CN102467605A - Pollution source tracking and positioning information processing method of sudden water supply pipe network pollution accident - Google Patents
Pollution source tracking and positioning information processing method of sudden water supply pipe network pollution accident Download PDFInfo
- Publication number
- CN102467605A CN102467605A CN201010534558XA CN201010534558A CN102467605A CN 102467605 A CN102467605 A CN 102467605A CN 201010534558X A CN201010534558X A CN 201010534558XA CN 201010534558 A CN201010534558 A CN 201010534558A CN 102467605 A CN102467605 A CN 102467605A
- Authority
- CN
- China
- Prior art keywords
- node
- water supply
- pollution source
- pollution
- information processing
- 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.)
- Granted
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a pollution source tracking and positioning information processing method of a sudden water supply pipe network pollution accident, which comprises the following steps that: 1) acquisition devices acquire pollutant concentration information monitored by monitoring points and transmit the pollutant concentration information to a processor; 2) the processor works out the maximum pollutant concentration of all nodes in a pipe network and the occurrence time of the maximum pollutant concentration according to the linear rule model of a node relation tree, and by conducting solution step by step, the maximum pollutant concentration of all nodes in the pipe network and the occurrence time of the maximum pollutant concentration can be worked out; and 3) the processor judges a node with the highest pollutant concentration and determines the node to be a pollutant injection node. Compared with the prior art, the method has the advantages that the pollution resource in the water supply pipe network can be rapidly and accurately positioned and the like.
Description
Technical field
The present invention relates to a kind of pollution source tracing and positioning information processing method, especially relate to a kind of pollution source tracing and positioning information processing method of water supply network burst contamination accident.
Background technology
The water quality safety of water supply network receives urban government day by day, the concern of the water undertaking and even the whole society.When pollution source got into water system from entrance, pollutant can spread all over whole pipe network rapidly, to residents ' health and life security formation grave danger.At present, China's water-supply systems water quality monitoring technology is backward relatively, and most of city does not still possess the ability of perfect online water quality monitoring.Yet the threat of water supply network burst contamination accident exists constantly.
China city is main with central water supply mostly; Because the pipe network complex structure is huge, actual water supply pipe net system constitutes complicated and is laid in undergroundly, and is interconnected; In case accident takes place; Cause pollutant to get into the distribution system of water supply, water quality is unusual or when user's complaint is recognized water pollution information when monitoring from the monitoring point, pollutes injection phase and contaminated zone if can't locate fast; Then can't be pointedly pollution source be taken measures, the pipe network of pollution range is isolated and contaminant water is handled, but only depend on the artificial experience decision-making to be difficult to accurate judgement is made in the position of pollution source and the spread condition of pollutant.In recent years Song Hua River chemical pollution, Taihu Lake blue-green algae had taken place and incident such as had broken out in China once, caused cutting off the water for a long time of cities such as Harbin, Wuxi on a large scale, had caused serious social influence.Therefore, under the situation that the pipe network contamination accident has taken place, need utilize limited Monitoring Data and information to carry out the quick location of pollution source, thereby the rapid and precise decision support is provided for influence evaluation, control and the elimination of accident.
Water supply network pollution source tracing and positioning method is meant after accident takes place; Water pollution data according to the online Water-quality Monitoring Points monitoring that is provided with in the pipe network; In conjunction with the topological sum hydraulic structure of pipe network, utilize certain method or technology to confirm the position of pollution source and the time that pollution is injected.
Because the water supply network complex structure is not at present still by the pollution source recognition technology method of extensive approval and application.
A kind of new method (simulation optimization method) of having set up Jiabao Guan etc. solves pollution source and the historical definite problem of release thereof in the complicated distribution system of water supply.This method is talked about to analyze with optimum and is the basis, and utilization EPANET (Environmental Protection Agency's water supply network simulation softward is annotated: famous software in the industry, there is not the corresponding title of ready-made Chinese, more than be given the free translation title) middle water distribution system model.In the method, through confirming to be arranged in the discharging history that distribution system of water supply potential quality is appointed as the pollution source node, use the pollutant levels that generate the monitoring point of forcing appointment with EPANET.These information are used to continuous optimization prediction correction algorithm and confirm pollution source and discharge historical.In simulation process, suppose that the water distribution system hydraulics that is studied is known.The historical optimal model of discharging in order to proofread and correct pollution source mainly is designed to discern the similarity between analog result and the monitoring point measured data.This message exchange is carried out with convergence algorithm in the closed hoop system.
When adopting said method, all nodes in the distribution system of water supply are no matter its complexity how, all will be selected as possible pollution source point and monitoring point, so the cost that calculates is than higher.
Authors such as Cristiana Di Cristo have described a straightforward procedure of the pollution source position of the accident contamination in the distribution system of water supply of location.This algorithm begins from the solute concentration data that measure, and comes the at first scope of limit pollution source decanting point through analyzing the pollution matrix, confirms that is selected a node fully.Confirm that through making a linear optimization problem desired value minimum most possible in institute's reconnaissance is the point of pollution source.Wherein linear optimization problem is to set up according to the analogue value and the minimum principle of measured value difference.
This method is calculated comparatively simple, but what the influence of the examined point of its result's reliability is very big, and not good to the locating effect of the situation of polluting (non-water source node) in the middle of the pipe network and many pollution source.
Summary of the invention
The object of the invention is exactly the pollution source tracing and positioning information processing method that a kind of water supply network burst contamination accident is provided for the defective that overcomes above-mentioned prior art existence.
The object of the invention can be realized through following technical scheme:
A kind of pollution source tracing and positioning information processing method of water supply network burst contamination accident is characterized in that, may further comprise the steps:
1) collector is gathered the pollutant levels information that the monitoring point monitors, and sends it to processor;
2) processor is obtained the greatest contamination substrate concentration of its associated upstream node and the time that the greatest contamination substrate concentration occurs through the linear programming model of node relationships tree; Find the solution down step by step, the greatest contamination substrate concentration that can obtain all nodes in the pipe network obtains the time of greatest contamination substrate concentration with it;
3) processor judges that the maximum node of pollutant levels is predicted as pollutant and injects node.
The linear programming modelling process of described node relationships tree is following:
1) structure node relational tree;
2) generate training dataset;
3) data clusters;
4), construct a linear programming that is used for representing the relational tree between the relative upstream node of node to each relational tree;
5) utilize the linear programming of the relational tree in the step 4), obtain the greatest contamination substrate concentration of all relevant upstream nodes of each node and the time of greatest contamination substrate concentration appearance;
6) judge whether to tally with the actual situation, if yes, judge that this model can use, if not, return step 1).
Relation between described node relationships tree description node and its associated upstream node.
Described generation training dataset step is following:
Inject through the EPANET simulating pollution, given at random pollutant injects node, sets certain pollutant then and continues injection length and simulation process duration, provides and injects the concentration range that pollutant injects, and obtains training dataset.
Described data clusters is classified training dataset for adopting the Kmeans algorithm.
The linear programming of described relational tree comprises feasible zone and linear regression formula, and linear programming is changed into the If-Then form, and If partly is a feasible zone, and Then partly is the linear regression formula.
Node in the described step 5) is according to the condition of linear programming and the concentration value of this node; Selection rule carries out the maximum contaminant level value that the upstream node relevant with node found the solution in linear programming; If the rule of choosing is tried to achieve not have and is separated; Just the suitable rule of alternative is found the solution in Else Rule, if strictly all rules does not satisfy solving condition or tries to achieve not have and separate, abandons so finding the solution its upstream node value by the present node value.
Compared with prior art, the present invention has and locatees pollution source in the water supply network quickly and accurately, for the dispatcher confirms pollution source and formulate and implement solution strong foundation is provided quickly and accurately.
Description of drawings
Fig. 1 is a model process flow diagram of the present invention;
Fig. 2 is a hardware configuration synoptic diagram of the present invention;
Fig. 3 connects the tree graph that concerns between input and the output for the relation between the upstream and downstream node of the present invention can be converted into;
Fig. 4 is a pipe network exemplary plot of the present invention;
Fig. 5 is relational tree (RTa) figure of node a of the present invention;
Fig. 6 is relational tree (RTb) figure of node b of the present invention;
Fig. 7 is relational tree (RTc) figure of node c of the present invention;
Fig. 8 is that the linear programming of utilizing relational tree of the present invention is by the anti-synoptic diagram that pushes away input value of output valve;
Fig. 9 is the water supply network model synoptic diagram of embodiment 2;
Figure 10 is upstream and downstream relation (relational tree) figure of each node of embodiment 2;
Figure 11 is the pollution source tracing process process flow diagram of embodiment 2.
Embodiment
Below in conjunction with accompanying drawing and specific embodiment the present invention is elaborated.
Like Fig. 1, shown in Figure 2, a kind of pollution source tracing and positioning information processing method of water supply network burst contamination accident may further comprise the steps:
1) collector A gathers the pollutant levels information that the monitoring point monitors, and sends it to processor B;
2) processor B is obtained the greatest contamination substrate concentration of its associated upstream node and the time that the greatest contamination substrate concentration occurs through the linear programming model of node relationships tree; Find the solution down step by step, the greatest contamination substrate concentration that can obtain all nodes in the pipe network obtains the time of greatest contamination substrate concentration with it;
3) processor B judges that the maximum node of pollutant levels is predicted as pollutant and injects node.
The linear programming modelling process of described node relationships tree is following:
1) structure node relational tree;
2) generate training dataset;
3) data clusters;
4), construct a linear programming that is used for representing the relational tree between the relative upstream node of node to each relational tree;
5) utilize the linear programming of the relational tree in the step 4), obtain the greatest contamination substrate concentration of all relevant upstream nodes of each node and the time of greatest contamination substrate concentration appearance;
6) judge whether to tally with the actual situation, if yes, judge that this model can use, if not, return step 1).
The detailed process of algorithm is following:
The first step:
Pollution source tracing and positioning problem is changed into an oriented relational tree that connects input and output, and relational tree has been described the relation between node and its associated upstream node, like Fig. 3.The rule that concerns between the upstream and downstream node through relational tree can be predicted output according to input.
1, tectonic relationship tree
In the POLLUTION SIMULATION process, need the structure node relational tree.If upstream node has water directly to flow to node i and is the node relevant with node i, the relational tree RTi of node i has provided other all nodes adjacent and relevant with node i.Accompanying drawing 4 is simple pipe network models, and the direction of arrow among the figure is represented the direction of current in the pipeline, and node a and node b have arrow points node c, and the water of expression node c derives from node a and node b, and then relevant with node c upstream node is node a and b.According to the direction of current, can construct the relational tree (RTi) of each node, like accompanying drawing 5, Fig. 6, shown in Figure 7.The input/output relation of the relational tree of each node is seen table 1.
Input of the relational tree of each node of table 1 and output
2, generate training dataset
Inject through the EPANET simulating pollution, given at random pollutant injects node, sets certain pollutant then and continues injection length and simulation process duration, provides and injects the concentration range that pollutant injects, and obtains training dataset.
Inject simulation process in primary pollution, write down all nodes maximum contaminant level that can reach and the time that obtains maximum contaminant level, so just formed a sample of training data.In order to obtain the more data sample, used random function srand () and rand () here, as long as generate a pollution concentration C and an injection length T at random, just obtained a sample.
3, data clusters
Kmeans algorithm (K-mean algorithm) is adopted in this invention, is divided into k disjoint point set to data set.Adopt the average (formula 1) in each group to define between input point apart from d as cluster centre and with Euclidean distance (formula 2); With the quadratic sum of cluster centre distance interior difference (formula 3),, difference searches for the cluster C of measured value x in Euclidean space through being minimized as cluster.Be defined as the distance (formula 4) between cluster centre to the cluster differences.
R
k＝∑x/n
k(x∈C
k) (1)
R
k-cluster centre;
The value of x-sample;
wc(C)＝wc(C
k)＝∑[d(x，R
k)]
2(k＝1，…，K；x∈C
k) (3)
Difference in the wc-cluster;
D (x, R
kThe Euclidean distance of)-sample point and cluster centre;
bc(C)＝∑d(R
j，R
k)·d(R
j，R
k)(1≤j＜k≤K) (4)
Bc (C)-cluster differences
The Kmeans algorithm has a variety of variants, and the present invention adopts basic version.K cluster centre begin from picking at random, is assigned to each point in the cluster near its average according to Euclidean distance again, calculates the mean vector of the point that is assigned to each cluster then.And carry out recurrence as new center.
Specific algorithm is following, tentation data point D={x1 ..., xn}, task be find K cluster C1 ..., Ck}.
For k=1 ..., K begins the point of r (k) for picked at random from D;
While changes in cluster Ck do takes place
Form cluster:
for?k＝1，...，K?do
Ck={x ∈ D | d (Rk, x)≤d (Rj, x) to all j=1 ..., K, j ≠ k};
end；
Calculate new cluster centre:
for?k＝1，...，K?do
The mean vector of point in the Rk=Ck
end；
end；
Second step:
To each relational tree, construct a linear programming that is used for representing the relation between the relative upstream node of node, linear programming is made up of feasible zone and linear regression formula two parts.
Rule is changed into the If-Then form, and If partly is a feasible zone, and Then partly is the linear regression formula, can utilize this rule to come predicted target values.The situation of the relational tree of two input values (X1, X2) is concrete to be represented as follows:
Rule 1:
if：X
11≤X
1≤X
12，X
13≤X
2≤X
14
Then：Y＝a
1·X
1+b
1·X
2+c
1
…
Rule i:
if：X
i1≤X
1≤X
i2，X
i3≤X
2≤X
i4
Then：Y＝a
i·X
1+b
i·X
2+c
i
…
Rule n:
if：X
n1≤X
1≤X
n2，X
n3≤X
2≤X
n4
Then：Y＝a
n·X
1+b
n·X
2+c
n
X wherein
I1, X
I2, X
I3, X
I4＞0
i＝1，2，...，n
When x and y satisfy wherein certain regular condition, then can use this regular linear regression formula predicted target values.In finding the solution the process of pollution source, when known concentration value therein in certain regular target range, can adopt the linear programming algorithm, and utilize this rule, oppositely obtain the value of all associated upstream nodes.
The 3rd step:
A given output valve is oppositely found the solution input value.
Concentrate the relational tree that looks for node by empirical data, in the process of finding the solution the pollution source node, utilize the linear programming of these relational trees, obtain the value of all upstream nodes relevant with node.Reverse solution procedure is seen accompanying drawing 8.
This step is through finding the solution the linear programming problem (LP, Linear Program) of the corresponding belt restraining condition of each rule.The linear programming problem here is linear standard approximation problem, makes that promptly the linear regression formula predicted value of rule and the difference between the actual measurement (or calculated value) are minimum, specific as follows:
Converting rule 1 in the relational tree into linear standard approaches:
Min‖a
1·X
1+b
1·X
2+c
1-c‖
s.t：X
11≤X
1≤X
12
X
13≤X
2≤X
14
Converting regular i in the relational tree into linear standard approaches:
Min‖a
i·X
1+b
i·X
2+c
i-c‖
s.t：X
i1≤X
1≤X
i2
X
i3≤X
2≤X
i4
Converting regular n in the relational tree into linear standard approaches:
Min‖a
n·X
1+b
n·X
2+c
n-c‖
s.t：X
n1≤X
1≤X
n2
X
n3≤X
2≤X
n4
Accompanying drawing 3 expression tectonic relationship trees also form linear programming, and accompanying drawing 8 expressions utilize the linear programming of relational tree oppositely to find the solution the pollution source node; Forward and reverse two processes that they are described are two key steps of algorithm.
Certain node pollutant levels value is known, according to linear programming (node pollutant levels value satisfies rule condition), uses the method for solving-simplicial method of linear programming, finds the solution the pollutant levels value of upper reaches interdependent node.The If condition of linear programming constitutes the constraint condition of linear programming problem, and its objective function is predicted value and the minimum of the difference between the actual value that makes the linear regression formula of rule.The expression formula of linear programming problem is:
Min‖a1x+b1y+c1-c‖(5)
s.t：X1≤x≤X2
Y1≤y≤Y2
With certain water supply network illustraton of model is that illustration is an example, further sets forth process, pipe network model such as accompanying drawing 9. that pollution source tracing and positioning algorithm and programming thereof realize
1, the generation of training dataset
Inject through the EPANET simulating pollution, obtain training dataset, satisfied condition:
1) polluting injection node (being pollution source) can be any one node of pipe network;
2) polluting the injection duration is one hour;
3) polluting the whole simulation process duration is 72 hours;
4) pollute implantation concentration at 2.5mg/L to random variation between the 25mg/L, pollute inject the start time can be in one day any one constantly.
Inject simulation process in primary pollution, write down all nodes maximum contaminant level that can reach and the time that obtains maximum contaminant level, so just formed a sample of training data.In order to obtain the more data sample, used random function srand () and rand () here, as long as generate a pollution concentration C and an injection length T at random, just obtained a sample.
If each node simulating pollution is injected num time, just generated 11 * num training data sample (11 nodes altogether).
Follow the trail of the result for the pollution source that obtain high accuracy, instance of the present invention has obtained 1,000 sample datas through simulation, lists 2 samples below.The data ordering of each sample is in proper order:
Sample k:T1, C1, T2, C2, T3, C3, T4, C4, T5, C5, T6, C6, T7, C7, T8, C8, T9, C9, T10, C10, T11, C11.
Sample 1:20.750000 22.500000 22.083334 19.333464 23.083334 17.753242 25.583334 12.388585 22.916666 16.500341 25.166666 13.600124 27.416666 4.900297 25.333334 11.328960 26.583334 5.605895 0.000000 0.000000 23.666666 0.512918
Sample 2:19.083334 20.000000 20.333334 17.246368 20.916666 15.532732 25.083334 5.327957 21.166666 14.771476 22.500000 11.589097 29.333334 2.979820 24.000000 9.532548 25.250000 4.486727 0.000000 0.000000 21.916666 0.599407
……
K represent sample sequence number (k=1,2 ..., 20 ... .)
2, the structure of relational tree
In the simulating pollution injection process, according to every ducted water (flow) direction, can confirm the upstream and downstream relation of node, the expression upstream node has direct influence to the pollutant levels of downstream node.The relational tree element of each node comprises all upstream nodes relevant with it.The node relationships tree is with integer two-dimension array Tree [50] [5] expression in program code, and array element is the node call number, representes node by the node call number, and call number is since 0.Like Tree [10] [2]=3, the 3rd upstream node of expression node 11 is nodes 3.
Accompanying drawing 10 has been described in the pipe network model (Fig. 9) relation of the upstream and downstream between each node, and the direction of arrow representes to connect the flow direction that current can take place in the pipeline of two nodes.The input/output relation of the relational tree of each node is seen table 2.
The input/output relation of each node relationships tree of table 2
| RT 10 | - | - | - | - | - | - | C 10，T 10 |
| RT 11 | T 3 | C 3 | - | - | - | - | C 11，T 11 |
3, the structure of rule set
Set corresponding data from concentrated the selecting of training data, be combined into new data set with node relationships.Again with the kmeans clustering algorithm with the classification of new data set, to linear programming of each subclass structure, each node generates number and data set the same number of linear programming of classifying at last, each node all need carry out rule and constructs in the pipe network.
The rule set construction process is realized code (c speech encoding, as follows) as follows:
For (i=0; I＜nnodes; I++) // to each joint structure rule set
{
if(Upnodenum[i]！＝0)
contam.LoadPatterns(i)；
contam.InitClusters()；
contam.RunKMeans()；
for(j＝0；j＜5；j++)
{
contam.ProRule(j)；
Rule[i][j]＝contam.RuleTemp[j]；
}
}
Nnodes is the node number, and Upnodenum [i] is the number of the relational tree middle and upper reaches node of node i+1, and contam is an instance of the pollution source recognition category of establishment.Function LoadPatterns (i) among the class contam loads the training dataset (only loading the data with node i+1 relational tree middle and upper reaches node) of node i+1; Function InitClusters () initialization class bunch is for the Kmeans cluster does homework; Function R unKMeans () realizes the Kmeans cluster, is divided into five types to data set; Function ProRule (j) is to j+1 class data subset create-rule, and the regular RuleTemp [j] of generation is stored among the Rule [i] [j].
J+1 rule of Rule [i] [j] expression node i+1, its element type is the rule structure, structure is following:
struct?rule {
Double bound [MAXVECTDIM] [2]; The bound of // each element
Double a [MAXVECTDIM]; The coefficient that // equation is every
Double b; // constant term coefficient
}；
Wherein MAXVECTDIM is the maxitem of a linear equation of program setting.
4, pollution source confirms
Anti-when pushing away the pollution concentration value of the upstream node relevant,, choose suitable rule and carry out linear programming and find the solution according to the condition of linear programming and the concentration value of this node with node.If the rule of choosing is tried to achieve not have and separated, just the suitable rule of alternative is found the solution in other four rules.If five rules do not satisfy solving condition or try to achieve not have and separate; Abandon so finding the solution its upstream node value by the present node value; And the node that utilizes next value to try to achieve is found the solution; Find the solution down by this, all tried to achieve up to the pollutant value of all nodes, wherein the maximum node of pollution concentration value is exactly that pollution source inject node.
It is following that rule is chosen code:
Int ChooseRule (int index, float c) //index is the node call number, c is a node pollution concentration value
{
int?i；
float?max，min；
for(i＝0；i＜5；i++)
{
Max=Rule [index-1] [i] .bound [Upnodenum [index-1] * 2] [0]; // upper limit of concentration
Min=Rule [index-1] [i] .bound [Upnodenum [index-1] * 2] [1]; // concentration limit
(c≤max&&c>=min) // node concentration c is between the concentration bound time, and this rule is suitable for if
return?i；
}
}
Suppose that node 4 in the accompanying drawing 10 monitored pollutant at 9.42 hours concentration is 5.87mg/L, Input Monitor Connector is to the node number that pollutes, and monitoring time (h) and pollution concentration value (mg/L) can be carried out pollution source then and found the solution.
The process that pollution source are found the solution is described below step by step, and the tracing process flow process is seen accompanying drawing 11.
1): the moment and the concentration that occur the greatest contamination substrate concentration by node 4 solution nodes 3 and node 7:
C4=5.87 satisfies rule 2, and service regeulations 2 are found the solution.
The rule 2 of node 4: condition 2.59≤C4≤11.63
If?2.16≤T3≤10.25，3.72≤C3≤15.09
8.06≤T7≤20.67，0.58≤C7≤9.56
Then?C4＝0.25T3+0.11C3+0.02T7+1.02C7+0.528
Trying to achieve optimum solution is: T3=5.93, C3=8.26, T7=17.01, C7=2.56.
Similarly, by node 7 solution node 4 (known) and nodes 6 ...,
By node 1 (by before step tried to achieve time and concentration) solution node 10:
C1=12.74 satisfies rule 1, and service regeulations 1 are found the solution.
The rule 1 of node 1: condition 3.77≤C1≤25.35
If2.43≤T10≤13.27，7.59≤C10≤22.34
Then?C1＝1.21T10+0.53C10+1.26
Trying to achieve optimum solution is: T10=3.60, C10=12.87.
Finally, obtain the greatest contamination substrate concentration and the moment that the greatest contamination substrate concentration occurs of all nodes in the pipe network.The concentration 12.87mg/L of node 10, maximum in the pollutant levels intermediate value of all nodes, can decision node 10 be pollution source decanting points.
Claims (7)
1. the pollution source tracing and positioning information processing method of a water supply network burst contamination accident is characterized in that, may further comprise the steps:
1) collector is gathered the pollutant levels information that the monitoring point monitors, and sends it to processor;
2) processor is obtained the greatest contamination substrate concentration of its associated upstream node and the time that the greatest contamination substrate concentration occurs through the linear programming model of node relationships tree; Find the solution down step by step, the greatest contamination substrate concentration that can obtain all nodes in the pipe network obtains the time of greatest contamination substrate concentration with it;
3) processor judges that the maximum node of pollutant levels is predicted as pollutant and injects node.
2. the pollution source tracing and positioning information processing method of a kind of water supply network burst contamination accident according to claim 1 is characterized in that, the linear programming modelling process of described node relationships tree is following:
1) structure node relational tree;
2) generate training dataset;
3) data clusters;
4), construct a linear programming that is used for representing the relational tree between the relative upstream node of node to each relational tree;
5) utilize the linear programming of the relational tree in the step 4), obtain the greatest contamination substrate concentration of all relevant upstream nodes of each node and the time of greatest contamination substrate concentration appearance;
6) judge whether to tally with the actual situation, if yes, judge that this model can use, if not, return step 1).
3. the pollution source tracing and positioning information processing method of a kind of water supply network burst contamination accident according to claim 2 is characterized in that, the relation between described node relationships tree description node and its associated upstream node.
4. the pollution source tracing and positioning information processing method of a kind of water supply network burst contamination accident according to claim 2 is characterized in that described generation training dataset step is following:
Inject through the EPANET simulating pollution, given at random pollutant injects node, sets certain pollutant then and continues injection length and simulation process duration, provides and injects the concentration range that pollutant injects, and obtains training dataset.
5. the pollution source tracing and positioning information processing method of a kind of water supply network burst contamination accident according to claim 2 is characterized in that, described data clusters is classified training dataset for adopting the Kmeans algorithm.
6. the pollution source tracing and positioning information processing method of a kind of water supply network burst contamination accident according to claim 2; It is characterized in that; The linear programming of described relational tree comprises feasible zone and linear regression formula; Linear programming is changed into the If-Then form, and If partly is a feasible zone, and Then partly is the linear regression formula.
7. the pollution source tracing and positioning information processing method of a kind of water supply network burst contamination accident according to claim 2; It is characterized in that; Node in the described step 5) is according to the condition of linear programming and the concentration value of this node; Selection rule carries out the maximum contaminant level value that the upstream node relevant with node found the solution in linear programming, separates if the rule of choosing tries to achieve not have, and just the suitable rule of alternative is found the solution in Else Rule; If strictly all rules does not satisfy solving condition or tries to achieve not have and separate, abandon so finding the solution its upstream node value by the present node value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010534558.XA CN102467605B (en) | 2010-11-08 | 2010-11-08 | Pollution source tracking and positioning information processing method of sudden water supply pipe network pollution accident |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010534558.XA CN102467605B (en) | 2010-11-08 | 2010-11-08 | Pollution source tracking and positioning information processing method of sudden water supply pipe network pollution accident |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102467605A true CN102467605A (en) | 2012-05-23 |
| CN102467605B CN102467605B (en) | 2015-04-15 |
Family
ID=46071236
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010534558.XA Expired - Fee Related CN102467605B (en) | 2010-11-08 | 2010-11-08 | Pollution source tracking and positioning information processing method of sudden water supply pipe network pollution accident |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102467605B (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102818884A (en) * | 2012-08-15 | 2012-12-12 | 长沙理工大学 | Method for positioning illegal sewage outfall |
| CN105353099A (en) * | 2015-10-26 | 2016-02-24 | 中国地质大学（武汉） | Water supply network pollution source positioning method based on multi-population co-evolutionary algorithm |
| CN107066831A (en) * | 2017-05-19 | 2017-08-18 | 君晟合众（北京）科技有限公司 | A kind of regional complex environmental assessment techniques, apparatus and system |
| CN108664935A (en) * | 2018-05-14 | 2018-10-16 | 中山大学新华学院 | The method for tracking target and system of depth Spatial-temporal Information Fusion based on CUDA |
| CN109063071A (en) * | 2018-07-24 | 2018-12-21 | 江苏卓易信息科技股份有限公司 | Water pollution tracing method and equipment based on topological correlation |
| CN109886830A (en) * | 2019-01-02 | 2019-06-14 | 同济大学 | A kind of water supply network pollution sources tracking positioning method based on customer complaint information |
| CN111832792A (en) * | 2020-01-10 | 2020-10-27 | 吉林建筑大学 | Method and system for arranging pipe network water quality monitoring points based on sudden pollution events |
| CN112083132A (en) * | 2019-06-14 | 2020-12-15 | 深圳市振瀚信息技术有限公司 | Sewage pollution tracing method |
| CN112381369A (en) * | 2020-11-02 | 2021-02-19 | 河海大学 | Water body pollution tracing and risk prediction evaluation method based on online spectrum identification |
| CN113947033A (en) * | 2021-12-22 | 2022-01-18 | 深圳市水务工程检测有限公司 | Artificial intelligence based drainage pipe network pollutant tracing system and method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1591008A (en) * | 2003-08-28 | 2005-03-09 | 湖南力合科技发展有限公司 | Pollution source on line monitoring network system |
| CN101349687A (en) * | 2007-07-20 | 2009-01-21 | 中国科学院生态环境研究中心 | Method and system for monitoring water contamination |
| CN101424678A (en) * | 2008-11-25 | 2009-05-06 | 烟台迪特商贸有限公司 | Water quality mutation biology behavioral indicator monitoring system and monitoring method |
-
2010
- 2010-11-08 CN CN201010534558.XA patent/CN102467605B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1591008A (en) * | 2003-08-28 | 2005-03-09 | 湖南力合科技发展有限公司 | Pollution source on line monitoring network system |
| CN101349687A (en) * | 2007-07-20 | 2009-01-21 | 中国科学院生态环境研究中心 | Method and system for monitoring water contamination |
| CN101424678A (en) * | 2008-11-25 | 2009-05-06 | 烟台迪特商贸有限公司 | Water quality mutation biology behavioral indicator monitoring system and monitoring method |
Non-Patent Citations (3)
| Title |
|---|
| 《Journal of Water Resources Planning and Management》 20060701 Ami Preis Contamination Source Identification in Water Systems: A Hybrid Model Trees-Linear Programming Scheme 第132卷, 第4期 * |
| AMI PREIS: "Contamination Source Identification in Water Systems: A Hybrid Model Trees–Linear Programming Scheme", 《JOURNAL OF WATER RESOURCES PLANNING AND MANAGEMENT》, vol. 132, no. 4, 1 July 2006 (2006-07-01) * |
| 吕谋等: "供水管网突发污染试验模拟及污染源定位研究", 《青岛理工大学学报》, no. 06, 31 December 2009 (2009-12-31) * |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102818884A (en) * | 2012-08-15 | 2012-12-12 | 长沙理工大学 | Method for positioning illegal sewage outfall |
| CN102818884B (en) * | 2012-08-15 | 2015-03-04 | 长沙理工大学 | Method for positioning illegal sewage outfall |
| CN105353099A (en) * | 2015-10-26 | 2016-02-24 | 中国地质大学（武汉） | Water supply network pollution source positioning method based on multi-population co-evolutionary algorithm |
| CN105353099B (en) * | 2015-10-26 | 2017-04-05 | 中国地质大学（武汉） | A kind of water supply network polluter localization method based on synergetic on multiple populations |
| CN107066831A (en) * | 2017-05-19 | 2017-08-18 | 君晟合众（北京）科技有限公司 | A kind of regional complex environmental assessment techniques, apparatus and system |
| CN108664935A (en) * | 2018-05-14 | 2018-10-16 | 中山大学新华学院 | The method for tracking target and system of depth Spatial-temporal Information Fusion based on CUDA |
| CN109063071A (en) * | 2018-07-24 | 2018-12-21 | 江苏卓易信息科技股份有限公司 | Water pollution tracing method and equipment based on topological correlation |
| CN109063071B (en) * | 2018-07-24 | 2022-05-13 | 江苏卓易信息科技股份有限公司 | Water pollution tracing method and equipment based on topological correlation |
| CN109886830A (en) * | 2019-01-02 | 2019-06-14 | 同济大学 | A kind of water supply network pollution sources tracking positioning method based on customer complaint information |
| CN109886830B (en) * | 2019-01-02 | 2023-07-04 | 同济大学 | Water supply network pollution source tracking and positioning method based on user complaint information |
| CN112083132A (en) * | 2019-06-14 | 2020-12-15 | 深圳市振瀚信息技术有限公司 | Sewage pollution tracing method |
| CN111832792A (en) * | 2020-01-10 | 2020-10-27 | 吉林建筑大学 | Method and system for arranging pipe network water quality monitoring points based on sudden pollution events |
| CN112381369A (en) * | 2020-11-02 | 2021-02-19 | 河海大学 | Water body pollution tracing and risk prediction evaluation method based on online spectrum identification |
| CN112381369B (en) * | 2020-11-02 | 2022-09-23 | 河海大学 | Water body pollution tracing and risk prediction evaluation method based on online spectrum identification |
| CN113947033A (en) * | 2021-12-22 | 2022-01-18 | 深圳市水务工程检测有限公司 | Artificial intelligence based drainage pipe network pollutant tracing system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102467605B (en) | 2015-04-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102467605B (en) | Pollution source tracking and positioning information processing method of sudden water supply pipe network pollution accident | |
| CN113128129A (en) | Forward and backward coupling tracing method and system for sudden water pollution | |
| Qu et al. | The influence of geological events on the endemism of East Asian birds studied through comparative phylogeography | |
| Burchard-Levine et al. | A hybrid evolutionary data driven model for river water quality early warning | |
| CN101299834B (en) | Method for checking base station position | |
| CN110232434A (en) | A kind of neural network framework appraisal procedure based on attributed graph optimization | |
| Dumedah et al. | Selecting model parameter sets from a trade-off surface generated from the non-dominated sorting genetic algorithm-II | |
| CN110895878B (en) | Traffic state virtual detector generation method based on GE-GAN | |
| Kang et al. | Optimal meter placement for water distribution system state estimation | |
| Aboutalebi et al. | Multiobjective design of water-quality monitoring networks in river-reservoir systems | |
| CN105353099B (en) | A kind of water supply network polluter localization method based on synergetic on multiple populations | |
| Glinskiy et al. | The assessment methods of the level of countries environmental safety | |
| Eliades et al. | Iterative deepening of Pareto solutions in water sensor networks | |
| CN117787795A (en) | Method for qualitatively evaluating running level of low carbon in water works | |
| Heijnen et al. | A method for designing minimum‐cost multisource multisink network layouts | |
| Łangowski et al. | Optimised robust placement of hard quality sensors for robust monitoring of quality in drinking water distribution systems | |
| Puleo et al. | Water and energy saving in urban water systems: the ALADIN project | |
| Xu et al. | Multi-Watershed nonpoint source pollution management through coupling Bayesian-based simulation and mechanism-based effluent trading optimization | |
| Dave et al. | Extending the use of bio-inspiration for water distribution networks to urban settings | |
| Shao et al. | Wetland Ecotourism Development Using Deep Learning and Grey Clustering Algorithm from the Perspective of Sustainable Development | |
| Mzembegwa et al. | A Comparison of Fully-Linear Deep Learning Methods for Pipe Burst Localization in Water Distribution Networks | |
| Moradian et al. | Providing a model for assessing and analyzing the military power of countries | |
| Wang et al. | A quick algorithm of counting flow accumulation matrix for deriving drainage networks from a DEM | |
| Li et al. | Global optimization by small-world optimization algorithm based on social relationship network | |
| Li et al. | A bootstrap regional model for assessing the long-term impacts of climate change on river discharge |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150415 Termination date: 20171108 |