CN110081851A - A kind of Monitoring Pinpelines and method for early warning based on association analysis - Google Patents
A kind of Monitoring Pinpelines and method for early warning based on association analysis Download PDFInfo
- Publication number
- CN110081851A CN110081851A CN201910201840.7A CN201910201840A CN110081851A CN 110081851 A CN110081851 A CN 110081851A CN 201910201840 A CN201910201840 A CN 201910201840A CN 110081851 A CN110081851 A CN 110081851A
- Authority
- CN
- China
- Prior art keywords
- sensor
- time window
- group
- pipeline
- degree
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B21/00—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant
- G01B21/32—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant for measuring the deformation in a solid
Abstract
The present invention relates to a kind of Monitoring Pinpelines and method for early warning based on association analysis.This method comprises: step 1, monitors pipeline by sensor unit and obtains pipeline strain data；Step 2, the pipeline strain data in original time window are obtained, and solve the expectation E and variances sigma of data correlation characteristic and associate feature distribution function by the pipeline strain data in original time window2；Step 3, the support and availability that monitor value K- closes on are calculated, to judge sensor reliability and pipe damage situation.The present invention analyzes and extracts characteristic information using the pipe stress data correlation of monitoring in the case where not needing to establish system analysis model case, realize to the monitoring and early warning of pipe stress state and the evaluation of sensor reliability.
Description
Technical field
The invention belongs to process industries to be safely operated safeguards technique field, be specifically related to a kind of pipe based on association analysis
Road monitoring and method for early warning.
Background technique
Pressure pipeline is the important component in process industry, and safety operation level directly determines petroleum chemical enterprise
Safety operation level.With the expansion of petroleum chemical enterprise's scale, the limitation of conventional detection means is also increasingly apparent, such as: it is daily
It is too many to consume manpower, specific position has certain risk, is easy the presence of detection dead angle.It can only be based on for equipment safety evaluation
Test data under shutdown status can not reflect the truth under service state, can not carry out to in-service device structure state
Real-time detection, whether safety can only be estimated according to testing result before during detecting twice, but this be estimated usually due to pipe
The influence of many uncertain factors in line running environment and there is very big error.
In order to ensure its safety operation level, domestic and international researcher proposes to carry out real-time monitoring to pipeline strain state,
To grasp pipeline configuration level of security.It is based primarily upon analytical model algorithm about the thinking of Monitoring Pinpelines at present, that is, solves pipeline
Strain value under typical condition assesses pipeline strain state as monitoring threshold.This method is in complicated pipeline system
There are limitations for system long period monitoring aspect: 1) complicated pipe-line system is difficult to be described with analytic modell analytical model, the monitoring threshold of solution with
Actual conditions differ greatly；2) long period lower sensor performance will be deteriorated unavoidably, cause monitor value unreliable, no
It can reflect pipeline strain state.
Summary of the invention
In order to solve the above technical problem, the present invention provides a kind of Monitoring Pinpelines and method for early warning based on association analysis.
In order to achieve the object of the present invention, the invention adopts the following technical scheme:
A kind of Monitoring Pinpelines and method for early warning based on association analysis, comprising the following steps:
Step 1, pipeline is monitored by sensor unit and obtains pipeline strain data, the arrangement of the sensor unit
It is divided into according to conduit types following several:
1st kind, for straight tube: the multiple sensors of the same circumferentially spaced-apart setting of the straight tube simultaneously constitute sensor group, described
Sensor group is arranged multiple along the length of straight pipe direction；
2nd kind, for elbow: the multiple sensors of the same circumferentially spaced-apart setting of the elbow simultaneously constitute sensor group, described
Sensor group is arranged multiple along the elbow length direction；
3rd kind, for threeway: the threeway is formed by being responsible for branch pipe welding, and the supervisor is equipped with along weld interval
Multiple sensors for being arranged simultaneously constitute sensor group, and the branch pipe is equipped with the multiple sensors being arranged along weld interval and constitutes
Sensor group；
Step 2, the pipeline strain data in original time window are obtained, and pass through the pipeline strain data in original time window
Solve the expectation E and variances sigma of data correlation characteristic and associate feature distribution function2, the original time window is current time
L time window before window；
Step 2-1 uses K- nearest neighbor algorithm to find out vector distance to describe associate feature between pipeline strain data:
Assuming that SxIt for any one group of sensor group and include n sensor, in each time window T, each sensor is adopted
Integrate interval time as Δ t, then each sensor acquires m data in each time window T, wherein m=T/ Δ t；At l
Between in window, each sensor acquires N=ml data altogether；U, v is two sensors in any one group of sensor group, then
In i-th of time window, the strain data vector set difference of the sensor u, sensor v acquisition is as follows:
εu,i={ εu,m×(i-1)+1,εu,m×(i-1)+2,…,εu,m×(i-1)+m} (1)
εv,i={ εv,m×(i-1)+1,εv,m×(i-1)+2,…,εv,m×(i-1)+m} (2)
In i-th of time window, the Euclidean distance between described sensor u, v is indicated are as follows:
In l time window, average Euclidean distance is indicated between sensor u and sensor v are as follows:
Found out in same sensor group with the smallest K sensor of the average Euclidean distance of sensor u, that is, think
It is that K=2-4 is typically chosen according to pipeline internal pressure situation with sensor u associate feature highest K sensor；
Step 2-2 closes on vector distance set based on K- and establishes associate feature number distribution function, and solves parameter:
In the case where pipeline does not occur degree of impairment, for l time window, the sensor u and the highest sensing of associate feature
Set d (the ε of Euclidean distance between device vu,εv) it is expressed as follows:
d(εu,εv)={ d (εu,1,εv,1),d(εu,2,εv,2),...,d(εu,l,εv,l)} (5)
Assuming that when time window quantity l is sufficiently large, d (εu,εv) meet normal distribution, and think d (εu,εv) expectation such as
Under:
D (the εu,εv) variance it is as follows:
σ2=s2(d(εu,εv)) (7)
In formula,
Step 3, the support and availability that monitor value K- closes on are calculated, to judge sensor reliability and pipe damage
Situation:
Current time window is i.e. in the l+1 time window inner sensor u and the highest sensor v of any one associate featurej
The Euclidean distance of the strain data vector of (1≤j≤K) acquisition is expressed as d (εu,l+1,εv,l+1)j, with season:
Z=| d (εu,l+1,εv,l+1)j-E(d(εu,εv))| (8)
Think | z | when≤3 σ, monitoring point u validity is by the highest sensor v of associate featurejSupport；
In the l+1 time window, sensor u is by the highest sensor v of associate featurejSupport indicate are as follows:
In formulaρ in formulaAFor ρujConfidence threshold, set ρA=0.7-0.9；
In the l+1 time window, the availability γ of sensor uuIt is the highest sensor of K associate feature to sensor
The weighted sum of the support of u, it may be assumed that
ω in formulajFor weight,The highest sensor vj support of associate feature is higher, and weight is bigger, institute
State ωjIt is expressed as follows:
Wherein
P=j in formula (12), the d (εu,i,εv,i)jIt indicates to be associated with spy with jth in i-th of time window inner sensor u
Euclidean distance between the highest sensor v of property；
Availability threshold gamma=0.8-0.9 of setting sensor u, in the time window of l+1, the sensor u is closed
Join the highest sensor v of characteristicjSupport should be greater than ρA, the availability γ of the sensor uuγ should be greater than, otherwise it is assumed that
Damage or sensor performance failure occur for pipeline configuration.
Further technical solution: for straight tube, the circumferential angle in sensor group between two adjacent sensors is 30
Degree or 45 degree, the distance between two neighboring sensor group be 3D-4D, wherein D for straight tube outer diameter；
For elbow, the circumferential angle in sensor group between phase two adjacent sensors is 30 degree or 45 degree, two neighboring
Section angle between sensor group is 30 degree or 45 degree；
For threeway, two adjacent sensors are 30 degree relative to the angle that Weld pipe mill location point is formed in sensor group
Or 45 degree.
Further technical solution, the time window l=90, each time window T is 1 day, and the acquisition interval time is Δ t
=1min.
The beneficial effects of the present invention are:
The present invention is analyzed in the case where not needing to establish system analysis model case using the pipe stress data correlation of monitoring
With extraction characteristic information, realize to the monitoring and early warning of pipe stress state and the evaluation of sensor reliability.
Detailed description of the invention
Fig. 1,2 for sensor group on straight tube arrangement schematic diagram.
Fig. 3,4 for sensor group on elbow arrangement schematic diagram.
Fig. 5,6 are that the supervisor of threeway goes up the arrangement schematic diagram of sensor group.
Fig. 7,8 for threeway branch pipe on sensor group arrangement schematic diagram.
Fig. 9 is the strain data vector change curve that sensor u, sensor v are monitored in part-time window.
Figure 10 is the Euclidean distance distribution map between 90 time window inner sensor u, sensor v.
Figure 11 is the sensor u availability curve that window changes at any time in normal state.
Figure 12 is the sensor u availability curve that window changes at any time under abnormal condition.
Attached meaning marked in the figure is as follows:
100- sensor, 200- supervisor, 300- branch pipe.
Specific embodiment
More specific detail is made to technical solution of the present invention below with reference to embodiment:
A kind of Monitoring Pinpelines and method for early warning based on association analysis, comprising the following steps:
Step 1, pipeline is monitored by sensor unit and obtains pipeline strain data, the arrangement of the sensor unit
It is divided into according to conduit types following several:
1st kind, for straight tube: as shown in Figure 1, 2, the multiple sensors of the same circumferentially spaced-apart setting of the straight tube simultaneously constitute biography
Sensor group, the sensor group is multiple along length of straight pipe direction setting, in sensor group between two adjacent sensors
Circumferential angle is 30 degree or 45 degree, and the distance between two neighboring sensor group is 3D-4D, and wherein D is the outer diameter of straight tube；
2nd kind, for elbow: as shown in Figure 3,4, the multiple sensors of the same circumferentially spaced-apart setting of the elbow simultaneously constitute biography
Sensor group, the sensor group is multiple along elbow length direction setting, in sensor group between phase two adjacent sensors
Circumferential angle is 30 degree or 45 degree, and the section angle between two neighboring sensor group is 30 degree or 45 degree；
3rd kind, for threeway: as shown in Fig. 5,6,7,8: the threeway is formed by being responsible for branch pipe welding, the supervisor
It is equipped with multiple sensors be arranged along weld interval and constitutes sensor group, the branch pipe along weld interval equipped with being arranged
Multiple sensors simultaneously constitute sensor group, the folder that two adjacent sensors are formed relative to Weld pipe mill location point in sensor group
Angle is 30 degree or 45 degree；
The quantity of the sensor group inner sensor and setting angle can according to need and selected, in this example with
It include 12 sensors in sensor group for straight tube, the circumferential angle in sensor group between phase two adjacent sensors is 30
Degree, the distance between two neighboring sensor group L=3D.
Step 2, the pipeline strain data in original time window are obtained, and pass through the pipeline strain data in original time window
Solve the expectation E and variances sigma of data correlation characteristic and associate feature distribution function2, the original time window is current time
L (90) a time window before window；Pipeline strain data in original time window are to sentence for current time window flexible innerduct structure
Disconnected or sensor reliability calculating basis and reference standard；
Step 2-1 uses K- nearest neighbor algorithm to find out vector distance to describe associate feature between pipeline strain data:
Assuming that SxIt for any one group of sensor group and include 12 sensors, in each time window T=1440min, often
A sensor acquisition interval time is Δ t=1min, then each sensor acquires m=1440 number in each time window T
According to；In l=90 time window, each sensor acquires N=ml=1440*90 data altogether；U, v is any one group of sensor
Two sensors in group, then in i-th of time window, the strain data vector set of the sensor u, sensor v acquisition
It closes as follows respectively:
εu,i={ εu,m×(i-1)+1,εu,m×(i-1)+2,…,εu,m×(i-1)+m} (1)
εv,i={ εv,m×(i-1)+1,εv,m×(i-1)+2,…,εv,m×(i-1)+m} (2)
Fig. 8,9 are illustrated respectively in the 2-6 time window, the strain data vector of the sensor u, sensor v acquisition
Change curve.
In i-th of time window, the Euclidean distance between described sensor u, v is indicated are as follows:
Euclidean distance d (the εu,i,εv,i) for describing the associate feature between sensor u, v, in the Europe is several
Obtain distance d (εu,i,εv,i) smaller, then data vector set εu,iWith εv,iSimilarity is higher；
In 90 time windows, Euclidean distance distribution is as shown in Figure 10 between the sensor u and sensor v；
In 90 time windows, average Euclidean distance is indicated between sensor u and sensor v are as follows:
Its variances sigma2=s2(d(εu1,εu2)), it may be assumed that
Found out in same sensor group with the smallest K sensor of the average Euclidean distance of sensor u, that is, think
Be with the highest K sensor of sensor u associate feature, according to pipeline internal pressure situation, the present embodiment selects K=3；
Step 2-2 closes on vector distance set based on K- and establishes associate feature number distribution function, and solves parameter:
In the case where pipeline does not occur degree of impairment, in time windows, the highest two sensor u of associate feature, sensing
Euclidean distance between device v meets normal distribution.For l time window, the sensor u and the highest biography of associate feature
Set d (the ε of Euclidean distance between sensor vu,εv) it is expressed as follows:
d(εu,εv)={ d (εu,1,εv,1),d(εu,2,εv,2),…,d(εu,l,εv,l)} (5)
Assuming that when time window quantity l is sufficiently large, d (εu,εv) meet normal distribution, and think d (εu,εv) expectation such as
Under:
D (the εu,εv) variance it is as follows:
σ2=s2(d(εu,εv)) (7)
In formula,
Step 3, the support and availability that monitor value K- closes on are calculated, to judge sensor reliability and pipe damage
Situation:
Current time window is i.e. in the l+1 time window inner sensor u and the highest sensor v of any one associate featurej
The Euclidean distance of the strain data vector of (1≤j≤3) acquisition is expressed as d (εu,l+1,εv,l+1)j, with season:
Z=| d (εu,l+1,εv,l+1)j-E(d(εu,εv))| (8)
In the l+1 time window, sensor u is by the highest sensor v of associate featurejSupport indicate are as follows:
In formulaρ in formulaAFor ρujConfidence threshold, take ρ in this exampleA=0.8；
In the l+1 time window, the availability γ of sensor uuIt is the highest sensor of K associate feature to sensor
The weighted sum of the support of u, it may be assumed that
ω in formulajFor weight,The highest sensor v of associate featurejSupport is higher, and weight is bigger, institute
State ωjIt is expressed as follows:
Wherein
P=j in formula (12), the d (εu,i,εv,i)jIt indicates in i-th of time window inner sensor u and j-th of associate feature
Euclidean distance between highest sensor v；
For l time window, Euclidean distance between sensor u and the highest sensor v of j-th of associate feature
Set d (εu,i,εv,i)jIt is expressed as follows:
That is d (εu,i,εv,i)j={ d (εu,1,εv,1)j,d(εu,2,εv,2)j,...,d(εu,l,εv,l)j}
Availability threshold gamma=0.8 of sensor u described in this example, it can be seen from figure 11 that in preceding 360 time
The availability γ of window inner sensor uuAlways above availability threshold value, it is believed that sensor u property retention is effective, and pipeline does not have at this
There is recurring structure damage, as can be seen from Figure 12, be gradually reduced in the availability of 780 time windows or so, sensor u, and lower than having
Validity threshold value, it is believed that sensor u sensor performance failure this moment or pipeline recurring structure damage at this need further existing
Field artificial judgment, judges via scene, it was demonstrated that damages for pipeline recurring structure.
Claims (3)
1. a kind of Monitoring Pinpelines and method for early warning based on association analysis, it is characterised in that the following steps are included:
Step 1, pipeline is monitored by sensor unit and obtains pipeline strain data, the arrangement of the sensor unit according to
Conduit types are divided into following several:
1st kind, for straight tube: the multiple sensors of the same circumferentially spaced-apart setting of the straight tube simultaneously constitute sensor group, the sensing
Device group is arranged multiple along the length of straight pipe direction；
2nd kind, for elbow: the multiple sensors of the same circumferentially spaced-apart setting of the elbow simultaneously constitute sensor group, the sensing
Device group is arranged multiple along the elbow length direction；
3rd kind, for threeway: the threeway is formed by being responsible for branch pipe welding, and the supervisor is equipped with to be arranged along weld interval
Multiple sensors and constitute sensor group, the branch pipe is equipped with the multiple sensors being arranged along weld interval and constitutes sensing
Device group；
Step 2, the pipeline strain data in original time window are obtained, and are solved by the pipeline strain data in original time window
The expectation E and variances sigma of data correlation characteristic and associate feature distribution function2, the original time window be current time window it
L preceding time window；
Step 2-1 uses K- nearest neighbor algorithm to find out vector distance to describe associate feature between pipeline strain data:
Assuming that SxIt for any one group of sensor group and include n sensor, in each time window T, each sensor acquisition interval
Time is Δ t, then each sensor acquires m data in each time window T, wherein m=T/ Δ t；In l time window
Interior, each sensor acquires N=ml data altogether；U, v is two sensors in any one group of sensor group, then i-th
In a time window, the strain data vector set difference of the sensor u, sensor v acquisition is as follows:
εu,i={ εu,m×(i-1)+1,εu,m×(i-1)+2,…,εu,m×(i-1)+m} (1)
εv,i={ εv,m×(i-1)+1,εv,m×(i-1)+2,…,εv,m×(i-1)+m} (2)
In i-th of time window, the Euclidean distance between described sensor u, v is indicated are as follows:
In l time window, average Euclidean distance is indicated between sensor u and sensor v are as follows:
Found out in same sensor group with the smallest K sensor of the average Euclidean distance of sensor u, that is, be considered with
The highest K sensor of sensor u associate feature is typically chosen K=2-4 according to pipeline internal pressure situation；
Step 2-2 closes on vector distance set based on K- and establishes associate feature number distribution function, and solves parameter:
In the case where pipeline does not occur degree of impairment, for l time window, the sensor u and the highest sensor v of associate feature it
Between Euclidean distance set d (εu,εv) it is expressed as follows:
d(εu,εv)={ d (εu,1,εv,1),d(εu,2,εv,2),...,d(εu,l,εv,l)} (5)
Assuming that when time window quantity l is sufficiently large, d (εu,εv) meet normal distribution, and think d (εu,εv) expectation it is as follows:
D (the εu,εv) variance it is as follows:
σ2=s2(d(εu,εv)) (7)
In formula,
Step 3, the support and availability that monitor value K- closes on are calculated, to judge sensor reliability and pipe damage feelings
Condition:
In the current time window i.e. the l+1 time window inner sensor u and highest sensor v of any one associate featurej(1≤j
≤ K) Euclidean distance of strain data vector of acquisition is expressed as d (εu,l+1,εv,l+1)j, with season:
Z=| d (εu,l+1,εv,l+1)j-E(d(εu,εv))| (8)
In the l+1 time window, sensor u is by the highest sensor v of associate featurejSupport indicate are as follows:
In formulaρ in formulaAFor ρujConfidence threshold, set ρA=0.7-0.9；
Define the availability γ of sensor uuFor the weighted sum of K closest sensor supports, it may be assumed that
ω in formulajFor weight,The highest sensor v of associate featurejSupport is higher, and weight is bigger, the ωj
It is expressed as follows:
Wherein
P=j in formula (12), the d (εu,i,εv,i)jIt indicates in i-th of time window inner sensor u and j-th of associate feature highest
Sensor v between Euclidean distance；
Availability threshold gamma=0.8-0.9 of setting sensor u, in the time window of l+1, the availability of the sensor u
γuγ should be greater than, otherwise it is assumed that damage or sensor performance failure occur for pipeline configuration.
2. Monitoring Pinpelines and method for early warning based on association analysis as described in claim 1, it is characterised in that: for straight tube,
Circumferential angle in sensor group between two adjacent sensors is 30 degree or 45 degree, the distance between two neighboring sensor group
For 3D-4D, wherein D is the outer diameter of straight tube；
For elbow, the circumferential angle in sensor group between phase two adjacent sensors is 30 degree or 45 degree, two neighboring sensing
Section angle between device group is 30 degree or 45 degree；
For threeway, two adjacent sensors are 30 degree or 45 relative to the angle that Weld pipe mill location point is formed in sensor group
Degree.
3. Monitoring Pinpelines and method for early warning based on association analysis as described in claim 1, it is characterised in that: the time window
L=90, each time window T is 1 day, and the acquisition interval time is Δ t=1min.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910201840.7A CN110081851A (en) | 2019-03-18 | 2019-03-18 | A kind of Monitoring Pinpelines and method for early warning based on association analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910201840.7A CN110081851A (en) | 2019-03-18 | 2019-03-18 | A kind of Monitoring Pinpelines and method for early warning based on association analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110081851A true CN110081851A (en) | 2019-08-02 |
Family
ID=67413248
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910201840.7A Pending CN110081851A (en) | 2019-03-18 | 2019-03-18 | A kind of Monitoring Pinpelines and method for early warning based on association analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110081851A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150206285A1 (en) * | 2014-01-21 | 2015-07-23 | Nvidia Corporation | Efficient approximate-nearest-neighbor (ann) search for high-quality collaborative filtering |
| US20170181621A1 (en) * | 2013-11-08 | 2017-06-29 | Precision Ocular Metrology, L.L.C. | Systems and methods for mapping the ocular surface |
| CN106959608A (en) * | 2017-02-27 | 2017-07-18 | 同济大学 | A kind of water supply network seepage optimal control method based on cluster particle cluster algorithm |
| CN107316057A (en) * | 2017-06-07 | 2017-11-03 | 哈尔滨工程大学 | Based on the nuclear power unit method for diagnosing faults being locally linear embedding into K nearest neighbor classifiers |
-
2019
- 2019-03-18 CN CN201910201840.7A patent/CN110081851A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20170181621A1 (en) * | 2013-11-08 | 2017-06-29 | Precision Ocular Metrology, L.L.C. | Systems and methods for mapping the ocular surface |
| US20150206285A1 (en) * | 2014-01-21 | 2015-07-23 | Nvidia Corporation | Efficient approximate-nearest-neighbor (ann) search for high-quality collaborative filtering |
| CN106959608A (en) * | 2017-02-27 | 2017-07-18 | 同济大学 | A kind of water supply network seepage optimal control method based on cluster particle cluster algorithm |
| CN107316057A (en) * | 2017-06-07 | 2017-11-03 | 哈尔滨工程大学 | Based on the nuclear power unit method for diagnosing faults being locally linear embedding into K nearest neighbor classifiers |
Non-Patent Citations (2)
| Title |
|---|
| 顾亚雄等: "基于K-最近邻算法的管道系统传感器数据有效性分析研究", 《传感技术学报》 * |
| 黄晓微等: "基于K-最近邻算法的挠度传感器有效度研究", 《仪器仪表学报》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103968256B (en) | Piping for tank farm leakage detection method | |
| CN109854953B (en) | Crude oil conveying pipeline leakage detection system and method under special working condition | |
| KR20140092385A (en) | System and method for identifying related events in a resource network monitoring system | |
| CN103412220B (en) | A kind of ship power station fault diagnosis system method based on data fusion | |
| CN107906375B (en) | Pipeline leakage detection method and system based on weighted permutation entropy | |
| CN113124327A (en) | Monitoring method and device for natural gas pipeline and monitoring and scheduling system | |
| CN103032626B (en) | System and method for diagnosing fault of adjusting valve | |
| CN108506732A (en) | Oil pipeline anti-leak monitors system | |
| CN114544040B (en) | Pile group node stress monitoring system based on neural network algorithm and early warning method thereof | |
| CN110081851A (en) | A kind of Monitoring Pinpelines and method for early warning based on association analysis | |
| CN106656618B (en) | The recognition methods of communication iron tower sensor abnormality and system based on traffic analysis | |
| CN111275936B (en) | Underwater oilfield facility safety protection monitoring system and method thereof | |
| CN202065685U (en) | Oil filed pipeline corrosion early warning system | |
| CN207005755U (en) | A kind of field oil recovery monitoring and alarming system | |
| CN116221626A (en) | Heating power pipeline leakage detection positioning method and system based on mathematical model | |
| CN105628351A (en) | Electronic expansion valve detection method and device | |
| CN115062891A (en) | Method for quickly positioning leakage of long-distance pressure water delivery pipeline | |
| CN110617403B (en) | Long-distance pipeline leakage detection method based on multi-sensor information fusion | |
| CN104266536B (en) | A kind of air cooling tower liquid level controlling method and device | |
| CN106647563B (en) | A kind of oil recovery site safety monitoring device | |
| CN208456669U (en) | A kind of well control classification linkage alarm device | |
| CN113029055A (en) | Explosion-proof anti-interference pipeline wall thickness ultrasonic online monitoring device | |
| AU2017330313B2 (en) | Anomaly detection and neural network algorithms for PST hydrocyclone condition monitoring | |
| KR102484932B1 (en) | Method and apparatus for detecting anomalies in industrial control system using status information of industrial facilities | |
| CN204900076U (en) | Offshore oil field internal -combustion engine monitoring system based on GPRS network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190802 |
|
| WD01 | Invention patent application deemed withdrawn after publication |