Data management system and method of industrial personal computer based on Internet of Things

Abstract

The invention discloses a data management system and method for an industrial personal computer based on the Internet of Things, and relates to the technical field of industrial personal computer data, and the management method comprises the following steps: detecting the operation conditions of operation equipment and the industrial personal computer, and generating an operation detection record set and a self-inspection record set; carrying out abnormal condition analysis on any operation detection record and any self-detection record; analyzing abnormal change conditions among different operation detection records, and extracting features; identifying abnormal features based on the abnormal condition of the randomly selected self-checking record; for any abnormal feature, constructing a feature function to present a change condition; establishing a self-checking triggering condition, and judging whether the industrial personal computer is subjected to self-checking or not; carrying out abnormity identification on the operation detection record generated in real time; and abnormal feature extraction is carried out aiming at the abnormal condition, whether self-inspection is carried out on the industrial personal computer is judged, and abnormal equipment is confirmed and abnormal reminding is carried out.

Description

Technical Field

[0001] The present invention relates to the field of industrial computer data technology, and in particular to a data management system and method for an industrial computer based on the Internet of Things. Background Art

[0002] Industrial computers are computers designed for industrial environments. They are mainly used to detect and control equipment and systems in industrial automation processes. Industrial computers can collect and detect the operating conditions of various operating equipment in the industrial production process in real time, detect abnormal conditions in time, and ensure the safety and stability of the production process.

[0003] However, in the process of the industrial computer detecting the running equipment, since the industrial computer itself is also a device with the risk of abnormality, if the industrial computer has an abnormality, it will lead to wrong judgment of the detection results of the running equipment, and miss the running equipment that actually has abnormalities, which seriously affects the efficient development of the production process and hinders production efficiency. Summary of the invention

[0004] The purpose of the present invention is to provide a data management system and method for industrial computers based on the Internet of Things to solve the problems raised in the prior art.

[0005] To achieve the above object, the present invention provides the following technical solution: a data management method for an industrial computer based on the Internet of Things, the management method comprising the following steps:

[0006] Step S100: Detect the running status of the running equipment through the industrial computer, and regularly perform self-check on the industrial computer, and respectively generate corresponding operation detection record sets and self-check record sets; perform abnormal situation analysis on the running status of any operation detection record and any self-check record;

[0007] Step S200: Analyze the abnormal changes between different operation detection records, and extract the features that affect the operation status of the operation equipment; based on the abnormal conditions of any selected self-test record, identify the abnormal features of a specific operation detection record;

[0008] Step S300: for any abnormal feature, construct a feature function to present the change of the abnormal feature; analyze the deviation of all abnormal features in a specific operation detection record, establish a self-check trigger condition, and determine whether to perform a self-check when an abnormality occurs in the operation detection record;

[0009] Step S400: Whenever an operation detection record is generated in real time, anomaly identification is performed on the operation detection record generated in real time; in case of anomaly, abnormal features are extracted from the operation detection record generated in real time, and it is determined whether to perform self-inspection on the industrial computer, confirm the equipment with abnormality and issue an abnormal reminder.

[0010] Furthermore, step S100 includes the following steps:

[0011] Step S101: It is assumed that several types of sensors are installed on the industrial computer and the running equipment respectively. Whenever the running equipment is working, the data acquisition module on the industrial computer reads the data of the sensors on the running equipment to generate a corresponding running detection record; the running detection records generated during each running process of the running equipment are summarized to obtain a set of running detection records;

[0012] Step S102: Detect the operation status of the industrial computer itself every unit period, collect the operation status of the industrial computer through the sensor installed on the industrial computer, and generate a corresponding self-test record; summarize the self-test records generated by each self-test process of the industrial computer to obtain a self-test record set; analyze the self-test records regularly performed by the industrial computer, which can help identify abnormal records in the operation test records, which is conducive to providing a data basis for subsequent self-test trigger conditions;

[0013] Step S103: randomly select an operation detection record or self-check record and set the selected record as the target record; divide the operation data stored in the target record into operation data of several dimensions, preset corresponding evaluation rules for each dimension, and obtain abnormal values ​​under any dimension; the dimensions of operation data division include equipment temperature, pressure, flow rate and vibration, etc., which can be used to determine whether there is an abnormality in the operation of the equipment;

[0014] Step S104: Set an importance level for each dimension, and set the importance level of the ath dimension to L a ; Get the outlier value of the ath dimension as Y a , according to the formula:

[0015]

[0016] Where a and b are both positive integers and a∈(1,z), b∈(1,z), z is the total number of dimensions, L b is the importance level of the bth dimension; the outlier value Y of the target record is calculated.

[0017] Further, step S200 includes the following steps:

[0018] Step S201: Preset an abnormal threshold Y for determining whether there is an abnormality in the running detection record max , select any one running test record from the running test record set, and obtain the abnormal value of the selected running test record as Y. If Y>Y max , then the selected operation detection record is set as the abnormal detection record;

[0019] Step S202: Divide the operation detection record set into a normal detection record set and an abnormal detection record set, arbitrarily select a dimension, obtain the operation data of each normal detection record under the dimension in the normal detection record set, and obtain the normal value range of the dimension;

[0020] Step S203: randomly select an abnormal detection record from the abnormal detection record set, obtain the operation data under the dimension, and if the operation data is not within the normal value range, extract features from the dimension; perform feature extraction on each abnormal detection record to obtain a feature set; feature extraction on the operation detection record mainly extracts the dimensions that do not meet the normal data interval, and the changes presented by these abnormal data can provide a data basis for subsequent abnormal feature extraction;

[0021] Step S204: Obtain a self-test record set, sort all self-test records and operation detection records in the self-test record set and the operation detection record set from front to back according to the record generation time; select any self-test record, preset a self-test abnormality threshold Y ’ max , if the abnormal value Y of the self-test record ’ >Y ’ max , then the self-check record is marked as abnormal;

[0022] Step S205: randomly select a self-check record, and simultaneously obtain the operation detection record located before the selected self-check record. If the operation detection record is an abnormal detection record, the obtained operation detection record is set as a specific operation detection record; extract a number of features from the specific operation detection record and set them as abnormal features.

[0023] Further, step S300 includes the following steps:

[0024] Step S301: arbitrarily select an abnormal feature, and if a specific operation detection record is obtained that contains the selected abnormal feature, then the obtained specific operation detection record is set as a target detection record;

[0025] Step S302: Obtain the change of the operation data corresponding to the abnormal feature selected in the target detection record over time, use the time point as the horizontal coordinate and the operation data as the vertical coordinate, and build a two-dimensional plane rectangular coordinate system to present the change of the selected abnormal feature; construct a time window with a window length of N, set the operation data at time t as D(t), according to the formula:

[0026]

[0027] Where i1 and i2 are positive integers and i1∈(0,2t), i2∈(tN / 2,t+N / 2), D(i1) is the operation data at time i1, D(i2) is the operation data at time i2, and T is the total duration of the target detection record; the average operation data D at time t is calculated. ave (t);

[0028] Step S303: extract the target dimension corresponding to the selected abnormal feature, and obtain the normal value range of the target dimension; set the normal value range of the target dimension to (D1, D2), and construct the feature function F(t):

[0029]

[0030] The deviation degree of the selected abnormal feature at different times is calculated by the characteristic function F(t); by analyzing the deviation degree of the abnormal feature at different times, the data change of the abnormal feature can be known, and the normal deviation during the normal operation process can also be used to obtain the abnormal deviation, so that the existing abnormal situation can be identified;

[0031] Step S304: From the characteristic function of the selected abnormal feature, arbitrarily select two adjacent time points, obtain two offset degrees F(t1) and F(t2), and obtain the offset difference ΔF=F(t2)-F(t1); obtain the offset differences of all two adjacent time points, and calculate the average value to obtain the average difference ΔF ave , if ΔF>ΔF ave , then the offset difference is set as the abnormal offset value, the number of abnormal offset values ​​in the statistical feature function is m, and the number of offset differences contained in the feature function is set to m total , the abnormal deviation ratio of the characteristic function is obtained as η = m / m total There are two ways to identify anomalies in the detection records: one is that the offset amplitude is too large, and the other is that the difference in the offset amplitude between two adjacent time points does not conform to the law and has large fluctuations. Therefore, by formulating corresponding conditions for these two methods, it is possible to identify possible error detections of industrial computers.

[0032] Step S305: Obtain c adjacent and continuous operation detection records before the target detection record, obtain the abnormal offset ratio of each operation detection record, and obtain an offset ratio range (η 1 ,η 2 ), if η∈(η 1 ,η 2 ), then the difference in the proportion of abnormal deviation is extracted as Δη=|η-η 1 |, otherwise, extract the offset ratio range (η1 ,η 2 );

[0033] Step S306: Obtain all specific operation detection records containing the selected abnormal features, extract the offset ratio range or abnormal offset ratio difference in each specific operation detection record, and select the minimum value η in each offset ratio range min and the maximum value η max Get an abnormal proportion range (η min ,η max ) and use it as a trigger condition for the selected abnormal feature to perform self-check on the industrial computer; select an abnormal offset ratio difference with the smallest value as another trigger condition for the selected abnormal feature to perform self-check on the industrial computer; by establishing corresponding self-check trigger rules for different features, the industrial computer can detect the running equipment while also discovering its own abnormal conditions in a timely manner, thereby reducing the probability of false detection and improving efficiency.

[0034] Furthermore, step S400 includes the following steps:

[0035] Step S401: Obtain a real-time generated operation detection record and set it as a real-time detection record, obtain the operation data of the real-time detection record in each dimension, and calculate the abnormal value Y of the real-time detection record. now ;

[0036] Step S402: Obtaining an abnormality threshold Y for determining whether the running detection record has an abnormality max , if Y now >Y max , then the running data of each dimension is compared with the corresponding normal value range to obtain several abnormal features;

[0037] Step S403: arbitrarily select an abnormal feature, obtain the presentation of the abnormal feature-related operating data in the real-time monitoring record over time, and obtain the abnormal deviation proportion η of the abnormal feature at each moment now and the difference Δη of the abnormal deviation ratio now ; Set the two self-test trigger conditions of the selected abnormal features to be the abnormal proportion range (η min ,η max ) and the difference in the abnormal deviation ratio Δη min , if η now ∈(η min ,η max ) or Δη now >η min , a self-check reminder is sent to the industrial computer;

[0038] Step S404: after the industrial computer is self-checked, if there is no abnormality in the industrial computer, an abnormality reminder is sent to the real-time detection record, and an abnormality maintenance reminder is sent to the running equipment.

[0039] In order to better implement the above method, a data management system for industrial computers is also proposed. The management system includes a detection data acquisition module, a detection anomaly analysis module, an anomaly self-detection recognition module and a real-time anomaly judgment module;

[0040] The detection data acquisition module is used to detect the operating status of the running equipment through the industrial computer, and to perform self-checks on the industrial computer regularly, and to generate corresponding operation detection record sets and self-check record sets respectively; to analyze the abnormal conditions of any operation detection record and any self-check record;

[0041] The detection anomaly analysis module is used to analyze the abnormal changes between different operation detection records and extract the features that affect the operation status of the running equipment; based on the abnormal conditions of any selected self-test records, the abnormal features of specific operation detection records are identified;

[0042] The abnormal self-check identification module is used to construct a feature function for any abnormal feature to present the change of the abnormal feature; analyze the deviation of all abnormal features in a specific operation detection record, establish a self-check trigger condition, and determine whether to perform a self-check when an abnormality occurs in the operation detection record;

[0043] The real-time anomaly judgment module is used to identify anomalies of the real-time generated operation detection record whenever an operation detection record is generated in real time; in the case of anomalies, the abnormal features of the real-time generated operation detection record are extracted to determine whether to perform self-inspection on the industrial computer, confirm the equipment with abnormalities and issue an abnormal reminder.

[0044] Further, the detection data acquisition module includes a detection record generation unit and an abnormal situation analysis unit;

[0045] The detection record generation unit is used to detect the operating status of the running equipment through the industrial computer, and regularly perform self-inspection on the industrial computer to generate corresponding operation detection record sets and self-inspection record sets respectively; the abnormal situation analysis unit is used to perform abnormal situation analysis on the operating status of any operation detection record and any self-inspection record.

[0046] Further, the detection anomaly analysis module includes a device feature extraction unit and an abnormal feature recognition unit;

[0047] The equipment feature extraction unit is used to analyze the abnormal changes between different operation detection records and extract the features that affect the operation status of the running equipment; the abnormal feature identification unit is used to identify the abnormal features of specific operation detection records based on the abnormal conditions of any selected self-test records.

[0048] Further, the abnormal self-check identification module includes a characteristic function presentation unit and a self-check trigger setting unit;

[0049] The feature function presentation unit is used to construct a feature function for any abnormal feature to present the changes of the abnormal feature; the self-test trigger setting unit is used to analyze the deviation of all abnormal features in a specific operation detection record, establish a self-test trigger condition, and determine whether to perform a self-test when an abnormality occurs in the operation detection record.

[0050] Furthermore, the real-time abnormality judgment module includes an abnormality real-time identification unit and an abnormality judgment reminder unit;

[0051] The real-time abnormality identification unit is used to identify the abnormality of the real-time generated operation detection record whenever a real-time operation detection record is generated; the abnormality judgment and reminder unit is used to extract abnormal features of the real-time generated operation detection record in response to the existence of abnormalities, determine whether to perform self-inspection on the industrial computer, confirm the equipment with abnormalities and issue abnormal reminders.

[0052] Compared with the prior art, the present invention has the following beneficial effects:

[0053] 1. The present invention simultaneously manages the abnormality of the industrial computer and the running equipment detected by the industrial computer, so as to help the staff to have an accurate grasp of the abnormal situation occurring in the production process, identify the specific cause of the abnormality as quickly as possible, and reduce the production risk caused by the abnormality of the equipment;

[0054] 2. The present invention not only detects abnormalities of the running equipment through the industrial computer, but also takes into account the possible abnormalities of the industrial computer itself; when the abnormal deviation of the running equipment is large, the industrial computer is self-checked to help identify the cause of the abnormality in time, avoid the occurrence of false detection by the industrial computer, quickly solve the abnormality and improve the continuity of the production process;

[0055] 3. The present invention sets corresponding self-check trigger conditions in different dimensions, which can help determine whether it is necessary to perform self-check on the industrial computer when an abnormality occurs in the running equipment. By setting the self-check conditions, it can avoid that too frequent self-checks hinder the development of the production process, and it can also ensure that the abnormal identification of the industrial computer can be discovered in time, thereby improving the overall production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1A schematic diagram of the steps of a data management method for an industrial computer based on the Internet of Things;

[0057] Figure 2 The figure is a structural diagram of a data management system for industrial computers based on the Internet of Things. DETAILED DESCRIPTION

[0058] 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.

[0059] Example: Figure 1 to Figure 2 As shown, the present invention provides a data management method for an industrial computer based on the Internet of Things, and the management method comprises the following steps:

[0060] Step S100: Detect the running status of the running equipment through the industrial computer, and regularly perform self-check on the industrial computer, and respectively generate corresponding operation detection record sets and self-check record sets; perform abnormal situation analysis on the running status of any operation detection record and any self-check record;

[0061] Wherein, step S100 includes the following steps:

[0062] Step S101: It is assumed that several types of sensors are installed on the industrial computer and the running equipment respectively. Whenever the running equipment is working, the data acquisition module on the industrial computer reads the data of the sensors on the running equipment to generate a corresponding running detection record; the running detection records generated during each running process of the running equipment are summarized to obtain a set of running detection records;

[0063] Step S102: Detect the operation status of the industrial computer itself every unit period, collect the operation status of the industrial computer through the sensor installed on the industrial computer, and generate a corresponding self-test record; summarize the self-test records generated by each self-test process of the industrial computer to obtain a self-test record set;

[0064] Step S103: randomly select an operation detection record or a self-check record and set the selected record as a target record; divide the operation data stored in the target record into operation data of several dimensions, preset corresponding evaluation rules for each dimension, and obtain abnormal values ​​under any dimension;

[0065] Step S104: Set an importance level for each dimension, and set the importance level of the ath dimension to L a ; Get the outlier value of the ath dimension as Ya , according to the formula:

[0066]

[0067] Where a and b are both positive integers and a∈(1,z), b∈(1,z), z is the total number of dimensions, L b is the importance level of the bth dimension; calculate the outlier value Y of the target record;

[0068] Example 1: Set the operation detection record to be divided into three dimensions, and the preset importance levels are 1, 2, and 2 respectively; obtain the abnormal values ​​under the three dimensions as 5, 8, and 10 respectively, and calculate the abnormal value of the operation detection record Y=5×1 / 5+8×2 / 5+10×2 / 5=1+3.2+4=8.2; set the abnormal threshold Y max =10, so the running detection record is not an abnormal detection record.

[0069] Step S200: Analyze the abnormal changes between different operation detection records, and extract the features that affect the operation status of the operation equipment; based on the abnormal conditions of any selected self-test record, identify the abnormal features of a specific operation detection record;

[0070] Wherein, step S200 includes the following steps:

[0071] Step S201: Preset an abnormal threshold Y for determining whether there is an abnormality in the running detection record max , select any one running test record from the running test record set, and obtain the abnormal value of the selected running test record as Y. If Y>Y max , then the selected operation detection record is set as the abnormal detection record;

[0072] Step S202: Divide the operation detection record set into a normal detection record set and an abnormal detection record set, arbitrarily select a dimension, obtain the operation data of each normal detection record under the dimension in the normal detection record set, and obtain the normal value range of the dimension;

[0073] Step S203: randomly selecting an abnormality detection record from the abnormality detection record set, obtaining the operation data under the dimension, and if the operation data is not within the normal value range, extracting features from the dimension; performing feature extraction on each abnormality detection record to obtain a feature set;

[0074] Step S204: Obtain a self-test record set, sort all self-test records and operation detection records in the self-test record set and the operation detection record set from front to back according to the record generation time; select any self-test record, preset a self-test abnormality threshold Y ’max , if the abnormal value Y of the self-test record ’ >Y ’ max , then the self-check record is marked as abnormal;

[0075] Step S205: randomly select a self-check record, and simultaneously obtain the operation detection record located before the selected self-check record. If the operation detection record is an abnormal detection record, the obtained operation detection record is set as a specific operation detection record; extract a number of features from the specific operation detection record and set them as abnormal features.

[0076] Step S300: for any abnormal feature, construct a feature function to present the change of the abnormal feature; analyze the deviation of all abnormal features in a specific operation detection record, establish a self-check trigger condition, and determine whether to perform a self-check when an abnormality occurs in the operation detection record;

[0077] Wherein, step S300 includes the following steps:

[0078] Step S301: arbitrarily select an abnormal feature, and if a specific operation detection record is obtained that contains the selected abnormal feature, then the obtained specific operation detection record is set as a target detection record;

[0079] Step S302: Obtain the change of the operation data corresponding to the abnormal feature selected in the target detection record over time, use the time point as the horizontal coordinate and the operation data as the vertical coordinate, and build a two-dimensional plane rectangular coordinate system to present the change of the selected abnormal feature; construct a time window with a window length of N, set the operation data at time t as D(t), according to the formula:

[0080]

[0081] Where i1 and i2 are positive integers and i1∈(0,2t), i2∈(tN / 2,t+N / 2), D(i1) is the operation data at time i1, D(i2) is the operation data at time i2, and T is the total duration of the target detection record; the average operation data D at time t is calculated. ave (t);

[0082] Example 2: Set the total duration of the target detection record to 60 minutes, obtain the operation data every 1 minute, and set the window length of the time window to 10 minutes; obtain the operation data at 10 minutes. Because 10 minutes>5 minutes, the third row formula is called to calculate the average operation data to obtain the average operation data of 10 minutes as 10;

[0083] Step S303: extract the target dimension corresponding to the selected abnormal feature, and obtain the normal value range of the target dimension; set the normal value range of the target dimension to (D1, D2), and construct the feature function F(t):

[0084]

[0085] The deviation degree of the selected abnormal features at different times is calculated by the characteristic function F(t);

[0086] Step S304: From the characteristic function of the selected abnormal feature, arbitrarily select two adjacent time points, obtain two offset degrees F(t1) and F(t2), and obtain the offset difference ΔF=F(t2)-F(t1); obtain the offset differences of all two adjacent time points, and calculate the average value to obtain the average difference ΔF ave , if ΔF>ΔF ave , then the offset difference is set as the abnormal offset value, the number of abnormal offset values ​​in the statistical feature function is m, and the number of offset differences contained in the feature function is set to m total , the abnormal deviation ratio of the characteristic function is obtained as η = m / m total ;

[0087] Step S305: Obtain c adjacent and continuous operation detection records before the target detection record, obtain the abnormal offset ratio of each operation detection record, and obtain an offset ratio range (η 1 ,η 2 ), if η∈(η 1 ,η 2 ), then the difference in the proportion of abnormal deviation is extracted as Δη=|η-η 1 |, otherwise, extract the offset ratio range (η 1 ,η 2 );

[0088] Step S306: Obtain all specific operation detection records containing the selected abnormal features, extract the offset ratio range or abnormal offset ratio difference in each specific operation detection record, and select the minimum value η in each offset ratio range min and the maximum value η max Get an abnormal proportion range (η min ,η max ), and use it as a trigger condition for the selected abnormal feature to perform self-check on the industrial computer; select the abnormal offset ratio difference with the smallest value as another trigger condition for the selected abnormal feature to perform self-check on the industrial computer.

[0089] Step S400: Whenever a running detection record is generated in real time, an abnormality recognition is performed on the running detection record generated in real time; in case of an abnormality, an abnormal feature is extracted from the running detection record generated in real time, and it is determined whether to perform a self-check on the industrial computer, and the equipment with abnormality is confirmed and an abnormality reminder is issued;

[0090] Wherein, step S400 includes the following steps:

[0091] Step S401: Obtain a real-time generated operation detection record and set it as a real-time detection record, obtain the operation data of the real-time detection record in each dimension, and calculate the abnormal value Y of the real-time detection record. now ;

[0092] Step S402: Obtaining an abnormality threshold Y for determining whether the running detection record has an abnormality max , if Y now >Y max , then the running data of each dimension is compared with the corresponding normal value range to obtain several abnormal features;

[0093] Step S403: arbitrarily select an abnormal feature, obtain the presentation of the abnormal feature-related operating data in the real-time monitoring record over time, and obtain the abnormal deviation proportion η of the abnormal feature at each moment now and the difference Δη of the abnormal deviation ratio now ; Set the two self-test trigger conditions of the selected abnormal features to be the abnormal proportion range (η min ,η max ) and the difference in the abnormal deviation ratio Δη min , if η now ∈(η min ,η max ) or Δη now >η min , a self-check reminder is sent to the industrial computer;

[0094] Step S404: after the industrial computer is self-checked, if there is no abnormality in the industrial computer, an abnormality reminder is sent to the real-time detection record, and an abnormality maintenance reminder is sent to the running equipment.

[0095] A data management system for an industrial computer, the management system comprising a detection data acquisition module, a detection anomaly analysis module, an anomaly self-detection identification module and a real-time anomaly judgment module;

[0096] The detection data acquisition module is used to detect the operating status of the running equipment through the industrial computer, and to perform self-checks on the industrial computer regularly, and to generate corresponding operation detection record sets and self-check record sets respectively; to analyze the abnormal conditions of any operation detection record and any self-check record;

[0097] The detection anomaly analysis module is used to analyze the abnormal changes between different operation detection records and extract the features that affect the operation status of the running equipment; based on the abnormal conditions of any selected self-test records, the abnormal features of specific operation detection records are identified;

[0098] The abnormal self-check identification module is used to construct a feature function for any abnormal feature to present the change of the abnormal feature; analyze the deviation of all abnormal features in a specific operation detection record, establish a self-check trigger condition, and determine whether to perform a self-check when an abnormality occurs in the operation detection record;

[0099] The real-time anomaly judgment module is used to identify anomalies of the real-time generated operation detection record whenever an operation detection record is generated in real time; in the case of anomalies, the abnormal features of the real-time generated operation detection record are extracted to determine whether to perform self-inspection on the industrial computer, confirm the equipment with abnormalities and issue an abnormal reminder.

[0100] Among them, the detection data acquisition module includes a detection record generation unit and an abnormal situation analysis unit;

[0101] The detection record generation unit is used to detect the operating status of the running equipment through the industrial computer, and regularly perform self-inspection on the industrial computer to generate corresponding operation detection record sets and self-inspection record sets respectively; the abnormal situation analysis unit is used to perform abnormal situation analysis on the operating status of any operation detection record and any self-inspection record.

[0102] Among them, the detection anomaly analysis module includes a device feature extraction unit and an abnormal feature recognition unit;

[0103] The equipment feature extraction unit is used to analyze the abnormal changes between different operation detection records and extract the features that affect the operation status of the running equipment; the abnormal feature identification unit is used to identify the abnormal features of specific operation detection records based on the abnormal conditions of any selected self-test records.

[0104] Among them, the abnormal self-check identification module includes a characteristic function presentation unit and a self-check trigger setting unit;

[0105] The feature function presentation unit is used to construct a feature function for any abnormal feature to present the changes of the abnormal feature; the self-test trigger setting unit is used to analyze the deviation of all abnormal features in a specific operation detection record, establish a self-test trigger condition, and determine whether to perform a self-test when an abnormality occurs in the operation detection record.

[0106] Among them, the real-time abnormality judgment module includes an abnormality real-time identification unit and an abnormality judgment reminder unit;

[0107] The real-time abnormality identification unit is used to identify the abnormality of the real-time generated operation detection record whenever a real-time operation detection record is generated; the abnormality judgment and reminder unit is used to extract abnormal features of the real-time generated operation detection record in response to the existence of abnormalities, determine whether to perform self-inspection on the industrial computer, confirm the equipment with abnormalities and issue abnormal reminders.

[0108] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A data management method for industrial computers based on the Internet of Things, characterized in that: The management method comprises the following steps: Step S100: Detect the running status of the running equipment through the industrial computer, and regularly perform self-check on the industrial computer, and respectively generate corresponding operation detection record sets and self-check record sets; perform abnormal situation analysis on the running status of any operation detection record and any self-check record; Step S200: Analyze the abnormal changes between different operation detection records, and extract the features that affect the operation status of the operation equipment; based on the abnormal conditions of any selected self-test record, identify the abnormal features of a specific operation detection record; Step S300: for any abnormal feature, construct a feature function to present the change of the abnormal feature; analyze the deviation of all abnormal features in a specific operation detection record, establish a self-check trigger condition, and determine whether to perform a self-check when an abnormality occurs in the operation detection record; Step S400: Whenever an operation detection record is generated in real time, anomaly identification is performed on the operation detection record generated in real time; in case of anomaly, abnormal features are extracted from the operation detection record generated in real time, and it is determined whether to perform self-inspection on the industrial computer, confirm the equipment with abnormality and issue an abnormal reminder.

2. The data management method of an industrial computer based on the Internet of Things according to claim 1 is characterized in that: The step S100 includes the following steps: Step S101: It is assumed that several types of sensors are installed on the industrial computer and the running equipment respectively. Whenever the running equipment is working, the data acquisition module on the industrial computer reads the data of the sensors on the running equipment to generate a corresponding running detection record; the running detection records generated during each running process of the running equipment are summarized to obtain a set of running detection records; Step S102: Detect the operation status of the industrial computer itself every unit period, collect the operation status of the industrial computer through the sensor installed on the industrial computer, and generate a corresponding self-test record; summarize the self-test records generated by each self-test process of the industrial computer to obtain a self-test record set; Step S103: randomly select an operation detection record or a self-check record and set the selected record as a target record; divide the operation data stored in the target record into operation data of several dimensions, preset corresponding evaluation rules for each dimension, and obtain abnormal values ​​under any dimension; Step S104: Set an importance level for each dimension, and set the importance level of the ath dimension to L a ; Get the outlier value of the ath dimension as Y a , according to the formula: Where a and b are both positive integers and a∈(1,z), b∈(1,z), z is the total number of dimensions, L b is the importance level of the bth dimension; the outlier value Y of the target record is calculated.

3. The data management method of an industrial computer based on the Internet of Things according to claim 2 is characterized in that: The step S200 includes the following steps: Step S201: Preset an abnormal threshold Y for determining whether there is an abnormality in the running detection record max , select any one running test record from the running test record set, and obtain the abnormal value of the selected running test record as Y. If Y>Y max , then the selected operation detection record is set as the abnormal detection record; Step S202: Divide the operation detection record set into a normal detection record set and an abnormal detection record set, arbitrarily select a dimension, obtain the operation data of each normal detection record under the dimension in the normal detection record set, and obtain the normal value range of the dimension; Step S203: randomly selecting an abnormality detection record from the abnormality detection record set, obtaining the operation data under the dimension, and if the operation data is not within the normal value range, extracting features from the dimension; performing feature extraction on each abnormality detection record to obtain a feature set; Step S204: Obtain a self-test record set, sort all self-test records and operation detection records in the self-test record set and the operation detection record set from front to back according to the record generation time; select any self-test record, preset a self-test abnormality threshold Y ’ max , if the abnormal value Y of the self-test record ’ >Y ’ max , then the self-check record is marked as abnormal; Step S205: randomly select a self-check record, and simultaneously obtain the operation detection record located before the selected self-check record. If the operation detection record is an abnormal detection record, the obtained operation detection record is set as a specific operation detection record; extract a number of features from the specific operation detection record and set them as abnormal features.

4. The data management method of an industrial computer based on the Internet of Things according to claim 3 is characterized in that: The step S300 includes the following steps: Step S301: arbitrarily select an abnormal feature, and if a specific operation detection record is obtained that contains the selected abnormal feature, then the obtained specific operation detection record is set as a target detection record; Step S302: Obtain the change of the operation data corresponding to the abnormal feature selected in the target detection record over time, use the time point as the horizontal coordinate and the operation data as the vertical coordinate, and build a two-dimensional plane rectangular coordinate system to present the change of the selected abnormal feature; construct a time window with a window length of N, set the operation data at time t as D(t), according to the formula: Where i1 and i2 are positive integers and i1∈(0,2t), i2∈(tN / 2,t+N / 2), D(i1) is the operation data at time i1, D(i2) is the operation data at time i2, and T is the total duration of the target detection record; the average operation data D at time t is calculated. ave (t); Step S303: extract the target dimension corresponding to the selected abnormal feature, and obtain the normal value range of the target dimension; set the normal value range of the target dimension to (D1, D2), and construct the feature function F(t): The deviation degree of the selected abnormal features at different times is calculated by the characteristic function F(t); Step S304: From the characteristic function of the selected abnormal feature, arbitrarily select two adjacent time points, obtain two offset degrees F(t1) and F(t2), and obtain the offset difference ΔF=F(t2)-F(t1); obtain the offset differences of all two adjacent time points, and calculate the average value to obtain the average difference ΔF ave , if ΔF>ΔF ave , then the offset difference is set as the abnormal offset value, the number of abnormal offset values ​​in the statistical feature function is m, and the number of offset differences contained in the feature function is set to m total , the abnormal deviation ratio of the characteristic function is obtained as η = m / m total ; Step S305: Obtain c adjacent and continuous operation detection records before the target detection record, obtain the abnormal offset ratio of each operation detection record, and obtain an offset ratio range (η1, η2). If η∈(η1, η2), then extract the abnormal offset ratio difference Δη=|η-η1|, otherwise, extract the offset ratio range (η1, η2); Step S306: Obtain all specific operation detection records containing the selected abnormal features, extract the offset ratio range or abnormal offset ratio difference in each specific operation detection record, and select the minimum value η in each offset ratio range min and the maximum value η max Get an abnormal proportion range (η min ,η max) , and used as the selected abnormal features to perform self-check on the industrial computer one A trigger condition is selected; an abnormal offset ratio difference with the smallest value is selected as another trigger condition for the selected abnormal feature to perform self-check on the industrial computer.

5. The data management method of an industrial computer based on the Internet of Things according to claim 4 is characterized in that: The step S400 includes the following steps: Step S401: Obtain a real-time generated operation detection record and set it as a real-time detection record, obtain the operation data of the real-time detection record in each dimension, and calculate the abnormal value Y of the real-time detection record. now ; Step S402: Obtaining an abnormality threshold Y for determining whether the running detection record has an abnormality max , if Y now >Y max , then the running data of each dimension is compared with the corresponding normal value range to obtain several abnormal features; Step S403: arbitrarily select an abnormal feature, obtain the presentation of the abnormal feature-related operating data in the real-time monitoring record over time, and obtain the abnormal deviation proportion η of the abnormal feature at each moment now and the difference Δη between the abnormal deviation ratio now ; Set the two self-test trigger conditions of the selected abnormal features to be the abnormal proportion range (η min ,η max) and the difference Δη between the abnormal deviation ratio min , if η now ∈(η min ,η max) Or Δη now >η min , a self-check reminder is sent to the industrial computer; Step S404: after the industrial computer is self-checked, if there is no abnormality in the industrial computer, an abnormality reminder is sent to the real-time detection record, and an abnormality maintenance reminder is sent to the running equipment.

6. A data management system for an industrial computer, used to execute a data management method for an industrial computer based on the Internet of Things as claimed in any one of claims 1 to 5, characterized in that: The management system includes a detection data acquisition module, a detection anomaly analysis module, an anomaly self-detection identification module and a real-time anomaly judgment module; The detection data acquisition module is used to detect the operating status of the running equipment through the industrial computer, and regularly perform self-tests on the industrial computer, and respectively generate corresponding operation detection record sets and self-test record sets; perform abnormal situation analysis on the operating status of any operation detection record and any self-test record; The detection anomaly analysis module is used to analyze the abnormal changes between different operation detection records and extract the features that affect the operation status of the running equipment; Based on the abnormal conditions of any selected self-test records, identify the abnormal characteristics of specific operation test records; The abnormal self-check identification module is used to construct a feature function for any abnormal feature to present the change of the abnormal feature; analyze the deviation of all abnormal features in a specific operation detection record, establish a self-check trigger condition, and determine whether to perform a self-check when an abnormality occurs in the operation detection record; The real-time anomaly judgment module is used to identify anomalies of the real-time generated operation detection record whenever an operation detection record is generated in real time; in the case of anomalies, anomaly features are extracted from the real-time generated operation detection record, and it is determined whether to perform self-inspection on the industrial computer, confirm the equipment with abnormalities and issue an abnormal reminder.

7. The data management system for an industrial computer according to claim 6, characterized in that: The detection data acquisition module includes a detection record generation unit and an abnormal situation analysis unit; The detection record generation unit is used to detect the operating status of the operating equipment through the industrial computer, and regularly perform self-inspection on the industrial computer to generate corresponding operation detection record sets and self-inspection record sets respectively; the abnormal situation analysis unit is used to perform abnormal situation analysis on the operating status of any operation detection record and any self-inspection record.

8. The data management system for an industrial computer according to claim 6, characterized in that: The detection anomaly analysis module includes a device feature extraction unit and an abnormal feature recognition unit; The equipment feature extraction unit is used to analyze the abnormal changes between different operation detection records and extract the features that affect the operation status of the running equipment; the abnormal feature identification unit is used to identify the abnormal features of specific operation detection records based on the abnormal conditions of any selected self-test records.

9. The data management system for an industrial computer according to claim 6, characterized in that: The abnormal self-check identification module includes a characteristic function presentation unit and a self-check trigger setting unit; The feature function presentation unit is used to construct a feature function for any abnormal feature to present the change of the abnormal feature; the self-test trigger setting unit is used to analyze the deviation of all abnormal features in a specific operation detection record, establish a self-test trigger condition, and determine whether to perform a self-test when an abnormality occurs in the operation detection record.

10. The data management system for an industrial computer according to claim 6, characterized in that: The real-time abnormality judgment module includes an abnormality real-time identification unit and an abnormality judgment reminder unit; The real-time abnormality identification unit is used to identify the abnormality of the operation detection record generated in real time whenever an operation detection record is generated in real time; the abnormality judgment and reminder unit is used to extract abnormal features of the operation detection record generated in real time in response to the existence of abnormalities, determine whether to perform self-inspection on the industrial computer, confirm the equipment with abnormalities and provide abnormal reminders.

Images