Method for setting a run threshold, monitoring method, monitoring system

Abstract

The present application relates to the technical field of device monitoring, in particular to a running threshold setting method, a monitoring method, a monitoring system and the like, the running threshold setting method of the device comprises the following steps: S1, obtaining running data of the device; S2, clustering the running data through a classification algorithm; S3, finding out a target threshold through a dichotomy method; according to the obtained target threshold, then executing step S4 or S5; wherein: S4, executing a first verification process of performing reliability verification of the original set threshold of the device with the obtained target threshold; if the verification is passed, setting the original set threshold of the device as the running threshold; S5, setting the obtained target threshold as the running threshold of the device; through the application of the running threshold setting method of the device, the set running threshold has the characteristic of high reliability, and the monitoring and verification of the monitoring system on the production state of the device are accurate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of device monitoring, in particular to a running threshold setting method and monitoring method, monitoring system. BACKGROUND

[0002] Currently, the threshold for determining whether the production line of a pollution discharge unit is in a production state mainly relies on the power provided by the pollution discharge unit during device production. Once the on-site monitoring power is greater than or equal to the set threshold, it is determined to be on, otherwise it is determined to be off. In fact, the pollution discharge unit generally provides the threshold based on the electrical nameplate, but there is often a gap in actual use, which affects the supervision results.

[0003] Secondly, the on-site working condition is complex, and the equipment of the pollution discharge unit may have multiple working conditions. For some standby and heat preservation devices, turning on the device is not in a production state, and it is difficult to determine whether the pollution discharge unit is really producing pollution. In addition, with the increase of use time, the equipment of the pollution discharge unit will age to some extent after a long time of work, and the energy consumption, power and other factors will increase, making it difficult to guarantee the continuous availability of the threshold.

[0004] Currently, whether the threshold provided for the pollution discharge unit conforms to the actual situation on site is mainly found by technical manual verification during project operation, which has high requirements for the professional level, data sensitivity and business of technical personnel. However, in the case of a large number of pollution discharge units and equipment, manual verification negligence is difficult to avoid.

[0005] The main difficulties encountered in the current production state determination threshold are:

[0006] (1) The collection of threshold values requires the cooperation of pollution discharge units to some extent. If the pollution discharge unit has the idea of avoiding monitoring, there may be false reporting of power threshold values, which may lead to situations where the pollution discharge unit is in a pollution production state but the online monitoring does not match the on-site situation, thereby affecting the control of the pollution discharge unit by the environmental protection department.

[0007] (2) The on-site production working condition is complex, and the production equipment is not necessarily in a pollution production state. The threshold values collected in the early stage are mainly the threshold values of the device opening state, and it is necessary to find a scientific threshold value for the production equipment pollution production state.

[0008] (3) The professional level of the technical personnel for threshold review has certain requirements, which requires professional knowledge, high data sensitivity, and familiarity with business. Therefore, a large amount of energy and resources need to be invested in professional and business training in the early stage.

[0009] (4) When the sample size is small, manual verification of the threshold may be very effective in a short period of time. However, with the full coverage of online monitoring and the increase in the number of sewage discharge units, a lot of manpower and time costs are required for secondary verification. However, manual verification is not foolproof. Once there is an omission or oversight, it will affect the regulatory authorities' judgment on the status of sewage discharge units.

[0010] (5) As time goes by, production equipment ages, energy consumption and power increase, and the continuous availability of thresholds is difficult to guarantee. If the historical experience value obtained deviates from the current actual situation, it will also affect the later supervision, and the labor cost will also increase. Summary of the Invention

[0011] In view of the current situation of full-process monitoring of sewage discharge equipment, it is necessary to provide a complete method for setting equipment operating thresholds, identify the boundary between the equipment's production status operating data and non-production status operating data, thereby clarifying its scientific and reasonable operating thresholds and accurately verifying them to ensure the accuracy of the equipment's operating threshold settings.

[0012] Therefore, the purpose of this invention is to provide a method for setting the operating threshold of a device, and at the same time, to provide a monitoring method and a monitoring system to meet the application requirements of the monitoring system for monitoring the status of working equipment, especially sewage discharge equipment.

[0013] The method for setting the operating threshold includes the following steps: S1, acquiring the operating data of the device; S2, clustering the operating data using a classification algorithm; S3, finding the target threshold using a cyclic binary search method; and then executing step S4 or S5 based on the obtained target threshold. Specifically: S4, executing a first verification process to check the reliability of the original device threshold using the obtained target threshold; if the verification passes, setting the device to the original threshold as the operating threshold; and S5, setting the device's operating threshold using the obtained target threshold.

[0014] In steps S2 to S3, the target threshold is calculated as follows: S2-1, cluster the obtained running data and denote the number of cluster centers as k; S2-2, determine the case of k; when k = 2, execute the following step: S2-2-1, denote i = 1, and the cluster center is A. i and A i+1 Then execute S2-3; S2-3, with A i and A i+1 Using the boundary as the boundary, we obtain set D. i S3-1. Use the bisection method on set D. i Perform equal-width partitioning, specifically by taking A. i and A i+1 The mean (A) i +A i+1 ) / 2 pairs of Di Equal-width partitioning is performed to obtain two sets, and the set intervals are [A i , (A i +A i+1 ) / 2), [(A i +A i+1 ) / 2, A i+1 ], respectively; the sizes of the data quantities of the two sets are compared to obtain a set D mi with a smaller data quantity; mi S3-2, determining whether the data quantity of D i is smaller than a preset data proportion of the data quantity of D mi ; when the data quantity of D i is smaller than the preset data proportion of the data quantity of D mi , performing S3-3; when the data quantity of D i is not smaller than the preset data proportion of the data quantity of D mi , outputting D i , setting D m i, and then returning to perform S3-1; S3-3, calculating the mean a mi of the maximum value and the minimum value in the set D i , and obtaining the target threshold value a i .

[0015] Further, in step S2-2, when k is greater than 2, the following steps are performed: S2-2-2, initializing i = 1; S2-2-3, selecting the i th and (i+1) th clustering centers A i and A i+1 ; and then performing S2-3; after step S3-3 is performed, when k is greater than 2, step S6 is performed; S6, determining whether i+1 is smaller than k; when i+1 is smaller than k, setting i = i+1, and then returning to perform step S2-2-3; when i+1 is not smaller than k, the target threshold value is obtained.

[0016] According to the setting of the running threshold value in the device, the device can be associated with the corresponding monitoring system; when the device is in a corresponding working state and generates specific running data, the corresponding running threshold value setting can effectively enable the monitoring system to confirm the current working state of the device, thereby effectively monitoring the specific working state of the device.

[0017] Further, in step S1, the obtained running data is cleaned.

[0018] Further, in step S4, if the first check procedure fails, a second check procedure for checking the reliability of the target threshold is performed; if the second check procedure passes, the target threshold is set as the running threshold of the device; if the second check procedure fails, the device is marked for processing.

[0019] Further, in step S5, a second check procedure for checking the reliability of the target threshold is performed; if the check passes, the target threshold is set as the running threshold of the device.

[0020] Further, the first check procedure comprises verifying the first deviation rate of the target threshold and the original threshold of the device; if the first deviation rate is less than the first preset deviation value, it is determined that the check passes; the setting range of the first preset deviation value is 8% to 12%.

[0021] Further, the first check procedure comprises: statistically comparing the analysis results of the target threshold on the running data and the analysis results of the original threshold of the device on the running data to obtain a third deviation rate; if the third deviation rate is less than a third preset deviation value, it is determined that the check passes; the setting range of the third preset deviation value is 8% to 12%.

[0022] Further, the second check procedure comprises verifying the second deviation rate of the target threshold and the real-time running data of the device; if the second deviation rate is less than a second preset deviation value, it is determined that the check passes; the setting range of the second preset deviation value is 8% to 12%.

[0023] Further, the second check procedure comprises: in a fixed working state of the device, obtaining real-time running data of the device in a unit time period, and statistically comparing the proportion of the real-time running data of the device greater than the target threshold in the unit time period to the real-time running data of the device in the unit time period; if the proportion is greater than a check proportion value, it is determined that the check passes; the setting range of the check proportion value is 85% to 95%.

[0024] Further, the running data comprises historical running power data or historical running current data or real-time running power data or real-time running current data of the device.

[0025] Further, when k is greater than 2, the number of target thresholds obtained is not less than two; there is a preset screening standard, and the target thresholds are screened according to the screening standard, and the screened target thresholds are used as the basis for performing step S4 or S5.

[0026] When the number of target thresholds obtained is not less than two:

[0027] First, the device has multiple working state conditions, and the device has corresponding multiple original set thresholds for each working state. In step S3, the obtained target threshold has a corresponding relationship with the original set threshold of the working state of the device. When step S4 is executed, one or more target thresholds are checked for reliability with the first check process of one or more original set thresholds of each working state. When step S5 is executed, the obtained one or more target thresholds replace the original set threshold of one or more corresponding working states as the running threshold of the device.

[0028] Second, the device has multiple working state conditions. In step S3, the obtained target threshold has a corresponding relationship with the working state condition of the device. According to the running data of the target working state, the running threshold selection range is preset. When step S4 is executed, the target threshold in the running threshold selection range is selected to perform the first check process of the reliability check of the original set threshold of the device. When step S5 is executed, the target threshold in the running threshold selection range is selected as the running threshold of the device.

[0029] Further, the preset data ratio setting range is 8% to 12%.

[0030] For the case where the corresponding check process in the above process does not pass, without further instructions, the actual application usually selects to mark the device, and the marked device is rechecked one by one in the subsequent process. The rechecking process can be processed by experienced engineers or calculated and verified by other system programs.

[0031] A device monitoring method applies the running threshold setting method described above to set the running threshold of the device, and a monitoring system monitors the operation of the device. The monitoring system compares the obtained running data with the running threshold of the device to make a judgment. When the monitored running data of the device is not less than the set running threshold, the judgment result is that the working state is opened. When the monitored running data of the device is less than the set running threshold, the judgment result is that the working state is closed.

[0032] A device monitoring system includes a plurality of devices and a monitoring device, and the monitoring device is in communication with each of the devices. The monitoring system applies the monitoring method described above to monitor the device. When the monitored running data of the device is not less than the set running threshold of the production working state, the judgment result is that the production working state is opened, and the device is in a production state. When the monitored running data of the device is less than the set running threshold of the production working state, the judgment result is that the production working state is closed, and the device is in a non-production state.

[0033] The beneficial effects of the present application are:

[0034] 1. The operation threshold setting method application of the device can make the set operation threshold have high reliability characteristics, make the monitored device have reliable monitoring, ensure that the monitoring system accurately verifies the monitoring of the production state of the device, and effectively save manpower and material resources without the need for monitoring auditors to periodically and quantitatively manually confirm and investigate a large number of devices one by one.

[0035] 2. Through the setting of different verification processes in multiple stages, the application can effectively ensure the reliability of the target threshold and the original set threshold as the operation threshold of the device, avoiding the deviation of the operation threshold setting of the device.

[0036] 3. The operation threshold setting method application can effectively distinguish and identify the working state of different devices in different processes and different industries in the monitoring process of the monitoring system, and make targeted monitoring and research and judgment operations, which helps the supervisor to accurately control the working state of the monitored device; especially it can solve the problem that the pollution of the pollution unit is difficult to confirm in the existing pollution supervision field. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 The real-time power trend chart for the paint spraying equipment in the automotive repair industry;

[0038] Figure 2 The real-time power scatter plot of the paint spraying equipment;

[0039] Figure 3 The violin plot of the clustering result of the power data of the paint spraying equipment;

[0040] Figure 4 The violin plot of the threshold division of the power data of the paint spraying equipment;

[0041] Figure 5 The result graph of the threshold division of the power data of the paint spraying equipment;

[0042] Figure 6 The flowchart of the target threshold calculation method of the embodiment 1 of the present application;

[0043] Figure 7 The verification flowchart of the first verification process of the embodiment 2 of the present application;

[0044] Figure 8 The verification flowchart of the second verification process of the embodiment 3 of the present application;

[0045] Figure 9 The verification flowchart of the second verification process of the embodiment 4 of the present application;

[0046] Figure 10The first check flow chart of the embodiment 5 of the present application;

[0047] Figure 11 The device real-time power state trend chart when the original set threshold is inappropriate;

[0048] Figure 12 The power situation schematic diagram of the injection molding equipment in different working states;

[0049] Figure 13 The flow chart of the multi-target threshold obtaining method of the embodiment 6 of the present application;

[0050] Figure 14 The injection molding equipment power data clustering result violin chart;

[0051] Figure 15 The injection molding equipment power data threshold division violin chart;

[0052] Figure 16 The injection molding equipment power data threshold division result chart;

[0053] Figure 17 The injection molding equipment current data clustering violin chart of the embodiment 7 of the present application;

[0054] Figure 18 The injection molding equipment current data threshold division violin chart of the embodiment 7 of the present application. DETAILED DESCRIPTION

[0055] In order to make the technical solutions, objectives and advantages of the present application clearer, the present application is further explained and described below in combination with the drawings and embodiments.

[0056] Embodiment 1:

[0057] The present application provides a device monitoring system, which comprises a plurality of devices and a monitoring device, the monitoring device is in communication connection with each of the devices, so that the monitoring device can acquire and monitor the data of each device; the monitoring system applies a device monitoring method, which sets the running threshold of the device, and compares the running data obtained by monitoring with the running threshold, so that the monitoring system can make a judgment on the working state of the device.

[0058] In the device monitoring method, the running threshold of the device is first set, and then the device with the set running threshold is monitored. The monitoring system compares the obtained running data with the running threshold of the device to make a judgment; when the running data of the device obtained by monitoring is not less than the set running threshold, the judgment result is that the working state is opened; when the running data of the device obtained by monitoring is less than the set running threshold, the judgment result is that the working state is closed.

[0059] Specifically, the equipment monitoring system is a monitoring system for sewage discharge equipment to monitor whether the sewage discharge equipment is actually in a sewage-generating state; when the monitored operating data of the sewage discharge equipment is not less than the operating threshold of the sewage-generating working state, it is judged that the sewage-generating working state is turned on, that is, the sewage discharge equipment is in a sewage-generating production state; when the monitored operating data of the equipment is less than the operating threshold of the sewage-generating working state, it is judged that the sewage-generating working state is turned off, and the sewage discharge equipment is in a non-production state.

[0060] Taking the power operation data of production equipment in the production line of a pollution discharge unit as an example to determine whether the working status is polluting, this invention provides a method for obtaining a target threshold, which can be used to obtain the power target threshold of the production equipment.

[0061] The specific application settings for determining this target threshold are as follows:

[0062] Step 1: Acquisition and cleaning of equipment operation data.

[0063] The system acquires equipment power operation data and corresponding working status data, forming a dataset for production equipment monitoring with a periodic sampling frequency of 5 minutes. After the power operation data recorded in these 5 minutes is input into the system, the system will automatically perform data cleaning processing to remove outliers, such as removing records with negative power values.

[0064] Step 2: Automatic clustering using classification algorithms.

[0065] The power distribution of production equipment in wastewater discharge units exhibits characteristics during both operation and non-operation: for example, during operation, power values ​​fluctuate within a certain range. Therefore, based on the data distribution, the data can be divided into intervals.

[0066] like Figures 1 to 3 As shown. Taking paint spraying equipment in the auto repair industry as an example, under normal circumstances, the original power threshold set on site is 0.2KW. Once the real-time power is not less than 0.2KW, the corresponding switching conversion is judged to be turned on. From the data, it is not difficult to find that its power is mostly distributed around (0,0.5) and (3,5).

[0067] In the diagram, a status switch of 1 indicates that the device is on, and 0 indicates that it is not on.

[0068] From a practical business perspective, the power consumption of this unit during non-production periods is affected by other lighting equipment, causing the power data to fluctuate around 0.4KW instead of being zero. This could lead to misjudgments of the status during non-production periods. In fact, the power consumption of this equipment during normal production fluctuates between (3,5). Therefore, we find a value between (0.4,3) as the boundary between production and non-production periods.

[0069] In the classification algorithm, by looking for high-density regions separated by low-density regions in the data set, the separated high-density regions are taken as an independent class. Assuming that there are many points around point x, denoted as x1, x2…x j , we calculate the movement of point x to each point x j around it. The sum of the required offsets is averaged to obtain the average offset. In addition to the size, the offset also contains the direction of the surrounding dense distribution. In the next step, point x moves in the direction of the average offset, and this is taken as the new starting point, which is iterated until a certain condition is met.

[0070] The general process is as follows:

[0071] 1) Randomly select a point in the data set as the starting center point S;

[0072] 2) Take S as the center and r as the radius, find all data points in the region, and consider these points to belong to the same cluster C. At the same time, record the number of data points appearing in the cluster plus 1;

[0073] 3) Take S as the center point, calculate the vector from S to each element in the data set, and add it to get the vector s;

[0074] 4) S=S+s, that is, S moves in the direction of s, and the moving distance is ||s||;

[0075] 5) Repeat steps 2), 3), and 4) until s is small, and record S at this time. At this time, all points encountered in this iteration process should be classified into cluster C;

[0076] 6) If the distance between the current cluster C and the center of the existing cluster C2 is less than the set threshold (the set threshold is determined by the data set, which is the 50% quantile of the distance set of any pair of samples in the data set), then C2 and C are merged, and the data point occurrence frequency is also merged. Otherwise, C is taken as a new cluster;

[0077] 7) Repeat steps 1) to 5) until all points are marked as visited;

[0078] 8) Classification: According to each class, the access frequency of each point, take the class with the maximum access frequency as the class to which the current point set belongs.

[0079] As Figure 3 shown, based on the application of paint spraying equipment in the automotive repair industry, it can be seen that the results of the pollution unit have two categories, and the first cluster center and the second cluster center are 0.05 and 3.94, respectively.

[0080] Step 3: Find the appropriate power target threshold by binary search.

[0081] As Figures 4 to 5 shown, based on the clustering center results obtained in the previous step, taking two clustering centers A1 and A2 as boundary conditions, all data in this interval are proposed as data set D, and bisection method is used to divide D to obtain two equal-width data D1 and D2. Take the preset data proportion value as 8% to 12%, preferably 10%; compare D1 and D2, and take the one with smaller data amount as D m , if the data amount of D m is less than (D1+D2)*10%, take the average of the maximum and minimum values of this data as the target threshold, if the data amount of D m is greater than (D1+D2)*10%, further divide D m by bisection method until a data set D m with data amount less than (D1+D2)*10% is found, and take the average of the maximum and minimum values of this data as the power target threshold.

[0082] The power data between the two clustering centers obtained by the classification algorithm of the spray painting equipment in the automotive repair industry is divided into two equal-width parts by bisection method, and the one with the smallest data amount is selected as D m At this time, the data amount of the one with the smallest data amount accounts for less than 10% of the total data amount of D1 and D2, and based on D m , the average a of the maximum and minimum values is obtained, and the average a is output as the power target threshold of the equipment, which is 2.23KW. Obviously, the numerical application of the obtained power target threshold 2.23KW can better divide the production and non-production situations than the original power original threshold 0.2KW of the original equipment.

[0083] The flow of the steps of the above-mentioned target threshold obtaining method is as follows:

[0084] S1, obtaining the running data of the equipment;

[0085] S2, clustering the running data by classification algorithm;

[0086] S3, finding the target threshold by bisection method.

[0087] According to the obtained target threshold, as a preferred target threshold application method, the target threshold can be selected as the running threshold setting of the equipment.

[0088] Based on the obtained power target threshold 2.23KW, the original threshold 0.2KW of the original equipment is replaced, and the power running threshold of the actual production working state of the equipment is changed; that is, when the real-time power of the spray painting equipment is not less than 2.23KW, the corresponding switch conversion amount is determined to be on, so that the monitoring system can accurately and effectively monitor the actual production working state of the equipment.

[0089] The target threshold value obtaining method can obtain the power target threshold value of the device from the actual production environment of the device based on the actual operation data of the device, and the obtained corresponding target threshold value can have more accurate characteristics relative to the original set threshold value of the device out of the factory. Figure 6

[0090] The operation threshold value setting method of the device can directly use the obtained target threshold value to replace the original threshold value of the device as the operation threshold value of the device, and the adjustment and setting of the operation threshold value can meet the accurate judgment demand of the monitoring system for the working state of the device.

[0091] In the operation data obtaining and calculation of the device, real-time operation power data of the device in real time or historical operation power data of the device in the past can be selected and applied according to actual needs.

[0092] Specifically, the specific obtaining process of the power target threshold value is as follows:

[0093] S2-1, clustering the obtained operation data, and recording the number of clustering centers as k;

[0094] S2-2, judging the condition of k; in the case of a single working state, two clustering centers will be generated;

[0095] When k=2, the following step S2-2-1 is performed: recording i=1, and the clustering centers are A i and A i+1 ; and then performing S2-3;

[0096] S2-3, taking A i and A i+1 as boundaries to obtain set D i ;

[0097] S3-1, dividing set D i by bisection method, and the specific operation is to take the mean value (A i +A i+1 ) / 2 of A i and A i+1 to equally divide D i to obtain two sets, and the set intervals are [A i , (A i +A i+1 ) / 2) and [(A i +A i+1 ) / 2, A i+1 ]; comparing the data amounts of the two sets to obtain the set D mi ​;

[0098] S3-2, Determine D mi Whether the amount of data is less than the preset data ratio of the Di data amount, wherein the preset data ratio is set in the range of 8% to 12%;

[0099] When D mi The amount of data is less than D i When the preset data ratio is used, execute S3-3;

[0100] When D mi The amount of data is not less than D i When the preset data ratio is used, the output D is... mi , making D i =D mi Then return to execute S3-1;

[0101] S3-3, Calculate D mi The average of the maximum and minimum values ​​in the set, a i The result is a i The target threshold.

[0102] Example 2:

[0103] Based on the target threshold determination method of Embodiment 1 above, another preferred application of this solution is as follows: Given that the target threshold obtained by the algorithm is reliable, the original threshold of the device is verified using a first verification process through the obtained target threshold, thereby determining whether the original threshold of the device is reliable. The application of this first verification process is described in reference to... Figure 7 As shown.

[0104] As mentioned above, the operating threshold of production equipment is not a fixed value but fluctuates within a range. Therefore, by finding a value within this range, we can effectively distinguish between the on / off states. If the original threshold is within the operating threshold range, and the target threshold obtained through the algorithm is also within this range, the deviation between the state determined by the original threshold and the state determined by the target threshold will be very small. However, if the original threshold is not within the threshold range, the state determined by the target threshold will deviate somewhat from the state determined by the original threshold. Therefore, if the numerical difference between the calculated target threshold and the original threshold of the equipment is within a reasonable range, the original threshold of the equipment can be considered reliable. This is the basic design principle of this scheme.

[0105] If the original threshold of the device is confirmed to be reliable, the original threshold will continue to be used as the operating threshold setting of the device, thereby ensuring accurate application of monitoring work while reducing the operation process.

[0106] The first verification procedure can optionally include verifying the first deviation rate of the obtained target threshold value from the original set threshold value of the device, and when the obtained first deviation rate is less than a first preset deviation value, determining that the verification is passed.

[0107] Taking the application of the paint spraying device in the automotive repair industry as an example, if the original set threshold value of the device is 2KW, and the first preset deviation value is 12%, the theoretical selection range of the target threshold value of the original set threshold value is between 1.76KW and 2.24KW, and thus when the obtained target threshold value is 2.23KW, it can be considered that the original set threshold value is reliable. Thus, the paint spraying device can continue to use 2KW as the running threshold value, and the monitoring device can continuously monitor the production working state at 2KW.

[0108] The setting step of the above-mentioned first verification procedure is applied as follows:

[0109] S4-1, comparing the obtained target threshold value with the original set threshold value of the device, and calculating a first deviation rate;

[0110] S4-2, determining whether the first deviation rate is less than a first preset deviation value;

[0111] When the first deviation rate is less than the first preset deviation value, the verification of the first verification procedure is passed.

[0112] Therefore, the running threshold value setting method of the device can optionally execute the first verification procedure to verify the reliability of the original set threshold value of the device with the obtained target threshold value, and if the verification is passed, the device continues to use the original set threshold value as its own running threshold value setting, thereby meeting the accurate judgment requirement of the monitoring system for the working state monitoring of the device.

[0113] Embodiment 3:

[0114] In order to avoid the obtained target threshold value from having a large deviation due to other conditions (such as the data packet loss condition of the running data acquisition), the second verification procedure can be used to verify the reliability of the obtained target threshold value, and when the obtained target threshold value passes the verification, the obtained target threshold value is used as the running threshold value setting of the device or as the reliability verification reference setting of the original set threshold value of the device. The application of the verification procedure of the second verification procedure is shown in the following table. Figure 8

[0115] The second verification procedure can optionally include verifying the second deviation rate of the obtained target threshold value from the real-time running data of the device, and when the obtained second deviation rate is less than a second preset deviation value, determining that the verification is passed. The value range of the second preset deviation value is 8-12%, and it is preferably set to 10%.​

[0116] Specifically, continue to take the application of the above-mentioned paint spraying equipment in the automotive repair industry as an example: the real-time running data can be selected as the average power data within the periodic sampling time (5 min) of the equipment in the target working state.

[0117] For example: if the target threshold value in the pollution-producing working state is required, the running data of the equipment within the periodic time period is collected in the clear pollution-producing working state (confirmed by manual confirmation or comprehensive analysis), and the data mean value is calculated. In the above-mentioned paint spraying equipment, if the real-time running data of the obtained mean value in the pollution-producing working state is 2.5 KW, the second preset deviation value is set to 12%, the current calculated target threshold value is 2.23 KW, the second deviation rate between the obtained target threshold value and the real-time running data of the equipment is less than the second preset deviation value, and the verification is passed.

[0118] Under the application premise of the above-mentioned embodiment 2, if the verification result is not passed when the first verification process is executed, there may be a possible situation that the target threshold value is calculated incorrectly or the deviation between the original threshold value and the calculated target threshold value is large (the target threshold value is reasonable). Therefore, the second verification process for reliability verification of the obtained target threshold value is executed subsequently. If the second verification process is passed, it is judged that the deviation between the original threshold value and the calculated target threshold value is large. The target threshold value is reasonable, and the obtained target threshold value can be used as the running threshold value of the equipment. If the second verification process is not passed, there may be a possible situation that the target threshold value is calculated incorrectly, and the equipment needs to be marked for subsequent inspection.

[0119] Embodiment 4:

[0120] Based on the application of the above-mentioned embodiment 3, the embodiment 3 is further preferably set in this embodiment.

[0121] The second verification process in the above-mentioned embodiment 3 can be changed to: in the fixed working state of the equipment, the real-time running data of the equipment within a unit time period is obtained, the proportion of the data greater than the obtained target threshold value in the real-time running data of the equipment within the unit time period is calculated, and when the proportion is greater than the verification proportion value, it is considered that the setting of the target threshold value is reliable, and the verification is passed. The verification process application of the second verification process is shown in the following Figure 9 .

[0122] For example:

[0123] In the case of a clear device is in a stable open working state (such as production state), in 24 hours as a unit of time period, the real-time power running data of the open working state of the device is compared with the target threshold value, if the power running data obtained by monitoring is more than 90% in 24 hours, it is considered that the verification is passed, and the target threshold value obtained is reliable.

[0124] Embodiment 5:

[0125] Based on the application of each embodiment described above, this embodiment further optimizes the setting mode.

[0126] In the application of the above embodiments, the corresponding verification can be based only on the data generated by the device itself; and in the application of the scheme of this embodiment, a simulated research and judgment action setting can be further combined, so that the corresponding verification process can have more accurate and reliable characteristics in the whole life cycle.

[0127] The first verification process for verifying the reliability of the original threshold value of the device based on the obtained target threshold value is: simulate the research and judgment results of the target threshold value on the running data and the research and judgment results of the original threshold value of the device on the running data, compare the research and judgment results of the target threshold value on the running data with the research and judgment results of the original threshold value of the device on the running data, and obtain a third deviation rate; when the obtained third deviation rate is less than a third preset deviation value, it is determined that the verification is passed. The value range of the third preset deviation value is 8-12%, and the preferred setting is 10%.

[0128] In actual application process, after obtaining the target threshold value, the corresponding research and judgment result situation made in the corresponding running data generation process is correspondingly counted according to the operation situation of the selected running data (or historical running data or real-time running data). For the working state situation (such as the above-mentioned paint spraying device of the automobile repair industry) that we need to research and judge, the open situation of the pollution working state obtained after setting the corresponding target threshold value in the data sampling period (5min) is simulated and counted, and the open situation of the pollution working state obtained by setting the original threshold value. When the deviation rate between the number of research and judgment results obtained based on the target threshold value setting and the number of research and judgment results obtained based on the original threshold value setting is less than a third preset deviation value, it is determined that the verification is passed, and the reliability of the original threshold value is clear. The verification process application situation of the first verification process is shown in Figure 10

[0129] ​For example, use the obtained target threshold value as the running threshold value, the monitoring system judges the opening and closing state of the equipment at each monitoring time point, and then compares it with the result judged by the original preset value. When there are 100 time points preset, the original preset value is judged to be on, but the result of the judgment by the obtained target threshold value as the running threshold value is that only 70 points are on, and the error rate reaches 30%. At this time, it has exceeded the preset deviation of 10%, so it is considered that the original set threshold value is unreliable, and the first verification process fails, and the second verification process needs to be processed or the equipment is marked for further confirmation by manual processing.

[0130] As a real application case:

[0131] Based on the environmental protection monitoring data of the pollution discharge unit, a total of 978 production lines and about 17.5 million data, the above algorithm is used to obtain the power running threshold value of the equipment, and the original set threshold value of the equipment is verified. The deviation is within 10%, a total of 719, about 73.6%, which shows that the current state judgment of most production line equipment is reasonable. The deviation of 259 is greater than 10%, which is determined as the first verification process of the original set threshold value reliability verification.

[0132] For the above first verification process, the second verification process can be used to verify the reliability of the target threshold value, or directly through the monitoring system or manual method for review. After review, 203 of the 259 production lines generally have the same power value, but at different times, they may be judged as different states, accounting for about 20.8% of the total; In this case, it is considered that the data generated by the equipment has quality problems, and it is not considered that the target threshold value obtained by the scheme is accurate.

[0133] In addition, there are 33 production lines with obvious power fluctuations, and the identified target threshold value is not 0 or even larger, and the actual state is always closed, accounting for about 3.4% of the total; It shows that the original set threshold value of the equipment is a problem, and in this case, the obtained power target threshold value needs to be changed to set the original preset running threshold value of the equipment.

[0134] As shown in Figure 11 After 8 o'clock, the power of the equipment fluctuates obviously, but the state is always closed, and the original set threshold value of the equipment on the spot is unreasonable, which may be set too high, resulting in failure to accurately identify the true state of the equipment. Overall, this method can verify the existing state to determine whether the original set threshold value of the equipment of the current pollution discharge unit is reasonable, and the overall misjudgment rate is 2.4%, which is relatively small and has reliability.

[0135]

[0136]

[0137] Embodiment 6:

[0138] In the above embodiments, the power target threshold of the equipment with single production condition (pollution production working state) is obtained by taking the working condition of the paint spraying equipment in the automotive repair industry as an example. In this embodiment, the power target threshold of the corresponding multi-condition equipment is further described.

[0139] As shown in Figure 12 , taking the injection molding industry as an example, the film blowing process has a heating condition, and the production equipment has two different working conditions, i.e., a heating working condition and a production working condition, in addition to the shutdown state. In the film blowing process, the heating working process will produce a certain power output, but since no actual production is carried out, no waste gas will be generated. Therefore, from the perspective of environmental protection monitoring, we need to accurately find out the production working condition of the film blowing process equipment for monitoring.

[0140] However, in actual application, the original threshold of the film blowing process equipment is generally set low (or only set to be greater than 1KW), which may cause the monitoring system to generate a linkage alarm in the preheating state, which does not meet the actual monitoring requirements.

[0141] Therefore, in view of the above application, the algorithm for obtaining the target threshold is further supplemented in this embodiment:

[0142] In the case of having two working conditions, the obtained running data will have three cluster centers; therefore, in step S2-2, when k is greater than 2, the following steps are executed: S2-2-2, initializing i = 1; S2-2-3, selecting the i, i+1th cluster centers A i and A i+1 ; and then executing S2-3;

[0143] After step S3-3 is executed, when k is greater than 2, step S6 is executed to determine whether i+1 is less than k;

[0144] When i+1 is less than k, i = i+1 is output to step S2-2-3, and the above operation steps are sequentially repeated, and a plurality of target thresholds a i are obtained until i+1 is not less than k.

[0145] When i+1 is not less than k, all target thresholds are obtained. At this time, the number of target thresholds obtained is greater than or equal to 2.

[0146] In this embodiment, two target thresholds will be generated based on the application of three cluster centers with two different working conditions.

[0147] In the above each target threshold a iIn the process of obtaining the target threshold, different first verification processes or second verification processes in the above embodiments can be selected and combined according to different verification requirements.

[0148] The flow setting of the multi-target threshold obtaining method is shown in Figure 13

[0149] As shown in Figures 14 to 16 Through the above device target threshold obtaining algorithm, it is found that the power of the injection molding production equipment fluctuates in three intervals, corresponding to the shutdown concentrated interval, the preheating concentrated interval, and the production concentrated interval. At this time, two target thresholds are obtained, which are 1.6 KW and 31.2 KW, respectively. The obtained target thresholds are not unique, and the obtained target thresholds have a corresponding relationship with the number and value selection range of the working state of the equipment.

[0150] The preset screening criteria are used to screen the obtained target thresholds, and the screened target thresholds are used as the execution basis for the device working state monitoring.

[0151] First, according to actual experience, the preheating state of the equipment is a low power output state, and the equipment pollution state is a high power output state. The power running data in the application of the two working states will have a significant difference. Therefore, the screening criteria are used to identify the obtained target thresholds with a significant difference in value through big data or artificial intelligence, and to divide and match the running thresholds of different working states.

[0152] Therefore, according to the screening criteria, the embodiment selects the low power output state value 1.6 KW as the running threshold of the equipment preheating state, and the high power output state value 31.15 KW as the running threshold of the equipment pollution state. Based on 31.15 KW as the actual pollution state power running threshold of the equipment, the first verification process and the verification running threshold setting application after verification are executed or the obtained target threshold is used as the running threshold setting application of the equipment. The subsequent actual monitoring situation obtained by the monitoring system will better meet the actual needs.

[0153] On the other hand, the above screening criteria can also preset the selection range of the running threshold according to the obtained running data.

[0154] ​For example, in the application of the film blowing process equipment, through artificial confirmation or comprehensive analysis of other analysis system programs, it can be known that there is a reasonable range of running data of different working states of the equipment (such as the reasonable range of power output data of the preheating state, and the reasonable range of power output data of the equipment in the pollution production state, which can be set as a 10% range of the upper and lower floating of the average of the corresponding working state running data); therefore, on this basis, the running threshold value selection range of the equipment in the preheating state is below the reasonable range of the power output data of the preheating state, and the running threshold value selection range of the equipment in the pollution production state is below the reasonable range of the power output data of the pollution production state. According to the monitoring target we need, which is the pollution production state of the equipment, among the multiple target threshold values obtained by the above algorithm, the target threshold value within the reasonable range of the power output data of the equipment in the pollution production state is selected as the running threshold value setting of the equipment in the pollution production state.

[0155] In the application of different environments, for the simultaneous monitoring of multiple working states, the equipment will have multiple original threshold values of corresponding working states, and the target threshold values obtained by the setting method have a one-to-one correspondence with the original threshold values of the working states in terms of quantity and value setting range.

[0156] Therefore, among the multiple target threshold values obtained by the above algorithm, each obtained target threshold value will have a matching relationship with the reasonable range of the power output data of each working state under the pre-set corresponding working state.

[0157] Correspondingly, in the running threshold value setting process, the first verification process of verifying the reliability of the original threshold values of different working states of the equipment and / or the second verification process of verifying the reliability of each target threshold value can be selected to be executed, and the target threshold values that meet the verification passing conditions are set as the running threshold values of each working state of the equipment, or the original threshold values of each working state that pass the verification are set as the running threshold values of the equipment, so that the obtained target threshold values are effectively matched with the corresponding working state, and the monitoring device of the monitoring system has different monitoring alarm measures for different running threshold value standards, which can meet the monitoring application needs of the monitoring system for multiple working states of the equipment.

[0158] Embodiment 7:

[0159] In the description of the above embodiments, the running threshold value setting and monitoring application of the equipment in the power running data are described. In this embodiment, the running threshold value setting and monitoring application of the corresponding current running data are further described. For example, Figures 17 to 18As shown, by applying the above-mentioned target threshold value calculation algorithm to the corresponding historical current data of the paint spraying device, and performing algorithm operation, it can be concluded that when the production device performs pollution production in the production concentration interval, the current operation threshold value for dividing pollution is the calculated current target threshold value 54.61 A, which is more in line with the actual situation. In the algorithm operation process of the embodiment, the selected operation data includes historical operation current data or real-time operation current data of the device.

[0160] Embodiment 8:

[0161] Based on the application of the target threshold value calculation algorithm as described in the above embodiments, those skilled in the art can understand that in the field of approximate application, such as monitoring whether there is a violation of the use of electrical appliances in school dormitories, the device monitoring system of the present application can also accurately monitor and analyze the use behavior of the corresponding violation of electrical appliances, meeting the needs of other device monitoring applications in addition to the field of pollution monitoring.

[0162] The above is only a preferred embodiment of the present application. For those skilled in the art, the embodiments can still be modified without departing from the implementation principle of the present application, and the corresponding modification scheme should also be considered as the protection scope of the present application.

Claims

1. A method of operating a threshold setting, characterized by, The method comprises the following steps: S1, obtaining operation data of the equipment; S2, clustering the operation data by a classification algorithm; S3, finding a target threshold value by dichotomy; In steps S2 to S3, the target threshold value is calculated as follows: S2-1, clustering the obtained operation data, and recording the number of clustering centers as k; S2-2, judging the value of k; When k = 2, perform steps: S2-2-1, let i = 1, cluster center is A i and A i+1 ; then perform S2-3; When k is greater than 2, the following steps are executed: S2-2-2, initializing i = 1; S2-2-3, selecting the i th and i + 1 th clustering centers A i and A i + 1 ; and then executing S2-3; S2-3, with A i and A i+1 as boundaries, resulting in the set D i ; S3-1, divide the set D by half i Perform equal-width division to obtain two sets; compare the data sizes of the two sets to obtain the set D mi with the smaller data size S3-2, judging D mi whether the data amount is less than D i a preset data proportion of the data amount When the data amount of D mi is less than the preset data ratio of the data amount of D i , S3-3 is executed; When D mi is not less than D i , output D mi , let D i = D mi , and then return to execute S3-1; S3-3, compute D mi a, the average of the maximum and minimum values in the set i , the resulting a i is the target threshold value; After step S3-3 is executed, step S6 is executed; S6, judging whether i + 1 is less than k; When i + 1 is less than k, i = i + 1 is set, and then step S2-2-3 is executed again; When i + 1 is not less than k, the target threshold value is obtained; According to the obtained target threshold value, step S4 or S5 is executed; S4, executing a first verification process of verifying the reliability of the original threshold value of the equipment by using the obtained target threshold value; The first verification process comprises verifying a first deviation rate of the obtained target threshold value and the original threshold value of the equipment, and when the obtained first deviation rate is less than a first preset deviation value, it is determined that the verification is passed; Or, the first verification process comprises statistically comparing a judgment result of the operation data generated by the obtained target threshold value with a judgment result of the operation data generated by the original threshold value of the equipment, obtaining a third deviation rate, and when the obtained third deviation rate is less than a third preset deviation value, it is determined that the verification is passed; If the verification is passed, the original threshold value is set as the operation threshold value of the equipment; if the verification of the first verification process is not passed, a second verification process of verifying the reliability of the obtained target threshold value is executed; if the verification of the second verification process is passed, the obtained target threshold value is set as the operation threshold value of the equipment; S5, executing a second verification process of verifying the reliability of the obtained target threshold value; if the verification is passed, the obtained target threshold value is set as the operation threshold value of the equipment; The second verification process comprises: verifying a second deviation rate of the obtained target threshold value and real-time operation data of the equipment, and when the obtained second deviation rate is less than a second preset deviation value, it is determined that the verification is passed; Or, in a fixed working state of the equipment, real-time operation data of the equipment in a unit time period is obtained, and a proportion of data greater than the obtained target threshold value in the real-time operation data of the equipment in the unit time period is calculated; when the proportion is greater than a verification proportion value, it is determined that the verification is passed.

2. The run threshold setting method of claim 1, wherein, In step S4, if the verification of the second verification process is not passed, the equipment is marked.

3. The run threshold setting method of claim 1, wherein, The first preset deviation value is set in a range of 8% to 12%.

4. The run threshold setting method of claim 1, wherein, The third preset deviation value is set in a range of 8% to 12%.

5. The run threshold setting method of claim 1, wherein, The second preset deviation value in the second verification process is set in a range of 8% to 12%.

6. The run threshold setting method of claim 1, wherein, The verification proportion value in the second verification process is set in a range of 85% to 95%.

7. The run threshold setting method of claim 1, wherein, The operation data comprises historical operation power data or historical operation current data or real-time operation power data or real-time operation current data of the equipment.

8. The run threshold setting method of claim 1, wherein, The device has multiple working states, and the device has multiple original threshold values corresponding to each working state; in step S3, the obtained target threshold value has a corresponding relationship with the original threshold value of each working state of the device; when step S4 is executed, one or more target threshold values are subjected to a first verification process for reliability verification with one or more original threshold values corresponding to each working state. When step S5 is executed, the obtained one or more target threshold values replace the original threshold values corresponding to each working state as the running threshold value of the device.

9. The run threshold setting method of claim 1, wherein, The device has multiple working states; in step S3, the obtained target threshold value has a corresponding relationship with each working state of the device; according to the running data of the target working state, a running threshold value selection range is preset, and when step S4 is executed, the target threshold value in the running threshold value selection range is selected for the first verification process for reliability verification of the original threshold value of the device. When step S5 is executed, the target threshold value in the running threshold value selection range is selected as the running threshold value of the device.

10. The run threshold setting method of claim 1, wherein, When k is greater than 2, the number of obtained target threshold values is not less than two; a screening standard is preset, and the obtained target threshold values are screened according to the screening standard, and the screened target threshold values are used as the basis for executing steps S4 or S5.

11. The run threshold setting method of claim 1, wherein, The preset data ratio is set in the range of 8% to 12%.

12. A monitoring method, characterized by, The running threshold value setting method of any one of claims 1 to 11 is applied to the device to set the running threshold value of the device, and a monitoring system is used to monitor the device; the monitoring system compares the obtained running data with the running threshold value of the device to make a judgment; when the monitored running data of the device is not less than the set running threshold value, the judgment result is that the working state is opened; when the monitored running data of the device is less than the set running threshold value, the judgment result is that the working state is closed.

13. A monitoring system, characterized by The monitoring method of claim 12 is applied to the device by the monitoring system; when the monitored running data of the device is not less than the set running threshold value of the production working state, the judgment result is that the production working state is opened, and the device is in a production state; when the monitored running data of the device is less than the set running threshold value of the production working state, the judgment result is that the production working state is closed, and the device is in a non-production state.

Images