CN107092653A - A kind of landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis - Google Patents
A kind of landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis Download PDFInfo
- Publication number
- CN107092653A CN107092653A CN201710153276.7A CN201710153276A CN107092653A CN 107092653 A CN107092653 A CN 107092653A CN 201710153276 A CN201710153276 A CN 201710153276A CN 107092653 A CN107092653 A CN 107092653A
- Authority
- CN
- China
- Prior art keywords
- fuzzy
- class
- landslide
- rainfall threshold
- cluster analysis
- 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
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
- G06F16/2468—Fuzzy queries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Fuzzy Systems (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Probability & Statistics with Applications (AREA)
- Computational Linguistics (AREA)
- Automation & Control Theory (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis, developing algorithm Investment Models are determined to landslide critical excitation approaches threshold value, the selection of the fuzzy clustering factor are carried out by Delphi method；Based on the fuzzy cluster analysis of this progress landslide Critical Rainfall threshold value, create data matrix, standardized data matrix and set up fuzzy similarity matrix, finally carry out fuzzy clustering and determine Critical Rainfall threshold value.Landslide Critical Rainfall Threshold of the invention based on method of fuzzy cluster analysis, strict mathematical algorithm is derived and statistical technique and physical means, each factor of influence in region is taken into full account, so that it is determined that rainfall threshold value.This method degree of accuracy is high, can meet the requirement of current landslide precaution alarm, greatly improves threshold accuracy, the reasonable determination of Critical Rainfall threshold value is to ensureing that landslide precaution alarm accuracy plays an important role.
Description
Technical field
The invention belongs to landslide disaster forecasting technique field, and in particular to a kind of landslide based on method of fuzzy cluster analysis is faced
Boundary's rainfall Threshold.
Background technology
With Global climate change, the extreme catchment frequency is constantly raised, and causes global mountain flood event to rise, times
It is concerned；China's mountain area area is vast, the features such as landslide disaster has occurrence frequency height, has a very wide distribution, loses serious, in advance
Survey, forecast, early warning difficulty are big, and the main risk factor that landslide occurs is rainfall.Nationwide internal cause extra torrential rain in 1998
And it is 1157 people to induce substantial amounts of landslide, the death toll that causes of avalanche and mud-stone flow disaster, the wounded is more than 10,000.Face
Using more universal in the mountain flood prevention work at home and abroad of this index of boundary's rainfall, determine that method is also varied, respectively have
An eternal lasting.Rational rainfall threshold value index is the key point that Regional Heavy Rain mud-rock flow is prevented and reduced natural disasters.
Conventional rainfall threshold value research method mainly has 2 kinds at present, and the first is the statistical analysis side based on disaster data
Method, by carrying out statistical analysis to actual rainfall and landslide disaster data, draws corresponding early stage effective precipitation and triggering
Relation between rainfall, so as to draw rainfall threshold curve, this method often have ignored underlying surface vegetation state, geological conditions
Deng some other factorses of influence landslide disaster, and the landslide Critical Rainfall of gained is to represent a regional extent rather than one
Individual specific landslide point；Second is similar analogy method, the area for lacking rainfall and disaster data, when these local ground
Reason, geology, ecology etc. cause the area of calamity threshold limit value more similar to having determined that, can approximately think to cause calamity Critical Rainfall also phase
Seemingly, can suitably it be adjusted according to actual conditions.No matter but each region topography and geomorphology or factor of influence are not identical
, accuracy rate is greatly reduced.
The content of the invention
It is an object of the invention to provide a kind of landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis, solution
Statistical analysis technique and the similar analogy method of having determined are when calculating threshold value, the problem of accuracy is too low.
The technical solution adopted in the present invention is that a kind of landslide Critical Rainfall threshold value based on method of fuzzy cluster analysis is determined
Method, specifically implements according to following steps：
Step 1, the fuzzy clustering factor are determined；
The fuzzy cluster analysis of step 2, the Critical Rainfall threshold value that comes down；
Step 3, carry out fuzzy clustering and determine Critical Rainfall threshold value.
The beneficial effects of the invention are as follows,
(1) during the fuzzy clustering factor is determined, carried out using De Feier methods special for the purpose of determining landslide formative element
Family's investigation, the fuzzy clustering factor finally chosen.
(2) landslide Critical Rainfall Threshold is carried out from method of fuzzy cluster analysis, strengthens threshold accuracy, it is to avoid
Because misleading rate caused by threshold value is inaccurate, so as to effectively remind people's disaster prevention to occur in landslide disaster.
In summary, compared with the existing methods：The problem of Critical Rainfall threshold value that comes down determination is a complexity, to make really
Process is determined more comprehensively from Systems Theory, based on method of fuzzy cluster analysis, passes through developing algorithm Investment Models, strict mathematics
Algorithmic derivation and statistical technique and physical means, take into full account each factor of influence in region, so that it is determined that rainfall threshold value, the determination
Method is more comprehensive, as a result relatively reliable, mutually presses close to actual.
Brief description of the drawings
Fig. 1 is that heavy rain landslide is excited in the landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis of the invention
Rainfall pattern generally changes figure；
Fig. 2 is that landslide causes calamity critical in the landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis of the invention
Rainfall threshold value assessment technology flow chart；
Fig. 3 be in the landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis of the invention Liu family's ditch come down it is close
Degree and critical excitation approaches strength relationship figure.
Embodiment
The present invention is described in detail with reference to the accompanying drawings and detailed description.
Landslide Critical Rainfall Threshold of the invention based on method of fuzzy cluster analysis, it is specifically real according to following steps
Apply：
The Critical Rainfall threshold value that comes down is affected by many factors, and these factors are with including but is not limited to the landform in landslide basin
Looks, material resource deposit, physical geography position, condition of raining etc..Data are observed in part landslide basin without accurate rainfall, if
Other landslide basins can be found, it is essentially identical with the landslide basin surface condition, while Critical Rainfall threshold values is, it is known that can join
The Critical Rainfall value is examined, so as to obtain the value that the Critical Rainfall threshold values is close.
Step 1, the fuzzy clustering factor are determined
The factor for characterizing landslide surface condition feature is numerous, and interdependence between each factor, mutually restriction.In order to select
Take out the most representational fuzzy clustering factor, main following several principles：1) simplicity and operability principle；2)
Comprehensive and representativeness principle；3) leading factor principle；4) principle of combination of qualitative and quantitative analysis.
Take a broad survey in the wild, indoor synthetic analysis and with reference on the basis of existing research, enter using De Feier methods
Expert investigation of the row for the purpose of determining landslide formative element, the fuzzy clustering factor finally chosen.
The illegal specific implementation step of Dare is as follows：
(1) panel of expert is constituted.Expert's number number, can according to prediction problem size and be related to face width and
It is fixed, it is usually no more than 20 people.
(2) the problem of proposing to be predicted to all experts and relevant requirement, and enclose all back ofs the body about this problem
Scape material, while asking expert proposes what material also needed to.Then, written reply is done by expert.
(3) material that each expert is received according to them, proposes the prediction opinion of oneself, and illustrates it how sharp oneself is
With these materials and propose predicted value.
(4) every expert is judged that opinion collects for the first time, arranges into chart, contrasted, then be distributed to every expert, allowed
Expert compares oneself different opinions with other people, changes opinion and the judgement of oneself.Every expertise can also be subject to
Arrange, or other experts for asking identity higher are commented on, and these opinions are then distributed to every expert again, so that they join
The opinion of oneself is changed after examination.
The fuzzy cluster analysis of step 2, the Critical Rainfall threshold value that comes down
Existing disaster data shows that heavy rain landslide excites rainfall pattern main based on fast peak type, as shown in Figure 1.It is above-mentioned
, there is number of peaks when rainy in the rainfall of type；The time that rainfall occurs is also different, and early stage, mid-term, later stage are all present
Possibility；The generally only heavy showers of short period, but some heavy showers can be with last longer.
Every landslide basin geologic setting environment is different, and its Critical Rainfall threshold value is incomplete same, but notices simultaneously
The similar landslide basin difference of surface condition is little.Based on this understanding, the essentially identical landslide of two geological conditions represents three
The similitude of a variety of physical quantitys in dimension space, it is above-mentioned similar based on what is occurred under fuzzy concept, the phase of fuzzy mathematics can be relied on
Theory is closed to be solved.The Critical Rainfall threshold value estimating techniques flow that comes down is as shown in Figure 2.
1. data matrix is created
Assuming that the domain for being classified object is expressed as U={ x1,x2,…,xn, the character in any object passes through m index
It is described：Xi={ xi1,xi2,…,xim(i=1,2 ..., n), the domain of U presentation class objects, x represents each of set
Individual element, i, j distinguish the row and column of representing matrix, xijThe element that the i-th row j is arranged in representing matrix X.
Raw data matrix is expressed as follows：X=(xij)n×m
2. standardized data matrix
In actual mechanical process, data are different, and corresponding dimension has differences.By the way of conversion data, so it is right
The frequently amount in more all kinds of dimensions.Even so, the data obtained using which, do not ensure necessarily to all belong to [0,1] area
In., it is necessary to which compressed data is into [0,1] interval on the basis of fuzzy matrix demand.The main side of standardized data matrix
Method is as follows：
1) standard deviation is standardized
I-th of variable is standardized, the mode used is conversion xijFor xij', wherein,Jth column element average value is represented,
Represent the i-th row element average value, SjRepresent variance, MjThe intermediate value of jth row is represented, max { } is represented to take maximum, and min { } is represented
Take minimum value, xij' element after standardization is represented, it is as follows：
In formula：
2) standardization with extreme difference：
3) extreme difference is standardized：
4) maximum is standardized：Wherein：Mj=max (x1j, x2j..., xnj) (6)
3. fuzzy similarity matrix is set up
Set up xiWith xjSimilarity degree rij=R (xi,xj) method mainly have：
1) Similar operator
1 ° of Cosin method：
2 ° of correlation coefficient process：
Usually, r is takenij=1-c (d (xi,xj))2, wherein c, α is the parameter suitably chosen, d (xi,xj) represent xiWith xj
Distance, it causes 0≤rij≤1。rijSimilarity degree is represented, R (x) represents fuzzy matrix function, xikI-th row in representing matrix
Kth column element, xjkJth row kth column element in representing matrix.
The distance of use has：
1 ° of Hamming distance from：
2 ° of Euclid distances：
3 ° of hebyshev distances：d(xi,xj)=max1≤k≤n|xik-xjk| (11)
2) Study on similar degree method
1 ° of minimax method:
The minimum method of 2 ° of arithmetic averages:
The minimum method of 3 ° of geometric averages:
Step 3, carry out fuzzy clustering and determine Critical Rainfall threshold value
Obscured according to the fuzzy similarity matrix R of gained in step 2 using transitive closure computing or Direct Cluster Analysis
Cluster, last Cluster merging is a class U, stops cluster, and according to γiLocate the cluster of a classes, substitute into formula IhWith IminDetermine rainfall
Measure threshold value.
1) Transitive Closure Method
Under normal circumstances, R=(rij) fuzzy matrix is substantially fuzzy similarity matrix, itself not necessarily possesses equivalence, because
This R is also not necessarily fuzzy equivalent matrix, and using fuzzy transmission closure, it is mainly comprised the following steps：
A) cluster sample is divided into a classes according to one, obtains fuzzy similarity matrix R transitive closure, calculateUntil meeting R2n=Rn, at this moment fuzzy matrix Rn is one fuzzy etc.
Valency matrix, note will
B) clustered by the orders of parameter γ from big to small, willIt is arranged in order in sequence, γ=1 is used as initially,
According to order from big to small alongTakeObtain withCorresponding γ-interceptWhen element is equal to 1, represent
Two corresponding samples of division or variable belong to a class, and corresponding is classified as into equivalence class respectively.
C) γ values constantly reduce, and are analogized downwards by aforesaid way and obtain more merging, if
When, whole variables (or sample) are classified as a major class U, now γ=b stops cluster, are a classes by equivalence class in cluster process
γ values be defined as γi。
D) γ=γiWhen, corresponding classification situation is considered as cluster result, to all classes produce critical hour and point
Clock rainfall threshold trait is analyzed：
A1, a2 represent the classification I of clustering factor, II class, a respectively31, a32Respectively represent the IIIth class first factor and
Second factor, a1 '=0.89, a2 '=0.82 represents the coefficient of I, II classification respectively.
2) Direct Cluster Analysis
The basic conception of Direct Cluster Analysis, is completed after fuzzy similarity matrix establishment, by similar matrix, no demand transmission
Closure, so as to obtain dendrogram, here is main flow：
If a) γ1=1, to each xiMake similar class [xi]R
[xi]R={ xj|rij=1 } (15)
Represent in rijUnder conditions of 1, x is obtainediWith xjBelong to a class, as Similarity Class.Distinguish equivalence class to it is similar
Common element is not present in class, equivalence class, it is as follows：
At this moment need to merge the Similarity Class in all common elements, so as to obtain γ in level1All of equal value points of=1
Class.
If b) γ2Second largest value is represented, γ is searched in R2The all elements of similarity degree, are expressed as (xi,xj) element pair or
Person rij=γ2, with γ1=1 corresponding equivalence classification, merges xiAffiliated class and xjClass belonging to all, merges above-mentioned whole states
Afterwards, γ2Obtained value is represented, is equivalence classification.
If c) γ3Value is the third-largest value, is searched and γ in R3Similarity identical element pair, is expressed as (xi,xj), then
rij=γ3Merge and γ2The x of corresponding all equivalence classesjAffiliated class and xiAffiliated class, merges after above-mentioned whole states, acquisition
γ3For equivalence classification.
D) analogized downwards by aforesaid way, until all classes are merged into a class U, now γ=b stops cluster, will be poly-
Equivalence class is defined as γ for the γ values of a classes in class processi。
E) γ=γiWhen, corresponding classification situation is considered as cluster result, to all classes produce critical hour and point
Clock rainfall threshold trait is analyzed：
A1, a2 represent the classification I of clustering factor, II class, a respectively31, a32Respectively represent the IIIth class first factor and
Second factor, a1 '=0.89, a2 '=0.82 represents the coefficient of I, II classification respectively.
Embodiment 1
Step 1, the fuzzy clustering factor are determined
Take a broad survey in the wild, indoor synthetic analysis and the basis with reference to the research such as domestic Liu Xilin, Gong Xue, Chen Wei
On, the expert investigation for the purpose of determining landslide formative element, the mould that investigation result is finally chosen are carried out using De Feier methods
Paste clustering factor as shown in table 1, one has 10.And according to classifying importance, it is followed successively by the Ith important (rainfall erosivity, solid
Material supply mode, slump gross area rate, regional climate type), it is the IIth important (structural property, formation lithology, dissected depth because
Sub- accounting), the IIIth important (average slope of mountain slope, drainage area, the unreasonable activity of the mankind).Found in factual survey, region
This two investigation value relative difficulties of the unreasonable activity of climate type, the mankind, thus it is determined that not considering during fuzzy factor, most
8 fuzzy clustering factors are chosen eventually, are rainfall erosivity, solid matter supply mode, slump gross area rate, construction spy respectively
Property, formation lithology, dissected depth factor accounting, average slope of mountain slope, drainage area.
The fuzzy clustering factor of table 1
These three factors of solid matter supply mode, structural property and formation lithology are Qualitative factors, for ease of analysis meter
Calculate, a kind of new method proposed based on Tan Ping Yan expert-" the quantification method of the landslide basin order of severity " can be commented according to therein
Then Qualitative factor carries out Quantitative scoring to divider, and qualitative variable quantification is realized indirectly.Criteria for classification and score value are listed in table 2.
The criteria for classification of table 2 and grade form
The fuzzy cluster analysis of step 2, the Critical Rainfall threshold value that comes down
8 be have chosen with certain representativeness and oneself knows the landslide basin of Critical Rainfall, its fuzzy clustering factor parameter takes
Value is shown in Table 3.
The fuzzy clustering factor value of table 3
Raw data matrix sees below formula：
In general, it is necessary to make proper transformation to raw data matrix, with reference to data characteristicses, using standardization with extreme difference method,
Shown in formula is shown in.
Then, fuzzy resembling relation matrix r ij is constructed using correlation coefficient process, formula is as shown in 8.
After operation program, fuzzy similarity matrix R is obtained：
Step 3, carry out fuzzy clustering and determine Critical Rainfall threshold value
Fuzzy sample is divided into by four class a=4 according to classified adaptive factor standard, fuzzy similarity matrix R quadratic methods are transmitted
Closure operation, R2->R4：R4That is t (R)=R4=R*。
When taking γ=0.94, it can fall into 5 types, be respectively { A1, A3 }, { A2, A7 }, { A4 }, { A6 }, { A5, A8 }.
When taking γ=0.90,4 classes can be divided into, be respectively { A1, A3, A8 }, { A2, A7 }, { A4 }, { A6, A5 }, according to
The sample clustering constituted, then γi=0.90.
The like, as γ=0.68, all landslide basins are all gathered for a class, stop cluster.
The Critical Rainfall threshold trait of table 4 is obtained according to formula.
A1, a2 represent the classification I of clustering factor, II class, a respectively31, a32Respectively represent the IIIth class first factor and
Second factor, a1 '=0.89, a2 '=0.82 represents the coefficient of I, II classification respectively.
Constitute new sample using basin to be asked as new cluster sample and above-mentioned 8 landslides basin sample, then according to
Foregoing Fuzzy Cluster Analysis method is analyzed, in the level of γ=0.9, studies that basin to be asked and which kind of landslide basin are gathered
At one group, Critical Rainfall threshold estimation value in basin to be asked is used as using the Critical Rainfall threshold value that the group comes down.
The Critical Rainfall threshold trait of table 4
It is clear to illustrate, it is assumed that great Yu Gou landslides in Gansu are basin to be asked, and detailed process is described as follows.Obtain the density that comes down
Scatter diagram corresponding with rainfall intensity, the density that the comes down as seen from Figure 3 proportionate relationship approximate with the presentation of rainfall intensity,
The clustering factor on Gansu great Yu Gou streams landslide is shown in Table 5.
The clustering factor value of table 5
By Gansu great Yu Gou streams A9 and above 8 landslide bands constitute new cluster sample, divide according to foregoing fuzzy clustering
Analysis method is analyzed.As a result show, as γ=0.93, Dongyang Ping Gou is clustered at the Ith group, therefore, it is possible to it was initially believed that big
Valley ditch is similar to the Ith group of Critical Rainfall feature.To verify the validity of Fuzzy Cluster Analysis method, result of calculation and reality are listed
Survey Critical Rainfall to be contrasted, be shown in Table 6.
The Comparative result of table 6
| Ith group of Critical Rainfall threshold value | Survey Critical Rainfall threshold value |
| 10min raininess >=4mm | 10min raininess >=3.4mm |
| 1h raininess >=20mm | 1h raininess >=19mm |
From the results shown in Table 6, the predicted value of the prediction actual Critical Rainfalls of Gansu great Yu Gou takes the Ith group of Critical Rainfall
Threshold value, with certain feasibility.
Claims (8)
1. a kind of landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis, it is characterised in that it is specific according to
Lower step is implemented：
Step 1, the fuzzy clustering factor are determined；
The fuzzy cluster analysis of step 2, the Critical Rainfall threshold value that comes down；
Step 3, carry out fuzzy clustering and determine Critical Rainfall threshold value.
2. the landslide Critical Rainfall Threshold according to claim 1 based on method of fuzzy cluster analysis, its feature
It is, the fuzzy clustering factor is illegally determined using Dare in step 1.
3. the landslide Critical Rainfall Threshold according to claim 1 based on method of fuzzy cluster analysis, its feature
It is, the Critical Rainfall threshold value estimating techniques flow that come down in step 2 is：
1. data matrix is created
Assuming that the character that the domain for being classified object is expressed as in U={ x1, x2 ..., xn }, any object is entered by m index
Row description：(i=1,2 ..., n), the domain of U presentation class objects, x represents each of set to xi={ xi1, xi2 ..., xim }
Individual element, i, j distinguish the row and column of representing matrix, xijThe element that the i-th row j is arranged in representing matrix X.
Raw data matrix is expressed as follows：X=(xij)n×m
2. standardized data matrix
1) standard deviation is standardized
I-th of variable is standardized, the mode used is conversion xijFor xij', wherein,Jth column element average value is represented,Represent
I-th row element average value, SjRepresent variance, MjThe intermediate value of jth row is represented, max { } represents to take maximum, and min { } represents to take minimum
Value, xij' element after standardization is represented, it is as follows：
In formula：
2) standardization with extreme difference：
3) extreme difference is standardized：
4) maximum is standardized：Wherein：Mj=max (x1j, x2j..., xnj) (6)；
3. fuzzy similarity matrix is set up
Set up xi and xj similarity degrees rij=R (xi, xj), wherein main method has Similar operator and Study on similar degree method.
4. the landslide Critical Rainfall Threshold according to claim 3 based on method of fuzzy cluster analysis, its feature
It is, Similar operator is specially：
1 ° of Cosin method：
2 ° of correlation coefficient process：
Usually, r is takenij=1-c (d (xi,xj))2, wherein c, α is the parameter suitably chosen, d (xi,xj) represent xiWith xjAway from
From it causes 0≤rij≤1。rijSimilarity degree is represented, R (x) represents fuzzy matrix function, xikThe i-th row kth is arranged in representing matrix
Element, xjkJth row kth column element in representing matrix.
Hamming distance from：
Euclid distances：
Chebyshev distances：d(xi,xj)=max1≤k≤n|xik-xjk| (11)。
5. the landslide Critical Rainfall Threshold according to claim 3 based on method of fuzzy cluster analysis, its feature
It is, Study on similar degree method is specially：
1 ° of minimax method:
The minimum method of 2 ° of arithmetic averages:
The minimum method of 3 ° of geometric averages:
6. according to any described landslide Critical Rainfall Thresholds based on method of fuzzy cluster analysis of claim 3-5,
Characterized in that, using Transitive Closure Method or Direct Cluster Analysis according to the fuzzy similarity matrix R of gained in step 2 in step 3
Fuzzy clustering is carried out, and is a class U by last Cluster merging, stops cluster, and according to γiLocate the cluster of a classes, substitute into formula Ih
With IminDetermine rainfall threshold value.
7. the landslide Critical Rainfall Threshold according to claim 6 based on method of fuzzy cluster analysis, its feature
It is, Transitive Closure Method is specially：
A) cluster sample is divided into a classes according to one, obtains fuzzy similarity matrix R transitive closure, calculateUntil meeting R2n=Rn, at this moment fuzzy matrix Rn is one fuzzy etc.
Valency matrix, note will
B) clustered by the orders of parameter γ from big to small, willIt is arranged in order in sequence, γ=1 conduct is initial, according to
Order from big to small alongTakeObtain withCorresponding γ-interceptWhen element is equal to 1, represent to divide
Two corresponding samples or variable belong to a class, and corresponding is classified as into equivalence class respectively；
C) γ values constantly reduce, and are analogized downwards by aforesaid way and obtain more merging, if
When, whole variables (or sample) are classified as a major class U, now γ=b stops cluster, are a classes by equivalence class in cluster process
γ values be defined as γi；
D) γ=γiWhen, corresponding classification situation is considered as cluster result, the critical hour produced to all classes and minute rain
Amount threshold trait is analyzed：
A1, a2 represent the classification I of clustering factor, II class, a respectively31, a32Respectively represent the IIIth class first factor and second
The factor, a1 '=0.89, a2 '=0.82 represents the coefficient of I, II classification respectively.
8. the landslide Critical Rainfall Threshold according to claim 6 based on method of fuzzy cluster analysis, its feature
It is, Direct Cluster Analysis is specially：
If a) γ1=1, to each xiMake similar class [xi]R
[xi]R={ xj|rij=1 } (15)
Represent in rijUnder conditions of 1, x is obtainediWith xjBelong to a class, as Similarity Class.Equivalence class and Similarity Class are distinguished, etc.
Common element is not present in valency class, it is as follows：
At this moment need to merge the Similarity Class in all common elements, so as to obtain γ in level1=1 all equivalence classifications；
If b) γ2Second largest value is represented, γ is searched in R2The all elements of similarity degree, are expressed as (xi,xj) element pair or rij
=γ2, with γ1=1 corresponding equivalence classification, merges xiAffiliated class and xjClass belonging to all, merges after above-mentioned whole states, γ2
Obtained value is represented, is equivalence classification；
If c) γ3Value is the third-largest value, is searched and γ in R3Similarity identical element pair, is expressed as (xi,xj), then rij=
γ3Merge and γ2The x of corresponding all equivalence classesjAffiliated class and xiAffiliated class, merges after above-mentioned whole states, the γ of acquisition3For
Equivalence classification；
D) analogized downwards by aforesaid way, until all classes are merged into a class U, now γ=b stops cluster, by cluster process
Middle equivalence class is defined as γ for the γ values of a classesi；
E) γ=γiWhen, corresponding classification situation is considered as cluster result, the critical hour produced to all classes and minute rain
Amount threshold trait is analyzed：
A1, a2 represent the classification I of clustering factor, II class, a respectively31, a32Respectively represent the IIIth class first factor and second
The factor, a1 '=0.89, a2 '=0.82 represents the coefficient of I, II classification respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710153276.7A CN107092653A (en) | 2017-03-15 | 2017-03-15 | A kind of landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710153276.7A CN107092653A (en) | 2017-03-15 | 2017-03-15 | A kind of landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107092653A true CN107092653A (en) | 2017-08-25 |
Family
ID=59646209
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710153276.7A Pending CN107092653A (en) | 2017-03-15 | 2017-03-15 | A kind of landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107092653A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107609715A (en) * | 2017-10-11 | 2018-01-19 | 北京师范大学 | One kind is based on Rainstorm Feature mountain torrents Critical Rainfall computational methods |
| CN108009596A (en) * | 2017-12-26 | 2018-05-08 | 中国水利水电科学研究院 | The method and apparatus for determining characteristics of rainfall |
| CN109146144A (en) * | 2018-07-26 | 2019-01-04 | 西安工程大学 | Based on the landslide disaster prediction technique for improving Bayesian network |
| CN110378437A (en) * | 2019-07-26 | 2019-10-25 | 西安工业大学 | A kind of product module classification method for considering change propagation and influencing |
| CN110488352A (en) * | 2019-08-20 | 2019-11-22 | 中国科学院武汉岩土力学研究所 | A kind of determination method of the microseism parameter threshold for rock burst grade early warning |
| CN110633858A (en) * | 2019-09-18 | 2019-12-31 | 国网湖南省电力有限公司 | Clustering early warning method and system for rainstorm geological disasters of power transmission line |
| CN110968651A (en) * | 2019-11-18 | 2020-04-07 | 广州赛宝联睿信息科技有限公司 | Data processing method and system based on grey fuzzy clustering |
| CN111581793A (en) * | 2020-04-26 | 2020-08-25 | 中国人民解放军海军航空大学青岛校区 | Airplane structure partitioning method based on actually measured environmental data |
| CN112580871A (en) * | 2020-12-17 | 2021-03-30 | 杭州鲁尔物联科技有限公司 | Feature screening method and device |
| CN113887635A (en) * | 2021-10-08 | 2022-01-04 | 河海大学 | Basin similarity classification method and classification device |
| CN114510851A (en) * | 2022-04-20 | 2022-05-17 | 南京信息工程大学 | Assessment method for assessing landslide disaster loss induced by climate change and falling water |
| CN115170820A (en) * | 2022-05-13 | 2022-10-11 | 中铁西北科学研究院有限公司 | Feature extraction and boundary identification method applied to data curve transition stage |
| CN115587434A (en) * | 2022-09-28 | 2023-01-10 | 重庆市二零八地质环境研究院有限公司 | Evaluation method for quantitatively evaluating hysteresis effect of landslide deformation caused by rainfall |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1388122A1 (en) * | 2001-05-08 | 2004-02-11 | Sunflare Co. Ltd. | Fourier series-based automatic generation system and method for multi-variable fuzzy systems |
| CN104318103A (en) * | 2014-10-23 | 2015-01-28 | 中国科学院、水利部成都山地灾害与环境研究所 | Landslide hazard monitoring and early warning rainfall threshold judging method |
| CN106055631A (en) * | 2016-05-27 | 2016-10-26 | 成都成信高科信息技术有限公司 | Acupuncture and moxibustion prescription main acupuncture point mining method based on fuzzy combined clustering method |
-
2017
- 2017-03-15 CN CN201710153276.7A patent/CN107092653A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1388122A1 (en) * | 2001-05-08 | 2004-02-11 | Sunflare Co. Ltd. | Fourier series-based automatic generation system and method for multi-variable fuzzy systems |
| CN104318103A (en) * | 2014-10-23 | 2015-01-28 | 中国科学院、水利部成都山地灾害与环境研究所 | Landslide hazard monitoring and early warning rainfall threshold judging method |
| CN106055631A (en) * | 2016-05-27 | 2016-10-26 | 成都成信高科信息技术有限公司 | Acupuncture and moxibustion prescription main acupuncture point mining method based on fuzzy combined clustering method |
Non-Patent Citations (2)
| Title |
|---|
| 邓勇等: "模糊聚类及其在滑坡监测中的应用研究", 《测绘科学》 * |
| 陈又星等: "《管理科学研究方法 数据模型决策》", 28 February 2013, 同济大学出版社 * |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107609715A (en) * | 2017-10-11 | 2018-01-19 | 北京师范大学 | One kind is based on Rainstorm Feature mountain torrents Critical Rainfall computational methods |
| CN107609715B (en) * | 2017-10-11 | 2020-06-19 | 北京师范大学 | Torrential rain characteristic-based mountain torrent critical rainfall calculation method |
| CN108009596A (en) * | 2017-12-26 | 2018-05-08 | 中国水利水电科学研究院 | The method and apparatus for determining characteristics of rainfall |
| CN109146144B (en) * | 2018-07-26 | 2021-11-19 | 西安工程大学 | Landslide disaster prediction method based on improved Bayesian network |
| CN109146144A (en) * | 2018-07-26 | 2019-01-04 | 西安工程大学 | Based on the landslide disaster prediction technique for improving Bayesian network |
| CN110378437A (en) * | 2019-07-26 | 2019-10-25 | 西安工业大学 | A kind of product module classification method for considering change propagation and influencing |
| CN110378437B (en) * | 2019-07-26 | 2023-03-24 | 西安工业大学 | Product module classification method considering change propagation influence |
| CN110488352A (en) * | 2019-08-20 | 2019-11-22 | 中国科学院武汉岩土力学研究所 | A kind of determination method of the microseism parameter threshold for rock burst grade early warning |
| CN110633858A (en) * | 2019-09-18 | 2019-12-31 | 国网湖南省电力有限公司 | Clustering early warning method and system for rainstorm geological disasters of power transmission line |
| CN110968651A (en) * | 2019-11-18 | 2020-04-07 | 广州赛宝联睿信息科技有限公司 | Data processing method and system based on grey fuzzy clustering |
| CN111581793A (en) * | 2020-04-26 | 2020-08-25 | 中国人民解放军海军航空大学青岛校区 | Airplane structure partitioning method based on actually measured environmental data |
| CN112580871A (en) * | 2020-12-17 | 2021-03-30 | 杭州鲁尔物联科技有限公司 | Feature screening method and device |
| CN112580871B (en) * | 2020-12-17 | 2024-01-02 | 杭州鲁尔物联科技有限公司 | Feature screening method and device |
| CN113887635A (en) * | 2021-10-08 | 2022-01-04 | 河海大学 | Basin similarity classification method and classification device |
| CN113887635B (en) * | 2021-10-08 | 2022-07-01 | 河海大学 | Basin similarity classification method and classification device |
| CN114510851A (en) * | 2022-04-20 | 2022-05-17 | 南京信息工程大学 | Assessment method for assessing landslide disaster loss induced by climate change and falling water |
| CN114510851B (en) * | 2022-04-20 | 2022-06-17 | 南京信息工程大学 | Assessment method for assessing landslide disaster loss induced by climate change and falling water |
| US11796695B1 (en) | 2022-04-20 | 2023-10-24 | Nanjing University Of Information Science & Technology | Evaluation method for evaluating precipitation-induced landslide disaster loss under climate change |
| CN115170820A (en) * | 2022-05-13 | 2022-10-11 | 中铁西北科学研究院有限公司 | Feature extraction and boundary identification method applied to data curve transition stage |
| CN115587434A (en) * | 2022-09-28 | 2023-01-10 | 重庆市二零八地质环境研究院有限公司 | Evaluation method for quantitatively evaluating hysteresis effect of landslide deformation caused by rainfall |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107092653A (en) | A kind of landslide Critical Rainfall Threshold based on method of fuzzy cluster analysis | |
| Segoni et al. | Landslides triggered by rainfall: A semi-automated procedure to define consistent intensity–duration thresholds | |
| CN107610469B (en) | Day-dimension area traffic index prediction method considering multi-factor influence | |
| Guzzetti et al. | Estimating the quality of landslide susceptibility models | |
| Mwangi et al. | Forecasting droughts in East Africa | |
| Nathan et al. | Estimating low flow characteristics in ungauged catchments | |
| Han et al. | Drought forecasting based on the remote sensing data using ARIMA models | |
| Kim et al. | Urban traffic flow prediction system using a multifactor pattern recognition model | |
| Caillouet et al. | Ensemble reconstruction of spatio-temporal extreme low-flow events in France since 1871 | |
| DE112021001581T5 (en) | Failure early warning analysis method for regulation of operation for reservoir under the influence of climate change | |
| Fanos et al. | A hybrid model using machine learning methods and GIS for potential rockfall source identification from airborne laser scanning data | |
| CN109716414A (en) | A kind of multi-modal road traffic method for detecting abnormality | |
| CN105741549B (en) | A kind of traffic flow forecasting method theoretical based on space copula | |
| CN113487123B (en) | Hyperspectral monitoring and GIS (geographic information System) coupled torrential flood disaster dynamic risk assessment method | |
| CN104899437A (en) | Early-warning method for heavy-rainfall type landslide hazard | |
| CN111985389B (en) | Basin similarity discrimination method based on basin attribute distance | |
| Licznar et al. | Precipitation variability within an urban monitoring network via microcanonical cascade generators | |
| CN102184423B (en) | Full-automatic method for precisely extracting regional impervious surface remote sensing information | |
| CN105374209A (en) | Urban region road network running state characteristic information extraction method | |
| CN107169289A (en) | It is a kind of based on the Landslide Hazard Assessment method of optimal weights combination method can be opened up | |
| Potdar et al. | Toward predicting flood event peak discharge in ungauged basins by learning universal hydrological behaviors with machine learning | |
| CN107957982B (en) | Method and system for rapidly evaluating easiness of post-earthquake secondary geological disasters | |
| CN112233381B (en) | Debris flow early warning method and system based on mechanism and machine learning coupling | |
| Tang et al. | On missing traffic data imputation based on fuzzy C-means method by considering spatial–temporal correlation | |
| Raynaud et al. | Object‐based verification metrics applied to the evaluation and weighting of convective‐scale precipitation forecasts |
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 |
Application publication date: 20170825 |
|
| RJ01 | Rejection of invention patent application after publication |