CN107025382B - A kind of engineering system health analysis system and method based on critical phase transformation theory - Google Patents
A kind of engineering system health analysis system and method based on critical phase transformation theory Download PDFInfo
- Publication number
- CN107025382B CN107025382B CN201710300213.XA CN201710300213A CN107025382B CN 107025382 B CN107025382 B CN 107025382B CN 201710300213 A CN201710300213 A CN 201710300213A CN 107025382 B CN107025382 B CN 107025382B
- Authority
- CN
- China
- Prior art keywords
- signal
- analysis
- critical phase
- phase transformation
- engineering
- 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
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The engineering system health analysis system and method based on critical phase transformation theory that the present invention relates to a kind of, the system includes: signal acquisition unit, signal analysis unit and interactive display unit three parts, and this method comprises the following steps that signal acquisition, Signal Pretreatment, random fluctuation signal extraction, critical phase transformation pre-warning signal index analysis and signal analysis are shown.The present invention will be in the critical phase-change characteristic application project system of engineering system, new point of penetration is provided for engineering system health analysis, the transition of the system mode present invention can reflect thrashing comprehensively when, realize the diagnosis and prediction of the system failure, have preferable applicability.
Description
Technical field
The present invention relates to the health analysis fields to engineering system, and in particular to a kind of engineering based on critical phase transformation theory
System health analysis system and method.
Background technique
Currently, the research of engineering system health analysis is broadly divided into three aspects: fault diagnosis, failure predication and health pipe
Reason.In terms of fault diagnosis, existing research carries out fault diagnosis to Mechatronic Systems using the Intelligent Decision-making Method based on model, together
When also have research use case reasoning (Case-based reasoning, abbreviation CBR) intelligent method, such as tasted in air line
The scheme and genetic algorithm for pinging case-based reasioning carry out fault diagnosis to Boeing-747-700；In terms of failure predication, have
Research carries out forecast analysis to the failure of system using neural-fuzzy technology, also have part research predicted based on neural network be
Failure of uniting and assessment equipment remaining life, hydroelectric system has at present makes some taste in terms of neural network prediction failure
Examination, suspension mechanical system is analyzed by using the decision making approach based on model and the remaining life of assessment system component,
There is research to be absorbed in the service life of the method evaluation means driven using statistical data；In terms of health control, there is research to make
Health control is carried out to engineering system with machine learning and Methods of Knowledge Discovering Based, aircraft control system is by using based on model
Decision-making technique is attempted to carry out health control to system.
However, the studies above method is confined to single component or single change to the fault diagnosis of engineering system and prediction mostly
The state-detection and fault diagnosis of amount are associated with increasingly between each component since component and number of states are numerous in engineering system
Complexity, the fault diagnosis of component level are not easy to reflect system level state transition comprehensively.And health analysis system knows mechanism
It is larger to know dependency degree, and is difficult to use in real-time prediction.There is a small amount of research to make trial in terms of system-level health control at present, such as
For some researchers using the integrated approach of Bayesian network design early warning system safe condition, which is to answer with gas burning system
With, not only the safety of component is analyzed, is also made prediction to system-level hidden danger, obtains certain achievement, but for being
The research of irrespective of size status change and health control is relatively fewer, still in the immature stage.
Complex System Theory is a kind of theoretical foundation of system-level specificity analysis of progress emerging in recent years, M Scheffer
The theory that the critical transformation temperature of forecasting system in complication system is proposed on " Nature " causes all circles' concern, complicated at present
It is system-level to carry out that Systems Theory has been applied to economics, ecology, finance, medicine, statistical physics every subjects
Research.Complication system is the system that there is moderate number the intelligence of action, adaptivity main body are made based on local message.In
Usually there is the critical transformation temperature of system in complication system so that system be changed into suddenly at this time point from a state it is another
A state.There is the spontaneous system failure in terms of medicine, such as epilepsy or asthma attack；In financial sector, there are system crashes to show
As；In terms of weather, there is Observed Abrupt Climatic Change.When system closes on critical transformation temperature, system just plays pendulum.
It is published on " Nature " studies have shown that system would generally have some universal when complication system closes on critical transformation temperature
Characteristic: first, when complication system transformation temperature critical close to system, system is restored to normal condition after being interfered
Speed can slow down, i.e. critical slowing down (critical slowing down)；Second, in time series before and after system
State becomes more and more similar；Third, when closing on system critical point, the autocorrelation of system can gradually increase system.It is above-mentioned
Complex System Theory is applied in multiple fields, with the critical transformation temperature of searching system.Such as in terms of meteorology, pass through
Whether the autocorrelation for analyzing weather system enhances reflection system critical slowing down, to judge system close to critical phase transformation
Point, so that prejudging weather occurs change dramatically.In terms of ecology, by monitor Lakes state variance whether increase come
Predict this critical transformation temperature of lake eutrophication.In terms of sociology, there is critical slowing down by analysis system come pre-
It surveys population and collapses this system sudden change state.
But still expect for critical phase allergic effect being used in engineering system without researcher at present, this is because critical phase
Become theoretical mainly to apply to ecology, environment, society, finance etc. and has in the system of complex systematic dynamics characteristic, in these all
With biggish randomness, and its kinetics equation is difficult to parse description with kinetics equation, and engineering system is because be
Artificially designing, system dynamics is often believed to describe by accurate system dynamics equation, therefore, in view of
Engineering system causes related fields scientific research personnel not expect managing critical phase transformation with the huge difference of system above in form
By apply to engineering system health analysis research in.
Summary of the invention
For the transition and health analysis system pair that can not reflect system level state in engineering system health analysis comprehensively
Mechanism knowledge dependency is larger, is difficult to the problem of predicting in real time, and the system-level Stochastic Dynamics of utilizing works system of the present invention are special
Property, it is based on critical phase transformation theory, by engineering system pre-warning signal index analysis, realizing to system-level health analysis theory
Support.
Applicant of the present invention's discovery, although the systems such as engineering system and ecology, environment, society, finance have in form
There is huge difference, but critical phase transformation theoretical origin still has biggish applicability in engineering system, and can preferably carry out engineering
System health analysis.
Specifically, the present invention provides a kind of engineering system health analysis system based on critical phase transformation theory, feature
It is, the system comprises such as lower units:
Signal acquisition unit, the signal acquisition unit work for acquiring in the engineering system with the engineering system
The relevant signal with timing of state；
Signal analysis unit, the signal analysis unit are used to carry out pre-processing to signal collected and based on pretreatment
Signal afterwards carries out critical Phase Transition Analysis.
In a kind of preferred implementation, the signal analysis unit further include: interactive display unit, the interactive display
Unit is used to export or show the result of the critical Phase Transition Analysis.
In another preferred implementation, the signal analysis unit is used to carry out based on pretreated signal and institute
Engineering system is stated to fail the purification of relevant random fluctuation signal.
In another preferred implementation, carrying out pretreatment to signal collected includes carrying out to signal collected
Resampling.
In another preferred implementation, the random fluctuation purifying signal process includes: that removal thrashing is unwise
Trend signal under sense state and working state of system, only extraction system fails relevant random fluctuation signal.
On the other hand, the present invention provides a kind of engineering system health analysis method based on critical phase transformation theory, feature
It is, described method includes following steps:
1) signal in the engineering system with timing is acquired,
2) signal collected is pre-processed,
3) critical Phase Transition Analysis is carried out based on pretreated signal.
In another preferred implementation, the method also includes: export or show the result of critical Phase Transition Analysis.
In a kind of preferred implementation, the method also includes: it is carried out and the engineering based on pretreated signal
The purification of the relevant random fluctuation signal of thrashing.
In another preferred implementation, carrying out pretreatment to signal collected includes carrying out to signal collected
Resampling.
In another preferred implementation, the random fluctuation purifying signal process includes: that removal thrashing is unwise
Trend signal under sense state and working state of system, only extraction system fails relevant random fluctuation signal.Institute in the present invention
The method or apparatus that the signal acquisition unit mentioned can acquire timing signal with all carry out signal acquisitions, including but
Be not limited to manual measurement or install sensing device subsidiary additional etc..It should be noted that timing letter mentioned in the present invention
It number refers to: the sensor signal acquired according to time series.It is any to have the characteristics that the signal of timing, there is variation at any time
The signal of property is intended to be included within, such as: the operation of vibration signal, train collected in bearing rotation process
Axis temperature transducing signal in the process.
It should be noted that the insensitive state of removal failure, which refers to, analyzes coherent signal based on system features in the present invention
Lesser signal will be influenced on thrashing to reject.Trend signal under working state of system can be asked by trend learning method
It takes and removes system integrality trend.The trend learning method can be selected but be not limited to Gauss curve fitting mode, specifically can base
Fitting learning method is selected in system features.In the present invention, the critical phase transformation pre-warning signal analysis: refer to that selection suitably refers to
Mark indicates critical phase transition phenomena.It mainly characterizes including but not limited to critical slowing (Critical Slowing Down) and symmetrical
It deviates (Deviating Skewness).Wherein critical slowing down mainly includes auto-correlation (Autocorrelation) and becomes
Different (Variance) index signal enhancing.
The beneficial effects of the present invention are:
(1) present invention jumps out at present to the traditional approach of health analysis used by engineering system, is conceived to complicated electromechanics
The Stochastic Dynamics characteristic of system carries out health analysis research with new system-level perspective in research.It is a kind of for engineering system
General-purpose system and method.
(2) present invention will be originally it is believed that the critical phase transformation theory for being not particularly suited for engineering system be applied to engineering system
Health analysis.Critical phase transformation theory is applied in engineering system, new point of penetration is provided for engineering system health analysis；Meaning
Good analytical effect is had received outside.The present invention can reflect the transition of system level state comprehensively, realize prediction in real time, and have
Preferable applicability.
Detailed description of the invention
Fig. 1 is the schematic architectural diagram of engineering system health analysis system of the invention；
Fig. 2 is the implementation flow chart of engineering system health analysis method of the invention；
Fig. 3 is the collector signal figure of insulated gate bipolar transistor used in engineering system；
Fig. 4 is random fluctuation purifying signal process schematic in the present invention, and wherein a, b, c respectively represent removal thrashing
The process of insensitive state removes the process and the relevant random fluctuation signal of finally obtained thrashing of trend signal；
Fig. 5 is the critical phase transformation pre-warning signal index of insulated gate bipolar transistor in engineering system；
Fig. 6 is the vibration signal of the bearing in engineering system；
Fig. 7 is critical phase transformation pre-warning signal index example in the present invention；
Fig. 8 is the schematic flow chart of the analysis method in the embodiment of the present invention 3.
Specific embodiment
Based on critical phase transformation theory engineering system health analysis system including such as lower unit in the present embodiment: signal acquisition list
Member, signal analysis unit, interactive display unit.
Wherein signal acquisition unit mainly uses all methods that can acquire clock signal, and such as manual measurement installs additional
Sensing device subsidiary etc. obtains any signal with timing from engineering system.Signal analysis unit is mainly completed
The pretreatment of data and critical Phase Transition Analysis work.Interactive display unit mainly opens up critical phase change signal analysis result
Show.
Embodiment 1
Included the following steps: in the present embodiment based on critical phase transformation theory engineering system health analysis method
1) engineering system signal acquisition
Health analysis of the invention is mainly based upon the health analysis of physical signal, is based especially on the object that sensor measures
Manage the analysis that signal carries out.In engineering system, for example, in order to measure the revolving speed of aero-engine or pressure；Driving motor
Current or voltage, insulated gate bipolar transistor can be used.
Here, the detailed description of signal processing is only carried out with insulated gate bipolar transistor.It is brilliant to obtain insulated gate bipolar
Collector current (collector current) signal of body pipe (Insulated Gate Bipolar Translator), is obtained
The signal taken is as shown in Fig. 3.Signal by analyzing the transistor can analyze the strong of the engineering system where the transistor
Health state.
2) Signal Pretreatment,
The step is mainly divided into two parts, and first part is to obtain time series data by processing, and second part is random fluctuation
Purifying signal.
Sensing data can usually have inconsistent, missing and duplicate number because being influenced by a variety of environmental factors
According to.In order to the time series data of time intervals such as obtain, we remove first is repeated and inconsistent data, on this basis to lacking
It loses data and carries out interpolation, input of the resampling as random fluctuation signal extraction process finally is carried out to the data of interpolation.
Assuming that we have clock signal { x1,x2,x3,…,xn, respectively correspond moment { t1,t2,t3,…,tnObservation.
Any missing values present in clock signal are expressed as (tk,xk).We can be used linear interpolation (1) and calculate missing values.It can also
To use p order polynomial to be returned to obtain (2) to known clock signal, then missing values xk=f (tk).We can also make
With p spline interpolation, it is necessary first to which clock signal is divided into disjoint segmentation, s1,s2,s3,…sm, for each segmentation
Recurrence learning is carried out using p order polynomial and obtains (3), then missing values xk=g (tk).
(ta,xa),(tb,xb) it is missing point (tk,xk) before and after two adjacent points of observation.
Unlike signal is different to thrashing sensibility, some signals even can not sensory perceptual system failure, therefore in order to mention
Rise the intensity of critical phase transformation pre-warning signal, the modes such as priori knowledge, expertise and the correlation analysis of this example based on system
The relevant random fluctuation purifying signal of thrashing is carried out to the signal of acquisition.Decomposability based on time series data, Wo Menli
Signal is decomposed with effective method and extracts the random fluctuation signal highly relevant with system stability.
Random fluctuation purifying signal mainly includes the trend removed under the insensitive state of thrashing and working state of system
Signal, to extract the relevant random fluctuation signal of thrashing as analysis content.
Wherein the insensitive state of removal failure is mainly based upon system features analysis coherent signal and will influence on thrashing
Lesser signal is rejected.It is from as shown in Fig. 3 in signal shown in attached drawing 4a by taking the signal of insulated gate bipolar transistor as an example
Interception a segment, we using the relevant priori knowledge of visual analyzing combination system determine working condition number (such as Fig. 3,
4a, signal work in two operating conditions, high operating conditions and low operating conditions), and identify the operating conditions sensitive to thrashing (such as
Fig. 4 a, 4b, the low operating conditions of signal is sensitive state) therefore, the preprocessing module of signal analysis unit can be in high work
The signal of state is removed, and leaving the sensitive low operating conditions signal of failure is to scheme signal shown in b.
The trend signal under removal working state of system is also needed to sensitive signal, that is, rejects the son for meeting a certain given rule
For signal to obtain residual error, residual error herein namely reflects the random fluctuation signal of system stability.For regular subsignal
Extraction, general method, such as polynomial regression (formula 2), Gauss kernel regression (formula 4), autoregression can be used in we
The methods of (formula 5) obtains to be fitted clock signal.In addition, we can also use the relevant method of system, base in this example
Regular subsignal is obtained in expertise and meets equation (6), and based on this, we can be used (6) and extract regular subsignal (such as
Fig. 4 b solid line) and reject it to obtain residual error (such as Fig. 4 c) from original signal.Signal shown in Fig. 4 c is final after purifying
Random fluctuation signal and the basis that will be calculated as warning index.
K is the kernel function that bandwidth (bandwith) is h, and x is input, and y is output
yt=AR (p)=α0+α1yt-1+α2yt-2+…+αt-pyt-p (5)
yt=exp (α1t+α0) (6)
3) critical phase transformation pre-warning signal index analysis,
The step is realized by signal analysis unit.In this step, suitable index is selected to indicate critical phase transition phenomena.
Available characterization includes but is not limited to critical slowing (Critical Slowing Down) and symmetrical deflection (Deviating
Skewness).Wherein critical slowing down mainly includes auto-correlation (Autocorrelation) and variation (Variance) index
Signal enhancing.Fig. 5 gives critical phase transformation pre-warning signal index in this example.
We use xt, t=1,2 ..., n indicates the signal value of t moment.The equal of clock signal can be calculated according to (7) formula
Value, (8) formula calculate variance (variance), and (9) formula calculates the auto-correlation (lag- at 1 time point of lag
1autocorrelation), (10) formula calculates deflection (skewness).It should be noted that index herein is based on cunning
The form calculus of dynamic window.
Phenomena such as critical phase transformation index analysis can be carried out based on formula (7)-(10), and then find critical slowing.
For insulated gate bipolar transistor signal shown in Fig. 5 critical phase transformation pre-warning signal index analysis it is found that working as
Variation (Variance) index and auto-correlation (Autocorrelation) index become larger suddenly, that is, critical slowing down occur；When
Occur critical slowing (Critical Slowing Down) phenomenon and symmetrical deflection (Deviating Skewness) phenomenon simultaneously
When, that is, illustrate critical phase transformation occur, system can provide warning.
4) signal analysis is shown
The step is realized by signal analysis display unit, in the step, mainly according to the critical phase change signal of engineering system
Analysis is as a result, health analysis prediction and early warning for engineering system provide warning.
Embodiment 2
Included the following steps: in the present embodiment based on critical phase transformation theory engineering system health analysis method
1) engineering system signal acquisition
This example uses all methods that can acquire clock signal, including but not limited to manual measurement or installation sensing dress
Subsidiary is set, the vibration signal (vibration signal) of engineering system middle (center) bearing (bearing), the signal of acquisition are obtained
As shown in Fig. 6.
2) Signal Pretreatment,
Since the signal obtained in this example is excessively intensive, the signal obtained in this example is carried out in this step
Resampling.
3) critical phase transformation pre-warning signal index analysis,
In this step, suitable index is selected to indicate critical phase transition phenomena.Available characterization includes but is not limited to face
Boundary's slowing down (Critical Slowing Down) and symmetrical deflection (Deviating Skewness).Wherein critical slowing down
It mainly include auto-correlation (Autocorrelation) and variation (Variance) index signal enhancing.
Fig. 7 gives critical phase transformation pre-warning signal index in this example.
In the present embodiment, according to the critical phase transformation pre-warning signal index analysis of bearing signal shown in fig. 7 it is found that when becoming
Different (Variance) index and auto-correlation (Autocorrelation) index become larger suddenly, that is, critical slowing down occur；When same
When there is critical slowing (Critical Slowing Down) phenomenon and symmetrical deflection (Deviating Skewness) phenomenon
When, that is, illustrate critical phase transformation occur, system can provide warning.
4) signal analysis is shown
In the step, mainly according to the critical phase change signal analysis of engineering system as a result, being the health analysis of engineering system
Prediction and early warning provide warning.
Embodiment 3
Step 3 above is adjusted in the present embodiment, the method for critical Phase Transition Analysis is as shown in figure 8, this point
Analysis method includes the following steps:
1) it, is analyzed, is obtained by typical complex dynamics of mechanic electric system specificity analysis and system sensing signal random noise
The system-level Stochastic Dynamics characteristic of complex electromechanical systems；
2) the critical phase transformation theory of complication system is utilized, the critical phase-change characteristic of complex electromechanical systems Stochastic Dynamics is divided
Analysis；
3), on the basis of complex electromechanical systems Stochastic Dynamics attributive analysis, in conjunction with the complicated machine based on Physical Mechanism
Electric system base reference model and system operating signal log recording complete the study of base reference model parameter, from Physical Mechanism angle
Degree modeling, directly to reflect the relationship between system parameters；
4) dynamic graph model, which is based on, in conjunction with system operating signal log recording completes complex electromechanical systems random perturbation parameter
Study, using Dynamic Graph model modeling, preferably to describe relationship and parameter between the parameter changed at random itself because making an uproar
Sound fluctuates caused by existing.System-level Stochastic Dynamics model energy a more complete description based on the building of this two parts is complicated electromechanical
The operating status of system, and the basis of system-level critical phase transformation state model is established later
5) next, using noise excitation simulation and critical phase transformation state evaluation, the depth driven by noise is strong
Change mode of learning and complete critical phase transformation study, constructs system-level critical phase transformation state model.In learning process, pass through setting
System sensing signal parameter value simulates noise excitation, and continuous repetitive exercise makes system actively reach critical transition, leads to
The variation characteristic of overwriting system status parameters data analysis system state parameter, fortune that can be current according to system based on this model
The response trend of row Signal Analysis System, to complete system health status analysis.
Engineering system health analysis system and method provided by the invention based on critical phase transformation theory are based on engineering system
Transducing signal analyze critical phase-change characteristic, construct the critical phase transformation elasticity theory of engineering system Stochastic Dynamics.By to engineering
The research of the critical phase transformation elasticity theory of system Stochastic Dynamics, in the angle research engineering system for providing a kind of new slave system
The method of relationship between all parts provides a new visual angle for the health analysis of engineering system.
Although the principle of the present invention is described in detail above in conjunction with the preferred embodiment of the present invention, this field skill
Art personnel are it should be understood that above-described embodiment is only the explanation to exemplary implementation of the invention, not to present invention packet
Restriction containing range.Details in embodiment is simultaneously not meant to limit the scope of the invention, without departing substantially from spirit of the invention and
In the case where range, any equivalent transformation based on technical solution of the present invention, simple replacement etc. obviously change, and all fall within
Within the scope of the present invention.
Claims (6)
1. a kind of engineering system health analysis system based on critical phase transformation theory, which is characterized in that the system comprises as follows
Unit:
Signal acquisition unit, the signal acquisition unit are used to acquire in the engineering system and the engineering system working condition
The relevant signal with timing；
Signal analysis unit, the signal analysis unit are used to carry out pre-processing to signal collected and based on pretreated
Signal carries out critical Phase Transition Analysis, carries out and the engineering system wherein the signal analysis unit is based on pretreated signal
It fails the purification of relevant random fluctuation signal, the random fluctuation purifying signal process includes: that removal thrashing is insensitive
Trend signal under state and working state of system, only extraction system fails relevant random fluctuation signal,
Wherein, the signal analysis unit carries out critical Phase Transition Analysis by following manner:
3.1) it, is analyzed, is obtained multiple by typical complex dynamics of mechanic electric system specificity analysis and system sensing signal random noise
The system-level Stochastic Dynamics characteristic of miscellaneous Mechatronic Systems；
3.2) the critical phase transformation theory of complication system is utilized, the critical phase-change characteristic of complex electromechanical systems Stochastic Dynamics is divided
Analysis；
3.3) electromechanical in conjunction with the complexity based on Physical Mechanism, on the basis of complex electromechanical systems Stochastic Dynamics attributive analysis
System-based reference model and system operating signal log recording complete the study of base reference model parameter, from Physical Mechanism angle
Modeling, directly to reflect the relationship between system parameters；
3.4) dynamic graph model, which is based on, in conjunction with system operating signal log recording completes complex electromechanical systems random perturbation parametrics
It practises, using Dynamic Graph model modeling, preferably to describe relationship and parameter between the parameter changed at random itself because of noise
It is fluctuated in the presence of caused by；
3.5) next, using noise excitation simulation and critical phase transformation state evaluation, the deeply driven by noise
Mode of learning completes critical phase transformation study, constructs system-level critical phase transformation state model, in learning process, is by setting
System transducing signal parameter value modulo quasi-random noise excitation, continuous repetitive exercise make system actively reach critical transition, pass through
Record the variation characteristic of system status parameters data analysis system state parameter, operation that can be current according to system based on this model
The response trend of Signal Analysis System, to complete system health status analysis.
2. engineering system health analysis system according to claim 1, which is characterized in that the signal analysis unit also wraps
Include: interactive display unit, the interactive display unit are used to export or show the result of the critical Phase Transition Analysis.
3. engineering system health analysis system according to claim 1 or 2, which is characterized in that signal collected into
Row pretreatment includes carrying out resampling to signal collected.
4. a kind of engineering system health analysis method based on critical phase transformation theory, which is characterized in that the method includes as follows
Step:
1) signal in the engineering system with timing is acquired,
2) signal collected is pre-processed,
3) critical Phase Transition Analysis is carried out based on pretreated signal,
Wherein, the method also includes: relevant to engineering system failure random wave is carried out based on pretreated signal
The purification of dynamic signal, the random fluctuation purifying signal process include: the insensitive state of removal thrashing and system work shape
Trend signal under state, only extraction system fails relevant random fluctuation signal,
Wherein, the step 3) includes:
3.1) it, is analyzed, is obtained multiple by typical complex dynamics of mechanic electric system specificity analysis and system sensing signal random noise
The system-level Stochastic Dynamics characteristic of miscellaneous Mechatronic Systems；
3.2) the critical phase transformation theory of complication system is utilized, the critical phase-change characteristic of complex electromechanical systems Stochastic Dynamics is divided
Analysis；
3.3) electromechanical in conjunction with the complexity based on Physical Mechanism, on the basis of complex electromechanical systems Stochastic Dynamics attributive analysis
System-based reference model and system operating signal log recording complete the study of base reference model parameter, from Physical Mechanism angle
Modeling, directly to reflect the relationship between system parameters；
3.4) dynamic graph model, which is based on, in conjunction with system operating signal log recording completes complex electromechanical systems random perturbation parametrics
It practises, using Dynamic Graph model modeling, preferably to describe relationship and parameter between the parameter changed at random itself because of noise
It is fluctuated in the presence of caused by；
3.5) using noise excitation simulation and critical phase transformation state evaluation, the deeply mode of learning driven by noise
Critical phase transformation study is completed, system-level critical phase transformation state model is constructed, in learning process, is believed by setting system sensing
Number parameter value modulo quasi-random noise excitation, continuous repetitive exercise make system actively reach critical transition, pass through record system
The variation characteristic of state parameter data analysis system state parameter can be analyzed based on this model according to the current run signal of system
The response trend of system,
To complete system health status analysis.
5. engineering system health analysis method according to claim 4, which is characterized in that the method also includes: output
Or show the result of critical Phase Transition Analysis.
6. engineering system health analysis method according to claim 4, which is characterized in that carried out to signal collected pre-
Processing includes carrying out resampling to signal collected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710300213.XA CN107025382B (en) | 2017-05-02 | 2017-05-02 | A kind of engineering system health analysis system and method based on critical phase transformation theory |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710300213.XA CN107025382B (en) | 2017-05-02 | 2017-05-02 | A kind of engineering system health analysis system and method based on critical phase transformation theory |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107025382A CN107025382A (en) | 2017-08-08 |
| CN107025382B true CN107025382B (en) | 2019-11-26 |
Family
ID=59526687
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710300213.XA Active CN107025382B (en) | 2017-05-02 | 2017-05-02 | A kind of engineering system health analysis system and method based on critical phase transformation theory |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107025382B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107679733A (en) * | 2017-09-27 | 2018-02-09 | 国网上海市电力公司 | A kind of quantitative estimation method of stabilization of power grids situation |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102430221A (en) * | 2011-10-10 | 2012-05-02 | 山西通威消防电子有限公司 | Device for monitoring operation states of fire-fighting facilities |
| CN105046957A (en) * | 2015-07-02 | 2015-11-11 | 清华大学 | Balanced sampling method for accident analysis and safety assessment |
| CN106033890A (en) * | 2015-08-12 | 2016-10-19 | 华北电力大学（保定） | Space-based early warning method for critical phase transformation of electric power system |
-
2017
- 2017-05-02 CN CN201710300213.XA patent/CN107025382B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102430221A (en) * | 2011-10-10 | 2012-05-02 | 山西通威消防电子有限公司 | Device for monitoring operation states of fire-fighting facilities |
| CN105046957A (en) * | 2015-07-02 | 2015-11-11 | 清华大学 | Balanced sampling method for accident analysis and safety assessment |
| CN106033890A (en) * | 2015-08-12 | 2016-10-19 | 华北电力大学（保定） | Space-based early warning method for critical phase transformation of electric power system |
Non-Patent Citations (1)
| Title |
|---|
| 《基于临界相变的电力系统连锁故障早期预警机理研究》;范佳琪;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20160515;第C042-715页； * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107025382A (en) | 2017-08-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wang et al. | Remaining life prediction method for rolling bearing based on the long short-term memory network | |
| Li et al. | A deep learning driven method for fault classification and degradation assessment in mechanical equipment | |
| Cross et al. | Features for damage detection with insensitivity to environmental and operational variations | |
| CN106006344B (en) | Staircase On-line Fault early warning system and method for diagnosing faults | |
| CN112033463B (en) | Nuclear power equipment state evaluation and prediction integrated method and system | |
| CN109581871B (en) | Industrial control system intrusion detection method of immune countermeasure sample | |
| Martin-del-Campo et al. | Online feature learning for condition monitoring of rotating machinery | |
| CN101533549B (en) | Method of executing fire alarm by distributed optical fiber temperature sensor | |
| Siegel et al. | An auto-associative residual processing and K-means clustering approach for anemometer health assessment | |
| CN108256556A (en) | Wind-driven generator group wheel box method for diagnosing faults based on depth belief network | |
| CN105834835A (en) | Method for monitoring tool wear on line based on multiscale principal component analysis | |
| Wang et al. | Online bearing fault diagnosis using numerical simulation models and machine learning classifications | |
| Xiong et al. | A novel deep convolutional image-denoiser network for structural vibration signal denoising | |
| Sarmadi et al. | Probabilistic data self-clustering based on semi-parametric extreme value theory for structural health monitoring | |
| CN104035431A (en) | Obtaining method and system for kernel function parameters applied to nonlinear process monitoring | |
| Qin et al. | Remaining useful life prediction for rotating machinery based on optimal degradation indicator | |
| CN109443766A (en) | A kind of heavy-duty vehicle gearbox gear Safety Analysis Method | |
| Ulmer et al. | Early fault detection based on wind turbine scada data using convolutional neural networks | |
| Shao et al. | Fault prognosis and diagnosis of an automotive rear axle gear using a RBF-BP neural network | |
| CN107025382B (en) | A kind of engineering system health analysis system and method based on critical phase transformation theory | |
| Siegel | Prognostics and health assessment of a multi-regime system using a residual clustering health monitoring approach | |
| CN102749584A (en) | Prediction method for residual service life of turbine generator based on ESN (echo state network) of Kalman filtering | |
| Liu et al. | A divide and conquer approach to anomaly detection, localization and diagnosis | |
| Zhu et al. | Fault detection of offshore wind turbine gearboxes based on deep adaptive networks via considering Spatio-temporal fusion | |
| Ko et al. | Structural health monitoring and intelligent vibration control of cable-supported bridges: Research and application |
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 |