CN106960086A - A kind of method based on Dam Monitoring data assessment seepage proof curtain reliability - Google Patents
A kind of method based on Dam Monitoring data assessment seepage proof curtain reliability Download PDFInfo
- Publication number
- CN106960086A CN106960086A CN201710139271.9A CN201710139271A CN106960086A CN 106960086 A CN106960086 A CN 106960086A CN 201710139271 A CN201710139271 A CN 201710139271A CN 106960086 A CN106960086 A CN 106960086A
- Authority
- CN
- China
- Prior art keywords
- dam
- seepage
- monitoring data
- component
- reliability
- 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]
- G06F30/10—Geometric CAD
- G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- Theoretical Computer Science (AREA)
- Civil Engineering (AREA)
- Structural Engineering (AREA)
- Computational Mathematics (AREA)
- Architecture (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
It is the seepage monitoring data for first collecting dam engineering the invention discloses a kind of method based on Dam Monitoring data assessment seepage proof curtain reliability：Collect and obtain the seepage monitoring data of reservoir dam, including upstream reservoir level, downstream tailwater level, temperature, precipitation, and different monoliths Uplift or the macroscopical factor of seepage discharge seepage flow behind the scenes；Then monitoring data is pre-processed, sets up dam infiltration lag model；Then the proportion shared by each component is determined using PSO BP neural networks；Finally according to the reliability and degree of aging of the ratio-dependent seepage control of dam curtain shared by each monolith different time sections upstream reservoir level component.The present invention have can quantitative assessment seepage control of dam curtain reliability and degree of aging, to ensure dam safety, play the characteristics of greatest benefit provides scientific basis, meanwhile, easy to operate with cost-effective, the characteristics of reliability is high.
Description
Technical field
It is particularly a kind of to be based on Dam Monitoring data the present invention relates to a kind of method for assessing seepage control of dam curtain reliability
The method for assessing seepage proof curtain reliability.
Background technology
By 2012, China was completed all kinds of reservoir dams 8.78 ten thousand, average 22.87 dam ages, from the age,
China's majority dams have been enter into~aged stage, seepage proof curtain is the crucial component of reservoir dam impervious leakage-stopping, for a long time
In the presence of high-pressure osmosis current, hydro-geochemical environment is complicated, and seepage proof curtain will continue to suffer from corroding and aging declines
Subtract, the reliability of seepage control of dam curtain has important influence to the stability of the dam foundation with security.
The problem of for seepage proof curtain reliability, in real work, or using on-the-spot test, the method for physico-chemical analysis,
Or using methods such as coring laboratory experiment, numerical simulations, carry out and inquire into and have made some progress.Fatigability pressure is used as in situ
Water test method assesses the reliability of seepage control system, and this method is to press the impermeabilisation energy of the intensity of variation indirect assessment curtain body of water
Power, but this method does not consider time factor, it is impossible to reflect attenuation law of the curtain body seepage capability with the time；Tried in anti-erosion room
The relation of the stripping quantity and erosion time according to CaO is tested, the antiseepage timeliness of curtain body is estimated, in practice using the difficulty of this method
Point is using accelerated ageing in which kind of technology catalysis curtain body room；It is special according to the physics and chemistry of dam site Features of Water Environment and precipitate behind the scenes
Levy, with the triangular transforming relationship of Analysis on Numerical Simulation Method water-rock-curtain body, analyze precipitate ingredient origin, can
The decay mechanism of seepage proof curtain body is studied, the antiseepage time limit is calculated.
Above method operation is cumbersome, and cost is higher, and needs certain professional knowledge, and some engineering practice persons are using greatly
Dam build up after Monitoring Data, according to the dynamic of dam foundation percolating water behind the scenes and its relation with reservoir level, set up linear function, lead to
The barrier performance and its timeliness for finding catastrophe point and variation tendency assessment curtain body are crossed, this method is easily understood, but it does not consider
The hysteresis effect of reservoir level, influence of the factors such as precipitation to percolating water is not considered, without fully excavation dam seepage monitoring data
Hiding information, causes the waste of data message, and Dam Foundation Seepage Complex Nonlinear System is portrayed with simple linear function sometimes very
It can extremely draw the wrong conclusion.
The content of the invention
It is an object of the present invention to provide a kind of method based on Dam Monitoring data assessment seepage proof curtain reliability.This
Invention with can quantitative assessment seepage control of dam curtain reliability and degree of aging, to ensure dam safety, play greatest benefit
The characteristics of scientific basis is provided, meanwhile, easy to operate with cost-effective, the characteristics of reliability is high.
Technical scheme：A kind of method based on Dam Monitoring data assessment seepage proof curtain reliability, including with
Lower step：
(1) the seepage monitoring data of dam engineering are collected：The seepage monitoring data of reservoir dam is collected and obtains, including it is upper
Trip reservoir level, downstream tailwater level, temperature, precipitation, and the Uplift or seepage discharge seepage flow behind the scenes of different monoliths are macroscopical
The factor；
(2) monitoring data is pre-processed, sets up dam infiltration lag model：
A, the data for removing substantially exception, set up model sample collection；
B, with Hydrology Period T (year, season or the moon) be unit, divide at times monolith build dam infiltration lag model；
(3) proportion shared by each component is determined using PSO-BP neutral nets；
(4) reliability of the ratio-dependent seepage control of dam curtain according to shared by each monolith different time sections upstream reservoir level component
Property and degree of aging.
In the foregoing method based on Dam Monitoring data assessment seepage proof curtain reliability, the step (2), be remove by
In data substantially abnormal caused by monitoring instrument replacing, artificial maloperation.
In the foregoing method based on Dam Monitoring data assessment seepage proof curtain reliability, the step (2), divide at times
Monolith set up dam infiltration lag model be：
Y=f (YUH, YDH, YP, YT, YA)
In formula, Y is macroscopical fluid flow factor；YUH, YDH, YP, YT, YA represent the upstream reservoir level of macroscopical fluid flow factor respectively
Component, downstream tailwater level component, precipitation component, temperature components and timeliness component.
The foregoing method based on Dam Monitoring data assessment seepage proof curtain reliability, when upstream reservoir level component is current
Reservoir level and the temporal effect amount under its early stage moment reservoir level combined influence are carved, is expressed as：
YUH=adHd=φ (H (t), H (t-1), H (t-2) ..., H (t-nH), w1, w2 ... wnH+1)
In formula, ad is upstream reservoir level regression coefficient；Hd is equivalent upstream reservoir level；H (t) is the upstream Ku Shui of t
Position；NH is upstream reservoir level lag time；Wi is the corresponding weights of i-th of reservoir level；
The foregoing method based on Dam Monitoring data assessment seepage proof curtain reliability, downstream tailwater level component can be represented
For：
YDH=a1h (t)
In formula, a1 is downstream tailwater level regression coefficient, and h (t) is the downstream tailwater level of t.
The foregoing method based on Dam Monitoring data assessment seepage proof curtain reliability, precipitation component is represented by：
YP=apPd=ξ (P (t), P (t-1), P (t-2) ..., P (t-nP), u1, u2 ... unP+1)
In formula, ap is the regression coefficient of precipitation；Pd is equivalent precipitation；P (t) is the precipitation of t；NP is precipitation
Lag time；Ui is the corresponding weights of i-th of precipitation.
The foregoing method based on Dam Monitoring data assessment seepage proof curtain reliability, temperature components are represented by：
In formula, b1i and b2i is the regression coefficient of temperature factor, i=1,2.
The foregoing method based on Dam Monitoring data assessment seepage proof curtain reliability, timeliness component uses following form table
Show：
YA=d1t+d2lnt
In formula, d1 and d2 is the regression coefficient of time effect factor.
In the foregoing method based on Dam Monitoring data assessment seepage proof curtain reliability, the step (3), each component institute
The proportion accounted for determines to comprise the following steps：
A, determine BP neural network model structure：The seepage flow lag model set up according to the step (2) selects three layers of BP
Artificial Neural Network Structures, input layer number N, hidden layer number M and output layer number K, K=1, BP neural network model structure
For N-M-K；
B, the program parameter for determining population (PSO) algorithm, including inertia weight, dimensionality of particle, particle number, learn because
Son and end condition；
C, using MATLAB software training PSO-BP model datas, power is attached to BP neural network with PSO algorithms
The optimization of value, until convergence；If mij is the connection weight between input layer j and hidden layer node i, mi1 is hidden layer section
Connection weight between point i and output layer；
D, the degree of certainty for determining hidden layer node：BP networks are error backward propagation method, true by node layer is exported
Fixed degree is through connection weight backpropagation, and output layer number K=1, output node layer degree of certainty is 1, then hidden layer node i determination
Spend for 1 × mi1；
E, the degree of certainty for determining input layer：For each hidden layer node i, each input layer j, by mij with
Mi1 is multiplied, and obtains Pij=mij × mi1, Pij is normalized, obtainInput layer j degree of certainty is
F, the weighing factor for calculating each input factor pair output factor
G, upstream reservoir level component, downstream tailwater level component, precipitation component, temperature components and timeliness component wrapped respectively
The weight summation of each factor contained, you can obtain the ratio that each component accounts for the macroscopical factor of seepage flow.
Beneficial effects of the present invention：
1st, dam seepage monitoring data have the non-linear of complexity, and the present invention collects and pre-processes Dam Monitoring number first
According to, divide monolith to set up dam infiltration lag model at times, then with through particle cluster algorithm (PSO) optimize after BP nerve nets
Network models fitting day part seepage monitoring data, each monolith upstream reservoir level component of day part is determined finally according to the weights of network
Shared proportion, can be played with the reliability and degree of aging of quantitative assessment seepage control of dam curtain, to ensure dam safety accordingly
Greatest benefit provides scientific basis.
2nd, the present invention only needs dam seepage monitoring data, without carrying out the step such as packer permeability test, in-situ sampling, laboratory experiment
Suddenly, it is cost-effective；Water-rock-curtain body couple numerical approach need not be set up, principle is simple, suitable major part engineering practice
Person operates；The present invention point monolith sets up seepage flow lag model, it can be estimated that the reliability of the curtain body of different monoliths；Will be late by because
Son add oozes control model, fully excavate dam infiltration data message, with intelligent optimization algorithm (PSO-BP) be fitted seepage flow macroscopic view because
The complex nonlinear relation between environment parameter and timeliness amount such as son and upstream reservoir level, the level of tail water, precipitation, oozes with conventional basis
Leakage quantity is compared with coefficient correlation changing rule judges seepage control system reliability method between upper pond level, and evaluation result is relatively reliable；
The initial connection weight of BP neural network is assigned at random, and the proportion shared by final each component is heavily dependent on initially
The selection of weights, the present invention is optimized using PSO algorithms to BP neural network initial weight, each component proportion value of gained
It is more accurate.The present invention only need to pass through letter in pretreated seepage monitoring data input MATLAB programs after model convergence
Single mathematical computations are the reliability that can determine whether dam foundation curtain body, simple to operate.
Brief description of the drawings
Accompanying drawing 1 is the calculation flow chart based on Dam Monitoring data assessment seepage proof curtain reliability；
Accompanying drawing 2 is BP neural network structure chart.
Embodiment
With reference to embodiment, the present invention is further illustrated, but is not intended as to the foundation of the invention limited.
Embodiments of the invention
Embodiment 1, a kind of method based on Dam Monitoring data assessment seepage proof curtain reliability, as shown in Figure 1, including
Following steps：
(1) the seepage monitoring data of dam engineering are collected：The seepage monitoring data of reservoir dam is collected and obtains, including it is upper
Trip reservoir level, downstream tailwater level, temperature, precipitation, and the Uplift or seepage discharge seepage flow behind the scenes of different monoliths are macroscopical
The factor；
(2) monitoring data is pre-processed, set up dam infiltration lag model：
A, removal set up model sample collection due to data substantially abnormal caused by monitoring instrument replacing, artificial maloperation；
B, with Hydrology Period T (year, season or the moon) be unit, divide at times monolith build dam infiltration lag model：
Upstream reservoir level, downstream tailwater level, temperature, precipitation etc. are to influence the important factor of dam infiltration, upstream Ku Shui
Influence of the position with precipitation to Dam Foundation Seepage has certain hysteresis quality, divides monolith to set up dam infiltration lag model at times：
Y=f (YUH, YDH, YP, YT, YA)
In formula, Y is macroscopical fluid flow factor；YUH, YDH, YP, YT, YA represent the upstream reservoir level of macroscopical fluid flow factor respectively
Component, downstream tailwater level component, precipitation component, temperature components and timeliness component；
A. reservoir level component in upstream can regard current time reservoir level and the wink under its early stage moment reservoir level combined influence as
When effect quantity, be represented by：
YUH=adHd=φ (H (t), H (t-1), H (t-2) ..., H (t-nH), w1, w2 ... wnH+1)
In formula, ad is upstream reservoir level regression coefficient；Hd is equivalent upstream reservoir level；H (t) is the upstream Ku Shui of t
Position；NH is upstream reservoir level lag time；Wi is the corresponding weights of i-th of reservoir level；
B. tailwater level component in downstream is represented by：
YDH=a1h (t)
In formula, a1 is downstream tailwater level regression coefficient, and h (t) is the downstream tailwater level of t；
C. precipitation and seepage flow behind the scenes are in dynamically complicated non-linear relation, mainly with precipitation, Infiltration Condition, rainfall pattern, drop
The factors such as water lasts, vegetation are relevant, similar to hysteresis effect of the reservoir level to seepage flow, and precipitation component is represented by：
YP=apPd=ξ (P (t), P (t-1), P (t-2) ..., P (t-nP), u1, u2 ... unP+1)
In formula, ap is the regression coefficient of precipitation；Pd is equivalent precipitation；P (t) is the precipitation of t；NP is precipitation
Lag time；Ui is the corresponding weights of i-th of precipitation；
D. the change of temperature can cause the change of dam foundation bedrock fracture opening width, so as to cause seepage flow behind the scenes dynamically to become
Change, seepage flow lag model of the present invention only considers the change of day and night temperature using day as period of change.Temperature components are represented by：
In formula, b1i and b2i is the regression coefficient of temperature factor, i=1,2.
E. before timeliness component and dam mud laying, uplift pressure meter (weir) surrounding soil property or basement rock fractured zones
It is closely related, represented using following form：
YA=d1t+d2lnt
In formula, d1 and d2 is the regression coefficient of time effect factor.
(3) proportion shared by each component is determined using PSO-BP neutral nets：
A, determine BP neural network model structure：The seepage flow lag model set up according to the step (2) selects three layers of BP
Artificial Neural Network Structures (as shown in Figure 2), input layer number N, hidden layer number M and output layer number K, K=1, BP god
It is N-M-K through network architecture；
B, the program parameter for determining population (PSO) algorithm, including inertia weight, dimensionality of particle, particle number, learn because
Son and end condition；
C, using MATLAB software training PSO-BP model datas, power is attached to BP neural network with PSO algorithms
The optimization of value, until convergence；If mij is the connection weight between input layer j and hidden layer node i, mi1 is hidden layer section
Connection weight between point i and output layer；
D, the degree of certainty for determining hidden layer node：BP networks are error backward propagation method, true by node layer is exported
Fixed degree is through connection weight backpropagation, and output layer number K=1, output node layer degree of certainty is 1, then hidden layer node i determination
Spend for 1 × mi1；
E, the degree of certainty for determining input layer：For each hidden layer node i, each input layer j, by mij with
Mi1 is multiplied, and obtains Pij=mij × mi1, Pij is normalized, obtainInput layer j degree of certainty is
F, the weighing factor for calculating each input factor pair output factor
G, upstream reservoir level component, downstream tailwater level component, precipitation component, temperature components and timeliness component wrapped respectively
The weight summation of each factor contained, you can obtain the ratio that each component accounts for the macroscopical factor of seepage flow.
(4) reliability of the ratio-dependent seepage control of dam curtain according to shared by each monolith different time sections upstream reservoir level component
Property and degree of aging.
Embodiment 2, a kind of method based on Dam Monitoring data assessment seepage proof curtain reliability, as shown in Figure 1, including
Following steps：
(1) the seepage monitoring data of dam engineering are collected：The seepage monitoring data of reservoir dam is collected and obtains, including it is upper
Trip reservoir level, downstream tailwater level, temperature, precipitation, and the Uplift or seepage discharge seepage flow behind the scenes of different monoliths are macroscopical
The factor；
(2) monitoring data is pre-processed, sets up dam infiltration lag model：
A, the data for removing substantially exception, set up model sample collection；
B, with Hydrology Period T (year, season or the moon) be unit, divide at times monolith build dam infiltration lag model：
Upstream reservoir level, downstream tailwater level, temperature, precipitation etc. are to influence the important factor of dam infiltration, upstream Ku Shui
Influence of the position with precipitation to Dam Foundation Seepage has certain hysteresis quality, divides monolith to set up dam infiltration lag model at times：
Y=f (YUH, YDH, YP, YT, YA)
In formula, Y is macroscopical fluid flow factor；YUH, YDH, YP, YT, YA represent the upstream reservoir level of macroscopical fluid flow factor respectively
Component, downstream tailwater level component, precipitation component, temperature components and timeliness component；
A. reservoir level component in upstream can regard current time reservoir level and the wink under its early stage moment reservoir level combined influence as
When effect quantity, be represented by：
YUH=adHd=φ (H (t), H (t-1), H (t-2) ..., H (t-nH), w1, w2 ... wnH+1)
In formula, ad is upstream reservoir level regression coefficient；Hd is equivalent upstream reservoir level；H (t) is the upstream Ku Shui of t
Position；NH is upstream reservoir level lag time；Wi is the corresponding weights of i-th of reservoir level；
B. tailwater level component in downstream is represented by：
YDH=a1h (t)
In formula, a1 is downstream tailwater level regression coefficient, and h (t) is the downstream tailwater level of t；
C. precipitation and seepage flow behind the scenes are in dynamically complicated non-linear relation, mainly with precipitation, Infiltration Condition, rainfall pattern, drop
The factors such as water lasts, vegetation are relevant, similar to hysteresis effect of the reservoir level to seepage flow, and precipitation component is represented by：
YP=apPd=ξ (P (t), P (t-1), P (t-2) ..., P (t-nP), u1, u2 ... unP+1)
In formula, ap is the regression coefficient of precipitation；Pd is equivalent precipitation；P (t) is the precipitation of t；NP is precipitation
Lag time；Ui is the corresponding weights of i-th of precipitation；
D. the change of temperature can cause the change of dam foundation bedrock fracture opening width, so as to cause seepage flow behind the scenes dynamically to become
Change, seepage flow lag model of the present invention only considers the change of day and night temperature using day as period of change.Temperature components are represented by：
In formula, b1i and b2i is the regression coefficient of temperature factor, i=1,2.
E. before timeliness component and dam mud laying, uplift pressure meter (weir) surrounding soil property or basement rock fractured zones
It is closely related, represented using following form：
YA=d1t+d2lnt
In formula, d1 and d2 is the regression coefficient of time effect factor.
(3) proportion shared by each component is determined using PSO-BP neutral nets：
A, determine BP neural network model structure：The seepage flow lag model set up according to the step (2) selects three layers of BP
Artificial Neural Network Structures (as shown in Figure 2), input layer number N, hidden layer number M and output layer number K, K=1, BP god
It is N-M-K through network architecture；
B, the program parameter for determining population (PSO) algorithm, including inertia weight, dimensionality of particle, particle number, learn because
Son and end condition；
C, using MATLAB software training PSO-BP model datas, power is attached to BP neural network with PSO algorithms
The optimization of value, until convergence；If mij is the connection weight between input layer j and hidden layer node i, mi1 is hidden layer section
Connection weight between point i and output layer；
D, the degree of certainty for determining hidden layer node：BP networks are error backward propagation method, true by node layer is exported
Fixed degree is through connection weight backpropagation, and output layer number K=1, output node layer degree of certainty is 1, then hidden layer node i determination
Spend for 1 × mi1；
E, the degree of certainty for determining input layer：For each hidden layer node i, each input layer j, by mij with
Mi1 is multiplied, and obtains Pij=mij × mi1, Pij is normalized, obtainInput layer j degree of certainty is
F, the weighing factor for calculating each input factor pair output factor
G, upstream reservoir level component, downstream tailwater level component, precipitation component, temperature components and timeliness component wrapped respectively
The weight summation of each factor contained, you can obtain the ratio that each component accounts for the macroscopical factor of seepage flow.
(4) reliability of the ratio-dependent seepage control of dam curtain according to shared by each monolith different time sections upstream reservoir level component
Property and degree of aging.
Claims (9)
1. a kind of method based on Dam Monitoring data assessment seepage proof curtain reliability, it is characterised in that：Comprise the following steps：
(1) the seepage monitoring data of dam engineering are collected：Collect and obtain the seepage monitoring data of reservoir dam, including upstream storehouse
Water level, downstream tailwater level, temperature, precipitation, and different monolith Uplift or the macroscopical factor of seepage discharge seepage flow behind the scenes；
(2) monitoring data is pre-processed, sets up dam infiltration lag model：
A, the data for removing substantially exception, set up model sample collection；
B, with Hydrology Period T (year, season or the moon) be unit, divide at times monolith build dam infiltration lag model；
(3) proportion shared by each component is determined using PSO-BP neutral nets；
(4) reliability of the ratio-dependent seepage control of dam curtain according to shared by each monolith different time sections upstream reservoir level component with
Degree of aging.
2. the method according to claim 1 based on Dam Monitoring data assessment seepage proof curtain reliability, it is characterised in that：
In the step (2), removed due to data substantially abnormal caused by monitoring instrument replacing, artificial maloperation.
3. the method according to claim 1 based on Dam Monitoring data assessment seepage proof curtain reliability, it is characterised in that：
In the step (2), the dam infiltration lag model that monolith foundation is divided at times is：
Y=f (YUH, YDH, YP, YT, YA)
In formula, Y is macroscopical fluid flow factor；YUH, YDH, YP, YT, YA represent the upstream reservoir level point of macroscopical fluid flow factor respectively
Amount, downstream tailwater level component, precipitation component, temperature components and timeliness component.
4. the method according to claim 3 based on Dam Monitoring data assessment seepage proof curtain reliability, it is characterised in that：
Upstream reservoir level component is current time reservoir level and the temporal effect amount under its early stage moment reservoir level combined influence, is represented
For：
YUH=adHd=φ (H (t), H (t-1), H (t-2) ..., H (t-nH), w1, w2 ... wnH+1)
In formula, ad is upstream reservoir level regression coefficient；Hd is equivalent upstream reservoir level；H (t) is the upstream reservoir level of t；nH
For upstream reservoir level lag time；Wi is the corresponding weights of i-th of reservoir level.
5. the method according to claim 3 based on Dam Monitoring data assessment seepage proof curtain reliability, it is characterised in that：
Downstream tailwater level component is represented by：
YDH=a1h (t)
In formula, a1 is downstream tailwater level regression coefficient, and h (t) is the downstream tailwater level of t.
6. the method according to claim 3 based on Dam Monitoring data assessment seepage proof curtain reliability, it is characterised in that：
Precipitation component is represented by：
YP=apPd=ξ (P (t), P (t-1), P (t-2) ..., P (t-nP), u1, u2 ... unP+1)
In formula, ap is the regression coefficient of precipitation；Pd is equivalent precipitation；P (t) is the precipitation of t；NP is that precipitation is delayed
Time；Ui is the corresponding weights of i-th of precipitation.
7. the method according to claim 3 based on Dam Monitoring data assessment seepage proof curtain reliability, it is characterised in that：
Temperature components are represented by：
In formula, b1i and b2i is the regression coefficient of temperature factor, i=1,2.
8. the method according to claim 3 based on Dam Monitoring data assessment seepage proof curtain reliability, it is characterised in that：
Timeliness component is represented using following form：
YA=d1t+d2lnt
In formula, d1 and d2 is the regression coefficient of time effect factor.
9. the method according to claim 1 based on Dam Monitoring data assessment seepage proof curtain reliability, it is characterised in that：
In the step (3), the proportion shared by each component determines to comprise the following steps：
A, determine BP neural network model structure：The seepage flow lag model set up according to the step (2) selects three layers of BP nerves
Network architecture, input layer number N, hidden layer number M and output layer number K, K=1, BP neural network model structure is N-
M-K；
B, the program parameter for determining population (PSO) algorithm, including inertia weight, dimensionality of particle, particle number, Studying factors with
End condition；
C, using MATLAB software training PSO-BP model datas, weights are attached to BP neural network with PSO algorithms
Optimization, until convergence；If mij is the connection weight between input layer j and hidden layer node i, mi1 is hidden layer node i
Connection weight between output layer；
D, the degree of certainty for determining hidden layer node：BP networks are error backward propagation method, will export node layer degree of certainty
Through connection weight backpropagation, output layer number K=1, output node layer degree of certainty is 1, then hidden layer node i degree of certainty is
1×mi1；
E, the degree of certainty for determining input layer：For each hidden layer node i, each input layer j, by mij and mi1 phases
Multiply, obtain Pij=mij × mi1, Pij is normalized, obtainInput layer j degree of certainty is
F, the weighing factor for calculating each input factor pair output factor
G, respectively upstream reservoir level component, downstream tailwater level component, precipitation component, temperature components and timeliness component are included
The weight summation of each factor, you can obtain the ratio that each component accounts for the macroscopical factor of seepage flow.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710139271.9A CN106960086A (en) | 2017-03-09 | 2017-03-09 | A kind of method based on Dam Monitoring data assessment seepage proof curtain reliability |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710139271.9A CN106960086A (en) | 2017-03-09 | 2017-03-09 | A kind of method based on Dam Monitoring data assessment seepage proof curtain reliability |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106960086A true CN106960086A (en) | 2017-07-18 |
Family
ID=59471277
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710139271.9A Pending CN106960086A (en) | 2017-03-09 | 2017-03-09 | A kind of method based on Dam Monitoring data assessment seepage proof curtain reliability |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106960086A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110677310A (en) * | 2018-07-03 | 2020-01-10 | 百度在线网络技术（北京）有限公司 | Traffic attribution method, device and terminal |
| CN112611353A (en) * | 2020-12-03 | 2021-04-06 | 中国水利水电科学研究院 | Dam monitoring alarm system and method |
| CN113742939A (en) * | 2021-09-15 | 2021-12-03 | 水利部南京水利水文自动化研究所 | Construction method of concrete dam effect quantity temperature component model in operation period |
| CN115062391A (en) * | 2022-07-12 | 2022-09-16 | 汉江水利水电(集团)有限责任公司丹江口水力发电厂 | Main passenger-water separation method for measured data of dam measuring weir |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104636801A (en) * | 2013-11-08 | 2015-05-20 | 国家电网公司 | Transmission line audible noise prediction method based on BP neural network optimization |
| CN104715282A (en) * | 2015-02-13 | 2015-06-17 | 浙江工业大学 | Data prediction method based on improved PSO-BP neural network |
| CN105096718A (en) * | 2015-07-03 | 2015-11-25 | 中国电建集团贵阳勘测设计研究院有限公司 | Test method and model for simulating anti-seepage curtain of reservoir dam |
| CN106295066A (en) * | 2016-08-24 | 2017-01-04 | 水利部交通运输部国家能源局南京水利科学研究院 | The method for numerical simulation of dam foundation curtain barrier performance decay |
| CN106324660A (en) * | 2016-09-29 | 2017-01-11 | 南京南瑞集团公司 | Device and method for dynamic dam safety monitoring under seismic condition |
-
2017
- 2017-03-09 CN CN201710139271.9A patent/CN106960086A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104636801A (en) * | 2013-11-08 | 2015-05-20 | 国家电网公司 | Transmission line audible noise prediction method based on BP neural network optimization |
| CN104715282A (en) * | 2015-02-13 | 2015-06-17 | 浙江工业大学 | Data prediction method based on improved PSO-BP neural network |
| CN105096718A (en) * | 2015-07-03 | 2015-11-25 | 中国电建集团贵阳勘测设计研究院有限公司 | Test method and model for simulating anti-seepage curtain of reservoir dam |
| CN106295066A (en) * | 2016-08-24 | 2017-01-04 | 水利部交通运输部国家能源局南京水利科学研究院 | The method for numerical simulation of dam foundation curtain barrier performance decay |
| CN106324660A (en) * | 2016-09-29 | 2017-01-11 | 南京南瑞集团公司 | Device and method for dynamic dam safety monitoring under seismic condition |
Non-Patent Citations (3)
| Title |
|---|
| 蔡婷婷等: "综合考虑渗流滞后效应和库水位变化速率影响的大坝渗流统计模型", 《水利水电技》 * |
| 袁理: "西霞院水库复合坝渗流安全评价", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 * |
| 钱文江等: "基于改进BP神经网络的库区渗漏量敏感性分析", 《三峡大学学报(自然科学版)》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110677310A (en) * | 2018-07-03 | 2020-01-10 | 百度在线网络技术（北京）有限公司 | Traffic attribution method, device and terminal |
| CN112611353A (en) * | 2020-12-03 | 2021-04-06 | 中国水利水电科学研究院 | Dam monitoring alarm system and method |
| CN112611353B (en) * | 2020-12-03 | 2021-11-02 | 中国水利水电科学研究院 | Dam monitoring alarm system and method |
| CN113742939A (en) * | 2021-09-15 | 2021-12-03 | 水利部南京水利水文自动化研究所 | Construction method of concrete dam effect quantity temperature component model in operation period |
| CN115062391A (en) * | 2022-07-12 | 2022-09-16 | 汉江水利水电(集团)有限责任公司丹江口水力发电厂 | Main passenger-water separation method for measured data of dam measuring weir |
| CN115062391B (en) * | 2022-07-12 | 2023-08-15 | 汉江水利水电(集团)有限责任公司丹江口水力发电厂 | Main-passenger water separation method for measured data of water measuring weir behind dam |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| McMillan | Linking hydrologic signatures to hydrologic processes: A review | |
| Chawla et al. | Isolating the impacts of land use and climate change on streamflow | |
| Chen et al. | Modeling spatially and temporally varied hydraulic behavior of a folded karst system with dominant conduit drainage at catchment scale, Hochifen–Gottesacker, Alps | |
| Dessie et al. | Water balance of a lake with floodplain buffering: Lake Tana, Blue Nile Basin, Ethiopia | |
| CN106960086A (en) | A kind of method based on Dam Monitoring data assessment seepage proof curtain reliability | |
| Zhu et al. | Sediment flux sensitivity to climate change: A case study in the Longchuanjiang catchment of the upper Yangtze River, China | |
| Qin et al. | Integrated hydrological modeling of the North China Plain and implications for sustainable water management | |
| Ahmed et al. | Groundwater flow modelling of Yamuna-Krishni interstream, a part of central Ganga Plain Uttar Pradesh | |
| Fabre et al. | Simulating past changes in the balance between water demand and availability and assessing their main drivers at the river basin scale | |
| Miao et al. | Spatio-temporal variability of streamflow in the Yellow River: possible causes and implications | |
| He et al. | An integrated uncertainty and ensemble-based data assimilation approach for improved operational streamflow predictions | |
| Wang et al. | Groundwater response to leakage of surface water through a thick vadose zone in the middle reaches area of Heihe River Basin, in China | |
| Chiamsathit et al. | Inflow forecasting using Artificial Neural Networks for reservoir operation | |
| CN104500051A (en) | Method for determining mixed stratum water macroscopic capture cross section of water flooded layer | |
| Weiß et al. | Model-based scenarios of Mediterranean droughts | |
| Liu et al. | Application and analysis of debris-flow early warning system in Wenchuan earthquake-affected area | |
| Baymani-Nezhad et al. | Hydrological modeling using effective rainfall routed by the Muskingum method (ERM) | |
| Hall et al. | Understanding flood regime changes in Europe: a state of the art assessment | |
| Hani et al. | On the use of finite‐difference and neural‐network models to evaluate the impact of underground water overexploitation | |
| Wang et al. | A modified Xin'anjiang model and its application for considering the regulatory and storage effects of small-scale water storage structures | |
| Li et al. | A modified simple dynamic model: Derived from the information embedded in observed streamflows | |
| Saifullah et al. | Hydrological response of the kunhar river basin in pakistan to climate change and anthropogenic impacts on runoff characteristics. Water 2021, 13, 3163 | |
| Makungo | Development of risk-based groundwater operating rules: A case study of Siloam Village, South Africa | |
| Rwigi | Comparative case study of rainfall-runoff models over the Nyando River Basin | |
| Fabre et al. | Simulating long-term past changes in the balance between water demand and availability and assessing their main drivers at the river basin management scale |
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: 20170718 |