Summary of the invention
The purpose of this invention is to provide a kind of method that adopts Fuzzy Pattern Recognition prediction tunnel gas, specifically be to carry out gas abnormity early warning prediction, overcome the influence of the unsafe factor that the gas state mutation brought according to the field monitoring data characteristic of gas density in the gas tunnel.
The present invention is a kind of method that adopts Fuzzy Pattern Recognition prediction tunnel gas, lay computer in ground central station, data communication interface and substation, face in the tunnel, lining cutting, add the broadband and monitoring point, return air inlet place is laid the face sensor respectively, the lining cutting sensor, add broadband sensor and return air inlet sensor, described four monitoring points are laid left arch springing sensor respectively, right arch springing sensor and dome sensors, all the sensors is connected with substation by communication cable, substation is connected with communication interface by the transmission data wire, data communication interface is connected with computer, it is characterized in that, comprise the steps:
(1) sticks with paste the mode identification technology model at the built-in formwork erection of described computer;
(2) described various sensor is transformed into the signal of telecommunication to the gas density that is monitored and is transferred to described substation;
(3) described substation carries out analyzing and processing respectively to the gas density data that difference detects the position, and the gas density after will handling simultaneously passes to described computer;
(4) computer is drawn the time graph figure according to gas density, determines 3 important abnormity point positions in the curvilinear figure: unusual starting point, slowly rise to the slow rising of fast rise/be climbed to/slowly drop to fast rise/drop to fast turning point, the peak point of slow decline.
(5) computer is determined the parameter of Fuzzy Pattern Recognition technology model according to described time graph figure and abnormity point position, calls the Fuzzy Pattern Recognition technology model gas density that monitors is carried out prediction.
Abnormity point is meant that gas density is 0.1% point among the present invention, slowly rise to the intensity of variation that the slow rising of fast rise/be climbed to/slowly drop to fast rise/drop to fast " slowly " in the slow decline, " fast " refer to gas density function curve mean curvature, curvature value is " slowly rising " or " slowly descending " between 0~0.5; Curvature value is " fast rise " or " descending fast " greater than 0.5.
The present invention carries out the early warning prediction by the gas monitor data are carried out gas monitor to gas distribution trend in the tunnel, provides reference frame for formulating control gas technical measures, simultaneously, guarantee constructor's safety, therefore, the present invention is significant to the gas tunnel construction safety.
The specific embodiment
Disclosed all features in this manual, or the step in disclosed all methods or the process except mutually exclusive feature and/or step, all can make up by any way.
Disclosed arbitrary feature in this manual (comprising any accessory claim, summary and accompanying drawing) is unless special narration all can be replaced by other equivalences or the alternative features with similar purpose.That is, unless special narration, each feature is an example in a series of equivalences or the similar characteristics.
As shown in Figure 1, the Fuzzy Pattern Recognition technology is used in a plurality of monitoring points, certain tunnel Gas Distribution is carried out the trend prediction forecast.
The system that comprises in this method: lay computer, data communication interface and substation in ground central station, face, lining cutting in the tunnel, add the broadband and monitoring point, return air inlet place is laid face sensor, lining cutting sensor respectively, added broadband sensor and return air inlet sensor, described four monitoring points are laid left arch springing sensor, right arch springing sensor and dome sensors respectively, all the sensors is connected with substation by communication cable, substation is connected with data communication interface by the transmission data wire, and data communication interface is connected with computer.
Control and measuring software and Fuzzy Pattern Recognition technology model are housed on the computer, and monitoring mode is selected to have: 1) single-point is surveyed:
Show face place, four monitoring points, lining cutting place respectively, add the monitored data of each monitoring point in broadband place and the return air inlet place; 2) branch is surveyed: the monitored data that can select to show arbitrary sensor; Substation have to the monitoring point sensor acquisition to data analyze, handle, shield the function of interfere information; The gas density of each corresponding monitoring point of sensor monitors.
Start computer when in the tunnel, constructing, orders such as computer sends to substation 3 simultaneously by data communication interface and disposes, patrols and examines, control, substation is given various sensors command transfer by communication cable again, the sensor that is distributed in each monitoring point begins the gas density of corresponding monitoring point is monitored in real time, sensor becomes the signal of telecommunication with the gas density transformation of data that monitors, be transferred to substation 3 by communication cable simultaneously, substation is passed to computer with analysis processing result by data-interface, calculates and finishes functions such as demonstration, printing, warning.
Choosing the data of 3 monitoring points in the present embodiment comes gas density following situation a middle or short term is carried out the trend prediction forecast.3 formed curvilinear figures of monitoring point monitored data are as follows:
1. seven days the per hour gas monitor data time curvilinear figure of 07.11.20-07.11.26 that arrives of face sensor monitors, wherein the A left side is for being used for the actual measurement part of modeling, and the A right side is an actual measurement part of proving modeling.
2. 30 gas monitor data time curvilinear figures of four days of 07.12.07-07.12.10 of arriving of k16+650 monitoring point left side arch springing sensor monitors, wherein the B left side is for being used for the actual measurement part of modeling, and the B right side is an actual measurement part of proving modeling.
3. the per day monitored data time graph figure of the 05.11.19-05.12.13 that arrives of K16+993 monitoring point sensor monitors, wherein the C left side is for being used for the actual measurement part of modeling, and the C right side is the actual measurement part that is used for proving modeling.
With Fuzzy Pattern Recognition Method it is carried out trend prediction, the steps include:
(1) extract following feature:
Feature 1: gas concentration F
1, promptly the amplitude of curve can be described as: " low " μ
11, " medium " μ
12" height " μ
13Feature 2: increment F
2, i.e. the increment of curve, the slope of curve is measured, and can be described as: " decline " μ
21, " gently " μ
22" rising " μ
23Feature 3: peak F
3, the spike number for the exception curve occurs can be described as: " no peak " μ
31, " unimodal " μ
32" multimodal,, μ
33
The initial data that obtains is realized the conversion of initial data to characteristic vector according to the extraction principle of characteristic value, and sample modeling to be checked characteristic vector partly is as follows:
{F
1,F
2,F
3}={0.08,0.06,0}
{F
1,F
2,F
3}={0.55,0.52,1}
{F
1,F
2,F
3}={0.09,0.06,2}
(2) obfuscation of characteristic parameter
Characteristic vector is blured division according to experiment and experience, determine the degree of membership μ of each parameter
Ij(i=1,2,3; J=1,2,3), the degree of membership division principle of each characteristic parameter:
??F
1 |
??μ
11 |
??μ
12 |
??μ
13 |
??F
1≥05%
|
??0 |
??0 |
??1 |
??03%≤F
1<05%
|
??0 |
??0.25 |
??0.75 |
??01%≤F
1<03%
|
??0 |
??0.75 |
??0.25 |
??F
1<01%
|
??1 |
??0 |
??0 |
??F
2 |
??μ
21 |
??μ
22 |
??μ
33 |
??F
2≥03%
|
??0 |
??0 |
??1 |
??01%≤F
2<03%
|
??0 |
??0.25 |
??0.75 |
??0≤F
2<01%
|
??0 |
??1 |
??0 |
??F
2<0
|
??1 |
??0 |
??0 |
??F
3 |
??μ
31 |
??μ
32 |
??μ
33 |
??F
3=0
|
??1 |
??0 |
??0 |
??F
3=1
|
??0 |
??1 |
??0 |
??F
3>1
|
??0 |
??0 |
??1 |
The unusual fuzzy set A of various typical cases
iUse F
iMembership function mui
IjCombination express, promptly be expressed as the form A of fuzzy vector
i=(μ
11, μ
12, μ
13, μ
21, μ
22, μ
23, μ
31, μ
32, μ
33), every kind of typical abnormal waveforms all is A
iOn the fuzzy subset, be expressed as follows:
Mild normal type: A
1=(1,0,0; 0,1,0; 1,0,0)
The normal shape of triangle: A
11=(0,0.5,0.5; 0,1,0; 0,1,0)
Increment normal type: A
31=(0,1,0; 0,1,0; 0,1,0)
Increment ectype: A
32=(0,1,0; 0,1,0; 1,0,0)
Multimodal is unusual: A
2=(0,0.5,0.5; 0,0.5,0.5; 0,0,1)
Be translated into characteristic vector in the domain according to the division principle of degree of membership
??μ
ij |
??μ
11 |
??μ
12 |
??μ
13 |
??μ
21 |
??μ
22 |
??μ
22 |
??μ
31 |
??μ
32 |
??μ
33 |
??B
1 |
??1 |
??0 |
??0 |
??0 |
??1 |
??0 |
??1 |
??0 |
??0 |
??B
2 |
??0 |
??0 |
??1 |
??0 |
??0 |
??1 |
??0 |
??1 |
??0 |
??B
3 |
??1 |
??0 |
??0 |
??0 |
??1 |
??0 |
??0 |
??0 |
??1 |
(3) decision rule
The approach degree of selecting Euclidean distance is as the decision rule on the sample training collection described in (two).The operational formula of Euclidean approach degree is as follows:
X in the formula
iBe the feature that comprises, A, B are the fuzzy subset, u
A, u
BFor passing judgment on the object factor.The approach degree that calculates according to the operational formula of Euclidean approach degree:
??ρ
i |
??ρ(A
1,B
1)
|
??ρ(A
11,B
1)
|
??ρ(A
31,B
1)
|
??ρ(A
32,B
1)
|
??ρ(A
2,B
1)
|
??B
1 |
??1 |
??0.376 |
??0.333 |
??0.528 |
??0.333 |
??ρ
i |
??ρ(A
1,B
2)
|
??ρ(A
11,B
2)
|
??ρ(A
31,B
2)
|
??ρ(A
32,B
2)
|
??ρ(A
2,B
2)
|
??B
2 |
??0.1835 |
??0.473 |
??0.333 |
??0.423 |
??0.451 |
??ρ
i |
??ρ(A
1,B
3)
|
??ρ(A
11,B
3)
|
??ρ(A
31,B
3)
|
??ρ(A
32,B
3)
|
??ρ(A
2,B
3)
|
??B
3 |
??0.528 |
??0.376 |
??0.333 |
??0.333 |
??0.514 |
The future trend that the monitored data of these 3 actual monitoring points is made with Fuzzy Pattern Recognition predicts the outcome and is respectively: mild normal type, triangle normal type and mild normal type.This predicts the outcome and conforms to the sample measured data, and sample measured data figure is shown in the right side branch of A, B, C.
The present invention is not limited to the aforesaid specific embodiment.The present invention expands to any new feature or any new combination that discloses in this manual, and the arbitrary new method that discloses or step or any new combination of process.