A rapid configuration method for alarm tasks based on intelligent monitoring of the Internet of Things

Abstract

The present invention belongs to the technical field of monitoring alarm, and specifically discloses a method for quickly configuring an alarm task based on intelligent monitoring of the Internet of Things. The method performs abnormal analysis of different alarm configuration information by retrieving historical alarm records of various monitoring objects in industrial production, thereby identifying the monitoring objects to be adjusted and their corresponding configuration information to be adjusted, so that the alarm configuration can be adjusted in a targeted manner, which is not limited to the adjustment of the alarm triggering threshold, greatly reduces the incidence of missed reports, and improves the accuracy of alarms. At the same time, by performing correlation analysis on the monitoring objects existing in industrial production with the help of historical alarm records, a related monitoring object group is screened out, and then alarm merging configuration is performed based on the related monitoring object group, so that repeated alarms generated by the related monitoring object group can be merged into one alarm information, which greatly reduces the alarm frequency rate, simplifies the alarm processing flow, and improves the alarm processing efficiency.

Description

Technical Field

[0001] The invention belongs to the technical field of monitoring alarms, and specifically discloses a method for quickly configuring alarm tasks based on intelligent monitoring of the Internet of Things. Background Art

[0002] With the continuous improvement of informatization, networking and automation, industrial production systems have become increasingly complex and large. In order to improve the stability of the system and ensure its safe operation, modern industrial production has widely adopted monitoring and alarm technology. Due to the high complexity and continuity of the industrial production environment, any failure may cause serious economic losses or even safety accidents. Therefore, through monitoring and alarm, production anomalies can be discovered in time, emergency plans can be quickly launched, and losses can be reduced. Therefore, monitoring and alarm have become an indispensable and important tool in industrial automation production.

[0003] In order to ensure the accuracy and effectiveness of the monitoring system, reduce false alarms and missed alarms, and improve the quality of alarm information, before using monitoring alarm technology to implement alarms, it is necessary to configure the alarms for the monitored objects, including but not limited to the monitoring period, alarm trigger threshold, alarm level, etc. However, the initial configuration may not fully capture all abnormal situations, so it is necessary to make quick adjustments after the initial configuration to ensure that the monitoring system always maintains the best state, reduce false alarms and missed alarms, and improve the stability and reliability of the system. However, in practice, the configuration adjustment of monitoring alarms often only focuses on the adjustment of effects such as alarm trigger thresholds and alarm levels, while the actual situation does not always require adjustment of thresholds. Ignoring other configuration items may lead to missed alarms, failure to discover real problems in a timely manner, and the risk of not meeting the actual configuration adjustment needs; in addition, it is easy to cause improper resource allocation, which may lead to resource waste or insufficient resources, affecting the overall effectiveness of the monitoring system.

[0004] In addition, the existing alarm configuration adjustment of the monitored object is generally a single adjustment, and there is a lack of attention to the correlation of the alarms of the monitored objects. This may cause another monitored object to have an abnormality when one monitored object is running abnormally, forming a chain reaction. If the alarm merging mechanism is not considered during the adjustment, it is easy to cause multiple monitored objects to generate alarms at the same time, causing a large amount of alarm information to occupy system resources and affect the normal operation of the monitoring system. Summary of the invention

[0005] To this end, one purpose of an embodiment of the present application is to provide a method for quickly configuring alarm tasks based on intelligent monitoring of the Internet of Things, which effectively solves the problems mentioned in the background technology by making targeted adjustments to the alarm configuration of the monitored objects in industrial production and adding alarm merging configuration with the help of analysis of the correlation of the monitored objects.

[0006] The purpose of the present invention can be achieved through the following technical solutions: a method for quickly configuring alarm tasks based on intelligent monitoring of the Internet of Things, comprising the following steps: statistics are taken on existing monitoring objects in industrial production, and alarm configuration information of the monitoring objects is extracted, wherein the alarm configuration information includes the monitoring terminal, the monitoring period, and the alarm trigger threshold.

[0007] Retrieve the alarm records of each monitoring object within the specified historical period, extract the alarm time and monitoring data, and evaluate the effectiveness of the alarm configuration information based on the alarm time and monitoring data, and then identify the monitoring objects that need to be adjusted.

[0008] Determine the configuration items that need to be adjusted for the monitored objects and adjust them.

[0009] Conduct correlation analysis on the monitoring objects in industrial production and identify the related monitoring object groups based on the analysis results.

[0010] Configure alarm merging based on associated monitoring object groups.

[0011] In combination with all the above-mentioned technical solutions, the positive effects of the present invention are as follows: 1. The present invention retrieves the historical alarm records of each monitoring object in industrial production to perform anomaly analysis of different alarm configuration information, thereby identifying the monitoring objects that need to be adjusted and their corresponding configuration items that need to be adjusted, so that the alarm configuration can be adjusted in a targeted manner, which is not limited to the adjustment of the alarm trigger threshold, greatly reduces the incidence of missed reports, improves alarm accuracy, and can also optimize resource allocation and enhance the overall efficiency of the monitoring system.

[0012] 2. The present invention performs correlation analysis on the monitoring objects existing in industrial production with the help of historical alarm records, thereby screening out related monitoring object groups, and then performing alarm merging configuration based on the related monitoring object groups. It can merge the repeated alarms generated by the related monitoring object groups into one alarm information, thereby greatly reducing the alarm frequency rate, simplifying the alarm processing process, and improving the alarm processing efficiency. At the same time, it also avoids a large amount of alarm information occupying system resources, which is conducive to continuously ensuring the normal operation of the monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The present invention is further described using the accompanying drawings, but the embodiments in the accompanying drawings do not constitute any limitation to the present invention. A person skilled in the art can obtain other drawings based on the following drawings without creative work.

[0014] Figure 1 The present invention is a flowchart of the steps for implementing the method.

[0015] Figure 2 The present invention is a schematic diagram of the alarm configuration adjustment operation of the monitoring object in industrial production.

[0016] Figure 3 It is a schematic diagram of the associated operation of monitoring objects in industrial production of the present invention. DETAILED DESCRIPTION

[0017] 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.

[0018] See also Figure 1 As shown, the present invention proposes a method for quickly configuring alarm tasks based on intelligent monitoring of the Internet of Things, comprising the following steps: performing statistics on existing monitoring objects in industrial production, and extracting alarm configuration information of the monitoring objects, wherein the alarm configuration information includes a monitoring terminal, a monitoring period, and an alarm trigger threshold.

[0019] Applied to the above implementation, industrial production can be an industrial production workshop, and the monitoring objects in the industrial production workshop can be the operating parameters of various production equipment, for example, the on / off status, operating temperature, operating speed, etc. of the machine tool. By monitoring these parameters, the status of the production equipment can be understood in real time.

[0020] It should be added that the alarm configuration information of the monitored object is saved in the configuration file after the initial configuration, and the alarm configuration information of the monitored object can be extracted from the configuration file.

[0021] It should be understood that the alarm configuration information of the monitored object actually covers the various alarm configuration contents of the monitored object, and these contents correspond to the operating parameters of the production equipment to be monitored. In an industrial production environment, the monitoring terminal is not only a camera, but also includes various types of sensors. In the example where the production equipment to be monitored is a machine tool, taking the machine tool as an example, the monitoring terminal can be a camera, where the operating parameters that the camera can monitor include the on / off status; the monitoring terminal can also be a non-camera monitoring terminal. For example, the non-camera monitoring terminal can be a temperature sensor for monitoring the temperature inside the machine tool or key components. In another example, the non-camera monitoring terminal can be a speed sensor for monitoring the speed of the machine tool spindle or moving parts. Therefore, according to the specific parameters to be monitored, the most suitable monitoring terminal can be selected, and through reasonable configuration of the monitoring terminal, the status of the production equipment can be fully and accurately monitored.

[0022] It should be noted that the monitoring period in the alarm configuration information refers to the working period of the monitoring terminal, which is specifically composed of a start monitoring time and an end monitoring time.

[0023] Once again, it is important to know that the alarm trigger threshold refers to the value or condition set for the monitored object. When the monitored object reaches or exceeds these predefined values ​​or conditions, an alarm will be triggered. Specifically, when the monitored object is the power on / off status of the machine tool, the power on / off status can be converted into the number of indicator lights that light up when the machine is turned on when configuring the alarm trigger threshold. This more intuitively reflects the status of the machine tool and makes it easier to set the alarm trigger threshold; when the monitored object is the operating temperature of the machine tool, the operating temperature threshold can be set when configuring the alarm trigger threshold.

[0024] Retrieve the alarm records of each monitoring object within the specified historical period, extract the alarm time and monitoring data, and evaluate the effectiveness of the alarm configuration information based on the alarm time and monitoring data, and then identify the monitoring objects that need to be adjusted.

[0025] It should be added that the specified historical period should be limited to the current time. This is to make the screened alarm records more timely and to exclude alarm records from earlier periods that may no longer be applicable to the current environment or configuration. In addition, the length of the specified historical period should not be too short or too long. This is because a period that is too short may not provide enough data samples, resulting in a lack of representativeness in the evaluation results, while a period that is too long will generate a large amount of data, increasing the difficulty and computing load of data processing. By way of example, the length of the specified historical period is 6 months.

[0026] In the specific implementation of the above scheme, the effectiveness of the alarm configuration information is evaluated as follows: retrieve the processing and troubleshooting records associated with the alarm records, compare the monitoring data of each monitored object in the alarm records with the troubleshooting data in the processing and troubleshooting records, and calculate the accuracy of the monitoring data of each monitored object in each alarm record.

[0027] It is important to know that after an alarm is triggered, the system will automatically generate an alarm record, which includes the alarm time, monitoring data, alarm level, etc. The monitoring data refers to the data of the corresponding production parameters of the monitored production equipment obtained during the monitoring process of the monitored object using the monitoring terminal; subsequently, processing and troubleshooting must be carried out to confirm the validity of the alarm and eliminate false alarms caused by system false alarms or other factors. Through careful investigation, the root cause of the alarm can be accurately identified, so that the problem can be quickly located, and effective corrective measures can be taken in time to prevent the problem from further deteriorating. After completing the processing and troubleshooting, in order to record the processing results and provide subsequent reference, the system will generate a detailed processing and troubleshooting record, which contains key information such as troubleshooting time, troubleshooting data, and processing measures. This process ensures that each alarm record can be matched with the corresponding processing and troubleshooting record, thus forming a complete closed-loop event management.

[0028] In the optimized implementation of the above scheme, the accuracy of the monitoring data of each monitored object in each alarm record is calculated by referring to the following process: the monitoring data of each monitored object in the alarm record is subtracted from the investigation data in the processing investigation record, and the absolute value is divided by the investigation data to obtain the relative difference of the monitoring data, and an exponential function is constructed with the reciprocal of the set constant as the base and the relative difference of the monitoring data as the exponent to obtain the accuracy of the monitoring data of each monitored object in each alarm record.

[0029] It should be added that the constant set above should be greater than 1. For example, the constant set may be 3.

[0030] The accuracy of the monitoring data of each monitored object in each alarm record is compared with the set accuracy threshold, and the alarm records that meet the accuracy threshold are screened out as available alarm records, where the available alarm records actually refer to the alarm records whose monitoring data when the alarm is triggered is consistent with the troubleshooting data in the processing and troubleshooting records. Such records reflect the accuracy and effectiveness of the alarm triggering, indicating that the alarm is caused by a real abnormal situation. Alarm records other than the available alarm records are recorded as data abnormality alarm records.

[0031] Divide the number of data anomaly alarm records of each monitored object by the total number of alarm records to obtain the proportion of data anomaly alarm records corresponding to each monitored object.

[0032] It is important to understand that the accuracy of the monitoring data in the alarm record is closely related to the monitoring terminal. When the alarm record frequently shows inaccurate monitoring data (i.e. the larger the proportion of data abnormality alarm records), this largely reflects that there may be an abnormality in the monitoring terminal.

[0033] Compare the alarm time of each monitored object in each available alarm record with the start and end monitoring time of the monitoring period in the alarm configuration information corresponding to the corresponding monitored object, and calculate the proximity of the alarm time of each monitored object in each available alarm record and the endpoint of the monitoring period. The specific calculation is as follows: Compare the alarm time of each monitored object in each available alarm record with the start and end monitoring time of the monitoring period in the alarm configuration information corresponding to the corresponding monitored object, and obtain the interval duration between the alarm time and the start monitoring time, and the interval duration between the alarm time and the end monitoring time of each monitored object in each available alarm record, and select the shortest interval duration as the alarm endpoint interval duration of the available alarm record.

[0034] In the example of the above solution, assuming that the alarm time of a certain monitored object in a certain available alarm record is 6:05, and the monitoring period of the monitored object is 6:00-18:00, then the alarm endpoint interval of the available alarm record is 5 minutes.

[0035] Compare the alarm endpoint interval of the available alarm record with the duration of the monitoring period, and use the expression Calculate the proximity between the alarm time of each monitored object in each available alarm record and the endpoint of the monitoring period, where Indicates the proximity of the alarm time in the available alarm records to the end point of the monitoring period. Indicates the alarm endpoint interval length of available alarm records. Indicates the duration of the monitoring period.

[0036] It can be seen from the above formula that the shorter the alarm endpoint interval of a certain available alarm record, the closer the alarm time of the available alarm record is to the endpoint time of the monitoring period, which may mean that the setting of the monitoring period is too narrow and does not cover the high-risk operation period of the production equipment, indicating that there is an abnormality in the configuration of the monitoring period.

[0037] The proximity between the alarm time and the monitoring period endpoint of each monitored object in each available alarm record is compared with the set proximity threshold. Exemplarily, the proximity threshold is 0.8, and the alarm records that reach the proximity threshold are selected as time anomaly alarm records.

[0038] The number of time anomaly alarm records of each monitored object is divided by the total number of available alarm records to obtain the proportion of time anomaly alarm records corresponding to each monitored object.

[0039] The monitoring data of each monitored object in each available alarm record is used to draw a histogram, and the number of available alarm records in each interval is extracted from the histogram.

[0040] The number of available alarm records in each interval is compared, the maximum and minimum values ​​are determined, and the ratio of the difference between the two to the maximum value is calculated to obtain the alarm distribution difference.

[0041] The number of available alarm records corresponding to each monitored object is summarized, and then the alarm frequency is calculated by combining it with the length of the historical period. The specific calculation process is: obtain the alarm time difference of each monitored object in the adjacent available alarm records, and calculate their average value to obtain the average alarm time difference.

[0042] The average alarm time difference of each monitored object is combined with the duration of the historical period into the calculation formula Get the alarm frequency of each monitored object, where Indicates the alarm frequency of the monitored object. Indicates the average alarm time difference, Indicates the length of the historical period.

[0043] It can be seen from the above formula that the smaller the average alarm time difference is, the greater the alarm frequency is.

[0044] It needs to be explained that the alarm distribution difference and alarm frequency can reflect the effectiveness of the alarm trigger threshold. When the alarm trigger threshold is set properly, it usually presents the characteristics of uniform alarm distribution and normal alarm frequency. Specifically, 1. The alarm distribution difference refers to the distribution of alarms in time. When the alarm distribution difference is low, it means that the alarm distribution in time is relatively uniform, that is, the alarms will not occur in a certain time period. The uniform alarm distribution indicates that the alarm trigger threshold is set reasonably, and abnormalities can be discovered in time without too many false alarms or missed alarms. 2. The alarm frequency reflects the frequency of alarm occurrence. When the alarm frequency is moderate, it means that the number of alarm triggers is neither too many nor too few, reflecting the normal state of the system, indicating that the alarm trigger threshold is set reasonably, which can detect abnormalities in time without too many false alarms.

[0045] It is further necessary to explain that when the alarm distribution difference is greater and the alarm frequency is greater, it can reflect the abnormality of the alarm trigger threshold configuration to a certain extent.

[0046] The proportion of data anomaly alarm records, the proportion of time anomaly alarm records, the alarm distribution difference and the alarm frequency corresponding to each monitored object are multiplied to obtain the comprehensive anomaly index corresponding to each monitored object.

[0047] The comprehensive anomaly index corresponding to each monitored object is compared with the configured allowable anomaly index. If the comprehensive anomaly index of a monitored object is less than or equal to the allowable anomaly index, the alarm configuration information corresponding to the monitored object is evaluated to be valid; otherwise, the alarm configuration information corresponding to the monitored object is evaluated to be invalid.

[0048] It should be pointed out that since the values ​​of the proportion of data abnormality alarm records, the proportion of time abnormality alarm records, the difference in alarm distribution and the alarm frequency are all between 0 and 1, the cumulative comprehensive abnormality index is also between 0 and 1. Only when the values ​​of the proportion of data abnormality alarm records, the proportion of time abnormality alarm records, the difference in alarm distribution and the alarm frequency are all smaller, the obtained comprehensive abnormality index will be smaller. This is because each smaller value will further reduce the result of the cumulative multiplication. By way of example, the set allowable abnormality index can be 0.001.

[0049] Further, the following operations are performed to identify the monitoring objects that need to be adjusted: the comprehensive abnormality index corresponding to each monitoring object is compared with the configured allowable abnormality index, and the monitoring objects with abnormality index greater than the allowable abnormality index are screened out as the monitoring objects that need to be adjusted.

[0050] Determine the configuration items that need to be adjusted for the monitored objects and make adjustments accordingly.

[0051] In the manner in which the above scheme can be implemented, the following process is used to determine the configuration items that need to be adjusted for the monitoring object that needs to be adjusted: sort the data abnormality alarm record ratio, time abnormality alarm record ratio, alarm distribution difference and alarm frequency of the monitoring object that needs to be adjusted in order from small to large, and then remove the smallest data from the sorting results in turn, recalculate the comprehensive abnormality index, and compare the recalculated comprehensive abnormality index with the allowable abnormality index. If the newly calculated comprehensive abnormality index is less than or equal to the allowable abnormality index, the configuration information corresponding to the removed data will not be used as the configuration item that needs to be adjusted for the control object that needs to be adjusted, and the configuration information corresponding to other data will be used as the configuration item that needs to be adjusted for the control object that needs to be adjusted; if the newly calculated comprehensive abnormality index is greater than the allowable abnormality index, continue to remove the next smallest data, and repeat the above steps until the last data is calculated.

[0052] In the example of the above implementation method, it is assumed that the data abnormal alarm record ratio, time abnormal alarm record ratio, alarm distribution difference and alarm frequency of the monitored object to be adjusted are 0.6, 0.2, 0.3 and 0.1 respectively. The cumulative multiplication of the comprehensive abnormal index at this time is 0.0036, which is greater than the allowable abnormal index of 0.001. At this time, the minimum data is 0.1, and then the cumulative multiplication result after removing the alarm frequency of 0.1 is 0.036, which is still greater than the allowable abnormal index of 0.001. After removing the next minimum data of the time abnormal alarm record ratio of 0.2, the cumulative multiplication result is 0.18, which is still greater than the allowable abnormal index of 0.001. After removing the next minimum data of the alarm distribution difference of 0.3, only the data abnormal alarm record ratio of 0.6 remains, which is still greater than the allowable abnormal index of 0.001. Since all the data cannot make the comprehensive abnormal index less than the allowable abnormal index after being removed, it means that the current configuration needs to be adjusted, that is, the monitoring terminal, monitoring period and alarm trigger threshold need to be adjusted.

[0053] In the specific adjustment, for the adjustment of the monitoring terminal, the monitoring terminal can be calibrated or replaced, for the adjustment of the monitoring period, the coverage of the monitoring period can be expanded, and for the adjustment of the alarm trigger threshold, the following method can be used: if the alarm frequency is large, the alarm trigger threshold can be appropriately increased to reduce false alarms. If the alarm distribution difference is large and the alarm frequency is not high, the threshold can be appropriately lowered to increase the sensitivity. More specifically, a dynamic threshold adjustment mechanism can be used to automatically adjust the threshold based on historical data. For example, a dynamic threshold can be calculated based on the statistical data of the past week, and the effect after adjustment can be observed. If the effect does not meet expectations, targeted adjustments can be made based on the original adjustment.

[0054] It should be added that the adjustment of the alarm trigger threshold is applicable to parameters such as the operating temperature and operating speed of the production equipment being monitored, but not to parameters such as the on / off status. This is because the operating temperature and operating speed are continuously changing parameters, and the changes are usually gradual. The alarm trigger threshold can be adjusted to capture anomalies more accurately, while the on / off status is a discrete parameter with only a few states: on, faulty, and off, and no gradual process. For example, the number of indicator lights is fixed when the machine is turned on and running normally, and the alarm trigger threshold cannot be adjusted by increasing or decreasing the number of indicator lights. Therefore, the alarm trigger threshold is usually not adjusted for parameters such as the on / off status, but the monitoring terminal and monitoring period can be adjusted.

[0055] The present invention provides a systematic way to identify and adjust configuration items. By gradually eliminating the minimum data and recalculating the comprehensive abnormality index, it is possible to accurately locate which configuration items have the greatest impact on the comprehensive abnormality index, and then quickly identify the configuration items that need to be adjusted, avoid blind adjustments, and improve the accuracy and efficiency of adjustments. At the same time, by determining the configuration items that need to be adjusted through a systematic data-driven method, it is possible to reduce misoperations caused by subjective misjudgment. This method makes decisions based on data and indicators, reduces the influence of human factors, and improves the objectivity and accuracy of decision-making.

[0056] For the above industrial production monitoring object alarm configuration adjustment, see Figure 2 shown.

[0057] Conduct correlation analysis on the monitoring objects existing in industrial production, and identify the related monitoring object groups based on the analysis results.

[0058] Optimally, the specific identification process of the associated monitoring object group is as follows: the monitoring objects existing in industrial production are randomly paired in pairs to obtain a number of monitoring object groups, so that each monitoring object can form a monitoring object group with other monitoring objects.

[0059] In the above optimization example, assuming that the monitoring objects in industrial production are A, B, C, D, E, F, G, and H, the monitoring object groups formed at this time can be (A, B), (B, C), (C, D), (D, E), (E, F), (F, G), (G, H), (A, C), (A, D)...

[0060] The available alarm records of the two monitoring objects in each monitoring object group are sorted in chronological order according to the alarm time to form respective available alarm record sets.

[0061] An element is extracted from each set in turn to form a control group of available alarm records.

[0062] In the example applied to the above implementation, taking the formed monitoring object group as (A, B) as an example, the available alarm record set formed by monitoring object A is , the available alarm record set formed by monitoring object B is At this time, the available alarm record control groups consisting of one element extracted from each of the two sets are a1 and b1, a1 and b2, a1 and b3, a1 and b4, a1 and b5, a2 and b1, a2 and b2...a6 and b5.

[0063] In another example, the available alarm record set formed by the monitoring object C in the monitoring object group (B, C) is , then the available alarm record control groups composed of the monitored object groups (B, C) are b1 and c1, b1 and c2, b1 and c3, b1 and c4, b1 and c5, b2 and c1, b2 and c2...b5 and c5.

[0064] In another example, the available alarm record control groups constituted by the monitoring object groups (A, C) are a1 and c1, a1 and c2, a1 and c3, a1 and c4, a1 and c5, a2 and c1, a2 and c2, ... a6 and c5.

[0065] For example, the available alarm record set formed by the monitoring object D in the monitoring object group (D, E) is , the available alarm record set formed by the monitoring object E is , the available alarm record control groups composed of the monitoring object groups (D, E) are d1 and e1, d1 and e2, d1 and e3, d1 and e4, d2 and e1, d2 and e2...d6 and e4.

[0066] The alarm times in each available alarm record control group constituted by each monitoring object group are identified to be similar, and the available alarm record control groups with similar alarm times are recorded as alarm similar control groups. Thus, the number of alarm similar control groups in each monitoring object group is divided by the number of available alarm record control groups to obtain the ratio value of alarm similar control groups, and compared with the set effective ratio value. Exemplarily, the effective ratio value is 0.8. If the ratio value of the alarm similar control group corresponding to a monitoring object group reaches the effective ratio value, the monitoring object group is recorded as the alarm similar monitoring object group.

[0067] In particular, the alarm times in each available alarm record control group consisting of each monitoring object group are identified to be similar as follows: two alarm times in the available alarm record control group are compared to obtain the alarm interval duration, and compared with the set similar alarm interval duration, wherein the similar alarm interval duration is initially set, exemplarily, the similar alarm interval duration is 2s, if the alarm interval duration of an available alarm record control group is less than or equal to the similar alarm interval duration, then the available alarm record control group is identified to have similar alarm times.

[0068] It should be understood that the more control groups with similar alarms in a certain monitoring object group, it means that when an alarm occurs in one monitoring object in the monitoring object group, another monitoring object also occurs successively, and there is a greater correlation.

[0069] The number of monitored object groups with similar alarms that have been screened out is counted, and the monitored objects in each monitored object group with similar alarms are compared. If it is found that two monitored object groups with similar alarms have the same monitored objects, the monitored objects in the two monitored object groups are deduplicated to form an associated monitored object group.

[0070] For the association of monitoring objects in industrial production mentioned above, see Figure 3 shown.

[0071] In the example of the above operation, assume that the screened monitoring object groups with similar alarms are (A, B), (A, C), (B, C), and (D, E). At this time, (A, B), (A, C), and (B, C) have the same monitoring objects, then the monitoring objects in these monitoring object groups are deduplicated to form an associated monitoring object group (A, B, C), and (D, E) does not have the same monitoring objects as other monitoring object groups with similar alarms, so the associated monitoring object group formed is (D, E).

[0072] The alarm merging configuration is performed based on the associated monitoring object group. The specific operation is as follows: the alarm similar control groups corresponding to the associated monitoring object group are retrieved to form an associated control group, and the alarm time of each monitoring object is extracted from it, and then the monitoring objects are sorted in the order of the alarm time.

[0073] In the example of the above operation, when the monitoring object groups (A, B, C) are associated, the alarm similar control groups retrieved are the alarm similar control groups corresponding to (A, B), (A, C), and (B, C), respectively. Specifically, the alarm similar control groups corresponding to (A, B) are a1 and b1, a1 and b2, a1 and b3; the alarm similar control groups corresponding to (A, C) are a1 and c4, a1 and c5, a2 and c1; the alarm similar control groups corresponding to (B, C) are b1 and c3, b3 and c2, b2 and c5. At this time, the associated control groups formed by the combination are a1 and b1, a1 and b2, a1 and b3, a1 and c4, a1 and c5, a2 and c1, b1 and c3, b3 and c2, b2 and c5.

[0074] In a further example of the above operation, the order of monitoring objects arranged in chronological order of alarm time extracted based on the associated control group is: assuming that the order of monitoring objects corresponding to a1 and b1 is A>B.

[0075] The order of monitored objects corresponding to a1 and b2 is A>B.

[0076] The order of monitored objects corresponding to a1 and b3 is A>B.

[0077] The order of monitoring objects corresponding to a1 and c4 is A <C。

[0078] The order of monitored objects corresponding to a1 and c5 is A>C.

[0079] The order of monitored objects corresponding to a2 and c1 is A>C.

[0080] The order of monitored objects corresponding to b1 and c3 is B>C.

[0081] The order of monitoring objects corresponding to b3 and c2 is B>C.

[0082] The order of monitored objects corresponding to b2 and c5 is B>C.

[0083] The monitoring object rankings obtained in each associated control group are compared, and repeated monitoring object rankings are extracted as associated alarm links corresponding to the associated monitoring object group.

[0084] In the above ordering, for (A, B), A>B appears most frequently, so A>B is the final order; for (A, C), A>C appears most frequently, so A>C is the final order; for (B, C), B>C appears most frequently, so B>C is the final order. On the whole, when the associated monitoring object group is (A, B, C), the monitoring object order is A>B>C, that is, monitoring object A alarms first, then the associated monitoring object B alarms, and then the associated monitoring object C alarms. The associated alarm chain formed by this monitoring object ordering is A→B→C.

[0085] It should be noted that when there is a sorting conflict in the associated sorting of the associated monitoring object group (A, B, C) (for example, A C exist simultaneously), the associated monitoring object group cannot form a consistent alarm link, and this associated monitoring object group needs to be excluded.

[0086] In the comprehensive example of the associated monitoring object group (D, E), the alarm-similar control groups extracted are d1 and e1, d1 and e2, and d2 and e1, all of which are used as associated control groups. The sorting of the monitoring objects arranged in the order of the alarm time sequence extracted from the associated control groups is that for the monitoring objects corresponding to d1 and e1, the sorting is D > E.

[0087] For the monitoring objects corresponding to d1 and e2, the sorting is D > E.

[0088] For the monitoring objects corresponding to d2 and e1, the sorting is D > E.

[0089] For (D, E), the number of times the sorting D > E appears repeatedly is the most. Therefore, when the associated monitoring object group is (D, E), the sorting of the monitoring objects is D > E, that is, the monitoring object D alarms first and then the associated monitoring object E alarms. The associated alarm link formed by this sorting of the monitoring objects is D → E.

[0090] Use the associated alarm link corresponding to the associated monitoring object group to define the alarm merging condition, and thus perform the alarm merging configuration.

[0091] In the example where the associated monitoring object group is (A, B, C), the alarm merging condition is as follows: When the monitoring data of the monitoring object A reaches the alarm trigger threshold, start the judgment of alarm merging, and identify whether the monitoring object B is about to alarm. Specifically, check whether the monitoring data of the monitoring object B is close to its alarm trigger threshold (about to alarm). If the monitoring object B is about to alarm, then further check whether the monitoring data of the monitoring object C is close to its alarm trigger threshold (about to alarm); if both the monitoring objects B and C are about to alarm, then merge the alarms of A, B, and C. If the monitoring object B is about to alarm and the monitoring object C is not about to alarm, then merge the alarms of the monitoring objects A and B.

[0092] When the present invention identifies the associated monitoring object group for the formed monitoring object group, it uses the available alarm records as the identification data because the available alarm records are screened based on real alarms, which can ensure the reliability of the identification of the associated monitoring object group.

[0093] The above contents are merely examples and explanations of the structure of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the structure of the invention or exceed the scope defined by the present invention, they should all fall within the protection scope of the present invention.

Claims

1. A method for rapid configuration of alarm tasks based on intelligent monitoring of the Internet of Things, characterized in that: The following steps are involved: In industrial production, statistics are collected on existing monitoring objects, and alarm configuration information of the monitoring objects is extracted, where the alarm configuration information includes monitoring terminals, monitoring time periods, and alarm triggering thresholds; Retrieve the alarm records of each monitoring object in the specified historical period, extract the alarm time and monitoring data, and evaluate the effectiveness of the alarm configuration information based on the alarm time and monitoring data, and then identify the monitoring objects that need to be adjusted; Determine the configuration items that need to be adjusted for the monitored objects and adjust them; Conduct correlation analysis on the monitoring objects in industrial production and identify the related monitoring object groups based on the analysis results; Configure alarm merging based on associated monitoring object groups; The validity of the alarm configuration information is evaluated as follows: Retrieve the processing and troubleshooting records associated with the alarm records, compare the monitoring data of each monitored object in the alarm records with the troubleshooting data in the processing and troubleshooting records, and calculate the accuracy of the monitoring data of each monitored object in each alarm record; Compare the monitoring data accuracy of each monitored object in each alarm record with the set accuracy threshold, select the alarm records that meet the accuracy threshold as available alarm records, and record the alarm records other than the available alarm records as data abnormality alarm records; Count the percentage of abnormal data alarm records corresponding to each monitoring object; Compare the alarm time of each monitored object in each available alarm record with the start and end monitoring time of the monitoring period in the corresponding alarm configuration information of the corresponding monitored object, and calculate the proximity between the alarm time of each monitored object in each available alarm record and the end point of the monitoring period; Compare the proximity between the alarm time and the endpoint of the monitoring period of each monitored object in each available alarm record with the set proximity threshold, and select the alarm records that reach the proximity threshold as time anomaly alarm records; Count the proportion of time anomaly alarm records corresponding to each monitoring object; Draw a histogram using the monitoring data of each monitored object in each available alarm record, and extract the number of available alarm records in each interval from the histogram; Compare the number of available alarm records in each interval, determine the maximum and minimum values, and calculate the ratio of the difference between the two to the maximum value to obtain the alarm distribution difference; Summarize the number of available alarm records corresponding to each monitoring object, and then calculate the alarm frequency by combining it with the length of the historical period; The proportion of data anomaly alarm records, the proportion of time anomaly alarm records, the alarm distribution difference and the alarm frequency corresponding to each monitoring object are multiplied to obtain the comprehensive anomaly index corresponding to each monitoring object; The comprehensive anomaly index corresponding to each monitored object is compared with the configured allowable anomaly index. If the comprehensive anomaly index of a monitored object is less than or equal to the allowable anomaly index, the alarm configuration information corresponding to the monitored object is evaluated to be valid; otherwise, the alarm configuration information corresponding to the monitored object is evaluated to be invalid.

2. A method for rapid configuration of alarm tasks based on intelligent monitoring of the Internet of Things as claimed in claim 1, characterized in that: The calculation of the monitoring data accuracy of each monitoring object in each alarm record refers to the following process: The relative difference of monitoring data is obtained by subtracting the absolute value of the monitoring data of each monitored object in the alarm record from the troubleshooting data in the processing and troubleshooting record and dividing the result by the troubleshooting data. An exponential function is constructed with the reciprocal of the set constant as the base and the relative difference of monitoring data as the exponent to obtain the accuracy of monitoring data of each monitored object in each alarm record.

3. A method for rapid configuration of alarm tasks based on intelligent monitoring of the Internet of Things as claimed in claim 1, characterized in that: The calculation of the proximity between the alarm time of each monitored object in each available alarm record and the endpoint of the monitoring period refers to the following process: Compare the alarm time of each monitored object in each available alarm record with the start and end monitoring time of the monitoring period in the corresponding alarm configuration information of the corresponding monitored object, obtain the interval length between the alarm time and the start monitoring time, and the interval length between the alarm time and the end monitoring time of each monitored object in each available alarm record, and select the shortest interval length as the alarm endpoint interval length of the available alarm record; Compare the alarm endpoint interval of the available alarm record with the duration of the monitoring period, and use the expression Calculate the proximity between the alarm time of each monitored object in each available alarm record and the endpoint of the monitoring period, where Indicates the proximity of the alarm time in the available alarm records to the end point of the monitoring period. Indicates the alarm endpoint interval length of available alarm records. Indicates the duration of the monitoring period.

4. A method for rapid configuration of alarm tasks based on intelligent monitoring of the Internet of Things as claimed in claim 1, characterized in that: The alarm frequency is calculated as follows: Obtain the alarm time difference of each monitored object in adjacent available alarm records, and calculate their average value to obtain the average alarm time difference; Import the average alarm time difference of each monitored object into the calculation formula combined with the duration of the specified historical period Get the alarm frequency of each monitored object, where Indicates the alarm frequency of the monitored object. Indicates the average alarm time difference, Indicates the duration of the specified historical period.

5. The method for quickly configuring alarm tasks based on intelligent monitoring of the Internet of Things according to claim 1, characterized in that: The identification of the monitoring object to be adjusted is performed as follows: The comprehensive abnormality index corresponding to each monitored object is compared with the configured allowable abnormality index, and the monitored objects with abnormality index greater than the allowable abnormality index are selected as the monitored objects that need to be adjusted.

6. A method for rapid configuration of alarm tasks based on intelligent monitoring of the Internet of Things as claimed in claim 1, characterized in that: The process of determining the configuration items that need to be adjusted for the monitored object is as follows: The proportion of data abnormal alarm records, the proportion of time abnormal alarm records, the difference in alarm distribution and the alarm frequency of the monitoring object to be adjusted are sorted in ascending order, and then the smallest data is eliminated from the sorting results in turn, the comprehensive abnormality index is recalculated, and the recalculated comprehensive abnormality index is compared with the allowable abnormality index. If the newly calculated comprehensive abnormality index is less than or equal to the allowable abnormality index, the configuration information corresponding to the eliminated data is not used as the configuration item to be adjusted for the control object to be adjusted, and the configuration information corresponding to other data is used as the configuration item to be adjusted for the control object to be adjusted; if the newly calculated comprehensive abnormality index is greater than the allowable abnormality index, the next smallest data is continued to be eliminated, and the above steps are repeated until the last data is calculated.

7. The method for quickly configuring alarm tasks based on intelligent monitoring of the Internet of Things according to claim 1, characterized in that: The associated monitoring object group is identified as follows: Randomly pair the monitoring objects in industrial production in pairs to obtain several monitoring object groups; The available alarm records of the two monitoring objects in each monitoring object group are sorted in chronological order according to the alarm time to form respective available alarm record sets; Extract one element from each set in turn to form a control group of available alarm records; The alarm times in each available alarm record control group formed by each monitoring object group are similarly identified, and the available alarm record control groups with similar alarm times are recorded as alarm similar control groups, thereby calculating the proportion value of the alarm similar control group and comparing it with the set effective proportion value. If the proportion value of the alarm similar control group corresponding to a certain monitoring object group reaches the effective proportion value, the monitoring object group is recorded as the alarm similar monitoring object group; The number of monitored object groups with similar alarms that have been screened out is counted, and the monitored objects in each monitored object group with similar alarms are compared. If it is found that two monitored object groups with similar alarms have the same monitored objects, the monitored objects in the two monitored object groups are deduplicated to form an associated monitored object group.

8. A method for rapid configuration of alarm tasks based on intelligent monitoring of the Internet of Things as claimed in claim 7, characterized in that: The following operations are performed to identify the alarm times in each available alarm record control group formed by each monitoring object group: The two alarm times in the available alarm record control group are compared to obtain the alarm interval duration, and compared with the set similar alarm interval duration. If the alarm interval duration of an available alarm record control group is less than or equal to the similar alarm interval duration, it is recognized that the alarm time of the available alarm record control group is similar.

9. A method for rapid configuration of alarm tasks based on intelligent monitoring of the Internet of Things as claimed in claim 7, characterized in that: The following process is used to configure the alarm merging based on the associated monitoring object group: Retrieve the alarm-similar control groups corresponding to the associated monitoring object groups to form an associated control group, and extract the alarm time of each monitoring object from it, and then obtain the monitoring object ranking according to the order of alarm time; Compare the monitoring object rankings obtained in each associated control group, and extract the repeated monitoring object rankings as the associated alarm links corresponding to the associated monitoring object groups; The alarm merging conditions are defined using the associated alarm links corresponding to the associated monitoring object groups, thereby performing alarm merging configuration.

Images