CN106368816B - A kind of online method for detecting abnormality of marine low speed diesel engine based on baseline offset - Google Patents
A kind of online method for detecting abnormality of marine low speed diesel engine based on baseline offset Download PDFInfo
- Publication number
- CN106368816B CN106368816B CN201610959364.1A CN201610959364A CN106368816B CN 106368816 B CN106368816 B CN 106368816B CN 201610959364 A CN201610959364 A CN 201610959364A CN 106368816 B CN106368816 B CN 106368816B
- Authority
- CN
- China
- Prior art keywords
- diesel engine
- speed diesel
- baseline
- health
- low
- 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
- 238000000034 method Methods 0.000 title claims abstract description 50
- 230000005856 abnormality Effects 0.000 title claims abstract description 8
- 238000012544 monitoring process Methods 0.000 claims abstract description 78
- 238000001514 detection method Methods 0.000 claims abstract description 38
- 230000036541 health Effects 0.000 claims description 47
- 239000013598 vector Substances 0.000 claims description 34
- 239000011159 matrix material Substances 0.000 claims description 27
- 238000005070 sampling Methods 0.000 claims description 24
- 230000008569 process Effects 0.000 claims description 22
- 230000002159 abnormal effect Effects 0.000 claims description 21
- 238000007689 inspection Methods 0.000 claims description 10
- 230000003862 health status Effects 0.000 abstract 1
- 238000010276 construction Methods 0.000 description 7
- 238000001816 cooling Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 6
- 239000010687 lubricating oil Substances 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 239000000446 fuel Substances 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 230000007704 transition Effects 0.000 description 3
- 239000000498 cooling water Substances 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 230000005654 stationary process Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 240000004760 Pimpinella anisum Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000003921 oil Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 230000002000 scavenging effect Effects 0.000 description 1
- 239000002904 solvent Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
- F02B—INTERNAL-COMBUSTION PISTON ENGINES; COMBUSTION ENGINES IN GENERAL
- F02B77/00—Component parts, details or accessories, not otherwise provided for
- F02B77/08—Safety, indicating, or supervising devices
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Testing Of Engines (AREA)
- Combined Controls Of Internal Combustion Engines (AREA)
Abstract
The present invention relates to a kind of online method for detecting abnormality of marine low speed diesel engine based on baseline offset, includes the following steps：Failure-free data is obtained from the parameters history of marine low speed diesel engine monitoring system record, selects the healthy baseline of steady working condition data configuration dot-blur pattern form, the health status benchmark as low-speed diesel engine monitoring system；Real-time Monitoring Data is obtained from marine low speed diesel engine monitoring system, baseline offset calculating is carried out to the Real-time Monitoring Data under steady working condition, exports the online abnormality detection result based on baseline offset.
Description
Technical Field
The invention relates to the technical field of automation, in particular to a baseline deviation-based online anomaly detection method for a low-speed diesel engine of a ship.
Background
As the most important transportation means at sea, large ships have been playing an indispensable role in the revolution of human society, the intercommunication of resources and knowledge, and the exploration of the world since birth.
Particularly, in an ocean vessel, the most common main engine of the vessel is a low-speed diesel engine which is used as a dynamic 'heart' of the vessel, is an important guarantee for the safe operation of the vessel and is also the part with the highest occurrence rate of safety accidents of the vessel, and the ocean vessel has high safety requirements, weak aviation guarantee capability and other objective situations in navigation and also provides higher requirements for the aviation operation safety of the low-speed diesel engine.
The traditional method for guaranteeing the operation safety of the low-speed diesel engine comprises two methods: firstly, the monitoring system gives a threshold alarm, namely, the alarm of corresponding monitoring parameters is carried out when the threshold is exceeded, the method is a basic safety guarantee mode, and for the slow-change type fault, the abnormity which is long in time before the threshold is often occurred, the problem when the threshold is exceeded is very serious, and even a safety accident can occur; and the other is fault diagnosis based on mechanism, namely that the low-speed diesel engine has a fault, and the fault is diagnosed and checked through the fault position and the corresponding mechanism, which belongs to post diagnosis.
The safety of ocean going voyages is expected to find abnormal or possible problems early to avoid failures and even accidents. Under the background of the big data era, the intellectualization of ships becomes the inevitable trend of the development of the fields of ship manufacturing and shipping at present, and the traditional mode of 'making models through mechanisms, monitoring through thresholds and managing through experiences' is difficult to meet the development requirements of the ship industry in the future.
Disclosure of Invention
In view of the above analysis, the present invention aims to provide an online anomaly detection method for a marine low-speed diesel engine based on baseline deviation, so as to solve the problem that the existing method for ensuring the operation safety of the low-speed diesel engine cannot find anomalies or possibly occurs as soon as possible to avoid faults and even accidents.
The purpose of the invention is mainly realized by the following technical scheme:
a ship low-speed diesel engine online anomaly detection method based on baseline deviation is characterized by comprising the following steps:
s101, acquiring fault-free data from a parameter historical record of a ship low-speed diesel engine monitoring system;
s102, selecting stable working condition data from the acquired fault-free data;
s103, constructing a health baseline in a memory matrix form according to the stable working condition data, and using the health baseline as a health state reference of the low-speed diesel engine monitoring system;
step S201, acquiring real-time monitoring data from a ship low-speed diesel engine monitoring system;
s202, judging the stable working condition of the acquired real-time monitoring data;
step S203, calculating the baseline deviation of the real-time monitoring data under the stable working condition;
and step S204, outputting an online abnormity detection result based on the baseline deviation.
The step S101 further includes:
acquiring all fault-free data in the historical time length T in the historical record of the ship low-speed diesel engine monitoring system, and recording the fault-free data as N health samples, wherein each sample comprises a plurality of monitoring parameters.
The step S102 further includes:
and judging whether the current rotating speed time sequence is a stable process, namely whether the low-speed diesel engine corresponding to the time is in a stable working condition running state, and selecting stable working condition data from the N healthy samples.
The step S103 further includes:
using all steady state observation vectors in N healthy samplesAnd constructing a health baseline in a memory matrix form for recording all historical health states, and taking the health baseline as a health state reference of the monitoring process of the low-speed diesel engine.
The step S202 further includes:
judging whether the rotating speed time sequence of the current real-time monitoring sampling point data is a stable process:
if the current rotating speed time sequence is not stable, no abnormal detection is carried out, and the step S201 is returned;
and if the current rotating speed time sequence is stable, obtaining a stable working condition running state quantity set, and reconstructing an observation vector at the current monitoring sampling point under the stable running state of the low-speed diesel engine.
The step S203 further includes:
the health base line in the form of the memory matrix constructed in step S103 is used to obtain the health state estimation vector at the current real-time monitoring sampling point, and the base line deviation between the real-time monitored health state observation vector and the health state estimation vector is calculated.
The step S204 further includes:
judging whether the base line deviation is abnormal or not according to the base line deviation calculated in the step S203; if the judgment shows that the abnormality occurs, outputting an online abnormality detection result in an early warning mode; if the judgment result shows that no abnormity occurs, the method returns to the step S201, and continues to monitor the next sampling point in real time.
The method further comprises step S104:
and verifying the abnormal inspection precision of the health baseline constructed by the memory matrix according to the parameter historical records.
The invention has the following beneficial effects:
the method adopts the off-line health baseline construction facing the monitoring process and the on-line abnormity detection based on the baseline deviation, the off-line process considers the stability of the working condition and the comprehensiveness of the monitoring state, the on-line process utilizes the unsupervised learning and detection method of AAKR, the construction method is simple (no large amount of samples are needed), the calculated amount is small, the model is sensitive, the abnormity detection precision is high, and the engineering realization is facilitated.
The method can quickly and timely find the abnormal trend of each parameter before threshold value alarming, effectively avoid the over-limit fault and even further occurrence of accidents through an early warning mode, particularly for the current situations of high navigation safety requirements and weak navigation guarantee capability of ocean-going ships, can be used for effectively evaluating the safety performance of a low-speed diesel engine in actual navigation of the ships, greatly reducing faults and accidents, improving the ocean guarantee capability of a host, and is used for supporting the achievement of the ultimate goal of near-zero fault operation.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
The drawings are only for purposes of illustrating particular embodiments and are not to be construed as limiting the invention, wherein like reference numerals are used to designate like parts throughout.
Fig. 1 is a flowchart of a method for detecting an online anomaly of a marine low-speed diesel engine based on a baseline deviation according to an embodiment of the present invention;
FIG. 2 is a schematic illustration of a stable condition of an embodiment of the present invention;
FIG. 3 is a diagram illustrating comparative verification results according to an embodiment of the present invention;
FIG. 4 is an abnormal view of a cooling system according to an embodiment of the present invention.
Detailed Description
The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form a part hereof, and which together with the embodiments of the invention serve to explain the principles of the invention.
According to a specific embodiment of the invention, the invention discloses a baseline deviation-based online anomaly detection method for a low-speed diesel engine of a ship, which comprises the following steps:
and constructing a health baseline facing the monitoring process, namely acquiring healthy stable working condition data by cleaning historical monitoring data, and constructing the health baseline of the monitoring process of the ship low-speed diesel engine in a memory matrix form.
And (3) online anomaly detection based on baseline deviation, namely, calculating the deviation degree of the baseline and the healthy baseline in real time under the condition of stable working conditions by acquiring online monitoring data, thereby performing online anomaly detection and timely sending an anomaly early warning result to the anomaly so as to avoid further degree faults and even accidents of the low-speed diesel engine of the ship.
In particular, the amount of the solvent to be used,
the health baseline configuration facing the monitoring process comprises the following steps:
step S101, obtaining fault-free data from historical records of monitoring parameters of low-speed diesel engine of ship
For a ship low-speed diesel engine, a multi-parameter monitoring system is usually used for monitoring the state of the ship low-speed diesel engine in a running process in real time, and the monitoring key is whether to detect the abnormality of a monitoring parameter timely and accurately so as to judge whether the low-speed diesel engine has potential safety hazards or faults. The general online monitoring system for the low-speed diesel engine generally comprises parameter monitoring for a plurality of parts such as a cylinder system, a fuel system, a lubricating oil system, a cooling system, a supercharger system, a main bearing and the like. In the aspect of data source selection, the comprehensive performance of working condition data and data ranges is considered in the monitoring of the state parameters of the low-speed diesel engine, and objective safety states corresponding to different working condition operations (namely rotating speed states) in the monitoring process can be objectively reflected.
The observation vector X of the low-speed diesel engine monitoring system comprising a plurality of monitoring parameters is uniformly described as follows:
X(k)＝[s(k)，x1(k)，x2(k)，…，xm(k)]T(1)
in the formula, s (k) is the rotating speed of the kth monitoring sampling point and represents the current working condition of the low-speed diesel engine, and xi(k) I is 1, and m is the ith monitoring parameter and the number of the monitoring parameters.
The low-speed diesel engine online monitoring system stores the observation vector X into a historical record, and also stores the health and non-health states obtained by comparing the observation vector X with a threshold value. In a preferred embodiment, in the process that the low-speed diesel engine online monitoring system stores the observation vector X into the history record, the observation vector X is filtered to filter out obvious error data.
Acquiring all fault-free data in the historical time length T in the historical record of the ship low-speed diesel engine monitoring system, and recording the fault-free data as N health samples, namely health observation vectors, wherein an observation vector X at the nth monitoring sampling point can be described as follows according to (1):
X(n)＝[s(n)，x1(n)，x2(n)，…，xm(n)]T，n＝1，...，N (2)
step S102, selecting stable working condition data from the fault-free data
In this embodiment, stable operating condition data among N healthy samples is selected by using an ADF (extended dicky-Fuller) inspection method.
The rotating speed of the low-speed diesel engine of the ship is a judgment basis of the operating condition, so that the key point of judging stable condition data is judging whether the current rotating speed time sequence is a stable process, and the ADF inspection is a typical unit root inspection method for judging the stable process.
For N healthy samples, at the nth monitoring sampling point, a fixed sampling time series window N0Then the current speed time series is described as:
s(ni)＝μ+ρs(ni-1)+u(ni)，ni＝n-n0，...，n，n＞n0(3)
wherein u (n)i) Is a smooth reversible ARMA (autoregressive moving average) process and E (u (n)i) 0); when the assumption | ρ | < 1 is passed through the ADF inspection, the rotation speed time series s (n) in (3) can be determinedi) The method is a stable process, namely the low-speed diesel engine corresponding to time is in a stable working condition running state.
Here, the inspection process for obtaining the ADF inspection based on (3) is:
wherein, selectingTesting statistics to test hypotheses of stationary processes, anIs composed ofIs estimated.
Thus, s (n) can passi) Obtaining a set of all steady-state operating state quantities in N healthy samplesMeanwhile, based on (2) reconstructing an observation vector at the nth monitoring sampling point in the stable running state of the low-speed diesel engine
Wherein,the stable rotation speed at the nth monitoring sampling point is obtained.
Step S103, constructing a healthy baseline in the form of a memory matrix
Utilizing all steady state observation vectors in N healthy samplesAnd constructing a 'memory matrix' for recording all historical health states as a health state reference (also called a health baseline, and all called health baseline or baseline without special description below) of the monitoring process of the low-speed diesel engine.
Assuming that M stably operating and fault-free observation vectors exist in N health samples, the memory matrixCan be configured based on (5) as follows:
when the scale of M historical observation vectors can cover typical stable rotating speed (in the embodiment, 5RPM is used as an interval to cover the normal rotating speed range of sailing 60-90RPM, and each rotating speed interval is not less than 10 samples), the memory matrixA healthy baseline for the normal operation of the entire low speed diesel engine can be characterized; thus, by monitoring the observation data and the memory matrix while onlineThe real-time comparison can timely and effectively detect whether the monitoring process is abnormal or not.
Further, step S104 may be included to verify the accuracy of the healthy baseline abnormal test constructed by the memory matrix according to the historical data. The calculation difference between the historical data and the healthy baseline constructed by the memory matrix is calculated, and if the mean value of each parameter exceeds 10 percent of the mean value of each parameter of the historical data, the detection precision is not satisfied, the number of healthy samples is adjusted, so as to adjust the sample amount under the condition of stable rotating speed, namely, the number of vectors in the memory matrix is adjusted, so as to improve the precision.
Online anomaly detection based on baseline deviations
The method comprises the following specific steps:
step S201, acquiring real-time monitoring data
Through the low-speed diesel engine monitored control system of boats and ships, obtain observation vector X at the kth real-time supervision sampling point department, can describe as according to (1):
X(k)＝[s(k)，x1(k)，x2(k)，…，xm(k)]T(7)
step S202, judging stable working conditions
Fixed sampling time series window k0In step S102The ADF inspection method (4) of (1), judging the current time series s (k) of the rotation speedi)＝μ+ρs(ki-1)+u(ki)，ki＝k-k0，...，k，k＞k0Whether it is a stationary process:
if the current rotating speed time sequence is not stable, returning to the step S201, and continuing to the (k + 1) th real-time monitoring sampling point;
if the current rotating speed time sequence is stable, the time sequence can pass through s (k)i) Obtaining the first k stable working condition running state quantity setsMeanwhile, reconstruction of observation vectors at the kth monitoring sampling point in the stable running state of the low-speed diesel engine is carried out based on the step (7)
Wherein,the current stable rotation speed.
Due to the characteristic of the control process of the electric control system of the low-speed diesel engine, the transition process caused by the adjustment of the control system is carried out on the change of the working condition (namely the rotating speed) of the low-speed diesel engine every time, namely, a certain time is usually needed to stabilize the working condition, and the monitoring parameter state in the transition process cannot represent an objective state, so that the stable working condition can be effectively judged and the transition process can be eliminated by adopting a unit root method (ADF inspection), so that the accuracy and the effectiveness of the evaluation and the abnormal detection result can be ensured.
Step S203: real-time baseline deviation calculation based on the AAKR method. An Autocorrelation Kernel Regression (AAKR) model is a multivariate-based recursive kernel Regression method, and because of its non-parametric and data-driven modeling manner, historical non-fault (or called health state) observation data (i.e., the memory matrix constructed in step S103) can be better utilized in the present patent to calculate the baseline deviation between the real-time monitored output data and the memory matrix structure, so as to determine whether the real-time monitored output data is in a healthy or abnormal state.
The AAKR method is an unsupervised learning method, only a memory matrix with comprehensive working conditions is used as a historical sample, the calculation process is simple, the on-line calculated amount is small, the engineering is applicable, namely, in the on-line process, the observed quantity under the current stable working condition is directly used for calculating the healthy baseline estimated quantity (namely, if the observed quantity is not abnormal, the deviation between the estimated quantity and the observed quantity is not more than 10%), the abnormal detection result can be output on line through the calculation of the deviation between the observed quantity and the estimated quantity, the accuracy is high, ship operators can be effectively guided to pay attention to the abnormal part of the diesel engine as soon as possible, and corresponding safety measures are taken to avoid further occurrence of fault events.
The specific method comprises the following steps:
recording observation vector at kth monitoring sampling point and stable working conditionIs composed ofFirst, it needs to calculateAnd a memory matrix constructed by (6)The Euclidean distance (Euclidean distance) of each memory vector in (A) is as follows:
thereby, for the firstNew observation vectors at k monitoring sampling pointsThe M × 1 distance vector d (k) ═ d can be calculated1(k)，d2(k)...，dM(k)]T。
The distance vectors may then be converted to similarity weights using a Gaussian Kernel function (Gaussian Kernel) calculation
Wherein h is the kernel bandwidth (k), and w (k) is the k-th monitoring sampling point in the memory matrixMedium M × 1 similarity weight vector. According to the weight vector w (k), it can pass through the memory matrix(i.e., healthy baseline) to obtain a health state estimate vector at the kth monitoring sample point
The sum of the weights is denoted as a scalarThen (11) can be converted into a more compact matrix form:
thenAndthe calculated difference of (c) is recorded as the baseline deviation at the kth monitored sample point.
Step S204: and outputting an online abnormity detection result based on the baseline deviation.
Baseline deviation at the kth monitor sample point ofAndthe calculated difference is based on the 0.1 confidence coefficient commonly used in engineering statistics, ifAndin the calculated difference, the mean value of each parameter exceeds the baseline estimateIf the average value of each parameter is 10%, the parameter is considered to be abnormal, namely, the online abnormal detection result is output in an early warning mode; otherwise, the monitoring is considered to be normal, the step S201 is returned, and the k +1 th real-time monitoring sampling point is continued.
The monitoring system of the diesel engine comprises typical monitoring parameters such as 6 cylinder exhaust temperature and piston cooling oil outlet temperature (a cylinder system), fuel inlet temperature and pressure (a fuel system), lubricating oil inlet temperature (a lubricating oil system), scavenging temperature before and after an air cooler, inlet and outlet cooling water temperature (a cooling system) of the air cooler, turbocharger rotating speed and inlet and outlet exhaust temperature (a supercharger), and lubricating oil temperature of a middle shaft bearing and rear bearing temperature (a main bearing) of a tail pipe.
For the comprehensiveness of the healthy baseline construction, the data from 2015/08/02 to 2015/08/09 during the pilot voyage are utilized to carry out the healthy baseline construction (including the construction of an extraction and memory matrix of stable working conditions) of the low-speed diesel engine, and the main engine rotating speed range of the data is 60RMP to 90RMP, and all working conditions of the ship during normal voyage are covered. Wherein 2015/08/0514 is extracted: the steady state process during the period 00-2015/08/0616: 00 is shown in figure 2 of the accompanying drawings:
the accuracy of the healthy baseline anomaly test constructed from the memory matrix is verified as follows:
the healthy baseline accuracy estimate is made using the voyage data during the time period of the bulk carrier 2015/09/06-2015/09/15 (which is trouble-free during the voyage period and has a range of speeds from 70RPM to 85 RPM). At the kth monitoring sampling point, the navigation data vector is recorded asUsing (9) - (12) health state estimate vectors calculated from healthy baselineThe verification results of comparing the navigation data with the health state estimation quantity at 2500 monitoring sampling points during the navigation are shown in the attached figure 3.
According to calculation, the average value of all system health state estimation errors calculated by the method is 3.51% (the maximum error does not exceed 7.82%), which indicates that the health baseline abnormal detection constructed by the memory matrix has acceptable accuracy and can be used for online abnormal detection.
Based on the verified healthy baseline, the method is continuously used for carrying out abnormity detection on the navigation data of the bulk carrier 2015/09/15-2015/09/20, and the abnormity of the cooling system is detected at 2015/09/1621: 32:00 as shown in FIG. 4.
Through an online abnormity detection method, the fact that related parameters (the temperature of cooling water at an inlet and an outlet of an air cooler) of a cooling system deviate from a baseline of 10% is timely found on the premise of stable rotating speed. Therefore, the low-speed diesel engine cooling system abnormity early warning result is output to avoid further fault events.
In summary, the embodiment of the invention provides an online anomaly detection method for a low-speed diesel engine of a ship based on baseline deviation.
The method adopts the off-line health baseline construction facing the monitoring process and the on-line abnormity detection based on the baseline deviation, the off-line process considers the stability of the working condition and the comprehensiveness of the monitoring state, the on-line process utilizes the unsupervised learning and detection method of AAKR, the construction method is simple (no large amount of samples are needed), the calculated amount is small, the model is sensitive, the abnormity detection precision is high, and the engineering realization is facilitated.
The method can quickly and timely find the abnormal trend of each parameter before threshold value alarming, effectively avoid the over-limit fault and even further occurrence of accidents through an early warning mode, particularly for the current situations of high navigation safety requirements and weak navigation guarantee capability of ocean-going ships, can be used for effectively evaluating the safety performance of a low-speed diesel engine in actual navigation of the ships, greatly reducing faults and accidents, improving the ocean guarantee capability of a host, and is used for supporting the achievement of the ultimate goal of near-zero fault operation.
Those skilled in the art will appreciate that all or part of the flow of the method implementing the above embodiments may be implemented by a computer program, which is stored in a computer readable storage medium, to instruct related hardware. The computer readable storage medium is a magnetic disk, an optical disk, a read-only memory or a random access memory.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention.
Claims (7)
1. A ship low-speed diesel engine online anomaly detection method based on baseline deviation is characterized by comprising the following steps:
s101, acquiring all fault-free data within a historical time length T in a historical record of a ship low-speed diesel engine monitoring system, and recording the fault-free data as N health samples, wherein each sample comprises a plurality of monitoring parameters;
s102, selecting stable working condition data from the acquired fault-free data;
s103, constructing a health baseline in a memory matrix form according to the stable working condition data, and using the health baseline as a health state reference of the low-speed diesel engine monitoring system;
step S201, acquiring real-time monitoring data from a ship low-speed diesel engine monitoring system;
s202, judging the stable working condition of the acquired real-time monitoring data;
step S203, calculating the baseline deviation of the real-time monitoring data under the stable working condition;
and step S204, outputting an online abnormity detection result based on the baseline deviation.
2. The baseline deviation-based online anomaly detection method for the marine low-speed diesel engine according to claim 1, wherein the step S102 further comprises:
and judging whether the current rotating speed time sequence is a stable process, namely whether the low-speed diesel engine corresponding to the time is in a stable working condition running state, and selecting stable working condition data from the N healthy samples.
3. The online anomaly detection method for the marine low-speed diesel engine based on the baseline deviation as claimed in claim 1, wherein the step S103 further comprises:
and constructing a health baseline in a memory matrix form for recording all historical health states by using all observation vectors X of the stable running states in the N health samples, wherein the health baseline is used as a health state reference in the monitoring process of the low-speed diesel engine.
4. The online anomaly detection method for the marine low-speed diesel engine based on the baseline deviation as claimed in claim 1, wherein the step S202 further comprises:
judging whether the rotating speed time sequence of the current real-time monitoring sampling point data is a stable process:
if the current rotating speed time sequence is not stable, no abnormal detection is carried out, and the step S201 is returned;
and if the current rotating speed time sequence is stable, obtaining a stable working condition running state quantity set, and reconstructing an observation vector at the current monitoring sampling point under the stable running state of the low-speed diesel engine.
5. The baseline deviation-based online anomaly detection method for the marine low-speed diesel engine according to claim 1, wherein the step S203 further comprises:
the health base line in the form of the memory matrix constructed in step S103 is used to obtain the health state estimation vector at the current real-time monitoring sampling point, and the base line deviation between the real-time monitored health state observation vector and the health state estimation vector is calculated.
6. The baseline deviation-based online anomaly detection method for the marine low-speed diesel engine according to claim 1, wherein the step S204 further comprises:
judging whether the base line deviation is abnormal or not according to the base line deviation calculated in the step S203; if the judgment shows that the abnormality occurs, outputting an online abnormality detection result in an early warning mode; if the judgment result shows that no abnormity occurs, the method returns to the step S201, and continues to monitor the next sampling point in real time.
7. The baseline deviation-based online anomaly detection method for the marine low-speed diesel engine according to claim 1, further comprising the step S104 of:
and verifying the abnormal inspection precision of the health baseline constructed by the memory matrix according to the parameter historical records.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610959364.1A CN106368816B (en) | 2016-10-27 | 2016-10-27 | A kind of online method for detecting abnormality of marine low speed diesel engine based on baseline offset |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610959364.1A CN106368816B (en) | 2016-10-27 | 2016-10-27 | A kind of online method for detecting abnormality of marine low speed diesel engine based on baseline offset |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106368816A CN106368816A (en) | 2017-02-01 |
| CN106368816B true CN106368816B (en) | 2018-09-25 |
Family
ID=57893781
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610959364.1A Active CN106368816B (en) | 2016-10-27 | 2016-10-27 | A kind of online method for detecting abnormality of marine low speed diesel engine based on baseline offset |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106368816B (en) |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106991145B (en) * | 2017-03-23 | 2021-03-23 | 中国银联股份有限公司 | Data monitoring method and device |
| CN107909677B (en) * | 2017-11-28 | 2020-08-07 | 中国船舶工业系统工程研究院 | Method for extracting and judging stability of operation data of ship electromechanical equipment |
| CN108062586B (en) * | 2017-11-30 | 2020-03-27 | 中国船舶工业系统工程研究院 | Ship host machine associated component state monitoring method and system based on decline contribution degree |
| CN110874088B (en) * | 2018-09-03 | 2022-12-23 | 中国船舶集团有限公司第七一一研究所 | Monitoring method of ship key equipment system based on multi-dimensional vector model |
| CN109307553B (en) * | 2018-09-27 | 2020-10-09 | 潍柴重机股份有限公司 | Main bearing maintenance prompting method based on main bearing temperature deviation |
| CN111289256B (en) * | 2018-12-07 | 2021-04-30 | 北京大学 | Data-driven marine diesel engine fault detection method |
| CN110071913B (en) * | 2019-03-26 | 2020-10-02 | 同济大学 | Unsupervised learning-based time series anomaly detection method |
| CN112801434A (en) * | 2019-11-13 | 2021-05-14 | 中兴通讯股份有限公司 | Method, device, equipment and storage medium for monitoring performance index health degree |
| CN111563685B (en) * | 2020-05-09 | 2022-03-08 | 国网江苏省电力有限公司 | Power generation equipment state early warning method based on auto-associative kernel regression algorithm |
| CN111582274B (en) * | 2020-05-11 | 2021-05-18 | 南京航空航天大学 | Wire wiring virtual guiding method based on marker |
| CN113833567B (en) * | 2020-06-23 | 2022-07-26 | 北京大学 | Fault detection method for diesel engine turbocharger with fused mechanical data |
| CN111881176B (en) * | 2020-07-07 | 2021-09-07 | 中国人民解放军海军工程大学 | Marine nuclear power anomaly detection method based on logical distance characterization safety operation domain |
| CN112504505A (en) * | 2020-08-31 | 2021-03-16 | 中国能源建设集团安徽省电力设计院有限公司 | High-voltage tunnel cable overheating early warning method based on multivariate state estimation |
| CN112208728B (en) * | 2020-09-30 | 2021-11-09 | 震兑工业智能科技有限公司 | Ship fouling monitoring and early warning method |
| CN112610330B (en) * | 2020-12-08 | 2023-05-09 | 孚创动力控制技术（启东）有限公司 | Monitoring and analyzing system and method for running state of internal combustion engine |
| CN115711183B (en) * | 2022-10-31 | 2024-11-08 | 中国船舶集团有限公司第七一一研究所 | Diesel engine health management system, method and readable storage medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102226428A (en) * | 2011-04-29 | 2011-10-26 | 哈尔滨工程大学 | Method for forecasting gas turbine health |
| CN103245912B (en) * | 2013-03-23 | 2016-09-14 | 中国水利电力物资有限公司 | A kind of running of wind generating set state analysis diagnostic method and system |
| CN105043776A (en) * | 2015-08-12 | 2015-11-11 | 中国人民解放军空军勤务学院 | Aircraft engine performance monitoring and fault diagnosis method |
| CN105551549B (en) * | 2015-12-07 | 2018-05-29 | 中广核工程有限公司 | A kind of nuclear power generating equipment operation conditions on-line monitoring method and system |
| CN105928710B (en) * | 2016-04-15 | 2019-02-12 | 中国船舶工业系统工程研究院 | A kind of diesel engine fault monitoring method |
-
2016
- 2016-10-27 CN CN201610959364.1A patent/CN106368816B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN106368816A (en) | 2017-02-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106368816B (en) | A kind of online method for detecting abnormality of marine low speed diesel engine based on baseline offset | |
| EP2487553B1 (en) | Turbine fault analysis | |
| CN108062586B (en) | Ship host machine associated component state monitoring method and system based on decline contribution degree | |
| CN109469896B (en) | Industrial boiler fault diagnosis method and system based on time series analysis | |
| US6587737B2 (en) | Method for the monitoring of a plant | |
| CN111400961B (en) | Wind generating set blade fault judging method and device | |
| KR102206737B1 (en) | Method and system for fault diagnosis of pump-turbines based on machine learning | |
| CN109783903A (en) | A kind of industrial water pipeline fault diagnostic method and system based on time series | |
| CN114239377A (en) | Method and system for evaluating health state of urban rail electromechanical equipment and storage medium | |
| CN113847950A (en) | Intelligent ship equipment state monitoring system based on cloud computing and information interaction method | |
| CN108801641A (en) | The fault diagnosis of exhaust-driven turbo-charger exhaust-gas turbo charger and Reliability Prediction Method and its system | |
| CN115326147A (en) | Real-time monitoring method and system for desulfurization tower for ship | |
| Kang et al. | Hierarchical level fault detection and diagnosis of ship engine systems | |
| CN114738132B (en) | Supercharger rotation speed sensor fault diagnosis method and system based on real ship data | |
| JP6846896B2 (en) | Analysis of ship propulsion performance | |
| CN108062618A (en) | Low-speed diesel engine Economic Analysis Method and system based on biradical line | |
| KR102457752B1 (en) | Method of real time fault detection and diagnosis for onboard engine room and system for forperming the same | |
| Wei et al. | Multi-sensor monitoring based on-line diesel engine anomaly detection with baseline deviation | |
| CN113187674A (en) | Fault determination method and system for wind turbine generator pitch system | |
| Sun et al. | Anomaly detection of marine diesel engine valve system based on Mahalanobis distance | |
| KR102212022B1 (en) | Method of automatically determining condition of hydro turbine in hydroelectric power plant and system for the same | |
| CN113833567B (en) | Fault detection method for diesel engine turbocharger with fused mechanical data | |
| Feng et al. | Degradation State Assessment Modeling Using Causality Discovery | |
| Vuong et al. | Anomaly Detection and Breakdown Diagnosis for Condition Monitoring of Marine Engines | |
| CN115165044A (en) | Fuel flow abnormity identification method and device and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 |