CN105005688A - Water quality pollution judging method based on Bayesian network model - Google Patents
Water quality pollution judging method based on Bayesian network model Download PDFInfo
- Publication number
- CN105005688A CN105005688A CN201510387662.3A CN201510387662A CN105005688A CN 105005688 A CN105005688 A CN 105005688A CN 201510387662 A CN201510387662 A CN 201510387662A CN 105005688 A CN105005688 A CN 105005688A
- Authority
- CN
- China
- Prior art keywords
- probability
- accident
- risk
- pollution
- network model
- 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
Abstract
The invention relates to environmental risk assessment, and concretely provides a water quality pollution accident risk assessment method based on a Bayesian network model. The method aims at solving the problem that the influence of superposition and counteracting among root node factors on the assessment result is not considered in the conventional method. In order to achieve the goal, the method comprises the following steps of: determining the water transportation pollution probability; determining the land transportation pollution probability; determining the fixed pollution probability; determining the surface pollution probability; and estimating the water quality pollution accident risk probability on the basis of the probabilities. During the land transportation pollution probability determination, the method uses the Bayesian network model with a three-layer structure; the weight of an intermediate node is estimated according to the root node probability; and the land transportation pollution probability is estimated according to the weight of the intermediate node. The method has the advantages that during the Bayesian network model building, the superposition and/or counteracting effect(s) among all factors of root nodes are/is considered, so that the accuracy of the evaluation structure can be greatly improved.
Description
Technical field
The present invention relates to environmental risk assessment, a kind of water pollution determination methods based on Bayesian network model is specifically provided.
Background technology
The object of venture analysis is to provide decision support to user or decision-making platform.The reason that accident may be caused to occur has multiple, and complicacy, the uncertainty of the probability that accident occurs and culprit make risk profile have extremely strong uncertainty.Event tree, Markov chain, Dynamic contingency tree, Petrinets, Bayesian network are all the main approaches for accident occurrence risk.The risk assessment of pop-up threat is relevant with many factors such as environment, regional economy, traffic, information transmission.Environmental risk assessment great majority concentrate on the risk assessment of toxic and harmful substance, wherein topmost step be identify contamination accident potential risk resource, analyze accident pattern and reason, calculating contingency occurrence probability.At present, about the environmental risk assessment of water pollutions is mainly divided into: Heavy Metals in Waters is to the hydrobiological risk assessment such as health risk assessment, fish of human body; The ecological risk assessment of wet land system and the construction of hydropower stations are to the ecological risk assessment of reservoir area and Lower Reaches.The method of risk assessment has multiple, such as evaluates the risk of area, harbour water pollutions with risk exponential; Analytic hierarchy process fado is applied to natural science, society and the uncertainty of ecologic environment aspect, many standards, multi-objective problem.Event tree analysis is also a kind of methods well known of uncertainty analysis, and is widely used in prediction burst accident probability, and for decision support system (DSS) is prepared, but analytical hierarchy process has strong artificial subjective judgement effect, may cause predicting the outcome inaccurate.
Bayesian network has extremely strong accident reasoning and culprit diagnostic function, it is one of common method of venture analysis, it directly can not judge the generation of accident, but the mutual relationship of each form factor can be identified in complication system, the factor that can association be found the strongest in numerous reasons that accident may be caused to occur.Bayesian network, as a kind of method evaluating power system technology fault and socio-economic factor consequence, in a government office, academic research and industrial circle is used widely.Complex Ecological Systems has great uncertainty, and mass data is limited, and thus Bayesian network is for comparing the concentration of boron nitride in Fish, with this for the risk assessment of boron nitride in watershed management.Bayesian network also can be used to evaluate the factor such as weather, land characteristics to the threat of soil.
But there are some problems in the existing method utilizing Bayesian network to carry out environmental risk assessment.Specifically, these methods usually using direct for the several factors node as the final accident probability of impact, are not considered the mutual superposition between these factors and counteracting, are therefore caused assessment result not accurate enough.Correspondingly, this area needs a kind of new water pollution appraisal procedure to overcome above-mentioned defect.
Summary of the invention
The present invention is intended to solve the problem, and is namely intended to solve existing method and fails to consider the superposition between root node factor and/or the problem of counteracting on the impact of final assessment result.For this purpose, the invention provides a kind of water pollution determination methods based on Bayesian network model, comprise the following steps: the probability determining water transportation pollution source; Determine the probability of land transport pollution source; Determine the probability of stationary pollution source; Determine the probability of non point source of pollution; And based on the probability of the probability of described water transportation pollution source, the probability of described land transport pollution source, the probability of described stationary pollution source and described non point source of pollution, estimate the risk probability of described water pollution accident.The method is characterized in that, describedly determine that the probability of land transport pollution source comprises the risk probability of road transport accident that estimation vehicular bridge occurs, and the risk probability of the road transport accident that described estimation vehicular bridge occurs comprises the following steps: to obtain the factor relevant to the road transport accident that vehicular bridge occurs further, the Bayesian network model of the road transport accident that vehicular bridge occurs is built according to described factor, and the risk probability of the road transport accident that vehicular bridge occurs is estimated according to described Bayesian network model, described Bayesian network model has Three Tiered Network Architecture, described Three Tiered Network Architecture comprises root node, intermediate node and evaluation object, and the step of the risk of the described road transport accident according to Bayesian network model estimation vehicular bridge occurs comprises the probability determining root node further, the weight of intermediate node and the risk probability according to the road transport accident that the weight estimation vehicular bridge of described intermediate node occurs according to the Probability estimate of described root node.
In the preferred implementation of said method, described intermediate node comprises human factor, vehicle factor and road and environmental factor.
In the preferred implementation of said method, described human factor comprises driver's age and driver's sex.
In the preferred implementation of said method, described vehicle factor comprises type of vehicle and vehicle condition.
In the preferred implementation of said method, described road and environmental factor comprise visibility and weather conditions.
In the preferred embodiment of said method, described method also comprises, according to following formula, the risk probability of described road transport accident is converted to logarithm risk class: P
l=lgP+5, wherein, Pl represents logarithm risk class, and P represents the risk probability estimated according to described Bayesian network model.
In the preferred embodiment of said method, described logarithm risk class is higher, then the risk of the road traffic accident described vehicular bridge occurred can acceptance lower.
In the preferred embodiment of said method, described Bayesian network model adopts Hugin8.0 software building.
In the preferred embodiment of said method, described method adopts Hugin8.0 software to carry out sensitivity analysis to correlative factor after being also included in the risk probability estimating the road transport accident that vehicular bridge occurs.
In the preferred embodiment of said method, the probability of the probability of described water transportation pollution source, the probability of described stationary pollution source and described non point source of pollution is all zero, and the risk probability of described water pollution accident is only estimated according to the risk probability of the road transport accident that described vehicular bridge occurs.
Those skilled in the art it is easily understood that, Bayesian network model is built into and comprises root node by method of the present invention, the Three Tiered Network Architecture of intermediate node and evaluation object, when estimating the risk of water pollution accident, first determine the probability of root node, then according to the weight of the probability assessment intermediate node of root node, then according to the risk probability of the road transport accident (i.e. evaluation object) that the weight estimation vehicular bridge of intermediate node occurs, the basis of the risk probability of the last road transport accident occurred on vehicular bridge calculates the risk probability of water pollution accident.In this way, the superposition between each factor of root node and negative function are just being taken into full account by during root node Probability estimate intermediate node weight, and compared with prior art, this improvement makes the accuracy of assessment result be greatly improved.
Accompanying drawing explanation
Fig. 1 is the process flow diagram according to water pollution accident appraisal procedure of the present invention.
Fig. 2 is the process flow diagram used in the method for Fig. 1.
Fig. 3 is the schematic diagram of the Bayesian network model used in the method for Fig. 2.
Embodiment
The pollution source of water pollution accident mainly comprise water transportation pollution source, land transport pollution source, stationary pollution source and non point source of pollution four class.The pollution of leaking or accident causes is there is in water transportation pollution source when referring to shipping, land transport pollution source refer to that the vehicle on road such as to have an accident at the pollution caused, stationary pollution source refers to that the fixation means such as factory discharge the pollution caused, and non point source of pollution refers to the pollution that rainfall, short smokestack, commercial coal stove etc. cause.When water source is closed artificial canal for water conveyance, the possibility pollution source that uniquely directly can contact with the external world are vehicular bridge, i.e. land transport pollution source.Freight transportation vehicle has an accident on vehicular bridge, causes the leakage of the dangerous cargo of transport or haulage vehicle to tumble in trunk canal, causes the poisonous and harmful substance of transport to enter trunk canal and cause trunk canal burst water pollution accident.So for closed artificial conveyance canal, the road traffic accident of vehicle on vehicular bridge is almost unique pollution source.
Below in conjunction with preferred implementation, technical scheme of the present invention is described.First consult Fig. 1, the figure shows the process flow diagram according to water pollution accident appraisal procedure 10 of the present invention.Method 10 starts from step S10, in step slo, determines the probability of water transportation pollution source.Then in step S20, determine the probability of land transport pollution source, namely determine the probability of the road transport accident that vehicular bridge occurs.Then in step s 30, the probability of stationary pollution source is determined.Afterwards in step s 40, the probability of non point source of pollution is determined.Last in step s 50, based on the probability of the probability of described water transportation pollution source, the probability of described land transport pollution source, the probability of described stationary pollution source and described non point source of pollution, estimate the risk probability of described water pollution accident.It is to be noted, in the inventive solutions, except the probability of land transport pollution source, the probability of the described probability of water transportation pollution source, the probability of stationary pollution source and non point source of pollution can be determined by various different modes such as measured data, documents and materials, expertises, and after determining this little probability, the risk probability of described water pollution accident can draw based on the different weight of above-mentioned probability.Such as, when water source is closed artificial canal for water conveyance, the probability of the described probability of water transportation pollution source, the probability of stationary pollution source and non point source of pollution all can be set to zero, because be as mentioned above, almost unique pollution source for the road traffic accident of closed artificial conveyance canal vehicle on vehicular bridge.
Although Fig. 1 does not illustrate, feature according to water pollution accident appraisal procedure 10 of the present invention is: for closed artificial conveyance canals such as such as projects of South-to-North water diversion, determines described the risk probability that the probability of land transport pollution source is equal to the road transport accident that estimation vehicular bridge occurs.Next consult Fig. 2, this figure is the process flow diagram of the method 100 of the risk probability of the road transport accident that aforementioned estimation vehicular bridge occurs.As shown in Figure 2, method 100 of the present invention starts from step S1.In step sl, obtain the factor relevant to the road transport accident that vehicular bridge occurs as root node, such as but not limited to driver age C, driver sex D, weather conditions E, visibility F, vehicle condition G, type of vehicle B etc. (see Fig. 3).It should be pointed out that those skilled in the art can delete above-mentioned factor according to practical application or add, this does not deviate from protection scope of the present invention.Then, in step s 2, the Bayesian network model of the road transport accident that vehicular bridge occurs is built according to described factor.The operation of this structure model can be realized by multiple different mode, such as, Hugin8.0 can be adopted to build Bayesian network model required for the present invention.As described above, Bayesian network model of the present invention has Three Tiered Network Architecture, and this Three Tiered Network Architecture comprises: root node B, C, D, E, F, G; Intermediate node C1, C2, C3; And evaluation object A.Exemplarily, root node B, C, D, E, F, G represents type of vehicle, driver's age, driver's sex, weather conditions, visibility and vehicle condition respectively; Intermediate node C1, C2, C3 represent human factor, road and environmental factor and vehicle factor respectively; Evaluation object A represents the road traffic accident (see Fig. 3) that vehicular bridge occurs.
Next, in step s3, the probability of root node is determined.Those skilled in the art are it is easily understood that the probability of described root node can be determined by various different modes such as measured data, documents and materials, expertises.Exemplarily, the probability of the root node determined by measured data (measured datas based on Hebei province 2011 and 2012) is given in table 1 below:
Table 1
Then in step s 4 which, according to the weight of the Probability estimate intermediate node of root node.The weight of described intermediate node refers to the quantification proportion of this intermediate node to the influence degree of final probability.This also can be realized by the method that historical data, expertise etc. are known.Then in step s 5, according to the risk probability of the weight estimation road transport accident of described intermediate node.
Then in step s 6, the risk probability estimated is converted to logarithm risk class according to following formula by described method further:
P
l＝lgP+5
Wherein, Pl represents logarithm risk class, and P represents the risk probability estimated according to described Bayesian network model.Above-mentioned formula obtains by training empirical data, and compared with the risk probability estimated, what the risk log-rank after conversion can reflect particular risk intuitively can acceptance, therefore more has directive significance for decision process.
Exemplarily, table 2 below gives the logarithm risk class of the risk probability estimated based on above-mentioned table 1:
Table 2
Logarithm risk probability is higher, shows that accident risk is larger, and the acceptable degree of accident risk can represent with risk class, and risk class is higher, and reflection accident to external world environment is more responsive.Therefore, the logarithm risk probability of Unequal distance is more reasonable when risk class divides.Inventor thinks, the interval of risk probability when risk class is 3-5 is in table 2 less than risk probability when risk class is 1-2.
Finally, in the step s 7, after the risk probability estimating the road transport accident that vehicular bridge occurs, method of the present invention carries out sensitivity analysis to correlative factor, the responsive parameter analysis module of Hugin8.0 is such as adopted to carry out sensitivity analysis to the factor of interdependent node, to distinguish different node and factor to the Different Effects degree of accident probability.Exemplarily, following table 3 shows the data sensitivity analysis result of making based on above-mentioned table 1 and 2:
Table 3
As can be seen from Table 3, for intermediate level of nodes, human factor C1 is most sensitive nodes, represents that the judgement of driver on burst accident is the most sensitive factor affecting whether accident occur.Second Main Factors is vehicle factor C3, and that relatively weak is road and environment nodes C2.On root node layer, vehicle condition G, visibility F, weather conditions E are the most sensitive factors causing accident to occur.
Finally consult Fig. 3, this figure is the schematic diagram according to Bayesian network model of the present invention.As shown in Figure 3, Bayesian network model of the present invention is built into Three Tiered Network Architecture, and this Three Tiered Network Architecture comprises: root node B, C, D, E, F, G; Intermediate node C1, C2, C3; And evaluation object A.Exemplarily, root node B, C, D, E, F, G represents type of vehicle, driver's age, driver's sex, weather conditions, visibility and vehicle condition respectively; Intermediate node C1, C2, C3 represent human factor, road and environmental factor and vehicle factor respectively; Evaluation object A represents the road traffic accident that vehicular bridge occurs.It should be noted that; here root node and the content of intermediate node and quantity are only given as examples; under the prerequisite not departing from principle of the present invention; those skilled in the art can make adjustment to the content of described node and quantity as required, and the technical scheme after adjustment also will fall within protection scope of the present invention.Such as; driver health situation can be added as root node on human factor C1; also surface evenness can be added as root node on road and environmental factor C2; this interpolation does not change ultimate principle of the present invention, and the scheme after therefore adding also will fall within protection scope of the present invention.
Further, when estimating the risk of the road transport accident that vehicular bridge occurs, the present invention first determines root node probability, then according to the weight of the probability assessment intermediate node of root node, then according to the risk probability of the road transport accident (i.e. evaluation object) that the weight estimation vehicular bridge of intermediate node occurs.So, the superposition between each factor of root node and negative function are just being taken into full account by during root node Probability estimate intermediate node probability, and compared with prior art, this improvement makes the accuracy of final assessment be greatly improved.For human factor, the Bayesian network of prior art is usually using driver's sex and driver's age direct factor as the final probability of impact, but, this network structure does not consider the superposition that different driver's age and sex combine mutually and/or negative function, therefore can make finally to estimate that the accuracy of result is affected.Specifically, although suppose that driver age C is 60 years old, belong to the crowd of delay of response, but driver sex D is the male sex, counteract the older high accident probability caused to a certain extent, the result of both common superposition determines driver's comprehensive emergency capability in case of emergency, and using obviously more objective as the factor directly affecting final accident probability for the probability representing this comprehensive emergency capability, corresponding assessment result is also more accurate.On the contrary, if driver age C and driver sex D is revised final accident probability as direct acting factor, then have ignored the general character between driver age C and driver sex D and superposition and/or negative function between them to a great extent, therefore must affect the accuracy of net result.
Table 4 below gives the traffic flow data in Hebei province.
Table 4
| Year | Private car quantity | Commerial vehicle number | Vehicle fleet | Traffic hazard number | Traffic hazard |
| (ten thousand) | () | Probability % | |||
| 2011 | 510.61 | 93.44 | 604.05 | 5197 | 0.086 |
| 2012 | 624.04 | 103.22 | 727.26 | 5285 | 0.073 |
Based on the example data in table 1 above, according to the probability of Bayesian network model of the present invention and each node, the result of calculation of accident probability is: P (A)=0.06%.In table 4 data from China National Bureau of Statistics of China, be the magnitude of traffic flow measured data in Hebei province.In table 4, the accident probability of 2011 and 2012 is respectively 0.086% and 0.073%, and estimate with technical scheme according to the present invention 0.06% closely, apparently higher than other accident probability evaluation methods of the prior art.As can be seen here, three layers of Bayesian network model of the present invention have increased substantially the accuracy of the risk probability estimation of the road traffic accident that vehicular bridge occurs really.
So far, shown by reference to the accompanying drawings preferred implementation describes technical scheme of the present invention, but those skilled in the art are it is easily understood that protection scope of the present invention is obviously not limited to these embodiments.Under the prerequisite not departing from principle of the present invention, those skilled in the art can make equivalent change or replacement to correlation technique feature, and these changes or the technical scheme after replacing it all will fall within protection scope of the present invention.Such as, although the application describes in conjunction with the road traffic accident that vehicular bridge occurs, technical scheme of the present invention obviously also can be applied to the road traffic accident of other types.
Claims (10)
1., based on a water pollution determination methods for Bayesian network model, comprise the following steps:
Determine the probability of water transportation pollution source;
Determine the probability of land transport pollution source;
Determine the probability of stationary pollution source;
Determine the probability of non point source of pollution; And
Based on the probability of the probability of described water transportation pollution source, the probability of described land transport pollution source, the probability of described stationary pollution source and described non point source of pollution, estimate the risk probability of described water pollution accident,
It is characterized in that, describedly determine that the probability of land transport pollution source specifically estimates the risk probability of the road transport accident that vehicular bridge occurs, and the risk probability of the road transport accident that described estimation vehicular bridge occurs comprises the following steps: further
Obtain the factor relevant to the road transport accident that vehicular bridge occurs;
The Bayesian network model of the road transport accident that vehicular bridge occurs is built according to described factor; And
According to the risk probability of the road transport accident that described Bayesian network model estimation vehicular bridge occurs,
Described Bayesian network model has Three Tiered Network Architecture, described Three Tiered Network Architecture comprises root node, intermediate node and evaluation object, and the step of the described risk probability of road transport accident according to Bayesian network model estimation vehicular bridge occurs comprise further determine root node probability, according to the Probability estimate of described root node the weight of intermediate node and the risk probability according to the road transport accident that the weight estimation vehicular bridge of described intermediate node occurs.
2. method according to claim 1, is characterized in that, described intermediate node comprises human factor, vehicle factor and road and environmental factor.
3. method according to claim 2, is characterized in that, described human factor comprises driver's age and driver's sex.
4. method according to claim 3, is characterized in that, described vehicle factor comprises type of vehicle and vehicle condition.
5. method according to claim 4, is characterized in that, described road and environmental factor comprise visibility and weather conditions.
6. method according to any one of claim 1 to 5, characterized by further comprising, according to following formula, the risk probability of described road transport accident is converted to logarithm risk class:
P
l＝lg P+5
Wherein, P
lrepresent logarithm risk class, P represents the risk probability estimated according to described Bayesian network model.
7. method according to claim 6, is characterized in that, described logarithm risk class is higher, then the risk of the road traffic accident described vehicular bridge occurred can acceptance lower.
8. method according to any one of claim 1 to 5, is characterized in that, described Bayesian network model adopts Hugin8.0 software building.
9. method according to claim 6, characterized by further comprising and adopt Hugin8.0 software to carry out sensitivity analysis to correlative factor after the risk probability estimating the road transport accident that vehicular bridge occurs.
10. method according to any one of claim 1 to 5, it is characterized in that, the probability of the probability of described water transportation pollution source, the probability of described stationary pollution source and described non point source of pollution is all zero, and the risk probability of described water pollution accident is only estimated according to the risk probability of the road transport accident that described vehicular bridge occurs.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510387662.3A CN105005688A (en) | 2015-06-30 | 2015-06-30 | Water quality pollution judging method based on Bayesian network model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510387662.3A CN105005688A (en) | 2015-06-30 | 2015-06-30 | Water quality pollution judging method based on Bayesian network model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105005688A true CN105005688A (en) | 2015-10-28 |
Family
ID=54378360
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510387662.3A Pending CN105005688A (en) | 2015-06-30 | 2015-06-30 | Water quality pollution judging method based on Bayesian network model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105005688A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108022049A (en) * | 2017-12-07 | 2018-05-11 | 浙江省水利水电工程质量与安全监督管理中心 | Quality of Water Conservancy Projects safety supervision risk evaluating system and method |
| CN111667104A (en) * | 2020-05-29 | 2020-09-15 | 重庆工商大学 | Bayesian transfer learning method under sudden water pollution environment |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7774293B2 (en) * | 2005-03-17 | 2010-08-10 | University Of Maryland | System and methods for assessing risk using hybrid causal logic |
| CN101838991A (en) * | 2010-05-21 | 2010-09-22 | 上海建科建设监理咨询有限公司 | Deep excavation risk evaluation method based on network reasoning |
| US8306936B2 (en) * | 2010-03-16 | 2012-11-06 | Gansner Harvey L | Automated legal evaluation using bayesian network over a communications network |
| US8423482B2 (en) * | 2010-03-16 | 2013-04-16 | Harvey L. Gansner | Automated legal evaluation using a decision tree over a communications network |
| CN103077447A (en) * | 2013-01-24 | 2013-05-01 | 华中科技大学 | Subway construction safety management decision-making method based on Bayesian network |
| CN104408313A (en) * | 2014-11-26 | 2015-03-11 | 上海工程技术大学 | Rail transit energy consumption influence factor collaborative correlation modeling method |
-
2015
- 2015-06-30 CN CN201510387662.3A patent/CN105005688A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7774293B2 (en) * | 2005-03-17 | 2010-08-10 | University Of Maryland | System and methods for assessing risk using hybrid causal logic |
| US8306936B2 (en) * | 2010-03-16 | 2012-11-06 | Gansner Harvey L | Automated legal evaluation using bayesian network over a communications network |
| US8423482B2 (en) * | 2010-03-16 | 2013-04-16 | Harvey L. Gansner | Automated legal evaluation using a decision tree over a communications network |
| CN101838991A (en) * | 2010-05-21 | 2010-09-22 | 上海建科建设监理咨询有限公司 | Deep excavation risk evaluation method based on network reasoning |
| CN103077447A (en) * | 2013-01-24 | 2013-05-01 | 华中科技大学 | Subway construction safety management decision-making method based on Bayesian network |
| CN104408313A (en) * | 2014-11-26 | 2015-03-11 | 上海工程技术大学 | Rail transit energy consumption influence factor collaborative correlation modeling method |
Non-Patent Citations (2)
| Title |
|---|
| 王少群 等: "高速公路交通事故概率预测方法研究", 《价值工程》 * |
| 陆莹 等: "基于模糊贝叶斯网络的地铁运营安全风险预测", 《东南大学学报(自然科学版)》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108022049A (en) * | 2017-12-07 | 2018-05-11 | 浙江省水利水电工程质量与安全监督管理中心 | Quality of Water Conservancy Projects safety supervision risk evaluating system and method |
| CN111667104A (en) * | 2020-05-29 | 2020-09-15 | 重庆工商大学 | Bayesian transfer learning method under sudden water pollution environment |
| CN111667104B (en) * | 2020-05-29 | 2023-08-04 | 重庆工商大学 | Bayesian migration learning method under sudden water pollution environment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105045975A (en) | Bayesian network model based risk evaluation method for road transportation accident | |
| CN106650825B (en) | Motor vehicle exhaust emission data fusion system | |
| CN109886568B (en) | Risk chemical road transportation basin water environment risk assessment method and system | |
| Gandini et al. | A holistic and multi-stakeholder methodology for vulnerability assessment of cities to flooding and extreme precipitation events | |
| CN103093643B (en) | Public parking lot berth quantity confirming method | |
| CN106204392A (en) | Environment risk source early warning system | |
| Duan et al. | Multi-objective optimal design of detention tanks in the urban stormwater drainage system: uncertainty and sensitivity analysis | |
| Hou et al. | A real-time, dynamic early-warning model based on uncertainty analysis and risk assessment for sudden water pollution accidents | |
| CN105930670A (en) | Model parameter uncertainty-based dynamic prediction method for river emergency pollution accident | |
| CN106779320A (en) | A kind of gas pipeline damage from third-party methods of risk assessment based on fuzzy mathematics | |
| CN105512741A (en) | Bus passenger traffic combined prediction method | |
| CN107798418A (en) | A kind of traffic accident frequency Forecasting Methodology based on traffic analysis cell | |
| Tran et al. | Comparison of structural deterioration models for stormwater drainage pipes | |
| CN108615098A (en) | Water supply network pipeline burst Risk Forecast Method based on Bayesian survival analysis | |
| CN113191582B (en) | Road torrential flood vulnerability evaluation method based on GIS and machine learning | |
| CN105005688A (en) | Water quality pollution judging method based on Bayesian network model | |
| Karimian | Failure rate prediction models of water distribution networks | |
| Wang et al. | Application of the hidden Markov model in a dynamic risk assessment of rainstorms in Dalian, China | |
| Wang et al. | Exploring the relationship between urban flood risk and resilience at a high-resolution grid cell scale | |
| Xiao et al. | Prediction model of asphalt pavement functional and structural performance using PSO-BPNN algorithm | |
| CN104569340A (en) | Underground environment quality determination method and device | |
| Dawood et al. | Watermain's failure index modeling via Monte Carlo simulation and fuzzy inference system | |
| CN113158591B (en) | Method for determining bearing capacity of basin land development and utilization | |
| Asghari Moghaddam et al. | Optimization of DRASTIC model by artificial intelligence for groundwater vulnerability assessment in Maragheh-Bonab plain | |
| Nepveu et al. | Hoisting a red flag: an early warning system for exceeding subsidence limits |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20151028 |
|
| RJ01 | Rejection of invention patent application after publication |