CN107133376A - A kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model - Google Patents
A kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model Download PDFInfo
- Publication number
- CN107133376A CN107133376A CN201710185959.0A CN201710185959A CN107133376A CN 107133376 A CN107133376 A CN 107133376A CN 201710185959 A CN201710185959 A CN 201710185959A CN 107133376 A CN107133376 A CN 107133376A
- Authority
- CN
- China
- Prior art keywords
- mrow
- sequence
- value
- msup
- grey
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B63—SHIPS OR OTHER WATERBORNE VESSELS; RELATED EQUIPMENT
- B63G—OFFENSIVE OR DEFENSIVE ARRANGEMENTS ON VESSELS; MINE-LAYING; MINE-SWEEPING; SUBMARINES; AIRCRAFT CARRIERS
- B63G8/00—Underwater vessels, e.g. submarines; Equipment specially adapted therefor
- B63G8/08—Propulsion
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mechanical Engineering (AREA)
- Aviation & Aerospace Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention provides a kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model, in gray background value construction phase, and gray background value is used as close to integration by calculate Accumulating generation sequence；In the albefaction equation solution stage, the minimum point of prediction residual is determined by the difference between original series predicted value and original series actual value, using the minimum corresponding original series value of point of predicated error as albefaction solution of equation initial value；In forecasting sequence construction phase, the residual sequence based on forecasting sequence and original series carries out re prediction, and the residual sequence obtained based on re prediction is modified to the forecasting sequence of original series so that predicted the outcome with adjustable.The present invention solves the problem of predicated error that exists during the prediction weak fault degree of AUV propellers is larger, and the gray background value building method in traditional gray method, albefaction equation solution method, forecasting sequence building method are improved respectively.
Description
Technical field
Gray prediction mould is based on the present invention relates to autonomous type underwater robot fault diagnosis technology field, more particularly to one kind
The weak fault degree Forecasting Methodology of autonomous type underwater robot propeller of type.
Background technology
Autonomous type underwater robot (AUV) unmanned untethered is operated in marine environment complicated and changeable, and security is AUV
Key character, condition monitoring and fault diagnosis is basis and the key technology for ensureing AUV securities.AUV is made up of multiple parts,
Wherein propeller is that its critical component is also load most heavy part, and research propeller fault diagnosis technology is for improving AUV peaces
Full property is significant.Many scholars achieve good achievement in research in terms of AUV propeller fault diagnosis technologies, but greatly
Propeller hard fault and the more failure of output loss are all paid close attention to, and less research exerts oneself the extent of damage less than gross capability
10%, the less weak failure of fault degree, meanwhile, the weak failure of propeller is generally initial failure, and effective detection is with examining as early as possible
Disconnected initial failure plays an important roll for avoiding AUV from occurring " catastrophic failure ".The weak failure of AUV propellers that the present invention is studied
Degree forecasting problem, refer to according to propeller in the past and current failure degree, come speculate following fault degree of propeller and its
Development trend, predicts obtained fault degree, can provide failure propeller preferential service rating adjustment institute for AUV work plannings
The fault message needed, and provide decision-making foundation for AUV active tolerant control device dynamic restructurings.
Typical failure prediction method, mainly there is time series analysis, regression analysis, neutral net, supporting vector at present
Machine, gray model GM (1,1) method etc..Time series analysis and regression analysis have advantage that is simple, easily realizing, but exist outer
The uncertain influence on fault data sequence statistic rule is disturbed to cause the problem of precision of prediction is relatively low in portion；Neutral net
There is the of a relatively high advantage of precision of prediction with SVMs, but need to obtain a number of fault sample in advance and instructed
Can be just predicted after white silk, and AUV fault samples it is more difficult acquisition and quantity is relatively fewer so that neutral net and SVMs
There is Generalization Ability of Neural Network decline when directly applying to the prediction of AUV fault degrees and optimal separating hyper plane is difficult to divide
Problem.
The advantage of grey GM (1,1) method is：Required data volume is relatively fewer, it is only necessary to which 4 and sample above just can be with
Prediction is modeled, applying precision is of a relatively high, is comparatively adapted to the less AUV propellers event of prediction fault sample
Barrier degree, and obtained practical application in AUV fault diagnosises.
The present invention is had found in the experimental study that the weak fault degree of AUV propellers is predicted based on grey GM (1,1) method：Directly
When connecing fault degree weak using traditional grey GM (1,1) method prediction propeller, in gray background value construction phase, due to using
One-accumulate formation sequence approaches gray background value close to the power generation of grade so that gray background value exists inclined with actual value
Difference；In the albefaction equation solution stage, due to the initial point of known array to be set to the initial value of albefaction solution of equation so that albefaction
There is deviation in non trivial solution, increase predicated error；In forecasting sequence construction phase, due to lacking regulation mechanism so that prediction
Predicted the outcome when residual error is larger and also do not possess adjustable.These reasons cause to push away using traditional grey GM (1,1) method prediction
Predicated error is larger when entering device weak fault degree.
The predicated error existed during for fault degree weak using traditional grey GM (1,1) method prediction AUV propellers compared with
Big the problem of, the present invention proposes a kind of weak fault degree Forecasting Methodology of grey forecasting model AUV propellers, to traditional grey GM
(1,1) gray background value building method, albefaction equation solution method, forecasting sequence building method in method is changed respectively
Enter.Have in traditional grey GM (1,1) method it is multiple realize step, core procedure be gray background value construction, albefaction equation solution,
Forecasting sequence constructs these three steps, and the present invention is improved around these three steps, using these three improved steps of the present invention
Suddenly, and in traditional grey GM (1,1) method other unmodified steps are together, of the invention a kind of based on gray prediction mould to constitute
The weak fault degree Forecasting Methodology of AUV propellers of type.
The content of the invention
It is weak the invention aims to provide a kind of autonomous type underwater robot propeller based on grey forecasting model
Fault degree Forecasting Methodology, it is specifically a kind of based on grey forecasting model event weak to autonomous type underwater robot propeller
Hinder the data processing method that level data is predicted, solve using traditional weak event of grey GM (1,1) method prediction AUV propellers
During barrier degree, the problem of predicated error existed is larger.
The object of the present invention is achieved like this：Comprise the following steps：
Step one：Kernel-based methods historical data identification result obtains currentlyying propel the weak fault degree of device；
Step 2：Accumulating generation：The original number of the weak fault degree obtained using one-accumulate generation method to step one
According to being handled, Accumulating generation sequence X is obtained(1)(k)：
In formula：X(0)(i) it is original series, k represents at k-th point in data sequence, and n is the data length of initial data,
I represents at i-th point in initial data；
Step 3：Construct gray background value：The Accumulating generation sequence X obtained according to step 2(1)(k) gray background, is constructed
Value Z(1)(k)；
Step 4：Construct grey forecasting model：The gray background value Z obtained according to step 3(1)(k) gray prediction, is constructed
Model X(0)(k)+aZ(1)(k) in=u, formula：A is grey development coefficient, and u is grey actuating quantity；
Step 5：Solve the albefaction equation of grey forecasting model：The grey forecasting model obtained according to step 4, using most
Small least square method, obtains grey development coefficient a and grey actuating quantity u value, and by calculating forecasting sequence and original series
Deviation, using the minimum point of deviation as the initial value of albefaction solution of equation, obtaining albefaction equation is：
In formula：For the predicted value of k+1 moment cumulative sequence, X(0)(l) original series for being Serial No. l
Value, sequence number l is by formulaObtain, andFor the predicted value of original series, X(0)(i)
For the actual value of original series；
Step 6：Structure forecast sequence：Using the deviation between the forecasting sequence and cumulative sequence original value of cumulative sequence,
The predicated error of the cumulative sequence of prediction, is modified to original predictive result based on the predicated error, obtains the pre- of original series
Sequencing is classified as：
In formula：Sgn is sign function,Added up the difference of sequence prediction value and cumulative sequence for the k+1 moment, and by
FormulaGained.
Present invention additionally comprises some such architectural features：
1. the building method of the gray background value in step 3 is：
Calculate Accumulating generation sequence close to constructing gray background value by way of integration, the grey back of the body after being improved
Scape value is：
Compared with prior art, the beneficial effects of the invention are as follows：Compared with existing grey GM (1,1) Forecasting Methodology, this
Invention proposes a kind of new weak fault degree Forecasting Methodology, and this method had both efficiently solved existing grey GM (1,1) side
Method uses the strategy of close approximation, by calculating being used as gray background value close to the power generation of grade and cause of one-accumulate formation sequence
Background value offset issue, solve again existing grey GM (1,1) methods solve albefaction equation when by original series just
Initial value is carried out as forecasting problem caused by the initial value of albefaction solution of equation, and using the residual sequence of forecasting sequence and original series
Re prediction, the residual sequence obtained based on re prediction is modified to the forecasting sequence of original series so that predicted the outcome
With adjustable, the precision of prediction of the weak fault degree of autonomous type underwater robot propeller is improved.
Brief description of the drawings
Fig. 1 is weak fault degree Forecasting Methodology flow chart of the invention；
Fig. 2 is the gray background value building method flow comparison diagram of the invention with conventional method；
The effect contrast figure for the gray background value that Fig. 3 is constructed for the present invention with conventional method；
Fig. 4 is the albefaction equation solution method flow comparison diagram of the invention with conventional method；
The effect contrast figure for the albefaction equation that Fig. 5 is solved for the present invention with conventional method；
Fig. 6 is the forecasting sequence building method flow comparison diagram of the invention with conventional method；
The effect contrast figure for the forecasting sequence that Fig. 7 is constructed for the present invention with conventional method；
Fig. 8 is the weak fault degree prediction effect comparison diagram of the invention with conventional method.
Embodiment
The present invention is described in further detail with embodiment below in conjunction with the accompanying drawings.
With reference to Fig. 1 to Fig. 8, the weak fault degree Forecasting Methodology of autonomous type underwater robot propeller to the present invention is explained
State.Weak fault degree Forecasting Methodology in the present invention is directed to using traditional weak event of grey GM (1,1) method prediction AUV propellers
The problem of predicated error existed during barrier degree is larger and propose, the present invention first using Accumulating generation method to weak failure journey
Degrees of data is handled, then using the sequence structure gray background value after Accumulating generation；Grey is constructed based on gray background value
Forecast model, and the albefaction equation of the grey forecasting model is solved；Finally obtained forecasting sequence is carried out secondary pre-
Final forecasting sequence is obtained after survey.The accuracy and authenticity of weak fault degree prediction can be finally improved, is that one kind is new, has
The weak fault degree Forecasting Methodology of autonomous type underwater robot propeller of effect.
The Forecasting Methodology flow of the present invention is as shown in figure 1, its specific implementation step is as follows：
1st, Process History data identification result is primarily based on to obtain currentlyying propel the weak fault degree of device.
2nd, Accumulating generation process：Weak fault degree initial data is handled using one-accumulate generation method,X in formula(1)(k) it is Accumulating generation sequence, X(0)(i) it is original series, k is represented in data sequence
K-th point, n is the data length of initial data, and i represents at i-th point in initial data.
3rd, gray background value construction process：The Accumulating generation sequence X obtained according to previous step(1)(k), the construction grey back of the body
Scape value, calculate Accumulating generation sequence close to constructing gray background value by way of integration, the present invention improve after the grey back of the body
Scape value such as formulaIt is shown.
Flow of the inventive method with conventional method in gray background value construction phase is contrasted as shown in Fig. 2 gray background
Value construction Contrast on effect is as shown in Figure 3.In Fig. 3, the gray background value of traditional gray background value building method construction, with actual ash
Between color background value, there is an error, this error is due to the solution with linear first-order differential equation, goes close approximation to integrate
Caused by, the error is as shown in dash area in Fig. 3.
For the error in reduction Fig. 3 between traditional gray background value and true gray background value, the present invention proposes that one kind changes
The gray background value building method entered, is further solved, the inventive method is not used to the actual gray background value in Fig. 3
The strategy of conventional method approximate solution, but by calculate cumulative sequence it is adjacent 2 points between close to integration, to reduce in Fig. 3
Deviation shown in dash area.
4th, GM (1,1) forecast model construction process：The gray background value Z obtained according to previous step(1)(k) grey, is constructed
Forecast model X(0)(k)+aZ(1)(k) a is grey development coefficient in=u, formula, and u is grey actuating quantity.
5th, the albefaction equation solution process of forecast model：The grey forecasting model obtained according to previous step, using minimum
Least square method, obtains grey development coefficient a and grey actuating quantity u value, and according to a, u value, traditional method obtains X(1)Obtain pre-
Surveying formula isIn formulaFor the predicted value at k+1 moment, X(0)(1)
For the initial value of original series.
It is of the invention then be to be improved：By calculating the deviation of forecasting sequence and original series, the minimum point of deviation is made
For the initial value of albefaction solution of equation, the albefaction non trivial solution such as formula after improvement
It is shown.In formula：For the predicted value of k+1 moment cumulative sequence, X(0)(l) the original series value for being Serial No. l, changes
Sequence number l after entering is by formulaObtain.In formulaFor the predicted value of original series, X(0)(i) it is the actual value of original series.
Flow of the inventive method with conventional method in the albefaction equation solution stage is contrasted as shown in figure 4, albefaction equation is asked
Solve Contrast on effect as shown in Figure 5.
In Fig. 5, traditional albefaction equation solution method, setting forecasting sequence must pass through the initial points of original series, i.e. Fig. 5
In (k-3) place, its deviation predicted the outcome between original series future time instance value be Fig. 5 in traditionE；The present invention is carried
Improved albefaction equation solution method, calculates the deviation e between forecasting sequence and original series first1、e2、e3、e4, compare hair
Existing e3For deviation minimum value, so by e3Corresponding point (k-1) must pass through point, the inventive method prediction as forecasting sequence
As a result the deviation between original series future time instance is this paper in Fig. 5E。
It is qualitative compare conventional method and the inventive method in Fig. 5 predict the outcome it is inclined between original series future time instance value
Difference.In Fig. 5, the deviation tradition of conventional methodESpan, more than the inventive method deviation hereinE, illustrate conventional method
In albefaction equation solution method caused by error be more than the inventive method, the albefaction equation solution effect of the inventive method is excellent
In conventional method.
6th, forecasting sequence construction process：Cumulative sequence X is obtained according to previous step(1)Predicted value after, carry out regressive life
Into, you can obtain original series X(0)Predicted value.Concretely comprise the following steps：It is former using the forecasting sequence and cumulative sequence of cumulative sequence
Deviation between initial value, the predicated error of the cumulative sequence of prediction, is modified based on the predicated error to original predictive result, this
The forecasting sequence such as formula of original series after invention improvement
It is shown.In formula, sgn is sign function,For the difference of k+1 moment cumulative sequence prediction value and cumulative sequence.
In above formula, after the present invention is improvedAdded up the difference of sequence prediction value and cumulative sequence for the k+1 moment, by
FormulaGained.
The inventive method is contrasted as shown in fig. 6, forecasting sequence is imitated with conventional method in the flow of forecasting sequence construction phase
Fruit contrasts as shown in Figure 7.
In Fig. 7, add up sequence X(1)With its forecasting sequenceBetween there is deviationClassical forecast sequence structure
Method does not consider these deviations, inverse accumulated generating method is directly used, based on forecasting sequence k+1 moment values and cumulative sequence k moment
Value, obtains original series k+1 moment values；The forecasting sequence building method that is carried herein, calculate first Accumulating generation sequence (1) with
Its forecasting sequenceBetween deviation ξ (k-3), ξ (k-2), ξ (k-1), ξ (k), the k+1 moment is predicted based on the biased sequence
DeviationAnd the deviation is introduced to the forecasting sequence of original series, the introducing of cumulative ordering bias value causes grey
The result of prediction possesses controllability, reduces final predicated error, illustrates the inventive method in cumulative sequence prediction knot
Effect in terms of the error of fruit is better than Classical forecast sequence constructing method.
Fig. 8 is the weak failure real change degree of autonomous type underwater robot, the present invention is predicted the outcome, conventional method prediction is tied
Fruit comparison diagram.For evaluation and foreca effect, typically using average relative error, mean square deviation ratio, the degree of association, small error possibility, this is several
Individual evaluation index.Next, this above-mentioned four indexs are respectively adopted in the present invention, and combine accuracy test grade with reference to table to contrast
Analyze effect of the context of methods with traditional grey GM (1,1) methods in terms of weak fault degree prediction.(1) in average relative error
Aspect, the inventive method is 64.81% than conventional method relative reduction；(2) in terms of mean square deviation ratio, the inventive method is than passing
System method relative reduction 98.37%；(3) in terms of the degree of association, the inventive method is more relative than conventional method to improve 0.06%；
(4) in terms of small error possibility, the inventive method is more relative than conventional method to improve 64.61%；(5) in terms of accuracy class,
Its precision of prediction grade of the inventive method is two grades, and the precision of prediction grade of conventional method is only level Four.
Above-mentioned experimental result reflects, for the non-monotonic variation tendency of weak fault degree, in average relative error, mean square deviation
In terms of ratio, the degree of association, small error possibility, precision of prediction grade, the inventive method is superior to traditional grey GM (1,1) method.
In summary, the present invention is when solving fault degree weak using traditional grey GM (1,1) method prediction AUV propellers
The problem of predicated error of presence is larger, using improved grey forecasting model of the invention to the weak fault degree mistake of AUV propellers
Journey historical data is handled, in gray background value construction phase, and ash is used as close to integration by calculate Accumulating generation sequence
Color background value；In the albefaction equation solution stage, determined by the difference between original series predicted value and original series actual value
The minimum point of prediction residual, using the minimum corresponding original series value of point of predicated error as albefaction solution of equation initial value；
Forecasting sequence construction phase, the residual sequence based on forecasting sequence and original series carries out re prediction, is obtained based on re prediction
To residual sequence the forecasting sequence of original series is modified so that predict the outcome with adjustable.The present invention is solved
The problem of predicated error existed during the prediction weak fault degree of AUV propellers is larger, to the ash in traditional grey GM (1,1) method
Color structure method of background value, albefaction equation solution method, forecasting sequence building method are improved respectively, are promoted available for AUV
The fields such as the weak fault degree prediction processing of device, sensor fault degree prediction processing.
Claims (2)
1. a kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model, its feature exists
In：Comprise the following steps：
Step one：Kernel-based methods historical data identification result obtains currentlyying propel the weak fault degree of device；
Step 2：Accumulating generation：The initial data of the weak fault degree obtained using one-accumulate generation method to step one is entered
Row processing, obtains Accumulating generation sequence X(1)(k)：
<mrow>
<msup>
<mi>X</mi>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<mi>k</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msup>
<mi>X</mi>
<mrow>
<mo>(</mo>
<mn>0</mn>
<mo>)</mo>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>)</mo>
</mrow>
</mrow>
In formula：X(0)(i) it is original series, k represents at k-th point in data sequence, and n is the data length of initial data, i generations
I-th point in table initial data；
Step 3：Construct gray background value：The Accumulating generation sequence X obtained according to step 2(1)(k), construction gray background value Z(1)(k)；
Step 4：Construct grey forecasting model：The gray background value Z obtained according to step 3(1)(k) grey forecasting model, is constructed
X(0)(k)+aZ(1)(k) in=u, formula：A is grey development coefficient, and u is grey actuating quantity；
Step 5：Solve the albefaction equation of grey forecasting model：The grey forecasting model obtained according to step 4, using a most young waiter in a wineshop or an inn
Multiply method, obtain grey development coefficient a and grey actuating quantity u value, and by calculating the deviation of forecasting sequence and original series,
Using the minimum point of deviation as the initial value of albefaction solution of equation, obtaining albefaction equation is：
<mrow>
<msup>
<mover>
<mi>X</mi>
<mo>^</mo>
</mover>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mo>&lsqb;</mo>
<msup>
<mi>X</mi>
<mrow>
<mo>(</mo>
<mn>0</mn>
<mo>)</mo>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>u</mi>
<mo>/</mo>
<mi>a</mi>
<mo>&rsqb;</mo>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mi>a</mi>
<mrow>
<mo>(</mo>
<mi>k</mi>
<mo>-</mo>
<mi>l</mi>
<mo>+</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
</msup>
<mo>+</mo>
<mi>u</mi>
<mo>/</mo>
<mi>a</mi>
</mrow>
In formula：For the predicted value of k+1 moment cumulative sequence, X(0)(l) the original series value for being Serial No. l, sequence
Number l is by formulaObtain, andFor the predicted value of original series, X(0)(i) to be original
The actual value of sequence；
Step 6：Structure forecast sequence：Using the deviation between the forecasting sequence and cumulative sequence original value of cumulative sequence, prediction
The predicated error of cumulative sequence, is modified to original predictive result based on the predicated error, obtains the pre- sequencing of original series
It is classified as：
<mrow>
<msup>
<mover>
<mi>X</mi>
<mo>^</mo>
</mover>
<mrow>
<mo>(</mo>
<mn>0</mn>
<mo>)</mo>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mo>=</mo>
<mo>&lsqb;</mo>
<mn>1</mn>
<mo>-</mo>
<msup>
<mi>e</mi>
<mi>a</mi>
</msup>
<mo>&rsqb;</mo>
<mo>&lsqb;</mo>
<msup>
<mi>X</mi>
<mrow>
<mo>(</mo>
<mn>0</mn>
<mo>)</mo>
</mrow>
</msup>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mi>u</mi>
<mo>/</mo>
<mi>a</mi>
<mo>&rsqb;</mo>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mi>a</mi>
<mi>k</mi>
</mrow>
</msup>
<mo>-</mo>
<mi>sgn</mi>
<mrow>
<mo>(</mo>
<mover>
<mi>&xi;</mi>
<mo>^</mo>
</mover>
<mo>(</mo>
<mrow>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
<mo>*</mo>
<mo>|</mo>
<mover>
<mi>&xi;</mi>
<mo>^</mo>
</mover>
<mo>(</mo>
<mrow>
<mi>k</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
<mo>)</mo>
<mo>|</mo>
</mrow>
In formula：Sgn is sign function,Added up the difference of sequence prediction value and cumulative sequence for the k+1 moment, and by formulaGained.
2. the weak failure journey of a kind of autonomous type underwater robot propeller based on grey forecasting model according to claim 1
Spend Forecasting Methodology, it is characterised in that：The building method of gray background value in step 3 is：
Calculate Accumulating generation sequence close to constructing gray background value by way of integration, the gray background value after being improved
For：
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710185959.0A CN107133376A (en) | 2017-03-27 | 2017-03-27 | A kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710185959.0A CN107133376A (en) | 2017-03-27 | 2017-03-27 | A kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107133376A true CN107133376A (en) | 2017-09-05 |
Family
ID=59721400
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710185959.0A Pending CN107133376A (en) | 2017-03-27 | 2017-03-27 | A kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107133376A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108764337A (en) * | 2018-05-29 | 2018-11-06 | 江苏科技大学 | Underwater robot propeller fault degree discrimination method based on relative grey correlative degree boundary constraint |
| CN109143094A (en) * | 2018-06-29 | 2019-01-04 | 上海科列新能源技术有限公司 | A kind of abnormal deviation data examination method and device of power battery |
| CN109325207A (en) * | 2018-09-04 | 2019-02-12 | 温州大学 | Valve body assembling quality failure prediction method based on grey GM (1,1) model |
| CN109359388A (en) * | 2018-10-18 | 2019-02-19 | 北京仿真中心 | A kind of Complex simulation systems credibility evaluation method |
| CN109683591A (en) * | 2018-12-27 | 2019-04-26 | 江苏科技大学 | Underwater propeller fault degree discrimination method based on fusion signal time domain energy and time-frequency entropy |
| CN112257283A (en) * | 2020-10-30 | 2021-01-22 | 中国矿业大学 | Grey prediction model method based on background value and structure compatibility combination optimization |
| CN113269350A (en) * | 2021-04-28 | 2021-08-17 | 长春工业大学 | Transformer fault prediction method based on gray GM (1,1) model |
| CN113537627A (en) * | 2021-08-04 | 2021-10-22 | 华能(浙江)能源开发有限公司清洁能源分公司 | Operation and maintenance-oriented offshore wind turbine generator fault interval time prediction method |
| CN114217106A (en) * | 2021-12-14 | 2022-03-22 | 青岛理工大学 | Intelligent electric energy meter reading data secondary research and judgment method based on improved GM (1,1) model |
| CN114460526A (en) * | 2022-04-12 | 2022-05-10 | 华中科技大学 | Transformer substation current transformer error prediction method and system based on follow-up compensation |
| CN114647814A (en) * | 2022-05-23 | 2022-06-21 | 成都理工大学工程技术学院 | Nuclear signal correction method based on prediction model |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105737876A (en) * | 2014-12-08 | 2016-07-06 | 哈尔滨米米米业科技有限公司 | State diagnosis and signal recovery system of underwater autonomous diving vehicle sensor |
| CN105823503A (en) * | 2016-03-23 | 2016-08-03 | 哈尔滨工程大学 | Improved gray prediction GM(1,1) model-based autonomous underwater vehicle (AUV) sensor fault diagnosis method |
| CN106021888A (en) * | 2016-05-13 | 2016-10-12 | 福建师范大学 | Fault forecasting method combining intuitionistic fuzzy set with gray model |
-
2017
- 2017-03-27 CN CN201710185959.0A patent/CN107133376A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105737876A (en) * | 2014-12-08 | 2016-07-06 | 哈尔滨米米米业科技有限公司 | State diagnosis and signal recovery system of underwater autonomous diving vehicle sensor |
| CN105823503A (en) * | 2016-03-23 | 2016-08-03 | 哈尔滨工程大学 | Improved gray prediction GM(1,1) model-based autonomous underwater vehicle (AUV) sensor fault diagnosis method |
| CN106021888A (en) * | 2016-05-13 | 2016-10-12 | 福建师范大学 | Fault forecasting method combining intuitionistic fuzzy set with gray model |
Non-Patent Citations (3)
| Title |
|---|
| MINGJUN ZHANG等: "Weak thruster fault prediction method for autonomous underwater vehicle", 《OCEANS 2016 SHANGHAI》 * |
| 周媛 等: "水下机器人传感器故障诊断的灰色预测模型", 《中国造船》 * |
| 张彬 等: "基于背景值和边值修正的GM（1,1）模型优化", 《系统工程理论与实践》 * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108764337A (en) * | 2018-05-29 | 2018-11-06 | 江苏科技大学 | Underwater robot propeller fault degree discrimination method based on relative grey correlative degree boundary constraint |
| CN108764337B (en) * | 2018-05-29 | 2021-10-19 | 江苏科技大学 | Underwater robot propeller fault degree identification method based on relative gray correlation degree boundary constraint |
| CN109143094A (en) * | 2018-06-29 | 2019-01-04 | 上海科列新能源技术有限公司 | A kind of abnormal deviation data examination method and device of power battery |
| CN109325207A (en) * | 2018-09-04 | 2019-02-12 | 温州大学 | Valve body assembling quality failure prediction method based on grey GM (1,1) model |
| CN109359388A (en) * | 2018-10-18 | 2019-02-19 | 北京仿真中心 | A kind of Complex simulation systems credibility evaluation method |
| CN109683591B (en) * | 2018-12-27 | 2021-03-19 | 江苏科技大学 | Underwater propeller fault degree identification method based on fusion signal time domain energy and time-frequency entropy |
| CN109683591A (en) * | 2018-12-27 | 2019-04-26 | 江苏科技大学 | Underwater propeller fault degree discrimination method based on fusion signal time domain energy and time-frequency entropy |
| CN112257283A (en) * | 2020-10-30 | 2021-01-22 | 中国矿业大学 | Grey prediction model method based on background value and structure compatibility combination optimization |
| CN113269350A (en) * | 2021-04-28 | 2021-08-17 | 长春工业大学 | Transformer fault prediction method based on gray GM (1,1) model |
| CN113269350B (en) * | 2021-04-28 | 2023-09-05 | 长春工业大学 | Transformer fault prediction method based on gray GM (1, 1) model |
| CN113537627A (en) * | 2021-08-04 | 2021-10-22 | 华能(浙江)能源开发有限公司清洁能源分公司 | Operation and maintenance-oriented offshore wind turbine generator fault interval time prediction method |
| CN113537627B (en) * | 2021-08-04 | 2023-07-11 | 华能(浙江)能源开发有限公司清洁能源分公司 | Marine wind turbine generator set fault interval time prediction method oriented to operation and maintenance |
| CN114217106A (en) * | 2021-12-14 | 2022-03-22 | 青岛理工大学 | Intelligent electric energy meter reading data secondary research and judgment method based on improved GM (1,1) model |
| CN114460526A (en) * | 2022-04-12 | 2022-05-10 | 华中科技大学 | Transformer substation current transformer error prediction method and system based on follow-up compensation |
| CN114647814A (en) * | 2022-05-23 | 2022-06-21 | 成都理工大学工程技术学院 | Nuclear signal correction method based on prediction model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107133376A (en) | A kind of weak fault degree Forecasting Methodology of autonomous type underwater robot propeller based on grey forecasting model | |
| Zhao et al. | Deep multi-scale convolutional transfer learning network: A novel method for intelligent fault diagnosis of rolling bearings under variable working conditions and domains | |
| Almalaq et al. | A review of deep learning methods applied on load forecasting | |
| Jiang et al. | Two-stage structural damage detection using fuzzy neural networks and data fusion techniques | |
| Lu et al. | Deep model based domain adaptation for fault diagnosis | |
| Shao et al. | Rolling bearing fault diagnosis using an optimization deep belief network | |
| CN109947086A (en) | Mechanical breakdown migration diagnostic method and system based on confrontation study | |
| Vladislavleva et al. | Predicting the energy output of wind farms based on weather data: Important variables and their correlation | |
| CN103728551B (en) | A kind of analog-circuit fault diagnosis method based on cascade integrated classifier | |
| Nozari et al. | Model-based robust fault detection and isolation of an industrial gas turbine prototype using soft computing techniques | |
| Smith Stegen | Redrawing the geopolitical map: international relations and renewable energies | |
| CN111339712A (en) | Method for predicting residual life of proton exchange membrane fuel cell | |
| CN106022954A (en) | Multiple BP neural network load prediction method based on grey correlation degree | |
| Moradi et al. | Marine route optimization using reinforcement learning approach to reduce fuel consumption and consequently minimize CO2 emissions | |
| CN115622047B (en) | Power Transformer load prediction method based on Transformer model | |
| JP2011054200A (en) | Neural network learning method | |
| CN111899905A (en) | Fault diagnosis method and system based on nuclear power device | |
| CN107239850A (en) | A kind of long-medium term power load forecasting method based on system dynamics model | |
| Taffese | Case-based reasoning and neural networks for real estate valuation. | |
| Do et al. | A comparative study of hierarchical ANFIS and ANN in predicting student academic performance | |
| Santos et al. | Performance analysis of a DNN classifier for power system events using an interpretability method | |
| Zhou et al. | Functional networks and applications: A survey | |
| Mo et al. | Large language models for travel behavior prediction | |
| Zhang et al. | BP-PSO-based intelligent case retrieval method for high-rise structural form selection | |
| Zhang et al. | Fault diagnosis with bidirectional guided convolutional neural networks under noisy labels |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170905 |