CN106503326A - A kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount - Google Patents
A kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount Download PDFInfo
- Publication number
- CN106503326A CN106503326A CN201610907830.1A CN201610907830A CN106503326A CN 106503326 A CN106503326 A CN 106503326A CN 201610907830 A CN201610907830 A CN 201610907830A CN 106503326 A CN106503326 A CN 106503326A
- Authority
- CN
- China
- Prior art keywords
- grades
- water quality
- river
- mainstream
- river course
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A20/00—Water conservation; Efficient water supply; Efficient water use
- Y02A20/152—Water filtration
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Biomedical Technology (AREA)
- Genetics & Genomics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Physiology (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount, step 1, water quality optimizing model is set up, carry out simulation of water quality, determine decision variable and constraints, different operating modes are produced by changing the flow set in each two grades of river courses;Step 2, using the water quality model analog result under different operating modes as train samples, build neutral net and neutral net is trained, nonlinear function approximation is realized, enables the network after training to predict the water quality condition at one-level river course control section;Step 3, randomly generated by the use of genetic algorithm a number of individual and as initial population, judgement screening is carried out to individuality by the use of trained perfect neutral net as constraints, select to meet the high individuality of condition of water quality and fitness, new population is obtained by intersecting, making a variation, the optimization design scheme of each two grades of discharges of river regulation and control is obtained.The present invention proposes more perfect design to river water quality simulation technology.
Description
Technical field
The present invention relates to river course discharge capacity scheduling Reverse Design, more particularly to a kind of a plurality of two grades of river courses are optimum to be adjusted
The Reverse Design of degree discharge capacity, the method can achieve mainstream Water-quality control, mainstream water quality index is met in control section
Control is required.
Background technology
Environmental fluid dynamics model (EFDC) is the collective model that is developed according to multiple Mathematical Modeling Integrated Developments, catchments
Power plant module, Sediment Transport module, Contaminants Transport module and water quality prediction module one, can be used for river, lake, reservoir
Etc. one-dimensional, two and three dimensions physics, the simulation of chemical process.EFDC is extensive used in the research in river water quality simulation direction.
Genetic algorithm (GA) is a kind of based on natural selection principle and search (optimizing) algorithm of natural genetic mechanism, and it is
Life concern mechanism in simulation nature, realizes the optimization of specific objective in manual system.The essence of genetic algorithm is logical
Collective search technology is crossed, the principle according to the survival of the fittest was evolved by generation, finally gave optimal solution or quasi-optimal solution.It must do with
Lower operation:The generation of initial population, seek each individual fitness, defect individual, selected is selected according to the principle of the survival of the fittest
The defect individual for going out is matched two-by-two, raw after intersecting the gene of gene random variation some chromosomes of its chromosome at random
Into colony of future generation, colony is according to said method made to evolve by generation.Compared with other optimized algorithms, genetic algorithm is more easy to explore and entirely sets
Meter space, and find globally optimal solution.
Some complicated nonlinear systems are frequently encountered in engineer applied, and these system state equations are complicated, it is difficult to
Mathematically accurate modeling, such as constraints " relation between the flow in two grades of river courses and mainstream downstream water quality " cannot use letter
Number is represented.In such a case, it is possible to setting up BP neural network expresses these nonlinear systems, the method regards unknown system as
Be a black box, BP neural network is trained with system inputoutput data first, enable the network to express the unknown function, then
Just can be with the BP neural network forecasting system output for training.
Content of the invention
Prior art is based on, the present invention proposes the reverse engineer side that a kind of many two grades of river course maximums of determination enter mainstream amount
Method, the water quality index according to mainstream on control section are required, are determined using Reverse Design and optimized per two grades of river courses of bar
Maximum scheduling drain into mainstream flow, can achieve mainstream Water-quality control, make mainstream water quality index meet control in control section
Require.
The present invention is that a kind of many two grades of river course maximums of determination enter the Reverse Design that mainstream is measured, and the method includes following step
Suddenly:
Step 1, simulation of water quality is carried out to the river for having a plurality of two grades of river courses using environmental fluid dynamics, set up water quality
Optimized model, determines decision variable and constraints, and produces different operating modes by changing the flow set in each two grades of river courses;
Step 2, using the water quality model analog result under different operating modes as train samples, by each two grades of river courses
Used as input, the water quality parameter at one-level river course control section builds neutral net and neutral net is entered as output flow
Row training, realizes nonlinear function approximation, enables the network after training to predict the water quality condition at one-level river course control section;
Step 3, randomly generated using genetic algorithm a number of individual and as initial population, using step 2
In trained perfect neutral net carry out judgement screening to individuality as constraints, select to meet condition of water quality and adaptation
The high individuality of degree, obtains new population by crossing operation, mutation operator, repeats this process, filter out optimum individual eventually, obtain
Each two grades of river courses enter the optimization design scheme of the mainstream water yield, i.e., meeting before at the control section of mainstream downstream, condition of water quality is up to standard
Put, make that two grades of river sewages are as much as possible to be entered in mainstream.
Compared with prior art, the present invention combines two aspect technology of genetic algorithm and neutral net, to river water quality simulation
Technology proposes more perfect design.
Description of the drawings
Fig. 1 is that a kind of many two grades of river course maximums of the present invention enter the Reverse Design illustraton of model that mainstream amount determines;
Stream Water Quality Models schematic diagrames of the Fig. 2 for the specific embodiment of the invention;
Fig. 3 is change of the desired value with iterations in genetic algorithm computing;
Fig. 4-7 is water quality spatial distribution in the river course under optimal case.
Specific embodiment
A kind of many two grades of river course maximums of determination of the present invention enter the Reverse Design method of mainstream amount, using genetic algorithm with
EFDC combines, and water quality of river is simulated and is designed, and produces new individual by the intersection of genetic algorithm, mutation process, makes
With artificial neural network as genetic algorithm in constraints, judge new individual whether meet condition of water quality and calculate its adapt to
Degree, finally filters out optimum individual, obtains the best design that two grades of river courses enter the mainstream water yield.
Below in conjunction with the drawings and the specific embodiments, technical scheme is described in further detail.
Two grades of river course optimization designs, are simulated to mainstream section water quality using EFDC water quality models, simulated domain total length
17.4 kilometers, the stress and strain model of water quality model is as shown in Figure 2.Assume that the mainstream of survey region becomes a mandarin and enter mainstream water with two grades of river courses
Amount is steady state value.For probing into impact of each two grades of Drain contamination for river channel flows to mainstream water quality, enter mainstream by changing two grades of river courses
The water yield creates different simulated scenarios, calculates the water quality at the control section of mainstream under different operating modes, is probed into when each two grades of river courses then
Impact when the mainstream water yield discharges pollutants to mainstream water quality is entered with difference.Using principle is reverse-engineered, control is reached with mainstream
Water standard is target, and Converse solved each two grades of river courses enter the mainstream water yield, realize mainstream water quality reaching standard.
The specific embodiment of the invention is described as follows:
1st, EFDC simulations mainstream water quality
Under different operating modes, mainstream upstream flowrate is definite value Q=100m3/s, and each two grades of river courses enter the mainstream water yield as variable,
Its water quality boundary condition is shown in Table 1.For simplified model, the 1st article of two grades of river courses are entered the mainstream water yield, the 2nd article of two grades of river courses and enters mainstream
The water yield, the 3rd article of two grades of river courses enter the mainstream water yield, the 4th article of two grades of river courses and enter the mainstream water yield, the 5th article of two grades of river courses and enter the mainstream water yield
And the 6th article of two grades of river courses enter the mainstream water yield as 6 variable Xs 1, X2, X3, X4, X5, X6, each variable has 5 levels.
With reference to the test orthogonal arrage of 6 factor, 5 level, the model that sets up under 27 different operating modes is simulated and calculates.
Table 1, mainstream and each two grades of river water quality units:mg/L
2nd, neural network constraints
The method combined with genetic algorithm using neutral net is drained flow to two grades of river courses and is reverse-engineered, by six
Two grades of river courses drain flow as input, and 27 kinds of operating modes that EFDC is calculated are made by the water quality at the control section of mainstream as output
Neutral net is trained for learning sample, makes artificial neural network on the premise of known each two grades of river courses drain flow,
Every water quality at the control section of mainstream is predicted.
3rd, genetic algorithm is combined the reverse engineer for calculating optimal case with neutral net, EFDC
Determine decision variable and constraints.In this research, variable is six two grades of discharges of river X1~X6, constraints is
Four water standards at the control section of mainstream, the span of two grades of discharges of river of each bar are shown in Table respectively with mainstream water standard
2 with table 3.
2, two grades of discharge of river excursion units of table:m3/s
Table 3, mainstream control section water standard unit:mg/L
(1) hard objectives function.Object of this investigation is on the premise of mainstream condition of water quality is met, and makes in two grades of river courses
Storm runoff is drained in mainstream as much as possible, and that seeks each two grades of river courses enters mainstream amount optimization design scheme.Object function:
max f(x1, x2, x3, x4, x5, x6)=x1+x2+x3+x4+x5+x6.
(2) determine coding method.Six two grades of river courses, i.e. variable is had in this research for 6, using real coding.Really
Determine individual evaluation method.This research is maximum problem, and fitness is equal to target function value.That is individual adaptation degree
F=x1+x2+x3+x4+x5+x6.
(3) operational factor of design genetic operator and determination genetic algorithm:
1), wheel disc selection opertor is used
1-1, each chromosome U of calculatingkFitness value
eval(Uk)=f (x1, x2, x3, x4, x5, x6)
Wherein, K represents the numbering of individual in population, k=1,2 ... ... 49,50;
1-2, the summation of the fitness value of calculating colony
1-3, calculating are corresponding to each chromosome UkSelect probability Pk:
1-4, calculate each chromosome UkAccumulated probability Qk:
Select defect individual meet following condition:
I. random number r between [0,1] is generatedkGive birth to 50 r in (k=1~50), common property;
If ii. r Q1, with regard to selective staining body U1, otherwise, select k-th chromosome UkSo that Qk-1≦r≦Qk.
3) crossing operation uses single-point crossover operator
The node of one chromosome string of random selection, then exchanges the part of two parent one right-hand members producing filial generation.
Crossover probability Pc=0.8, i.e., the chromosome for having 80% in average level are intersected.
Select defect individual meet following condition:
I. random number r between [0,1] is generatedkGive birth to 50 r in (k=1~50), common property;
If ii. rk≦Pc, with regard to selective staining body UkFor the individual for intersecting.
3) mutation operator uses basic bit mutation operator
Mutation probability Pm=0.1
So far, the individual evolution of a generation is completed, repeats this process.In 200 iterative calculation, desired value is with iteration time
Several changes such as Fig. 4.The optimal case for obtaining is shown in Table 4.
Under table 4, optimal case, each two grades of river courses drain flow
Under optimal case, at the control section of mainstream, water quality the result is shown in Table 5.As can be seen that in mainstream control section,
Water quality meets control standard.
Table 5, mainstream control section water quality
Claims (4)
1. a kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount, it is characterised in that the method includes following
Step:
Step (1), simulation of water quality is carried out to the river for having a plurality of two grades of river courses using environmental fluid dynamics, set up water quality excellent
Change model, determine decision variable and constraints, and different operating modes are produced by changing the flow set in each two grades of river courses;
Step (2), using the water quality model analog result under different operating modes as train samples, by each two grades of river course stream
Used as input, the water quality parameter at one-level river course control section builds neutral net and neutral net is carried out as output amount
Training, realizes nonlinear function approximation, enables the network after training to predict the water quality condition at one-level river course control section;
Step (3), randomly generated using genetic algorithm a number of individual and as initial population, using step (2)
In trained perfect neutral net carry out judgement screening to individuality as constraints, select to meet condition of water quality and adaptation
The high individuality of degree, obtains new population by crossing operation, mutation operator, repeats this process, filter out optimum individual eventually, obtain
Each two grades of river courses enter the optimization design scheme of the mainstream water yield, i.e., meeting before at the control section of mainstream downstream, condition of water quality is up to standard
Put, make that two grades of river sewages are as much as possible to be entered in mainstream.
2. a kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount as claimed in claim 1, and its feature exists
In the selection in step (3) meets the process of the high individuality of condition of water quality and fitness and specifically includes below scheme:
Calculate each chromosome UkFitness value
eval(Uk)=f (x1, x2, x3, x4, x5, x6)
Wherein, K represents the numbering of individual in population, k=1,2 ... ... 49,50;
Calculate the summation of the fitness value of colony
Calculate corresponding to each chromosome UkSelect probability Pk:
Calculate each chromosome UkAccumulated probability Qk:
Select defect individual meet following condition:
Generate random number r between [0,1]kGive birth to 50 r in (k=1~50), common property;
If r is Q1, with regard to selective staining body U1, otherwise, select k-th chromosome UkSo that Qk-1≦r≦Qk.
3. a kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount as claimed in claim 1, and its feature exists
In the crossing operation in step (3) specifically includes below scheme:
The node of one chromosome string of random selection, then exchanges the part of two parent one right-hand members producing filial generation.Intersect
Probability P c=0.8, i.e., the chromosome for having 80% in average level are intersected;
Select defect individual meet following condition:
Generate random number r between [0,1]kGive birth to 50 r in (k=1~50), common property;
If rk≦Pc, with regard to selective staining body UkFor the individual for intersecting.
4. a kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount as claimed in claim 1, and its feature exists
In the mutation operator in step (3) specifically includes below scheme:
Using basic bit mutation operator, i.e. mutation probability Pm=0.1, the individual evolution of a generation so far, is completed, repeats this process.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610907830.1A CN106503326A (en) | 2016-10-18 | 2016-10-18 | A kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610907830.1A CN106503326A (en) | 2016-10-18 | 2016-10-18 | A kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106503326A true CN106503326A (en) | 2017-03-15 |
Family
ID=58293850
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610907830.1A Pending CN106503326A (en) | 2016-10-18 | 2016-10-18 | A kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106503326A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109034442A (en) * | 2018-05-25 | 2018-12-18 | 天津大学 | A kind of prediction technique of heavy rain period river mainstream eutrophic state response |
CN109272171A (en) * | 2018-10-16 | 2019-01-25 | 天津大学 | A method of the draining of prediction storm period second level river is to mainstream water quality dynamic effects |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101824807A (en) * | 2010-04-16 | 2010-09-08 | 同济大学 | Method for short-term prediction of salt tide at water intake in tidal river region |
CN103426027A (en) * | 2013-07-24 | 2013-12-04 | 浙江大学 | Intelligent normal pool level optimal selection method based on genetic neural network models |
CN105389442A (en) * | 2015-12-09 | 2016-03-09 | 天津大学 | Reverse design method for coupling genetic algorithm, neural network and numerical simulation |
-
2016
- 2016-10-18 CN CN201610907830.1A patent/CN106503326A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101824807A (en) * | 2010-04-16 | 2010-09-08 | 同济大学 | Method for short-term prediction of salt tide at water intake in tidal river region |
CN103426027A (en) * | 2013-07-24 | 2013-12-04 | 浙江大学 | Intelligent normal pool level optimal selection method based on genetic neural network models |
CN105389442A (en) * | 2015-12-09 | 2016-03-09 | 天津大学 | Reverse design method for coupling genetic algorithm, neural network and numerical simulation |
Non-Patent Citations (1)
Title |
---|
刘磊等: "暴雨径流对海河干流水质的影响及排干水质要求的确定", 《安全与环境学报》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109034442A (en) * | 2018-05-25 | 2018-12-18 | 天津大学 | A kind of prediction technique of heavy rain period river mainstream eutrophic state response |
CN109272171A (en) * | 2018-10-16 | 2019-01-25 | 天津大学 | A method of the draining of prediction storm period second level river is to mainstream water quality dynamic effects |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Afshar et al. | Particle swarm optimization for automatic calibration of large scale water quality model (CE-QUAL-W2): Application to Karkheh Reservoir, Iran | |
CN106779151B (en) | A kind of line of high-speed railway wind speed multi-point multi-layer coupling prediction method | |
Haghighi et al. | Uncertainty analysis of water supply networks using the fuzzy set theory and NSGA-II | |
CN107358021A (en) | DO prediction model establishment method based on BP neural network optimization | |
Karamouz et al. | Application of genetic algorithms and artificial neural networks in conjunctive use of surface and groundwater resources | |
CN108595905A (en) | A kind of erosion failure quantitative forecasting technique based on BP neural network model | |
CN103105246A (en) | Greenhouse environment forecasting feedback method of back propagation (BP) neural network based on improvement of genetic algorithm | |
Bozorg-Haddad et al. | A self-tuning ANN model for simulation and forecasting of surface flows | |
Saadatpour et al. | Waste load allocation modeling with fuzzy goals; simulation-optimization approach | |
Ghimire et al. | Development of stage-discharge rating curve in river using genetic algorithms and model tree | |
CN102222313A (en) | Urban evolution simulation structure cell model processing method based on kernel principal component analysis (KPCA) | |
Ning et al. | GA-BP air quality evaluation method based on fuzzy theory. | |
CN105184415A (en) | Power distribution network reconstruction design method | |
CN113449930A (en) | Optical fiber preform preparation quality prediction method based on BP neural network | |
CN115948964A (en) | Road flatness prediction method based on GA-BP neural network | |
CN106503326A (en) | A kind of many two grades of river course maximums of determination enter the Reverse Design of mainstream amount | |
CN105160422A (en) | Sewage treatment effluent total phosphorus prediction method based on self-organization cascade neural network | |
CN117350146A (en) | GA-BP neural network-based drainage pipe network health evaluation method | |
CN105372995A (en) | Measurement and control method for sewage disposal system | |
Zhang et al. | Effluent Quality Prediction of Wastewater Treatment System Based on Small-world ANN. | |
CN102663230B (en) | Method for land resource evaluation factor level classification based on genetic algorithm | |
Tayfur | GA-optimized model predicts dispersion coefficient in natural channels | |
CN113743003B (en) | Method for calculating intensity of high-voltage line to ground electric field by considering influence of temperature and humidity | |
CN113343601A (en) | Dynamic simulation method for water level and pollutant migration of complex water system lake | |
Rezapour et al. | Review of genetic algorithm model for suspended sediment estimation |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170315 |
|
WD01 | Invention patent application deemed withdrawn after publication |