CN102982374A - Pipe explosion positioning method based on back propagation (BP) artificial neural networks - Google Patents
Pipe explosion positioning method based on back propagation (BP) artificial neural networks Download PDFInfo
- Publication number
- CN102982374A CN102982374A CN 201210426108 CN201210426108A CN102982374A CN 102982374 A CN102982374 A CN 102982374A CN 201210426108 CN201210426108 CN 201210426108 CN 201210426108 A CN201210426108 A CN 201210426108A CN 102982374 A CN102982374 A CN 102982374A
- Authority
- CN
- China
- Prior art keywords
- booster
- pipe
- pipe explosion
- network
- artificial neural
- 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 discloses a pipe explosion positioning method based on back propagation (BP) artificial neural networks. Pipe explosion monitoring points is arranged at a panel point sensitive to pressure change in a target pipe network. The pipe explosion positioning method based on the back propagation (BP) artificial neural networks is capable of carrying out dynamic analysis of changing process of pressure running state of the pipe explosion monitoring points, establishing a non-linear relationship between the pressure change of three pipe explosion monitoring points and pipe explosion position and damaged condition through the BP artificial neural networks, achieving a research target that pipe explosion positioning is comparatively accurate and pipe explosion response time is shortened. Closely centering on the research target, an article consists of researches of pipe explosion reasons and characteristics, arrangement method of the pipe explosion monitoring points of a water pipe network, the pipe explosion positioning based on the artificial neural networks and the like. An ideal experiment result is achieved. Some novel methods and thinking of the pipe explosion are put forward.
Description
Technical field
The present invention relates to a kind of localization of bursted pipe method, particularly a kind of localization of bursted pipe method based on the BP artificial neural network.
Background technology
Pipe burst detects and different according to detected object of location technology, and present existing detection method is broadly divided into two classes: a class is based on the method for hardware, the another kind of method that is based on software.Hardware based method refers to leakage is carried out direct-detection, listens detection method, correlation detection, regional detection method etc. such as manual patrol method, sound; Refer to that based on the method for software these class methods have KLR signal approach, negative pressure wave method, real time dynamic model method, statistic law (comprising the state estimation method), pattern-recongnition method, artificial intelligence detection method (comprising BP artificial neural network method) etc. by detecting because the physical parameters such as flow that leakage causes, pressure, sound change to judge whether leakage occurs and definite leak position.
Achievement in research is the BP artificial neural network method more significantly at present.Its basic step is:
One, at first, to the summary achievement of the booster eigenwert of the different tubing of water supply network, combining target pipe network tubing operating position, damage type and the damaged condition of pipe explosion accident may occur in each pipeline section in the pipe network that sets objectives.
Two, then, utilize the pipe network model that various booster damaged conditions are carried out the booster simulation, obtain each pipeline section and in various degree pipe explosion accident occur to the pressure delta data of specific monitoring point, with these data as the training sample of setting up localization of bursted pipe and damaged condition model.
Three, last, based on these training samples, the Neural Network Toolbox that provides among the MATLAB is provided, by the requirement of BP artificial neural network technology localization of bursted pipe and damage assessment model are made up and train, make the nonlinear relationship between its grasp monitoring point pressure variation and booster damaged condition, the booster position, finally reach the fast purpose of location of pipe explosion accident.
The core of BP artificial neural network method is to set up a BP neural network model that is applicable to target pipe network pipe explosion accident location.Set up the BP neural network model and need a large amount of, the real service data of this water supply network when accident condition, but be difficult to provide the service data of abundance in the reality, the key of therefore setting up the BP neural network model is the method that need to seek suitable booster state simulation.
Mainly contain three methods for the booster state simulation: the one, set up physical model in the laboratory; Two are based on water supply network, discharge water the operating mode of simulation booster by the evacuated tube in hose saddle or pipe network; The 3rd, the method for employing computer simulation.First method is set up pipe network by similarity criterion, carries out the research of pipe network operating mode in the laboratory.Need expend a large amount of human and material resources, financial resources and time, and the experiment pipe network is generally small, the various situations in the time that actual booster can not being simulated.Second method is method the most effective, that be adjacent to actual conditions most, but owing to be to carry out the booster test at real water supply network, must cause the water supply network pressure drop, normal water influences the masses of the people, secondly the huge water yield that produces may cause the difficulty of draining, and therefore second method implements difficult.The 3rd method is the present scholar method the most frequently used to the booster state simulation, but mainly there are 3 deficiencies in computer simulation: the one, and pipe network model and actual pipe network differ larger, cause booster analog result and actual booster to have error; The 2nd, many scholars do not understand pipe explosion accident reason and correlation properties, and the physical characteristics according to pipeline section does not determine damage type and tube bursting and leakage amount, just increase a node flow at the simulation pipeline section simply, cause simulation not conform to the actual conditions; The 3rd, be based in computer simulation and carry out under the perfect condition of pipe network of zero noise, but in fact there is pressure surge in pipe network and produces data noise.
Summary of the invention
The object of the present invention is to provide a kind of localization of bursted pipe method based on the BP artificial neural network, solve booster state simulation out of true problem.
A kind of localization of bursted pipe method based on the BP artificial neural network, its concrete steps are:
The first step is built wisdom pipeline software platform system
Wisdom pipeline software platform system comprises: topomap database management module, landform and pipe network data update module, pipe network data management module, pipe network ancillary data administration module, pipe network auxiliary tube grooming tool module, subsidiary function module, pipe network component module, pipe network transverse and longitudinal fractograph analysis module, accident treatment module, off-line editing module, system management module and Web delivery system module.
Second step accident treatment module application pipe explosion accident is processed
Accident treatment mainly is to provide fast decision support for pipe explosion accident, fire failure.After water supply network burst booster or the accident such as leak, the user only need appointment accident nidus, system can automatic search goes out valve, the user that cuts off the water that need to close, cut off the water zone or nearest hose saddle information etc. on every side, and make rational processing scheme, and can automatically generate valve opening and closing advice note, on-site maintenance figure, the user assistance repair personnel such as advice note that cut off the water and construct.The core that pipe explosion accident is processed is the localization of bursted pipe method.The below is the localization of bursted pipe method of using based on the BP artificial neural network.
The 3rd step target pipe burst detects determining of minimum value;
In the pipeline exploding early warning subsystem of SCADA system, it is the leading indicators that detects as booster that hydraulic pressure value moment of monitoring point descends, but in the water factory's secondary pump station board switch, pipe network large user's water yield increase suddenly and pipe network in the abnormal pressure fluctuation all can cause monitoring point hydraulic pressure value moment to descend.For avoiding producing the booster wrong report, misrepresent deliberately, each pipeline exploding early warning subsystem will arrange in response to the present situation of water supply network a voltage drop value, when hydraulic pressure moment voltage drop value in monitoring point surpasses this voltage drop value, just can issue the booster alarm.Therefore this voltage drop value is exactly that booster detects minimum value.
Determining of the 4th step network input/output argument;
The purpose of training BP neural network be make its grasp that monitoring point pressure changes and booster damaged condition, booster position between nonlinear relationship.Based on this, the network input parameter is the change value of pressure of three booster monitoring points, and the network output parameter is booster position and booster damaged condition.
The foundation of the 5th step training sample;
Employing is carried out the booster simulation in conjunction with booster feature and the historical booster record of different tubing to the high damage type of booster fidelity factor, reaches the closing to reality application purpose.
The structure of the 6th step network
Plan to build vertical BP neural network and comprise three input parameters, four output parameters.In four output parameters, three is apart from variable, and one is pressure variety, and the dimension of two groups of amounts is different.For avoiding phase mutual interference between the output quantity, improve precision of prediction, two neural networks are set up in this research, are respectively applied to localization of bursted pipe and booster damage assessment.
After input data and target data pre-service are finished, next step will carry out to pipe network the training study of neural network.The Neural Network Toolbox that provides among this research and utilization MATLAB is come pipe network is trained and later trouble spot is positioned prediction.
The 6th step booster position decision method
The booster position is determined at the CAD graphical interfaces.
Though location model can not accurately be located the booster position, but satisfied the more accurately location of judging booster pipeline section and booster position fully, and in three monitoring points, it is less to ensure at least two monitoring point predicted values and actual value deviation, and reliability forecasting is higher.Therefore judge that location model is feasible.
Although damage assessment model error ratio is larger, by analysis, error mainly in the estimation of booster place hydraulic pressure, has nothing to do with model itself.And the booster damage assessment is to allow water undertaking have one roughly to understand to the pipe explosion accident damaged condition, and reference significance is greater than practical significance, therefore in this error range, and damage assessment model or available.
The prediction accuracy of two models changes decision by booster monitoring point pressure, and change value of pressure is larger, and then two model predictions are more accurate.It shows as: under the identical pipeline section, the little model prediction accuracy rate larger than the damaged condition of booster of the damaged condition of booster is low; Under the identical damaged condition, the model of heavy caliber pipeline section is lower than small-bore pipeline section predictablity rate.
Embodiment
First step target pipe burst detects determining of minimum value;
In the SCADA system, a pressure tap is arranged as the pressure monitoring reference point of target pipe network, this pressure tap was passed a pressure data in per 3 minutes back, and the number needs of collecting data satisfies the needs that the evaluating objects pipe burst detects minimum value.By collecting the data of pressure tap hydraulic pressure in former years, remove the data when having an accident, calculates 3 minutes change value of pressure and compare, to find the shared number percent in different pressures interval.Based on the compare of analysis to the pressure tap change value of pressure, the booster of target pipe network detects determines minimum value.
Determining of second step network input/output argument;
The booster position is represented by horizontal range between booster place and the booster monitoring point.Add sign by the difference of the X of pipe place coordinate and booster monitoring point X coordinate on the distance value, when the difference of booster position X coordinate and monitoring point X coordinate for negative, expression booster position just is being east in the west, monitoring point.This method can judge that the booster place is positioned at east or the west of booster monitoring point, thus the differentiation difficulty of reduction accident pipeline section and raising locating effect.
The booster damaged condition is to weigh with the size of tube bursting and leakage flow, but the flow measurement of actual conditions tube bursting and leakage is difficult, and water undertaking is generally with impaired area ratio expression booster damaged condition.Because change value of pressure is the amount of same dimension with impaired area than not, guarantee that the network precision just must strengthen training burden and training time.The appearance of considering the tube bursting and leakage flow can make the booster place produce larger change value of pressure, and booster place change value of pressure size also is to weigh the important evidence of booster damaged condition.Therefore, be Effective Raise network precision and training effectiveness, this research is adopted with the booster place change value of pressure identical with the input parameter dimension and is weighed the booster damaged condition.
The foundation of the 3rd step training sample;
(1) according to historical booster record, the booster damaged condition that the target pipe network may occur is estimated.The booster damage type is relevant with conduit running pressure, tubing physical property and local environment with damaged condition.In same water supply network, booster damage type and the damaged condition of identical tubing are similar.According to historical booster record damage type and the degree of different tubing are concluded, utilized relevant achievement that the target pipe network is carried out booster hypothesis, the damage type that each pipeline section may occur in the target setting pipe network and degree.
The used tubing of target pipe network is three kinds of spigot-and-socket ductile iron pipe, welded still pipe, HDPE pipes etc.; The system of laying of feedwater piping is divided into integrated pipe canal and installs and buried-pipe laying.Based on actual conditions, the booster situation is able to lower setting:
Carry out the booster hypothesis according to tubing booster feature under the feedwater piping of buried-pipe laying, because the target pipe network interior conduit age is shorter, it is aging to ignore pipeline under present pipe network operating mode.
The pipeline that lays at integrated pipe canal is steel pipe, connected mode is the ring flange interface, and in pipe trench, substantially getting rid of corrosion and ageing and people is to dig quick-fried possibility, to be that non-uniform settling and interface sealing rubber are aging leak potential booster factor, so the booster form is only considered linear cracking.
For characteristics and the historical booster feature of reference and fidelity factor that pipeline section in the target pipe network lays, combining target pipe network pipeline minimum detects the booster eigenwert, simulates the booster damaged condition for the pipeline section of buried-pipe laying in the target pipe network by the impaired area ratio; Simulate the booster damaged condition for the steel pipe that the flange that lays connects by the impaired area ratio in integrated pipe canal.
(2) for booster simulation is evenly distributed in each pipeline section of target pipe network, so the booster position is arranged in the middle of the simulation booster pipeline section.
(3) choose toward the nominal situation of the highest operating mode when per day of annual target tube net water yield fidelity factor before as booster.In the SCADA system, consult the pressure value of each pressure tap in the water supply network, carry out the pipe network adjustment in the ingress pipe pessimistic concurrency control, obtain the operating mode when pipe network normally moves before the booster, obtain the data such as each node flow, booster monitoring point.
(4) in the pipe network model, select to carry out the pipeline section of booster simulation, setting incident node b of booster position increase, draw a short tube at incident node b, another node of short tube is denoted as node c.The Pipeline damage area is scaled the short tube equivalent diameter, goes out the flow field simulation booster with short tube and go out stream.Be the closing to reality situation, short tube should be lacked as far as possible, to reduce the extra loss of flood peak.In addition, the coefficient of shock resistance of short tube is established by the orifice outflow coefficient.
(5) calculate booster state duty parameters such as obtaining tube bursting and leakage flow, destruction place's hydraulic pressure and the pressure variation of booster monitoring point.Its method is as follows: it is 9 that node c sets a leakage flow, carries out the pipe network adjustment by the pipe network operating mode before the booster again, obtains the free hydraulic pressure of node c
Because hole exits hydraulic pressure is 0, therefore continuous tentative calculation leakage flow extremely
When
After, the node flow of node c is the tube bursting and leakage flow, and the node pressure on the incident node b is destruction place hydraulic pressure, and the node pressure on the booster monitoring point is the booster feature hydraulic pressure of certain damaged condition on the accident pipeline section.
(6) because the span data of simulating is larger, for the ease of the training of network, prevent that problems such as " overfittings " from appearring in computation process, if input and the target vector of neural network carried out certain pre-service, can accelerate the training speed of network.Therefore this research is carried out normalized by the following method to training data.
The pressure variation represents namely with rate of change α:
In the formula:
H
0---the normal hydraulic pressure of the front monitoring point of booster;
The change value of pressure of monitoring point before and after Δ H---the booster;
The booster position represents namely with relative distance β:
In the formula:
Sign (x
t-x
c)---if the difference of booster position x coordinate and monitoring point x coordinate is for negative, and expression booster position just is being east in the west, monitoring point:
m
Tc---the distance of booster position and monitoring point:
m
CmaxOne---the monitoring point represents the maximum monitoring distance in monitoring point to the ultimate range of target pipe network;
After normalization, pressure change rate α span is in [0,1]. and relative distance β span is in [1,1].
According to above way, utilize the pipe network model to carry out the training sample that the booster simulation produces.
The structure of the 4th step network
(1) sets up three layers of BP neural network, comprise an input layer, hidden layer, output layer.Wherein input layer has 3 neurons, corresponds respectively to the variation in water pressure rate of 3 monitoring points; Output layer is respectively l, 3 neurons, and the distance and the booster place pressure that correspond respectively to 3 monitoring points and leakage point change.
(2) the case is extremely complicated about the setting of the neuron of hidden layer, not yet finds so far one well to determine neuron number purpose method.The hidden layer node number can cause learning time long too much, and the hidden layer node number is very little, poor fault tolerance, and identification is low without the ability sample of study.From present result of study, hidden layer node number and input-output unit number have direct relation, research and propose and calculate as follows the hidden layer node number:
In the formula:
N---hidden layer node number;
n
i, n
0---input, output node number;
A---constant is between l to 10;
By formula 3, the hidden layer neuron of objective network is 3 to 12.Find that in the network training process hidden layer neuron is at 3 to 12 o'clock, mean square deviation is near 0.005.In tentative calculation, find, neuron is at 13 when above, increasing mean square deviation with neuron number decreases, when neuron number is 31, be neuron number when equaling target tube webmaster hop count, the mean square deviation of two models is minimum, if neuron continues to increase, mean square deviation descends little, but the training time is multiplied.Therefore, find after tentative calculation no matter comparatively suitable in mean square deviation and training time 31 neurons are.
(3) select the TmirLlm function as the training function.Aspect the algorithm selection, in order to access stable training result, thereby select for medium scale network training and have very fast constringent Levenberg-Marquardt algorithm.
(4) the study factor is made as 0.3, and training circulation is 10000 times, and desired value and the mean square deviation of network output valve are as controlling target in the training sample.
Through training, booster damage assessment model training curve based on the BP artificial neural network, therefrom as seen, after 10000 training circulations, the mean square deviation of desired value and network output valve in the training sample, individualized training sample and network output valve maximum error, average error is less than the quantity of average error.
Localization of bursted pipe model training curve based on the BP artificial neural network, after 10000 training circulations, obtain that the mean square deviation of desired value and network output valve exists in the training sample, fail training sample desired value and network output valve maximum error for three groups, average error, most data error is under average error.Superimpose data is scaled distance, value and real-valued the differing below 100m of network output, and judge the booster position in the method rate of accuracy reached 100% of that direction of monitoring point with sign, and locating accuracy is higher, and training effect is just more satisfactory.
The 5th step booster position decision method
(1) the model output valve is the distance between corresponding monitoring point and the booster position, and take the monitoring point coordinate points as the center of circle, output distance is made semicircle for radius, and the output distance value is negative two, the three quadrant semicircles of doing, export distance value for just do one, the four-quadrant semicircle.
(2) having average error according to output valve and desired value, all is booster position fiducial intervals in the inside and outside 80m scope of circular arc.The zone that the booster position fiducial interval that three monitoring points are exported overlaps is decided to be the booster zone, and pipeline section is the booster pipeline section of location model prediction in the zone.
Claims (2)
1. localization of bursted pipe method based on the BP artificial neural network comprises that the target pipe burst detects the determining of the determining of minimum value, network input/output argument, the foundation of training sample, the structure of network, booster position decision method.
2. the localization of bursted pipe method based on the BP artificial neural network according to claim 1 is characterized in that foundation and the network struction of training sample.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201210426108 CN102982374A (en) | 2012-10-30 | 2012-10-30 | Pipe explosion positioning method based on back propagation (BP) artificial neural networks |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201210426108 CN102982374A (en) | 2012-10-30 | 2012-10-30 | Pipe explosion positioning method based on back propagation (BP) artificial neural networks |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102982374A true CN102982374A (en) | 2013-03-20 |
Family
ID=47856350
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201210426108 Pending CN102982374A (en) | 2012-10-30 | 2012-10-30 | Pipe explosion positioning method based on back propagation (BP) artificial neural networks |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102982374A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103574291A (en) * | 2013-07-02 | 2014-02-12 | 同济大学 | Tube bursting positioning method based on artificial immunity system |
| CN106874532A (en) * | 2016-12-23 | 2017-06-20 | 武汉智博创享科技有限公司 | A kind of grid method for analyzing cartridge igniter |
| CN108763464A (en) * | 2018-05-29 | 2018-11-06 | 杭州电子科技大学 | Water supply localization of bursted pipe method based on monitoring point cluster and abnormal area gravity model appoach |
| CN110688776A (en) * | 2019-10-16 | 2020-01-14 | 熊猫智慧水务有限公司 | Pipe burst identification method based on pipe network adjustment |
| CN110858315A (en) * | 2018-08-13 | 2020-03-03 | 西门子医疗有限公司 | Deep machine learning based magnetic resonance imaging quality classification considering less training data |
| CN112097126A (en) * | 2020-09-18 | 2020-12-18 | 同济大学 | Water supply network pipe burst pipeline accurate identification method based on deep neural network |
-
2012
- 2012-10-30 CN CN 201210426108 patent/CN102982374A/en active Pending
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103574291A (en) * | 2013-07-02 | 2014-02-12 | 同济大学 | Tube bursting positioning method based on artificial immunity system |
| CN103574291B (en) * | 2013-07-02 | 2016-11-23 | 同济大学 | Localization of bursted pipe method based on artificial immune system |
| CN106874532A (en) * | 2016-12-23 | 2017-06-20 | 武汉智博创享科技有限公司 | A kind of grid method for analyzing cartridge igniter |
| CN108763464A (en) * | 2018-05-29 | 2018-11-06 | 杭州电子科技大学 | Water supply localization of bursted pipe method based on monitoring point cluster and abnormal area gravity model appoach |
| CN108763464B (en) * | 2018-05-29 | 2021-08-03 | 杭州电子科技大学 | Water supply pipe burst positioning method based on monitoring point clustering and abnormal region gravity center method |
| CN110858315A (en) * | 2018-08-13 | 2020-03-03 | 西门子医疗有限公司 | Deep machine learning based magnetic resonance imaging quality classification considering less training data |
| CN110858315B (en) * | 2018-08-13 | 2023-11-03 | 西门子医疗有限公司 | Deep machine learning based magnetic resonance imaging quality classification with less training data considered |
| CN110688776A (en) * | 2019-10-16 | 2020-01-14 | 熊猫智慧水务有限公司 | Pipe burst identification method based on pipe network adjustment |
| CN110688776B (en) * | 2019-10-16 | 2023-01-20 | 熊猫智慧水务有限公司 | Pipe burst identification method based on pipe network adjustment |
| CN112097126A (en) * | 2020-09-18 | 2020-12-18 | 同济大学 | Water supply network pipe burst pipeline accurate identification method based on deep neural network |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107420743B (en) | Intelligent urban gas PE pipe network measurement and control system and measurement and control method | |
| CN102982374A (en) | Pipe explosion positioning method based on back propagation (BP) artificial neural networks | |
| CN110245411B (en) | Urban central heating system and leakage point detection method | |
| CN109578816B (en) | Water supply pipe network pipe explosion detection method and device, control method and device | |
| CN103939749A (en) | Oil pipeline network leakage intelligent self-adaptation monitoring system and method based on big data | |
| CN109442221B (en) | A kind of water supply network booster method for detecting extracted based on pressure disturbance | |
| CN106015951A (en) | Gas transmission pipeline leakage detection system and method suitable for multiple state changes | |
| CN108758357B (en) | A kind of localization of bursted pipe method based on the analysis of water supply network change value of pressure | |
| CN107013812B (en) | A kind of THM coupling line leakage method | |
| CN109886506B (en) | Water supply network pipe explosion risk analysis method | |
| CN113074324B (en) | Database based on urban water supply pipe network operation safety dynamic early warning and establishing method | |
| Karathanasi et al. | Development of a leakage control system at the water supply network of the city of Patras | |
| CN103559399A (en) | Water transportation pipe network leakage positioning method based on Bayesian decision theory and genetic algorithm | |
| He et al. | A framework of smart pipeline system and its application on multiproduct pipeline leakage handling | |
| CN110285330B (en) | Water network pipe burst detection method based on local outlier factor | |
| CN108388685A (en) | The prediction technique and device of leakage current amount in oil-gas pipeline | |
| CN107291997A (en) | A kind of cold rolling hydraulic AGC system design of Fault Diagnosis Strategy method | |
| CN103544527A (en) | Method of predicting water level of oxygen extractor in ultra-supercritical steam turbine under FCB conditions | |
| CN110543719B (en) | Water supply pipeline leakage prediction method based on space metering model | |
| CN103337000B (en) | A kind of safe monitoring and pre-alarming method of oil-gas gathering and transportation system | |
| CN112393125A (en) | Gas pipe network leakage detection system and method based on inverse problem analysis | |
| Wang et al. | Research on Intelligent Location Method of Water Supply Pipe Network Burst Based on BP Neural Network Deep Learning | |
| Hu et al. | Adaptive Dynamic Programming-Based Method for Signal Evaluation of Energy Transportation System | |
| Pengfei et al. | Burst analysis of water supply pipe based on hydrodynamic simulation | |
| Wu et al. | Leakage Location Method of Water Supply Pipe Network Based on Integrated Neural Network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20130320 |