A monitoring method for the operating state of a small package appearance detection device

Abstract

The present application relates to a kind of small package appearance detection equipment operating state monitoring method, by judging whether small package appearance detection equipment is normally started, system real-time acquisition small package appearance detection equipment's detection software's start state and rejection state, and the data information collected is transferred to cloud control platform, to judge whether the operation of small package appearance detection equipment is normal.The present application by the data of the packaging machine host auxiliary machine in the process of cigarette production, the heartbeat data of small package appearance detection equipment industrial computer, the start state and rejection state of detection software as variable, then using distributed computing method identifies whether small package appearance detection equipment is normally started.Then according to offline model, the fluctuation interval when the hardware resource of small package appearance detection equipment is normally operated is established.Real-time judgment production state, detection camera state, hardware resource use state, solve the problem that the process is complicated, workload is large and equipment cannot be found in time when failure occurs in the process of cigarette production.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of cigarette production monitoring, and particularly relates to a monitoring method for the running state of a small-pack appearance detection device. BACKGROUND

[0002] The small-pack appearance detection device is a device for detecting whether there is an obvious defect in the appearance of a small-pack product in the cigarette production process. In order to produce high-quality cigarettes, the appearance of each cigarette is detected in the cigarette production process. If the small-pack appearance detection device does not work normally, it may cause appearance quality accidents of the cigarette and reduce the core competitiveness of the enterprise.

[0003] In the actual production process, there are mainly three cases that cause the small-pack appearance detection device to work abnormally. One is that the small-pack appearance detection device does not work normally due to not being normally powered in the production process. Two is that the appearance defect detection function cannot normally play due to abnormal appearance detection software. Three is that the appearance detection function cannot normally play due to abnormal appearance detection software caused by abnormality of the hardware resources of the device.

[0004] At present, whether the small-pack appearance detection device works normally is mainly judged by special personnel who regularly conduct on-site inspection. Due to the large area and many devices in the production site, this manual inspection method has the problems of complicated inspection process, heavy workload and inability to timely find the faults of the devices, and quality accidents are prone to occur. SUMMARY

[0005] The present application aims to provide a monitoring method for the running state of a small-pack appearance detection device. The real-time signal data generated by the small-pack appearance detection device and the running data of the associated device are collected, a dynamic mathematical model is established, and distributed computing technology is used to identify whether the small-pack appearance detection device works normally, so as to solve the problems in the background technology.

[0006] To achieve the above-mentioned purpose, the present application is implemented by the following technical scheme:

[0007] A monitoring method for the running state of a small-pack appearance detection device adopts the following steps:

[0008] S1, judging whether the small-pack appearance detection device is normally started;

[0009] S2, if the small-pack appearance detection device is normally started, the system collects the start state and the rejection state of the detection software of the small-pack appearance detection device in real time, and transmits the collected data information to a cloud control platform to judge whether the small-pack appearance detection device works normally;

[0010] S3, in the step S2 judging the running normal condition of the small package appearance detection device, the system real-time collects the small package appearance detector detection quantity set [s0, s1, s2, …, s k ], the data after collection is transmitted to the cloud control platform through the network, its calculation formula calculates its difference:

[0011] s (i) = s (i) -s (i-1)

[0012] Wherein t takes the value [1, t], s (t) It is the difference value of each first setting time data, the cloud control platform calculates the sample data of each second setting time, through identifying each s (t) Greater than 0.

[0013] At the same time, the system real-time collects the network port acceptance rate of each detection camera of small package appearance detector eth0 [v e0 ,v e1 ,v e2 ,…,v ek ], eth1 [v e0 ,v e1 ,v e2 ,…,v ek ], eth2 [v e0 ,v e1 ,v e2 ,…,v ek ], eth3 [v e0 ,v e1 ,v e2 ,…,v ek ], the data after collection is transmitted to the cloud control platform using the calculation formula:

[0014]

[0015] Wherein Mark the network average acceptance rate of each camera network port, then use the formula:

[0016]

[0017] Through the formula, the s v Of the network port acceptance rate of each detection camera is obtained, the cloud control platform judges its s v Less than the setting value, then consider that in each second setting time sample data, the fluctuation of the rate of network port acceptance data of each detection camera is very small, then judge that each detection camera is in normal working state, at this time, the small package appearance detector is in normal running state.

[0018] Preferably, step S3 further comprises that the system collects the memory usage [u m1 m2 m3 mn ] and CPU usage [u c1 c2 c3 cn ] of the detection software of the small package appearance detector in real time, and establishes the confidence space of memory and CPU usage offline.

[0019] Preferably, the confidence space of memory and CPU usage is established offline using the following calculation formula:

[0020]

[0021]

[0022]

[0023]

[0024] wherein, represents the average usage rate of the device memory in normal operation, represents the average CPU usage rate of the device in normal operation, wherein σ m represents the standard deviation of the usage rate of the device memory in normal operation, σ c represents the standard deviation of the CPU usage rate of the device in normal operation;

[0025] Thus, the normal fluctuation range of memory usage is The normal fluctuation range of CPU usage is

[0026] Preferably, the collected data points of the set time are transmitted to the cloud control platform, and it is judged in real time whether the usage of each resource index of the small package appearance detector in operation is within the confidence space. If each detection camera is in a normal working state and the usage of each resource index is within the confidence space at the same time, it is judged that the small package appearance detection device is normal. If it cannot be satisfied at the same time, it is judged that the small package appearance detection device is abnormal.

[0027] Preferably, before step S1, it is judged whether the main machine and auxiliary machine of the roll package packaging machine are in a normal operation state, using the following steps:

[0028] S01, the system collects the auxiliary machine and main machine data of the roll package packaging machine;

[0029] ​​​​​​S02, the data collected in step S01 is transmitted to a cloud control platform through a network, the cloud control platform calculates the input data, and judges the running state of the main machine and the auxiliary machine of the packaging machine, and the calculation formula is:

[0030] Y m = vt m ;

[0031] Y s = vt s ;

[0032] Wherein v m is the running speed of the main machine of the packaging machine, v s is the running speed of the auxiliary machine of the packaging machine, Y m is the number of cigarette packs produced by the main machine of the packaging machine, Y s is the number of cigarette packs produced by the auxiliary machine of the packaging machine, simply referred to as yield, and t is the running time.

[0033] S03, the cloud control platform calculates the slope of the yield increase according to step S02 in real time, and judges the slope of the yield time window, if it meets the function monotone increasing characteristic, it is judged that the main machine and the auxiliary machine of the packaging machine are in normal running state.

[0034] Preferably, the data items collected in step S01 include running speed, yield and running time.

[0035] Preferably, in step S2, the system collects the heartbeat data of the small package appearance detection equipment industrial computer as [x 0, x1,x2,…,x t ], the cloud control platform calculates through the collected signal data, and the calculation formula is:

[0036]

[0037] Wherein y is the average value of the heartbeat state returned by the small package appearance industrial computer in t time, and the value is [0, 1], if y is equal to 0, it is considered that the small package appearance detection equipment is normally started.

[0038] Compared with the prior art, the beneficial effects of the present application are:

[0039] The application aims at the situation that the small package appearance detector in the cigarette production process cannot detect abnormity and find it in time, and proposes a monitoring method for the running state of the small package appearance detection equipment: taking the running rate, yield, running time of the main machine and auxiliary machine of the packaging machine in the cigarette production process, the heartbeat data of the small package appearance detection equipment industrial computer, the starting state and the rejection state of the detection software as variables, and then using the distributed computing method to identify whether the small package appearance detection equipment is normally started. Then, according to the offline model, the fluctuation interval of the normal running of the hardware resources of the small package appearance detection equipment is established. The yield state, detection camera state and hardware resource usage state are judged in real time. The problems that the inspection process is complicated and laborious in the cigarette production process and the equipment cannot be found in time when a fault occurs are solved. BRIEF DESCRIPTION OF DRAWINGS

[0040] Figure 1 It is a schematic diagram of the intelligent identification method for the running state of the small package appearance detection equipment.

[0041] Figure 2 It is a schematic diagram of another embodiment of the intelligent identification method for the running state of the small package appearance detection equipment.

[0042] Figure 3 It is a schematic diagram of the normal fluctuation range of the memory and CPU occupation established offline. DETAILED DESCRIPTION

[0043] The technical solutions of the application will be described in detail below with reference to the drawings. The following embodiments are only exemplary and can only be used to explain and illustrate the technical solutions of the application, but cannot be explained as a limitation of the technical solutions of the application.

[0044] The application relates to a monitoring method for the running state of a small package appearance detection equipment, which adopts the following steps:

[0045] The system collects the data of the main machine and auxiliary machine of the cigarette packaging machine, and the specific collected data items mainly include: running rate (packs / min), yield (packs), running time (minutes).

[0046] The system collects the signal data of the small package appearance detection equipment through the Ethernet and the small package appearance detection equipment industrial computer, and the data items mainly include: detection software state, detection rejection switch, heartbeat packet data, small package appearance detection quantity (packs), network rate (kbps) of each detection camera of the small package appearance detector, CPU usage rate (%), memory occupation (G).

[0047] The collection frequency of the above data is once every 2 seconds, and the big data stream computing technology can fully meet the requirements of the cigarette production scene.

[0048] The cloud inspection platform refers to using the distributed computing technology based on micro services to establish a distributed computing environment to process big data flow in real time or near real time, so as to obtain the required information and results close to real time.

[0049] Firstly, the system collects the signal data of the packaging machine equipment in real time through Ethernet (the specific data items are described above), and the collected data is transmitted to the cloud inspection platform through the network. The cloud inspection platform uses its distributed computing capability to process the input data on multiple server nodes of the cloud inspection platform. The calculation formula is:

[0050] Y m = vt m ;

[0051] Y s = vt s ;

[0052] Where v m is the running speed of the packaging machine host, v s is the running speed of the auxiliary machine of the packaging machine, Y m is the number of packs produced by the packaging machine host, Y s is the number of packs produced by the auxiliary machine of the packaging machine, and t is the running time. Through this calculation formula, it can be known that as the running time of the elevator increases, the number of products lifted increases. The cloud inspection platform calculates the slope of the increase in output in real time, and then judges the slope of the output time window. If it meets the characteristics of the monotonous increasing function, it is determined that the packaging machine host and auxiliary machine are in normal running state.

[0053] Therefore, when the packaging machine is working normally, the small pack appearance detection device as a product quality guarantee means must be in normal working state. This conclusion will be used as a sufficient and necessary condition for the next step of intelligently identifying whether the small pack appearance detection device should be started.

[0054] Secondly, the system collects the heartbeat data of the small pack appearance detection device industrial computer as [x 0, x1,x2,…,x t ] through Ethernet and the small pack appearance detection device industrial computer in real time. The collected data is transmitted to the cloud inspection platform through the network. The calculation formula is:

[0055]

[0056] Where is the average value of the heartbeat state returned by the small pack appearance industrial computer within t time, and the value is [0, 1]. The cloud inspection platform calculates the sample data transmitted every 10s, and identifies whether is equal to 0. If yes, it is considered that the small pack appearance detection device is started normally.

[0057] Thirdly, the system collects the start state and the rejection state of the small package appearance detection software in real time through the Ethernet and the small package appearance detection equipment, and transmits all the collected data to the cloud control platform to determine whether the operation is normal (the value range of the two data is shown in Table 1).

[0058] Table 1

[0059] Serial number Name Number sampling value format State value 1 Software state 1 Open as 1, off as 0 2 Culling state 1 Open as 1, off as 0

[0060] Fourthly, after the small package appearance detection equipment system is normally started, it also needs to identify whether its working state is normal in real time in the production process. The system collects the detection quantity set [s0, s1, s2, …, s k ] of the small package appearance detector in real time through the Ethernet and the small package appearance detection equipment. The collected data is transmitted to the cloud control platform through the network, and the difference is calculated by the following formula:

[0061] s (i) = s (i) -s (i-1) ,

[0062] wherein t is [1, t], s (t) is the difference value of every 2 seconds of data, the cloud control platform calculates the sample data of every 10 seconds, and identifies whether each s (t) is greater than 0.

[0063] At the same time, the system collects the network acceptance rate of each detection camera of the small package appearance detector in real time, eth0[v e0 , v e1 , v e2 , …, v ek ], eth1[v e0 , v e1 , v e2 , …, v ek ], eth2[v e0 , v e1 , v e2 , …, v ek ], and eth3[v e0 , v e1 , v e2 , …, v ek ]. The collected data is transmitted to the cloud control platform through the network, and the following formula is used:

[0064]

[0065] wherein indicates the network average acceptance rate of each camera network, and the following formula is used:

[0066]

[0067] The s of the network port receiving rate of each detection camera is calculated by formula v The cloud control platform judges whether s is small enough v If the s is small enough, it is considered that the rate of the network port receiving data of each detection camera in every 10s sample data fluctuates very little, and it is judged that each detection camera is in a normal working state. The small bag appearance detector is in a normal running state.

[0068] At the same time, under the normal running condition of the detection software, the resource utilization rate of the equipment should be in a stable and small fluctuation state for a long time. The system is connected with the small bag appearance detection equipment through Ethernet, and the memory occupation [u m1 , u m2 , u m3 , …, u mn ] and CPU utilization [u c1 , u c2 , u c3 , …, u cn ] of the detection software of the small bag appearance detector are collected in real time.

[0069] a Offline establishment of confidence interval of memory and CPU occupation:

[0070] Using formula:

[0071]

[0072]

[0073] Wherein represents the average utilization rate of the memory of the small bag appearance detection equipment when it is in normal operation, represents the average occupation rate of the small bag appearance detection equipment when it is in normal operation.

[0074] Using formula:

[0075]

[0076]

[0077] Wherein σ m represents the standard deviation of the utilization rate of the memory of the small bag appearance detection equipment when it is in normal operation, and σ c represents the standard deviation of the occupation rate of the small bag appearance detection equipment when it is in normal operation, so that the normal fluctuation range of the memory occupation is The normal fluctuation range of the CPU utilization rate is

[0078] b Confidence interval application:

[0079] The data points collected every 2s are transmitted to the cloud control platform. Real-time judgment is made on whether the occupancy of each resource index during operation is within the confidence interval range.

[0080] Table 2

[0081] Serial number Name Number value Unit Confidence interval 1 CPU occupancy 4.44 G [C tc ]]> 2 Memory occupancy 1.38 ％ C tm ]]>

[0082] If the above conditions are met at the same time, it is judged that the small package appearance detection equipment is working and functioning normally, and if the above conditions are not met at the same time, it is judged that the small package appearance detection equipment is functioning abnormally.

[0083] Specifically:

[0084] The running speed, output, running time and other data of the main machine and auxiliary machine of the packaging machine are collected. The collected data is collected at a frequency of 2 seconds per sample, and the data is transmitted to the cloud control platform through the network.

[0085] The cloud control platform uses the formula Y m =v m t; Y s =v s t; where v m is the running speed of the main machine of the packaging machine, v s is the running speed of the auxiliary machine of the packaging machine, Y m is the number of cigarette packs produced by the main machine of the packaging machine, and Y s is the number of cigarette packs produced by the auxiliary machine of the packaging machine, simply referred to as output, and t is the running time. The slope of the output data is calculated, and if the slope is greater than 0, it indicates that the packaging machine is producing normal cigarette products.

[0086] The heart beat data of the small package appearance detection equipment industrial computer is collected in real time as [x 0, x1,x2,…,x t ]. The collected data is transmitted to the cloud control platform through the network. The calculation formula is:

[0087]

[0088] Where is the average value of the heart beat state returned by the small package appearance industrial computer within t time, and the value is [0, 1]. The cloud control platform calculates the sample data transmitted every 10s, and identifies that is equal to 0, it is considered that the small package appearance detection equipment is started normally.

[0089] The start state and rejection state of the small package appearance detector detection software are collected in real time. The cloud control platform judges that the start state and rejection state of the small package appearance detector detection software are both open.

[0090] Real-time acquisition of small package appearance detector detection quantity set [s0, s1, s2, …, s k ], the collected data is transmitted to the cloud control platform through the network, and the difference is calculated by the formula:

[0091] s (i) = s (i) -s (i-1) ,

[0092] Where t takes the value [1, t], s (t) is the difference value of every 2 seconds of data, the cloud control platform calculates the sample data of every 10s, and identifies that s (t) is greater than 0.

[0093] From the above data, it is determined that the small package appearance detection device is normally enabled and is in a running state.

[0094] Real-time acquisition of small package appearance detector every detection camera network acceptance rate

[0095] eth0[v e0 ,v e1 ,v e2 ,…,v ek ], eth1[v e0 ,v e1 ,v e2 ,…,v ek ], eth2[v e0 ,v e1 ,v e2 ,…,v ek ], eth3[v e0 ,v e1 ,v e2 ,…,v ek ] The collected data is transmitted to the cloud control platform through the network, and the formula is:

[0096]

[0097] Where indicates the average network acceptance rate of each camera network card, and the formula is:

[0098]

[0099] The s v of the network acceptance rate of each detection camera is calculated by the formula, and the cloud control platform judges that s v is small enough, that is, the fluctuation of the network acceptance rate of each detection camera is small, and that each detection camera is in a normal working state.

[0100] The cigarette product production process 2 working days, a total of 57600 data, each historical data has memory occupation [u m1 ,u m2 ,u m3 ,…,u mn ] and CPU usage [u c1 ,u c2 ,u c3 ,…,u cn ]. The first 720 data as shown in Table 3:

[0101] Table 3 first 720 input data table

[0102]

[0103]

[0104]

[0105]

[0106]

[0107]

[0108]

[0109]

[0110]

[0111] As shown in Figure 3 , offline establish memory and CPU occupation confidence interval, using the formula:

[0112]

[0113]

[0114] Where represent the average use of the device in the normal operation of the memory, represent the average occupancy of the device in the normal operation.

[0115] Using the formula:

[0116]

[0117]

[0118] Where, σ m represent the use of the device in the normal operation of the memory standard deviation, σ cThe standard deviation of the occupancy rate of the representative device in the normal operation, so that the normal range of the memory occupancy fluctuation is The normal range of the CPU usage fluctuation is The normal range of the memory occupancy and the CPU usage fluctuation is calculated by substituting the data:

[0119] C tm =[4.449681-0.035898, 4.449681+0.035898],

[0120] C tc =[1.489458-0.723832, 1.489458+0.723832],

[0121] The memory and CPU occupancy rates of the small package appearance detector in the running state are collected in real time, and the cloud inspection and control platform judges whether the received real-time value is in the normal fluctuation range according to the received real-time value.

[0122] The cloud inspection and control platform strictly calculates and processes the received data according to the time sequence, and judges that the first-order difference value of each detection quantity of the small package appearance detector is greater than 0, the fluctuation of each detection camera is small, and the memory and CPU occupancy are in the confidence interval range. If the above conditions are met at the same time, it is considered that the small package appearance detector is in a normal running state, and if the above conditions cannot be met at the same time, it is considered that the small package appearance detector is in an abnormal running state.

[0123] Although the embodiments of the present application have been shown and described, it can be understood by those skilled in the art that various changes, modifications, replacements and variations can be made to the embodiments without departing from the principles and spirits of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A method for monitoring the operating status of a small package appearance inspection device, characterized in that, The following steps are adopted: S1. Determine if the small package appearance inspection equipment is started normally; S2. If the small package appearance inspection equipment starts normally, the system collects the start-up status and rejection status of the inspection software of the small package appearance inspection equipment in real time, and transmits the collected data information to the cloud inspection and control platform to determine whether the small package appearance inspection equipment is operating normally. S3. Under the condition that the small package appearance inspection equipment is operating normally as determined in step S2, the system collects the detection count set of the small package appearance detector in real time. The collected data is transmitted to a cloud-based monitoring platform via the network, and its difference is calculated using a specific formula: ， Where i takes the value [1, t]. The cloud-based monitoring platform calculates the difference between data points every 2 seconds, and then identifies each of the incoming sample data points every 10 seconds. Greater than 0; Simultaneously, the system collects the real-time network port receiving rate eth0 of each detection camera of the small packet appearance detector. ],eth1[ ],eth2[ ], eth3[ The collected data is transmitted to the cloud-based monitoring platform via the network and used for calculation using the following formula: , in Mark the average network receive speed of each camera's network port, then use the formula: ， The network port receiving speed of each detection camera is calculated using a formula. The cloud-based inspection and control platform judged its If the value is less than the set value, it is considered that the data rate fluctuation of each detection camera's network port in every 10 seconds of sample data is extremely small, and each detection camera is judged to be in normal working condition. At this time, the small packet appearance detector is in normal operating condition. Step S3 also includes the system collecting the memory usage of the small packet appearance detector's detection software in real time. ] and CPU utilization [, and establish a confidence space for memory and CPU usage offline.] 2. The method for monitoring the operating status of the small package appearance inspection equipment according to claim 1, characterized in that, Build a confidence space for memory and CPU usage offline using the following formula: ， ， ， ， in, This represents the average utilization rate of the device's memory during normal operation. This represents the average CPU utilization rate of the device during normal operation, where This represents the standard deviation of the device's memory utilization during normal operation. This represents the standard deviation of CPU utilization during normal operation of the device. Therefore, the normal range for fluctuations in memory usage is [ The normal range for CPU utilization fluctuations is []. ].

3. The method for monitoring the operating status of the small package appearance inspection equipment according to claim 1, characterized in that, The data points collected at the set time are transmitted to the cloud-based inspection and control platform, and the system judges in real time whether the occupation of each resource indicator is within the confidence space when the small package appearance detector is running. If both the normal working state of each inspection camera and the occupation of each resource indicator are within the confidence space are met, the small package appearance inspection equipment is judged to be functioning normally. If both conditions cannot be met, the small package appearance inspection equipment is judged to be functioning abnormally.

4. The method for monitoring the operating status of the small package appearance inspection equipment according to claim 1, characterized in that, Before step S1, it is necessary to determine whether the main and auxiliary machines of the roll packaging machine are in normal operating condition, by using the following steps: S01. The system collects data from the auxiliary and main units of the roll packaging machine; S02. The data collected in step S01 is transmitted to the cloud-based monitoring and control platform via the network. The cloud-based monitoring and control platform calculates the input data and determines the operating status of the main and auxiliary machines of the packaging machine. The calculation formula is as follows: ; ; in The operating speed of the packaging machine main unit, The operating speed of the packaging machine's auxiliary equipment. This represents the number of packs of cigarettes produced by the main packaging machine. The quantity of cigarette packs produced by the auxiliary equipment of the packaging machine, referred to as output, where t is the running time; S03. The cloud-based inspection and control platform calculates the slope of the output increase in real time according to step S02, and then judges the slope of the output time window. If it meets the monotonically increasing characteristic of the function, it is judged that the main machine and auxiliary machine of the packaging machine are in normal operation.

5. The method for monitoring the operating status of the small package appearance inspection equipment according to claim 4, characterized in that, The data items collected in step S01 include operating rate, output, and operating time.

6. The method for monitoring the operating status of the small package appearance inspection equipment according to claim 1, characterized in that, In step S2, the system collects the heartbeat data of the industrial control computer of the small package appearance inspection equipment in real time. The cloud-based monitoring platform performs calculations based on the collected signal data. The calculation formula is as follows: ， in The average value of the heartbeat status returned by the industrial control computer for the appearance of the small package within time t is taken as [0,1]. If the value is 0, the small package appearance inspection equipment is considered to have started normally.

Images