CN107390661B - A kind of method for early warning of process flow industry process abnormal state - Google Patents
A kind of method for early warning of process flow industry process abnormal state Download PDFInfo
- Publication number
- CN107390661B CN107390661B CN201710750153.1A CN201710750153A CN107390661B CN 107390661 B CN107390661 B CN 107390661B CN 201710750153 A CN201710750153 A CN 201710750153A CN 107390661 B CN107390661 B CN 107390661B
- Authority
- CN
- China
- Prior art keywords
- early warning
- trend
- limit value
- warning
- point
- 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.)
- Active
Links
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Abstract
The invention discloses a kind of method for early warning of process flow industry process abnormal state, set the warning information of trend and limit value first, trending early warning is divided into 4 classes according to operating condition trend, concurrently sets the bound section of duty parameter limit value early warning；It is then based on time shaft, p sampled point is chosen forward using current point as starting point and carries out recurrence calculating, obtain slope and related coefficient；It moves forward every time then according to above-mentioned steps and 1 sampled point and computes repeatedly total n times；The last identification process according to trending early warning, determines the trending early warning classification of current working, according to the identification process of limit value early warning, judges whether to generate limit value early warning, and the recognition result of integrative trend early warning and limit value early warning, judges the warning level of current operating condition.This method is practical, reliable, has not only reduced the probability for leading to erroneous judgement because of noise data, but also can give warning in advance to unusual service condition.This method has important application value to timely troubleshooting, reduction breakdown loss, guarantee production safety.
Description
Technical field
The present invention relates to the fault detection in process flow industry process field and warning aspects, are based on the two class early warning of trend and limit value
Information establishes the real-time detection and method for early warning of a kind of abnormality.
Background technique
Failure or abnormality in the process industries such as petroleum, petrochemical industry field are often reported using Distributed Control System (DCS)
It is alert.However, DCS generallys use fixed bound alarm at present, existing can not provide that early warning, alarm are not in time etc. many to ask
Topic.
Have the method for early warning of some failures or exception at present, is broadly divided into based on model and based on data-driven two greatly
Class.Method based on model needs profound understanding and grasps accurate technical process mechanism, is greatly limited in practical applications
System；Method based on data-driven is to acquire real-time and historical data from DCS, then carries out data processing and judgement.But data
The method of driving still has some problems in practical applications, for example, the data of measurement easily cause failure often with noise
Wrong report；And lag issues are readily incorporated in order to eliminate the filtering method that noise uses sometimes, cause alarm not in time.Therefore,
A practical, reliable abnormality early warning system how is established, balance can be found between accuracy, real-time, in recent years
Increasingly cause to pay close attention to.
Summary of the invention
For the present invention in view of the above-mentioned problems, combining the early warning identification information of trend and limit value, establishment process abnormal state is pre-
Alert judgment rule and process, can be violent to variation or gradual change sexual abnormality is measured in real time and early warning.
The invention adopts the following technical scheme:
A kind of method for early warning of process flow industry process abnormal state, this method combine the early warning identification letter of trend and limit value
Breath, the judgment rule and process of establishment process abnormal state early warning have follow steps:
(1) trending early warning is divided into 4 classes: trend rapid increase, trend rapid decrease, the big amplitude wave of trend according to operating condition trend
Dynamic and trend is normal；
(2) setting duty parameter operates normally the upper limit x in sectionhWith lower limit xl；
(3) real-time running data is extracted from monitoring system to return p history samples point using current sampling point as starting point
Return calculating, obtains slope k1And correlation coefficient r1；
(4) based on time shaft 1 sampled point of Forward, again to p sampled point digital simulation slope k2And correlation coefficient r2, weight
This multiple process, until obtaining k1~knAnd r1~rn；
(5) according to trending early warning identification process, the trending early warning classification of current working is determined: if trend is rapid increase, fastly
(6) speed decline or fluctuation, then follow the steps；If trend be it is normal, then follow the steps (7)；
(6), according to limit value early warning identification process, judge whether to generate limit value early warning；
(7) the recognition result of integrative trend and limit value, the warning level of the current operating condition of judgment means.
Specifically, process provides the identification process to trend rapid increase and rapid decrease early warning in current working:
(1) r is counted1~rnIn related coefficient early warning range q1l≤|r|≤q1rRatio w1, and with threshold value m1Compare, if
w1< m1The trending early warning for then carrying out fluctuation to current working identifies；If w1≥m1Then go to step (2)；
(2) k is counted1~knIn slope early warning range ul≤|k|≤urRatio w2, and with threshold value m2Compare, if just tiltedly
Rate section and w2≥m2Then it is determined as trend rapid increase；If between negative slope region and w2≥m2Then it is determined as trend rapid decrease；
If w2< m2Then it is determined as that trend is normal.
Specifically, process provides the identification process to trend fluctuation early warning in current working:
(1) r is counted1~rnIn related coefficient early warning range q2l≤|r|≤q2rRatio w3, and with threshold value m3Compare, if
w3< m3Then it is determined as normal trend, if w3≥m3Then go to step (2)；
(2), using current sampling point as starting point, 2 are chosen forward based on time shaftpA sampled point simultaneously calculates standard deviation v, with threshold value
m4Compare, if v >=m4Then it is judged as trend fluctuation, if v < m4Then it is judged as normal trend.
Specifically, being chosen forward process provides the identification process to limit value early warning in current working based on time shaft
3 sampled points including current point predict following 3 sampled points, the specific steps are as follows:
(1) current sampling point is x (p), and T is the sampling period, and weight coefficient α, β and γ is arranged, and makes alpha+beta+γ=1, extracts x
(p), x (p-1) and x (p-2) amounts to 3 sampled points；
(2) K is calculated according to following formula(0) 1、K(0) 2And K(0) 3
According to formulaIt calculatesAccording to formulaMeter
It calculates
(3) K is calculated according to following formula(1) 1、K(1) 2And K(1) 3
According to formulaIt calculatesAccording to formula
It calculates
(4) K is calculated according to following formula(2) 1、K(2) 2And K(2) 3
According to formulaIt calculatesAccording to formulaIt calculates
(5) ifWithIn have 2 or more be more than given duty parameter limit value, that is, produce
Raw limit value early warning.
Preferably, identification information of this method based on trend and limit value early warning establishes the pre- of a set of process status exception
Alert judgment rule, wherein judging result is divided into three warning levels: green, yellow and red, Green are indicating operating condition just
Often, yellow expression is pointed out, red for alarm state, and specific rules are as follows:
| Classification | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 |
| Trend rapid increase | It is | It is | It is no | It is no | It is no | It is no | It is no | It is no |
| Trend rapid decrease | It is no | It is no | It is | It is | It is no | It is no | It is no | It is no |
| Trend fluctuation | It is no | It is no | It is no | It is no | It is | It is | It is no | It is no |
| Trend is normal | It is no | It is no | It is no | It is no | It is no | It is no | It is | It is |
| It whether is more than limit value | It is | It is no | It is | It is no | It is | It is no | It is | It is no |
| Warning level | It is red | Yellow | It is red | Yellow | It is red | Yellow | It is red | Green |
The utility model has the advantages that
The present invention is based on trend for abnormal state real-time monitoring and the insufficient actual conditions of pre-alerting ability in process industry
The identification information of early warning and limit value early warning, establishes the early warning judgment rule of a set of process status exception, and provides current operation
The warning level of operating condition.This method is practical, reliable, has not only reduced the probability for leading to erroneous judgement because of noise data, but also can be to exception
Operating condition gives warning in advance.This method applies valence with important to timely troubleshooting, reduction breakdown loss, guarantee production safety
Value.
Detailed description of the invention
Fig. 1 is the early warning overview flow chart of process flow industry process abnormal state
Fig. 2 is the regression figure in case 1 comprising current sampling point
Fig. 3 is the Trend value prognostic chart of 1 trending early warning of case
Fig. 4 is the regression figure in case 2 comprising current sampling point
Fig. 5 is the Trend value prognostic chart of 2 trending early warning of case
Specific implementation process
With reference to the accompanying drawing and two specific examples, detailed calculating process and concrete operations process are provided, with right
The present invention is described further.The implementation case is implemented under the premise of the technical scheme of the present invention, but guarantor of the invention
Shield range is not limited to following case study on implementation.
Case 1
By taking primary distillation tower tower top pressure in the Atmospheric vacuum technique of certain Petrochemical Enterprises as an example, selection is the enterprise in November, 2016
Every the data of acquisition in 5 minutes in 23:25 to the 26 days 1:30 period on the 25th, according to the abnormal state early warning rule and stream established
Journey completes the anomalous identification of primary distillation tower tower top pressure.
1 implementing procedure of case is as shown in Figure 1, specific implementation steps are as follows:
(1) warning information of trend and limit value is set
Trending early warning is divided into 4 classes: trend rapid increase, trend rapid decrease, trend fluctuation according to operating condition trend
And normal trend, setting duty parameter upper limit x is required according to site techniqueh=343kPa and lower limit xl=208kPa.
(2) it returns and calculates slope and related coefficient
It is extracted 26 sample point datas first, data are as shown in table 1:
1 primary distillation tower tower top pressure data of table
The sampled point quantity p=7 of recurrence is set in present case, is based on time shaft, is selected forward using current sampling point as starting point
7 sampled points are taken to carry out least square regression.Regression result is as shown in Figure 2, it is seen then that regression equation y=-2.07x+208.93,
Fit slope k1=-2.07, correlation coefficient r1=-0.91.
1 sampled point of each Forward simultaneously calculates, and finally obtains the slope k summarized1~k20And correlation coefficient r1~r20, specifically
Numerical value is as shown in table 2:
2 regression coefficient of table summarizes
| Slope (kPa/h) | Related coefficient | |
| 1 | -2.07 | -0.91 |
| 2 | -3.12 | -0.93 |
| 3 | -4.21 | -0.96 |
| 4 | -5.20 | -0.99 |
| 5 | -5.93 | -1.00 |
| 6 | -6.21 | -1.00 |
| 7 | -6.21 | -1.00 |
| 8 | -6.21 | -1.00 |
| 9 | -6.21 | -1.00 |
| 10 | -6.21 | -1.00 |
| 11 | -6.21 | -1.00 |
| 12 | -5.98 | -1.00 |
| 13 | -5.33 | -0.99 |
| 14 | -4.43 | -0.97 |
| 15 | -3.45 | -0.95 |
| 16 | -2.54 | -0.95 |
| 17 | -1.86 | -0.98 |
| 18 | -1.59 | -1.00 |
| 19 | -1.59 | -1.00 |
| 20 | -1.59 | -1.00 |
(3) trending early warning identifies
The related coefficient range for the trending early warning that this case is set as 0.50≤| r |≤1.00, threshold value m1=0.60；Trend
The negative slope upper limit of early warning is -2.50kPa/h, threshold value m2=0.65.Correlation coefficient r1~r20Ratio in trending early warning section
Example:Slope k1~k20Belong to the ratio in trending early warning section:It is all larger than the threshold of setting
Value, and the related coefficient section of trending early warning is negative section, thus trending early warning identification the result is that: trend rapid decrease.
(4) limit value early warning identifies
The current sampling point of this case is x (p)=208.04kPa, and the sampling period isWeight is set
Following three sampled points are predicted based on x (p-2), x (p-1) and x (p) in factor alpha=0.5, β=0.3 and γ=0.2, specific to calculate
It is as follows:
1. calculating separatelyWith
It brings formula into and obtains predicted value
2. calculating separatelyWith
It brings formula into and obtains predicted value
3. calculating separatelyWith
It brings formula into and obtains predicted value
Calculated Trend value prediction is as shown in Figure 3.
4. countingWithIn beyond operate normally section have 3, generate limit value early warning.
(5) the early warning recognition result of integrative trend and limit value, the operating condition not only trend rapid decrease, but also generation limit value early warning, according to
According to the judgment rule of process status abnormity early warning, show that current point in time 01:30 operating condition warning level on November 26th, 2016 is
It is red.
Case 2
Again by taking normal pressure column overhead temperatures in enterprise's Atmospheric vacuum technique as an example, on 2 4th, 2017 11:15 to 13:20 are chosen
It is complete according to the fault pre-alarming rule and process established every the normal pressure column overhead temperatures data that sampling in 5 minutes is primary in period
At the anomalous identification of normal pressure column overhead temperatures parameter, specific implementation steps are as follows:
(1) the warning information of trend and limit value is set
Trending early warning is divided into 4 classes according to operating condition trend, setting duty parameter upper limit x is required according to site techniqueh=146
DEG C and lower limit xl=128 DEG C.
(2) return and calculate slope and related coefficient
26 sample point datas are extracted, as shown in table 3:
3 normal pressure column overhead temperatures data of table
| Serial number | Date | Time point | Temperature (DEG C) |
| 1 | 2017/2/4 | 11:15 | 141.79 |
| 2 | 2017/2/4 | 11:20 | 142.17 |
| 3 | 2017/2/4 | 11:25 | 142.29 |
| 4 | 2017/2/4 | 11:30 | 142.49 |
| 5 | 2017/2/4 | 11:35 | 142.58 |
| 6 | 2017/2/4 | 11:40 | 142.92 |
| 7 | 2017/2/4 | 11:45 | 143.33 |
| 8 | 2017/2/4 | 11:50 | 143.39 |
| 9 | 2017/2/4 | 11:55 | 143.53 |
| 10 | 2017/2/4 | 12:00 | 143.48 |
| 11 | 2017/2/4 | 12:05 | 143.30 |
| 12 | 2017/2/4 | 12:10 | 143.37 |
| 13 | 2017/2/4 | 12:15 | 143.42 |
| 14 | 2017/2/4 | 12:20 | 143.45 |
| 15 | 2017/2/4 | 12:25 | 143.60 |
| 16 | 2017/2/4 | 12:30 | 143.65 |
| 17 | 2017/2/4 | 12:35 | 143.69 |
| 18 | 2017/2/4 | 12:40 | 143.75 |
| 19 | 2017/2/4 | 12:45 | 143.71 |
| 20 | 2017/2/4 | 12:50 | 144.15 |
| 21 | 2017/2/4 | 12:55 | 144.18 |
| 22 | 2017/2/4 | 13:00 | 144.23 |
| 23 | 2017/2/4 | 13:05 | 144.27 |
| 24 | 2017/2/4 | 13:10 | 144.11 |
| 25 | 2017/2/4 | 13:15 | 144.79 |
| 26 | 2017/2/4 | 13:20 | 144.96 |
The sampled point quantity p=7 returned is set, and regression result is as shown in Figure 4.Digital simulation slope and related coefficient, obtain
To k1~k20And r1~r20, specific value is as shown in table 4:
4 regression coefficient of table summarizes
(3) trending early warning identifies
The related coefficient range for the trending early warning that this case is set as 0.50≤| r |≤1.00, threshold value m1=0.60；Trend
The positive slope lower limit of early warning is 0.80, threshold value m2=0.65.Correlation coefficient r1~r20Ratio in trending early warning section is
0.85, slope k1~k20Ratio in trending early warning section is 0.70, is all larger than the threshold value respectively set, and trending early warning
Related coefficient section is positive section, thus trending early warning identification the result is that: trend rapid increase.
(4) limit value early warning identifies
The current sampling point of this case is x (p)=144.96 DEG C, setting weight coefficient α=0.5, β=0.3 and γ=
0.2, following three sampled points are predicted based on x (p-2), x (p-1) and x (p), as shown in figure 5, specifically calculating as follows:
1. calculating separately slope obtains K(0) 1=2.04 DEG C/h, K(0) 2=8.16 DEG C/h and K(0) 3=5.10 DEG C/h, synthesis is
It obtains slope and weight coefficient calculatesIt brings formula into and obtains predicted value
2. calculating separately slope obtains K(1) 1=4.44 DEG C/h, K(1) 2=2.04 DEG C/h and K(1) 3=3.24 DEG C/h, synthesis is
It obtains slope and weight coefficient calculatesIt brings formula into and obtains predicted value
3. calculating separately slope obtains K(2) 1=3.48 DEG C/h, K(2) 2=4.44 DEG C/h and K(2)3=3.96 DEG C/h, synthesis is
It obtains slope and weight coefficient calculatesIt brings formula into and obtains predicted value
4. countingWithIn be more than that set interval has 0, do not generate limit value early warning.
(5) the early warning identification information of integrative trend and limit value, which is trend rapid increase, but does not generate limit value early warning,
According to judgment rule, show that on 2 4th, 2017 13:20 operating condition warning levels of current point in time are yellow.
Claims (4)
1. a kind of method for early warning of process flow industry process abnormal state, it is characterised in that have follow steps:
(1) trending early warning is divided into 4 classes according to operating condition trend: trend rapid increase, trend rapid decrease, trend fluctuation and
Trend is normal；
Following identification process is used to trend rapid increase in current working and trend rapid decrease early warning:
(1-1) counts r1~rnIn [q1l, q1r] ratio w1, q1lAnd q1rThe respectively lower limit of the first early warning of related coefficient and upper
Limit；w1And with threshold value m1Compare, if w1< m1The trending early warning for then carrying out trend fluctuation to current working identifies；If w1≥m1
Then go to step (2)；
(1-2) counts k1~knIn [ul, ur] ratio w2, ulAnd urThe respectively lower and upper limit of slope early warning；w2And and threshold
Value m2Compare, if between positive slope region and w2≥m2Then it is determined as trend rapid increase；If between negative slope region and w2≥m2Then sentence
It is set to trend rapid decrease；If w2< m2Then it is determined as that trend is normal；
(2) setting duty parameter operates normally the upper limit x in sectionhWith lower limit xl；
(3) real-time running data being extracted from monitoring system, recurrence meter is carried out to p history samples point using current sampling point as starting point
It calculates, obtains slope k1And correlation coefficient r1；
(4) based on time shaft 1 sampled point of Forward, again to p sampled point digital simulation slope k2And correlation coefficient r2, repeat this
Process, until obtaining k1~knAnd r1~rn；
(5) according to trending early warning identification process, the trending early warning classification of current working is determined: if under trend is rapid increase, is quick
(6) drop or fluctuation, then follow the steps；If trend be it is normal, then follow the steps (7)；
(6), according to limit value early warning identification process, judge whether to generate limit value early warning；
(7) the recognition result of integrative trend and limit value, the warning level of the current operating condition of judgment means.
2. a kind of method for early warning of process flow industry process abnormal state according to claim 1, it is characterised in that current
The early warning of trend fluctuation uses following identification process in operating condition:
(1) r is counted1~rnIn [q2l, q2r] ratio w3, q2lAnd q2rThe respectively lower and upper limit of the second early warning of related coefficient；
w3And with threshold value m3Compare, if w3< m3Then it is determined as normal trend, if w3≥m3Then go to step (2)；
(2), using current sampling point as starting point, choose 2p sampled point forward based on time shaft and calculate standard deviation v, v and threshold value m4Than
Compared with if v >=m4Then it is judged as trend fluctuation, if v < m4Then it is judged as normal trend.
3. a kind of method for early warning of process flow industry process abnormal state according to claim 1, it is characterised in that limit value is pre-
Alert identification process is chosen forward 3 sampled points including current point based on time shaft and carried out in advance to following 3 sampled points
It surveys, the specific steps are as follows:
(1) current sampling point is x (p), and T is the sampling period, and weight coefficient α, β and γ is arranged, and makes alpha+beta+γ=1, extracts x (p), x
(p-1) and x (p-2) amounts to 3 sampled points；
(2) K is calculated according to following formula(0) 1、K(0) 2And K(0) 3
According to formulaIt calculatesAccording to formulaIt calculates
(3) K is calculated according to following formula(1) 1、K(1) 2And K(1) 3
According to formulaIt calculatesAccording to formulaIt calculates
(4) K is calculated according to following formula(2) 1、K(2) 2And K(2) 3
According to formulaIt calculatesAccording to formulaIt calculates
?
(5) ifWithIn have 2 or more be more than given duty parameter limit value, that is, generate limit
It is worth early warning.
4. a kind of method for early warning of process flow industry process abnormal state according to claim 1, it is characterised in that be based on
The early warning identification information of gesture and limit value, the early warning judgment rule of establishment process abnormal state, wherein judging result is divided into 3 in advance
Alert rank: green, yellow and red, Green indicate that operating condition is normal, and yellow expression is pointed out, red for alarm shape
State, specific rules are as follows:
。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710750153.1A CN107390661B (en) | 2017-08-28 | 2017-08-28 | A kind of method for early warning of process flow industry process abnormal state |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710750153.1A CN107390661B (en) | 2017-08-28 | 2017-08-28 | A kind of method for early warning of process flow industry process abnormal state |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107390661A CN107390661A (en) | 2017-11-24 |
| CN107390661B true CN107390661B (en) | 2019-05-24 |
Family
ID=60346651
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710750153.1A Active CN107390661B (en) | 2017-08-28 | 2017-08-28 | A kind of method for early warning of process flow industry process abnormal state |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107390661B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108132867B (en) * | 2018-01-11 | 2021-05-25 | 合肥科博软件技术有限公司 | Equipment fault alarm method and computing equipment |
| CN108765889B (en) * | 2018-04-17 | 2020-08-04 | 中国石油集团安全环保技术研究院有限公司 | Oil and gas production operation safety early warning method based on big data technology |
| CN109063261B (en) * | 2018-06-29 | 2022-11-29 | 深圳市科列技术股份有限公司 | Method and device for judging aging trend of power battery |
| CN110764975B (en) * | 2018-07-27 | 2021-10-22 | 华为技术有限公司 | Early warning method and device for equipment performance and monitoring equipment |
| CN110134079B (en) * | 2019-03-26 | 2022-05-03 | 石化盈科信息技术有限责任公司 | Process parameter early warning method and system based on slope analysis |
| CN110349046A (en) * | 2019-05-29 | 2019-10-18 | 浙江陀曼云计算有限公司 | Produce method for early warning and device |
| CN112149957A (en) * | 2020-08-20 | 2020-12-29 | 汉威科技集团股份有限公司 | Risk trend deduction and grading early warning method based on online monitoring data |
| CN112633561A (en) * | 2020-12-09 | 2021-04-09 | 北京名道恒通信息技术有限公司 | Production risk intelligent prediction early warning method based on industrial big data |
| CN112508105B (en) * | 2020-12-11 | 2024-03-19 | 南京富岛信息工程有限公司 | Fault detection and retrieval method for oil extraction machine |
| CN113380411B (en) * | 2021-07-19 | 2024-03-01 | 苏州百孝医疗科技有限公司 | Method for improving accuracy of concentration change real-time trend in continuous monitoring process of animal analyte concentration |
| CN115599646A (en) * | 2022-10-14 | 2023-01-13 | 南京凯盛国际工程有限公司（Cn） | Method for improving early warning advance and accuracy of evaluation model |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102117731A (en) * | 2009-12-31 | 2011-07-06 | 中芯国际集成电路制造(上海)有限公司 | Method and device for monitoring measurement data in process production flow of semiconductor |
| CN103714238A (en) * | 2012-10-01 | 2014-04-09 | 福特全球技术公司 | System and method for rating computer model relative to empirical results for dynamic systems |
| CN103744389A (en) * | 2013-12-30 | 2014-04-23 | 中国石油天然气股份有限公司 | Operating state early-warning method of oil and gas production equipment |
| CN105389648A (en) * | 2015-10-21 | 2016-03-09 | 南京富岛信息工程有限公司 | Distinguishing method for steady-state operating condition of atmospheric and vacuum distillation device |
| CN106061370A (en) * | 2013-12-16 | 2016-10-26 | 美敦力迷你迈德公司 | Use of electrochemical impedance spectroscopy (EIS) in continuous glucose monitoring |
-
2017
- 2017-08-28 CN CN201710750153.1A patent/CN107390661B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102117731A (en) * | 2009-12-31 | 2011-07-06 | 中芯国际集成电路制造(上海)有限公司 | Method and device for monitoring measurement data in process production flow of semiconductor |
| CN103714238A (en) * | 2012-10-01 | 2014-04-09 | 福特全球技术公司 | System and method for rating computer model relative to empirical results for dynamic systems |
| CN106061370A (en) * | 2013-12-16 | 2016-10-26 | 美敦力迷你迈德公司 | Use of electrochemical impedance spectroscopy (EIS) in continuous glucose monitoring |
| CN103744389A (en) * | 2013-12-30 | 2014-04-23 | 中国石油天然气股份有限公司 | Operating state early-warning method of oil and gas production equipment |
| CN105389648A (en) * | 2015-10-21 | 2016-03-09 | 南京富岛信息工程有限公司 | Distinguishing method for steady-state operating condition of atmospheric and vacuum distillation device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107390661A (en) | 2017-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107390661B (en) | A kind of method for early warning of process flow industry process abnormal state | |
| US10545487B2 (en) | Interactive diagnostic system and method for managing process model analysis | |
| DE102014100855A1 (en) | Systems and methods for monitoring operational processes | |
| CN104677997B (en) | A kind of transformer oil chromatographic on-line monitoring differentiation method for early warning | |
| CN109933905A (en) | A kind of mechanical equipment health state evaluation method based on multidimensional early warning analysis | |
| EP2477088A1 (en) | System and method for identifying likely geographical locations of anomalies in a water utility network | |
| WO2016160910A1 (en) | Advanced data cleansing system and method | |
| CN104156473B (en) | LS-SVM-based method for detecting anomaly slot of sensor detection data | |
| CN102282552A (en) | System, method and computer program for pattern based intelligent control, monitoring and automation | |
| CN106447210A (en) | Distribution network equipment health degree dynamic diagnosis method involving credibility evaluation | |
| CN104091035A (en) | Health monitoring method for effective loads of space station based on data-driven algorithm | |
| CN104808651B (en) | The measuring method of the reasonable alarm limit of chemical plant installations key parameter | |
| CN107065834A (en) | The method for diagnosing faults of concentrator in hydrometallurgy process | |
| CN104615123B (en) | K-nearest neighbor based sensor fault isolation method | |
| CN112598144A (en) | CNN-LSTM burst fault early warning method based on correlation analysis | |
| CN103309347B (en) | A kind of multiple operating modes process method for supervising based on rarefaction representation | |
| CN113341087A (en) | Intelligent monitoring and analyzing method and system for river water pollution source and computer storage medium | |
| CN115077627A (en) | Multi-fusion environmental data supervision method and supervision system | |
| CN116991130A (en) | Intelligent automatic control system and method for petrochemical production | |
| CN115410342A (en) | Landslide disaster intelligent early warning method based on crack meter real-time monitoring | |
| CN111338310B (en) | Industrial process steady-state working condition identification and classification method | |
| CN106599201A (en) | Full life circle management method of gas transmission and distribution equipment | |
| CN110826024A (en) | Sampling discrimination method for sensor data abnormity | |
| CN113627278A (en) | Water use trend graphical analysis method based on water fingerprint algorithm | |
| CN114997313B (en) | Abnormality detection method for ocean on-line monitoring data |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |