. The sensor recognition event refers to the event of outputting corresponding data using the sensor recognition response. The sensor recognition event records the response time L and the recognition result Z. Use the formula:

[0026] Y=a 1 ×(1 / L 0 )+a 2 ×Z 0 ,

[0027] Calculate the effective utilization index Y, L of each sensor in the unit analysis object based on the sensor recognition event 0 It means to obtain the normalized average response time of multiple sensor recognition events recorded by the same sensor, Z 0 It means obtaining the normalized average value of the recognition results of multiple sensor recognition events recorded by the same sensor; a 1 、a 2 represents the corresponding reference coefficient;

[0028] The larger the Y is, the more accurate the sensor recognition is and the better the effect is;

[0029] Step S420: Sort the sensors in each unit analysis object from large to small according to the corresponding effective utilization index, and select the sensor with the first ranking as the optimal sensor for the corresponding unit analysis object.

[0030] A rail transit data automatic monitoring system based on big data, comprising a sensor monitoring event extraction module, a unit event inspection cycle analysis module, a necessary sensor data analysis module, a calibration and correction module and an optimal sensor early warning module;

[0031] The sensor monitoring event extraction module is used to extract the sensor monitoring events of the rail vehicle on the corresponding rail path based on the historical records of the rail transit monitoring system and the redundant sensor groups with different execution purposes are deployed;

[0032] The unit event inspection cycle analysis module is used to construct the unit event inspection cycle of the corresponding track path;

[0033] The necessary sensor data analysis module is used to output necessary sensor data of the sensor monitoring event record redundant sensor group within the first analysis period;

[0034] The verification and correction module is used to verify and correct the necessary sensor data of each redundant sensor group;

[0035] The optimal sensor early warning module is used to determine the optimal sensor source of the unnecessary sensor data of each unit analysis object.

[0036] Further, the unit event inspection period analysis module includes a coordinate system construction unit, a point coordinate identification unit, a monitoring period type determination unit and a unit event inspection period output unit;

[0037] The coordinate system construction unit is used to construct a coordinate system in which the horizontal axis records the path length of the track path and the vertical axis records the track vehicles at different departure times;

[0038] The point coordinate identification unit is used to identify the sensor monitoring events recorded by different rail vehicles on the same rail path with point coordinates in the coordinate axis;

[0039] The monitoring cycle type determination unit is used to determine the normal monitoring cycle and the sensing analysis cycle;

[0040] The unit event inspection period output unit is used to traverse all coordinate points with the same ordinate recorded in the coordinate system, generate a unit event inspection period corresponding to each rail vehicle, and construct a unit event inspection period set of the corresponding rail path by all rail vehicles.

[0041] Further, the necessary sensor data analysis module includes a parameter type set construction unit, a target type set output unit, a first intersection calculation unit and a sensor data output unit;

[0042] The parameter type set construction unit is used to construct a first parameter type set and a second parameter type set;

[0043] The target type set output unit is used to calculate the target type set by taking the intersection of the first parameter type set and the second parameter type set;

[0044] The first intersection calculation unit is used to traverse all analysis periods of the same type of analysis period records to obtain a target type set corresponding to each analysis period, and calculate a first intersection;

[0045] The sensor data output unit is used to output necessary sensor data and buffer sensor data.

[0046] Further, the optimal sensor warning module includes a non-essential data determination unit, an effective utilization index calculation unit, and a ranking output unit;

[0047] The non-essential data determination unit is used to mark other sensor data except necessary sensor data and cached sensor data as non-essential sensor data for each rail vehicle on the same rail path;

[0048] The effective utilization index calculation unit is used to calculate the effective utilization index of each sensor in the unit analysis object based on the sensor recognition event;

[0049] The sorting output unit is used to sort the sensors in each unit analysis object from large to small according to the corresponding effective utilization index, and select the sensor with the first ranking as the optimal sensor for the corresponding unit analysis object.

[0050] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: the present invention extracts and analyzes the historical record data of rail vehicles with redundant sensor groups, analyzes the necessary sensor data, realizes effective key monitoring of different rail vehicles designed with redundant sensor groups in the corresponding rail paths, and makes reasonable storage reminders for the resources required by the source sensors of the necessary sensor data, so as to ensure that when a large amount of similar sensor data is generated by setting up redundant sensor groups, the space utilization rate is improved while ensuring the effectiveness and accuracy of the analysis and decision-making by using sensors to obtain data; and after determining the necessary sensor data, the cached sensor data is screened, which can effectively distinguish the differences in the degree of data emphasis, and provide the selection and traceability within the range for the system to identify the sensor data required for different scene deviations; at the same time, the present application also considers the data characteristics under multiple types of scenes to verify the necessary sensor data that has been analyzed, thereby increasing the intelligence and comprehensiveness of the automatic monitoring of rail transit data. Comprehensive improvement is achieved through verification; the system improves the utilization rate of the corresponding data storage space of the redundant sensor group during application and ensures the effective identification of the necessary data when responding to the corresponding rail transit event, as well as the matching of data differences with environmental characteristics, so as to realize the effective traceability of cached data. The problem of space pressure in the process of realizing data monitoring by using redundant sensor groups is greatly reduced, and the practicability of using redundant sensor groups is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] 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:

[0052] Figure 1 It is a structural schematic diagram of a rail transit data automatic monitoring system and method based on big data of the present invention. DETAILED DESCRIPTION

[0053] 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.

[0054] See also Figure 1 , the present invention provides a technical solution:

[0055] A rail transit data automatic monitoring system based on big data, comprising a sensor monitoring event extraction module, a unit event inspection cycle analysis module, a necessary sensor data analysis module, a calibration and correction module and an optimal sensor early warning module;

[0056] The sensor monitoring event extraction module is used to extract the sensor monitoring events of the rail vehicle on the corresponding rail path based on the historical records of the rail transit monitoring system and the redundant sensor groups with different execution purposes are deployed;

[0057] The unit event inspection cycle analysis module is used to construct the unit event inspection cycle of the corresponding track path;

[0058] The necessary sensor data analysis module is used to output necessary sensor data of the sensor monitoring event record redundant sensor group within the first analysis period;

[0059] The verification and correction module is used to verify and correct the necessary sensor data of each redundant sensor group;

[0060] The optimal sensor early warning module is used to determine the optimal sensor source of the unnecessary sensor data of each unit analysis object.

[0061] The unit event inspection cycle analysis module includes a coordinate system construction unit, a point coordinate identification unit, a monitoring cycle type determination unit and a unit event inspection cycle output unit;

[0062] The coordinate system construction unit is used to construct a coordinate system in which the horizontal axis records the path length of the track path and the vertical axis records the track vehicles at different departure times;

[0063] The point coordinate identification unit is used to identify the sensor monitoring events recorded by different rail vehicles on the same rail path with point coordinates in the coordinate axis;

[0064] The monitoring cycle type determination unit is used to determine the normal monitoring cycle and the sensing analysis cycle;

[0065] The unit event inspection period output unit is used to traverse all coordinate points with the same ordinate recorded in the coordinate system, generate a unit event inspection period corresponding to each rail vehicle, and construct a unit event inspection period set of the corresponding rail path by all rail vehicles.

[0066] The necessary sensor data analysis module includes a parameter type set construction unit, a target type set output unit, a first intersection calculation unit and a sensor data output unit;

[0067] The parameter type set construction unit is used to construct a first parameter type set and a second parameter type set;

[0068] The target type set output unit is used to calculate the target type set by taking the intersection of the first parameter type set and the second parameter type set;

[0069] The first intersection calculation unit is used to traverse all analysis periods of the same type of analysis period records to obtain a target type set corresponding to each analysis period, and calculate a first intersection;

[0070] The sensor data output unit is used to output necessary sensor data and buffer sensor data.

[0071] The optimal sensor early warning module includes a non-essential data determination unit, an effective utilization index calculation unit, and a sorting output unit;

[0072] The non-essential data determination unit is used to mark other sensor data except necessary sensor data and cached sensor data as non-essential sensor data for each rail vehicle on the same rail path;

[0073] The effective utilization index calculation unit is used to calculate the effective utilization index of each sensor in the unit analysis object based on the sensor recognition event;

[0074] The sorting output unit is used to sort the sensors in each unit analysis object from large to small according to the corresponding effective utilization index, and select the sensor with the first ranking as the optimal sensor for the corresponding unit analysis object.

[0075] A method for automatic monitoring of rail transit data based on big data, comprising the following analysis steps:

[0076] Step S100: Extract the rail transit monitoring system history records based on sensor monitoring events of rail vehicles on corresponding rail paths with redundant sensor groups for different execution purposes. Sensor monitoring events refer to events that use redundant sensor groups to collect data for analysis and decision-making based on rail environmental characteristics; redundant sensor groups refer to data collection groups composed of multiple sensors with the same execution functions distributed at different positions; the execution purpose is determined by the execution function of the sensor. For example, if the function of the sensor is infrared recognition, then the corresponding execution purpose is to identify foreign objects on the track, and there are also redundant sensor groups with multiple execution functions corresponding to the execution purpose to be generated; decision-making refers to safety decisions such as emergency braking in the case of foreign objects on the track; analyze the sensor monitoring events recorded in the history of each rail path, and construct a unit event inspection cycle for the corresponding rail path;

[0077] Step S200: extracting sensor monitoring events that record the same execution purpose in various track paths and contain one redundant sensor group as the first monitoring event, using the unit event inspection period corresponding to the first monitoring event as the first analysis period, and outputting necessary sensor data of the redundant sensor group recorded by the sensor monitoring event in the first analysis period;

[0078] Step S300: Filtering sensor monitoring events recorded in various track paths with the same execution purpose and containing different numbers of redundant sensor groups as second monitoring events, and verifying and correcting necessary sensor data of each redundant sensor group based on the second monitoring events;

[0079] Step S400: Mark the non-essential sensor data based on the necessary sensor data, analyze the non-essential sensor data based on the source redundant sensor group, and determine the optimal sensor source of the non-essential sensor data for each unit analysis object; when the storage space alarm is triggered, retain the non-essential sensor data recorded by the optimal sensor in the redundant sensor group and eliminate other non-essential sensors.

[0080] Constructing a unit event investigation cycle corresponding to an orbital path includes the following steps:

[0081] Extract several sensor monitoring events of the same rail vehicle in the valid driving events recorded in each rail path history, where the valid driving event is the driving event when the number of sensor monitoring events of the rail vehicle in the corresponding rail path record is the maximum;

[0082] Construct a coordinate system, where the horizontal axis records the length of the track path, and the vertical axis records the rail vehicles at different departure times; the longer the horizontal axis, the longer the track path, and the longer the vertical axis, the closer the travel time of the rail vehicles in the historical record is to the real time;

[0083] The sensor monitoring events recorded by different rail vehicles on the same rail path are marked with point coordinates in the coordinate axis; each point coordinate stores the response time and end time of the sensor monitoring event recorded by the corresponding rail vehicle; the response time is the time when the redundant sensor group starts to collect data for decision-making and judgment based on the characteristics of the rail environment; the end time is the time when the decision-making and judgment are implemented;

[0084] The interval between the response times of adjacent sensor monitoring events marked on the same vertical coordinate is the normal monitoring period of the next recorded sensor monitoring event. The normal monitoring period and the sensor analysis period of the next recorded sensor monitoring event are used to construct the unit event inspection period of the corresponding track path; the sensor analysis period refers to the interval period from the response time to the end time of each sensor monitoring event;

[0085] All coordinate points with the same ordinate recorded in the coordinate system are traversed to generate a unit event inspection period corresponding to each rail vehicle, and a set of unit event inspection periods corresponding to the rail path is constructed by all rail vehicles.

[0086] As shown in the embodiment: Track path 1: records the track vehicle 1 departing at 10:00 and the track vehicle 2 departing at 16:00;

[0087] Rail vehicle 1 records the response time of the sensor monitoring event at a path length of 58 km at 11:13 and the end time at 11:15;

[0088] The response time of the sensor monitoring event at a path length of 196 km was recorded at 13:05, and the end time was recorded at 13:06;

[0089] Then the rail vehicle with the same vertical coordinate, such as 10:00, will correspond to the marked point coordinates (58km, 10:00) and (196km, 10:00).

[0090] Generating necessary sensor data of a redundant sensor group for recording sensor monitoring events in a first analysis period includes the following specific steps:

[0091] Step S210: Mark the first analysis cycles with the same execution purpose as a type of analysis cycle, obtain the sensor parameters recorded by the redundant sensor group corresponding to the normal monitoring cycle in the i-th first analysis cycle of each type of first analysis cycle record, and form a first parameter type set Q i , Q i ={q 1 ,q 2 ,......,q n} i , where q 1 ,q 2 ,......,q nrepresents the sensor parameters of the 1st, 2nd, ..., nth types in the first parameter type set, and the sensor parameters recorded by the redundant sensor group required when the sensor analysis cycle in the i-th first analysis cycle corresponds to the decision-making judgment, and constitutes the second parameter type set F i , F i ={f 1 ,f 2 ,...,f m} i ;f 1 ,f 2 ,...,f m represents sensor parameters of types 1, 2, ..., m in the second parameter type set;

[0092] Step S220: The first parameter type set Q formed in the i-th first analysis period i and the second parameter type set F i Perform intersection calculation and output the target type set w corresponding to the i-th first analysis cycle i , w i =Q i ∩F i ; Traverse all first analysis cycles of the same type of first analysis cycle records to obtain the target type set w corresponding to each first analysis cycle; calculate the first intersection c, c={w 1 ,w 2 ,...,w k}, k≥i, k represents the total number of first analysis cycle records of the same type; if c is the same as the target type set of each first analysis cycle, then the sensor parameters in the output target type set are the necessary sensor data; if c does not exist and is the same as the target type set of each first analysis cycle, then the sensor parameters in the output first intersection are the necessary sensor data;

[0093] Step S230: extract the target type set corresponding to the record when the number of sensor parameter types in the k first analysis cycles is the largest as the cache set w max , filter the cache sensor data set D, D=w max -c; store the cached sensor data in the cached sensor data set in correspondence with the track environment characteristics.

[0094] Analyzing the necessary sensor data is to effectively focus on monitoring the different rail vehicles designed with redundant sensor groups in the corresponding rail paths, and to make reasonable storage reminders for the resources required by the source sensors of the necessary sensor data, so as to ensure that when a large amount of similar sensor data is generated by setting up redundant sensor groups, the space utilization rate is improved while ensuring the effectiveness and accuracy of using sensors to obtain data for analysis and decision-making; and after determining the necessary sensor data, the cached sensor data is screened, which can effectively distinguish the differences in the degree of data emphasis, and provide a range of selection and traceability for the sensor data required for the system to identify deviations in different scenarios.

[0095] Step S300 includes the following specific steps:

[0096] Step S310: taking the unit event inspection period corresponding to the second monitoring event as the second analysis period, and replacing the first analysis period in steps S210 to S230 with the second analysis period, and obtaining necessary sensor data and cached sensor data corresponding to the second analysis period;

[0097] Step S320: Retrieve each sensor monitoring event recorded in each second analysis cycle as a verification event, extract the sensor monitoring events corresponding to the same execution purpose and recorded in the first analysis cycle as control events, compare the necessary sensor data recorded in the verification event and the control event, and when the number of necessary sensor data types recorded in the verification event is less than or equal to the number of necessary sensor data types recorded in the control event, output the necessary sensor data recorded in the control event as the necessary sensor data of the corresponding redundant sensor group; when the number of necessary sensor data types recorded in the verification event is greater than the number of necessary sensor data types recorded in the control event, correct the union of the necessary sensor data recorded in the verification event and the necessary sensor data recorded in the control event to be the necessary sensor data of the corresponding redundant sensor group.

[0098] Because it is impossible to effectively predict the types of rail transit events that may require sensor data for decision-making when actually collecting data from redundant sensor groups, it is necessary to increase the diversity of data types under different rail transit scenarios when considering data to determine the necessary sensor data, thereby increasing the intelligence and comprehensiveness of automated monitoring of rail transit data. Comprehensive improvement is achieved through verification. When there is a conflict between the necessary sensor data and cached sensor data recorded by the same type of redundant sensor group, it is retained as necessary sensor data.

[0099] Step S400 includes:

[0100] Step S410: marking the sensor data other than the necessary sensor data and the cached sensor data as non-essential sensor data for each rail vehicle on the same rail path;

[0101] Step S420: Obtain the sensor recognition event of the unit analysis object history record. The sensor recognition event refers to the event of outputting corresponding data by using the sensor recognition response. The sensor recognition event records the response time L and the recognition result Z. The recognition result is determined by the system and the output score is output. For example, when using infrared sensor recognition, there will be a measurement of recognition accuracy, and the clustering algorithm can be used for scoring. The formula is used:

[0102] Y=a 1 ×(1 / L 0 )+a 2 ×Z 0 ,

[0103] Calculate the effective utilization index Y, L of each sensor in the unit analysis object based on the sensor recognition event 0 It means to obtain the normalized average response time of multiple sensor recognition events recorded by the same sensor, Z 0 It means obtaining the normalized average value of the recognition results of multiple sensor recognition events recorded by the same sensor; a 1 、a 2 represents the corresponding reference coefficient;

[0104] The larger the Y is, the more accurate the sensor recognition is and the better the effect is;

[0105] Step S420: sort the sensors in each unit analysis object from large to small according to the corresponding effective utilization index, and select the sensor with the first ranking as the optimal sensor of the corresponding unit analysis object.

[0106] 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.

[0107] 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 automatic monitoring of rail transit data based on big data, characterized in that: The analysis steps include: Step S100: extracting the rail transit monitoring system history records based on sensor monitoring events of rail vehicles on corresponding rail paths with redundant sensor groups with different execution purposes, wherein the sensor monitoring events refer to events that use redundant sensor groups to collect data for analysis and decision-making based on rail environment characteristics; the redundant sensor groups refer to data collection groups composed of multiple sensors with the same execution functions distributed at different positions; Analyze the sensor monitoring events recorded in the history of each track path and construct a unit event inspection cycle for the corresponding track path; Step S200: extracting sensor monitoring events that record the same execution purpose in various track paths and contain one redundant sensor group as the first monitoring event, using the unit event inspection period corresponding to the first monitoring event as the first analysis period, and outputting necessary sensor data of the redundant sensor group recorded by the sensor monitoring event in the first analysis period; Step S300: Filtering sensor monitoring events recorded in various track paths with the same execution purpose and containing different numbers of redundant sensor groups as second monitoring events, and verifying and correcting necessary sensor data of each redundant sensor group based on the second monitoring events; Step S400: Mark the non-essential sensor data based on the necessary sensor data, analyze the non-essential sensor data based on the source redundant sensor group, and determine the optimal sensor source of the non-essential sensor data for each unit analysis object; when the storage space alarm is triggered, retain the non-essential sensor data recorded by the optimal sensor in the redundant sensor group and eliminate other non-essential sensors.

2. The method for automatic monitoring of rail transit data based on big data according to claim 1, characterized in that: The unit event investigation period for constructing the corresponding orbital path comprises the following steps: Extracting a number of sensor monitoring events of the same rail vehicle in valid driving events recorded in each rail path history, wherein the valid driving event is a driving event when the number of sensor monitoring events of the rail vehicle in the corresponding rail path record is the maximum; Constructing a coordinate system, wherein the horizontal axis of the coordinate system records the path length of the track path, and the vertical axis records the track vehicles at different departure times; The sensor monitoring events recorded by different rail vehicles on the same rail path are marked with point coordinates in the coordinate axis; each point coordinate stores the response time and end time of the sensor monitoring event recorded by the corresponding rail vehicle; the response time is the time when the redundant sensor group starts to collect data for decision-making and judgment based on the rail environment characteristics; the end time is the time when the decision-making and judgment are implemented; The time interval between the response times of adjacent sensor monitoring events marked on the same vertical coordinate is the normal monitoring period of the next recorded sensor monitoring event, and the normal monitoring period and the sensor analysis period of the next recorded sensor monitoring event are used to construct a unit event inspection period of the corresponding track path; the sensor analysis period refers to the interval period from the response time to the end time of each sensor monitoring event; All coordinate points with the same ordinate recorded in the coordinate system are traversed to generate a unit event inspection period corresponding to each rail vehicle, and a set of unit event inspection periods corresponding to the rail path is constructed by all rail vehicles.

3. The method for automatic monitoring of rail transit data based on big data according to claim 2 is characterized in that: Generating necessary sensor data of a redundant sensor group for recording sensor monitoring events in a first analysis period includes the following specific steps: Step S210: Mark the first analysis cycles with the same execution purpose as a type of analysis cycle, obtain the sensor parameters recorded by the redundant sensor group corresponding to the normal monitoring cycle in the i-th first analysis cycle of each type of first analysis cycle record, and form a first parameter type set Q i , Q i ={q1,q2,......,q n } i , where q1,q2,......,q n represents the sensor parameters of the 1st, 2nd, ..., nth types in the first parameter type set, and the sensor parameters recorded by the redundant sensor group required when the sensor analysis cycle in the i-th first analysis cycle corresponds to the decision-making judgment, and constitutes the second parameter type set F i , F i ={f1,f2,...,f m } i ; f1,f2,...,f m represents sensor parameters of types 1, 2, ..., m in the second parameter type set; Step S220: The first parameter type set Q formed in the i-th first analysis period i and the second parameter type set F i Perform intersection calculation and output the target type set w corresponding to the i-th first analysis cycle i , w i =Q i ∩F i ; Traverse all the first analysis cycles of the same type of first analysis cycle records to obtain the target type set w corresponding to each first analysis cycle; calculate the first intersection c, c={w1,w2,...,w k }, k≥i, k represents the total number of first analysis cycle records of the same type; if c is the same as the target type set of each first analysis cycle, then the sensor parameters in the output target type set are the necessary sensor data; if c does not exist and is the same as the target type set of each first analysis cycle, then the sensor parameters in the output first intersection are the necessary sensor data; Step S230: extract the target type set corresponding to the record when the number of sensor parameter types in the k first analysis cycles is the largest as the cache set w max , filter the cache sensor data set D, D=w max -c; store the cached sensor data in the cached sensor data set in correspondence with the track environment characteristics.

4. The method for automatic monitoring of rail transit data based on big data according to claim 3 is characterized in that: The step S300 includes the following specific steps: Step S310: taking the unit event inspection period corresponding to the second monitoring event as the second analysis period, and replacing the first analysis period in steps S210 to S230 with the second analysis period, and obtaining necessary sensor data and cached sensor data corresponding to the second analysis period; Step S320: Retrieve each sensor monitoring event recorded in each second analysis cycle as a verification event, extract the sensor monitoring event corresponding to the same execution purpose and recorded in the first analysis cycle as a control event, compare the necessary sensor data recorded in the verification event with the necessary sensor data recorded in the control event, and when the number of necessary sensor data types recorded in the verification event is less than or equal to the number of necessary sensor data types recorded in the control event, output the necessary sensor data recorded in the control event as the necessary sensor data corresponding to the redundant sensor group; When the number of necessary sensor data types of the verification event record is greater than the number of necessary sensor data types of the control event record, the union of the necessary sensor data of the verification event record and the necessary sensor data of the control event record is modified to be the necessary sensor data of the corresponding redundant sensor group.

5. The method for automatic monitoring of rail transit data based on big data according to claim 1 is characterized in that: The step S400 includes: Step S410: marking the sensor data other than the necessary sensor data and the cached sensor data as non-essential sensor data for each rail vehicle on the same rail path; Step S420: Acquire the sensor recognition event of the unit analysis object history record, wherein the sensor recognition event refers to an event that uses the sensor recognition response to output corresponding data, and the sensor recognition event records the response time L and the recognition result Z; using the formula: Y=a1×(1 / L0)+a2×Z0, The effective utilization index Y of each sensor based on the sensor recognition event in the unit analysis object is calculated, wherein L0 represents the normalized average value of the response time of multiple sensor recognition events recorded by the same sensor, and Z0 represents the normalized average value of the recognition results of multiple sensor recognition events recorded by the same sensor; a1 and a2 represent corresponding reference coefficients; Step S420: sort the sensors in each unit analysis object from large to small according to the corresponding effective utilization index, and select the sensor with the first ranking as the optimal sensor of the corresponding unit analysis object.

6. A rail transit data automatic monitoring system based on big data, such as using a rail transit data automatic monitoring method based on big data according to any one of claims 1 to 5, characterized in that: It includes sensor monitoring event extraction module, unit event inspection cycle analysis module, necessary sensor data analysis module, calibration and correction module and optimal sensor early warning module; The sensor monitoring event extraction module is used to extract sensor monitoring events of rail vehicles on corresponding rail paths based on redundant sensor groups with different execution purposes arranged in the rail transit monitoring system historical records; The unit event inspection cycle analysis module is used to construct a unit event inspection cycle corresponding to the track path; The necessary sensor data analysis module is used to output necessary sensor data of the sensor monitoring event record redundant sensor group within the first analysis period; The verification and correction module is used to verify and correct the necessary sensor data of each redundant sensor group; The optimal sensor early warning module is used to determine the optimal sensor source of the unnecessary sensor data of each unit analysis object.

7. The rail transit data automatic monitoring system based on big data according to claim 6 is characterized by: The unit event inspection period analysis module includes a coordinate system construction unit, a point coordinate identification unit, a monitoring period type determination unit and a unit event inspection period output unit; The coordinate system construction unit is used to construct a coordinate system in which the horizontal axis records the path length of the track path and the vertical axis records the track vehicles at different departure times; The point coordinate identification unit is used to identify the sensor monitoring events of different rail vehicles recorded on the same rail path with point coordinates in the coordinate axis; The monitoring cycle type determination unit is used to determine the normal monitoring cycle and the sensing analysis cycle; The unit event inspection period output unit is used to traverse all coordinate points with the same ordinate recorded in the coordinate system, generate a unit event inspection period corresponding to each rail vehicle, and construct a unit event inspection period set corresponding to the rail path by all rail vehicles.

8. The rail transit data automatic monitoring system based on big data according to claim 6 is characterized by: The necessary sensor data analysis module includes a parameter type set construction unit, a target type set output unit, a first intersection calculation unit and a sensor data output unit; The parameter type set construction unit is used to construct a first parameter type set and a second parameter type set; The target type set output unit is used to calculate the target type set by taking the intersection of the first parameter type set and the second parameter type set; The first intersection calculation unit is used to traverse all analysis periods of the same type of analysis period records to obtain a target type set corresponding to each analysis period, and calculate a first intersection; The sensor data output unit is used to output necessary sensor data and buffer sensor data.

9. The rail transit data automatic monitoring system based on big data according to claim 6 is characterized by: The optimal sensor early warning module includes a non-essential data determination unit, an effective utilization index calculation unit and a ranking output unit; The non-essential data determination unit is used to mark other sensor data except necessary sensor data and cached sensor data as non-essential sensor data for each rail vehicle on the same rail path; The effective utilization index calculation unit is used to calculate the effective utilization index of each sensor in the unit analysis object based on the sensor recognition event; The ranking output unit is used to sort the sensors in each unit analysis object from large to small according to the corresponding effective utilization index, and select the sensor with the first ranking as the optimal sensor of the corresponding unit analysis object.

Images