CN116305985A - Local intelligent ventilation method based on multi-sensor data fusion - Google Patents
Local intelligent ventilation method based on multi-sensor data fusion Download PDFInfo
- Publication number
- CN116305985A CN116305985A CN202310305069.4A CN202310305069A CN116305985A CN 116305985 A CN116305985 A CN 116305985A CN 202310305069 A CN202310305069 A CN 202310305069A CN 116305985 A CN116305985 A CN 116305985A
- Authority
- CN
- China
- Prior art keywords
- data
- parameter
- sensor
- environmental
- fusion
- 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
- 230000004927 fusion Effects 0.000 title claims abstract description 42
- 238000000034 method Methods 0.000 title claims abstract description 36
- 238000009423 ventilation Methods 0.000 title claims abstract description 21
- 230000007613 environmental effect Effects 0.000 claims abstract description 70
- 230000005641 tunneling Effects 0.000 claims abstract description 30
- 238000001514 detection method Methods 0.000 claims abstract description 13
- 238000004140 cleaning Methods 0.000 claims abstract description 4
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 claims description 23
- 229910002091 carbon monoxide Inorganic materials 0.000 claims description 23
- 239000000428 dust Substances 0.000 claims description 21
- 238000012544 monitoring process Methods 0.000 claims description 13
- 238000012360 testing method Methods 0.000 claims description 13
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000012549 training Methods 0.000 claims description 8
- 230000008878 coupling Effects 0.000 claims description 5
- 238000010168 coupling process Methods 0.000 claims description 5
- 238000005859 coupling reaction Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 230000007935 neutral effect Effects 0.000 claims description 4
- 230000001502 supplementing effect Effects 0.000 claims description 4
- 238000012795 verification Methods 0.000 claims description 4
- 208000025174 PANDAS Diseases 0.000 claims description 3
- 208000021155 Paediatric autoimmune neuropsychiatric disorders associated with streptococcal infection Diseases 0.000 claims description 3
- 235000016496 Panda oleosa Nutrition 0.000 claims description 3
- 238000007500 overflow downdraw method Methods 0.000 claims description 3
- 101100272279 Beauveria bassiana Beas gene Proteins 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 claims description 2
- 238000005259 measurement Methods 0.000 claims description 2
- 238000005457 optimization Methods 0.000 claims description 2
- 240000004718 Panda Species 0.000 claims 2
- 230000003044 adaptive effect Effects 0.000 claims 1
- 238000006243 chemical reaction Methods 0.000 claims 1
- 238000010606 normalization Methods 0.000 claims 1
- 230000009469 supplementation Effects 0.000 claims 1
- 238000004519 manufacturing process Methods 0.000 abstract description 6
- 239000003245 coal Substances 0.000 abstract description 5
- 230000008859 change Effects 0.000 abstract description 3
- 230000001276 controlling effect Effects 0.000 abstract description 3
- 230000001105 regulatory effect Effects 0.000 abstract description 3
- 238000007664 blowing Methods 0.000 abstract 1
- 230000006870 function Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 7
- 238000013139 quantization Methods 0.000 description 5
- 238000004088 simulation Methods 0.000 description 4
- 239000011159 matrix material Substances 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000004880 explosion Methods 0.000 description 2
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 240000000220 Panda oleosa Species 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 239000002817 coal dust Substances 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 208000035475 disorder Diseases 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- E—FIXED CONSTRUCTIONS
- E21—EARTH OR ROCK DRILLING; MINING
- E21F—SAFETY DEVICES, TRANSPORT, FILLING-UP, RESCUE, VENTILATION, OR DRAINING IN OR OF MINES OR TUNNELS
- E21F1/00—Ventilation of mines or tunnels; Distribution of ventilating currents
- E21F1/006—Ventilation at the working face of galleries or tunnels
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F04—POSITIVE - DISPLACEMENT MACHINES FOR LIQUIDS; PUMPS FOR LIQUIDS OR ELASTIC FLUIDS
- F04D—NON-POSITIVE-DISPLACEMENT PUMPS
- F04D27/00—Control, e.g. regulation, of pumps, pumping installations or pumping systems specially adapted for elastic fluids
- F04D27/004—Control, e.g. regulation, of pumps, pumping installations or pumping systems specially adapted for elastic fluids by varying driving speed
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mining & Mineral Resources (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Geometry (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geochemistry & Mineralogy (AREA)
- Geology (AREA)
- Mechanical Engineering (AREA)
- Feedback Control In General (AREA)
Abstract
The invention provides a local intelligent ventilation method based on multi-sensor data fusion, and belongs to the field of mine intelligent ventilation. The method comprises the following steps: step 1, acquiring environmental parameters in a tunneling roadway through sensors of different types, and establishing an environmental parameter detection system; step 2, cleaning the data of the environmental parameter data set; step 3, carrying out data level fusion, feature level fusion and decision level fusion on the processed data set, and establishing an environment parameter prediction system; step 4, predicting environmental parameters by using an environmental parameter prediction system model; step 5, comparing and analyzing according to the environmental parameters and the predicted parameters in the tunneling roadway, and correcting the model; step 6, setting an air quantity fuzzy controller; and 7, determining a fuzzy input value, operating an air quantity fuzzy controller, and intelligently regulating and controlling the air quantity. The intelligent air quantity control system can predict the environmental change in the tunneling roadway, realize intelligent air quantity control, change the traditional ventilation fan 'one air blowing' mode and save electric energy for coal mine production.
Description
Technical Field
The invention relates to the technical field of mine ventilation, in particular to a local intelligent ventilation method based on multi-sensor data fusion.
Background
In the actual production process, the conditions in the tunnel are continuously changed along with the increase of the tunneling length of the working face, the effective air quantity of the tunneling working face is reduced, the dust concentration is too high, gas is accumulated, the coal mine production environment is influenced, and even accidents such as gas explosion, coal dust explosion, mine fire and the like are caused. In order to avoid the conditions of wind flow disorder, insufficient air quantity and excessive ventilation of a tunneling working face, the air quantity of the working face needs to be continuously regulated so as to meet the air quantity requirements under different conditions. The effectiveness of air quantity regulation can directly influence the mine safety production, the air quantity redundancy can increase the power consumption of a main ventilator, however, the problems of overhigh temperature, dust hazard, gas accumulation and the like cannot be effectively solved by too small air quantity, and accidents are easy to occur.
The effectiveness of air volume adjustment depends on the variation of different environmental parameters of the working surface. Early air volume adjustments were typically determined by the skilled artisan based on actual data or working experience, lacking support from scientific theory. Due to the severe working environment of the working face, errors exist in collected data due to the fact that aging, damage and the like of a sensor detection element are easy to occur, and the collected data cannot accurately reflect the working environment. The data is analyzed by adopting a single-class sensor model, and only a certain environmental parameter is considered as a dependent variable for regulating and controlling the air quantity, so that the normal air quantity requirement of a working face can not be met, and the safety production of a coal mine is affected. The environment condition of the working face needs to be comprehensively mastered to accurately and timely adjust the air quantity, and multiple kinds of sensors are needed to acquire different environment parameters of the working face and conduct data fusion, so that the environment state of the working face is predicted, intelligent air quantity regulation and control is achieved, and normal production of a coal mine is ensured.
Disclosure of Invention
In order to improve the ventilation problem, the invention mainly provides a local intelligent ventilation method based on multi-sensor data fusion to realize local intelligent ventilation in a tunneling working face, and provides an effective implementation scheme for ventilation intellectualization.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
1. a local intelligent ventilation method based on multi-sensor data fusion comprises an environment parameter detection system, an environment parameter multi-data fusion system, an environment parameter prediction system and a local ventilation intelligent regulation and control system, wherein:
the environmental parameter detection system refers to various sensors in a tunneling roadway, including a wind speed sensor, a gas concentration sensor, a carbon monoxide sensor, a dust concentration sensor and a temperature and humidity sensor. The wind speed sensor is used for monitoring the average wind speed of the tunneling roadway; the gas concentration sensor is used for monitoring the gas concentration of the tunneling roadway; the carbon monoxide sensor is used for monitoring the carbon monoxide concentration of the tunneling roadway; the dust concentration sensor is used for monitoring the dust concentration in the tunneling roadway; the temperature and humidity sensor is used for monitoring the temperature and humidity of the tunneling tunnel and realizing the real-time acquisition of the environmental parameters such as wind speed, gas concentration, carbon monoxide concentration, dust concentration, temperature and humidity in the tunneling tunnel.
The environment parameter multi-data fusion system is responsible for processing data acquired by various sensors, integrating the data into an environment parameter set based on a time sequence, correcting and supplementing abnormal data and missing data existing in the parameter set, and carrying out data fusion on the processed multi-category environment parameter set.
The data fusion process is to adopt an arithmetic mean value recursive estimation algorithm to single-class environmental parameters, and then complete data-level fusion to the original environmental parameter set by using a self-adaptive weighting algorithm; and carrying out feature coupling on the original environmental parameter features by a feature polynomial fusion method to realize feature level fusion.
The environment parameter prediction system is responsible for carrying out data mining on the environment parameters in the tunneling roadway through the LSTM neural network algorithm on the parameter set after data fusion is completed, extracting common characteristics in the parameter set, predicting the environment parameter change of the tunneling working face, and providing data support for intelligent regulation and control of local ventilation.
The local ventilation intelligent regulation and control system mainly comprises an air volume fuzzy controller, and has the main functions of selecting environmental parameters of an environmental parameter prediction system to compare with environmental parameter thresholds, selecting the maximum item of the prediction parameters and the corresponding thresholds as the input quantity of the air volume fuzzy controller, and using the frequency of a frequency converter as the output quantity to finish intelligent control of local air volume.
2. A local intelligent ventilation method based on multi-sensor data fusion comprises the following steps:
Wherein a, b, c, d, e, f respectively represents carbon monoxide concentration, gas concentration, dust concentration, temperature, humidity and wind speed, s is the number of the sensor, s epsilon N, i represents the ith minute,
represents the concentration of carbon monoxide detected by the carbon monoxide sensor at the i-th minute s; a total set P of the environmental parameters in the roadway can be obtained.
(1) The missing values existing in the corresponding environmental parameter time sequence aggregate are supplemented by adopting a linear interpolation method, and the formula is as follows:
wherein x is p And x p+r The environmental parameter values at the p-th time and the p+r time, x p+q Is the missing environmental parameter value at the p+q time.
(2) And carrying out data fusion on the supplemented single-class environmental parameter time sequence by using an arithmetic mean value recurrence estimation method, wherein the specific steps are as follows:
1) The arithmetic mean of the environmental parameter set is calculated as follows:
wherein the method comprises the steps of
For the arithmetic mean of the collection of such sensor parameters, n is the number of elements in the collection, x i The parameter value at the i-th time.
2) Will be
As a recursion estimation initial value, consistency test is carried out on subsequent detection parameters, and the formula is as follows:
wherein the method comprises the steps of
Is an error requirement.
3) If the above formula is satisfied, consider x i+1 Compliance with consistency check, recalculate x i And x i+1 Is a recursive estimate x of (2) i ' +1 And x is taken as i ' +1 And carrying out consistency check and recursive estimation on the sensor detection value at the next moment. If not, calculate x i+2 And x i+3 Is a recursive estimate x of (2) i ' +2 、x i ' +3 For x respectively i ' +1 And x i ' +2 ，x i ' +2 And x i ' +3 Consistency checking, if both are satisfied, reserving x i+1 The method comprises the steps of carrying out a first treatment on the surface of the If not, reject x i+1 Supplementing the removed data by adopting a linear interpolation method, and
still as a recurrence initial.
For the estimated parameter x i ' +1 The calculation formula is as follows:
wherein, the liquid crystal display device comprises a liquid crystal display device,
is x i 、x i+1 Is a variance of (c).
For the estimated parameter x i ' +1 Variance of (2)
The calculation formula is as follows:
(2) And carrying out self-adaptive weighting algorithm processing on the parameter set after the arithmetic mean value recursive estimation method, wherein the specific steps are as follows:
1) Calculating the weighting factor theta of the s-th sensor data after j times of measurement s The calculation formula is as follows:
2) The fusion estimation value Y is calculated, and the calculation formula is as follows:
(3) Repeating the steps to obtain data fusion values of carbon monoxide concentration, gas concentration, dust concentration, wind speed, temperature and humidity.
wherein T is n Normalized value of the feature, T is the original feature value, T max T is the maximum characteristic value in the data set min Feature minima in the data set.
Step 4, predicting environmental parameters by using an environmental parameter detection system model; an LSTM multi-parameter data fusion environment prediction system is adopted to extract space-time characteristics affecting environmental changes from sensor monitoring data. The normalized data set is divided into non-test sets with 95% part and 5% test set according to time sequence, and is divided into training set with 95% of non-test set and verification set with 5%. And carrying out parameter optimization by a grid search method, and determining corresponding parameters of the LSTM model.
The LSTM model input gate formula is as follows:
i t ＝σ(W i ·[h t-1 ,x t ]+b i ) (11)
the LSTM model forget gate formula is:
f t ＝σ(W f ·[h t-1 ,x t ]+b f ) (13)
the LSTM model output gate formula is:
o t ＝σ(W o ·[h t-1 ,x t ]+b o ) (14)
h t ＝o t *tanh(C t ) (15)
cell state calculation formula:
wherein i is t Is an input door W i For inputting door weight parameter, b i Representing input gate bias term, f t Indicating forgetting the door,W f For forgetting gate weight parameter matrix, b f For forgetting door bias term, o t For outputting door W o To output the gate weight parameter matrix, b o Representing the output gate bias term, sigma is a sigmoid activation function, h t-1 Conceal layer state, x for time t t For the original data vector of the original data matrix corresponding to x at time t,
cell state at time t, C t-1 Is the cell state at the moment t and h t At time t at C t And calculating the output under control.
And step 5, inputting the normalized characteristic parameter training set, the normalized verification set and the normalized test set into a model, and training and testing the model to realize the prediction of environmental parameters.
And 6, analyzing and evaluating the prediction result by adopting a Mean Square Error (MSE), and verifying the validity of the prediction result.
The formula is as follows:
where MSE is the mean square error, n represents the number of samples in the training set, Y i As a function of the true value of the data,
is a data predictor.
Step 7, setting an air quantity fuzzy controller; according to the actual environmental parameter detected by the environmental parameter detection system and the environmental parameter predicted by the environmental parameter prediction system, the environmental error E and the error rate E are obtained c And obtaining corresponding fuzzy quantity according to the respective membership functions, performing fuzzy decision through a fuzzy control rule, and controlling the output quantity h of the frequency converter by adopting a discrete accurate quantity method to perform fuzzy judgment. The method comprises the following specific steps:
(1) The air volume fuzzy controller selects a single environmental parameter as an input quantity, and uses the frequency of the frequency converter as an output quantity, and the basic structure is shown in figure 3.
(2) Setting an input variable environmental error E and an environmental error rate E c And the frequency h physical domain of the output variable frequency converter.
Let the environmental error physical domain be [1, -1], the environmental error physical domain be [ -0.5,0.5], the output variable physical domain be [0,10].
(3) The physical domain is converted into a quantization domain according to a quantization formula, and the formula is as follows:
wherein a is i 、b i Take value for physical domain of theory, x H For the upper limit of the physical domain, x L Is the lower limit of the physical domain, k is the quantization factor,
according to the above, taking n=3 for the environmental error quantization domain as the available environmental error quantization domain
{ -3, -2, -1,0,1,2,3}, the environmental error rate physical argument taking n=6 may yield a quantized argument of { -3, -2, -1,0,1,2,3}, and the output variable physical argument taking n=1.67 may yield a quantized argument of { -3, -2, -1,0,1,2,3}.
(4) Establishing a fuzzy subset and a fuzzy language set.
The fuzzy subset of the environmental error is { NB, NS, ZO, PS, PB }, and the fuzzy language set is { oversized deviation, large deviation, general deviation, small deviation and no deviation }; the fuzzy subset of the environment error rate is { NB, NS, ZO, PS, PB }, the fuzzy language set is { negative big, negative small, zero, positive small, positive big }; the output h fuzzy subset is { NB, NS, ZOPS, PB }, and the fuzzy language set is { highest, very high, medium high, small high, none };
selecting a fuzzy control rule if-then conditional statement: { if e=a i and E c ＝B j then h＝C k The method comprises the steps of carrying out a first treatment on the surface of the i, j, k=1, 2 … … } and determines the control rules.
Wherein A is i As the environmental error EFuzzy subset, B j C is a fuzzy subset of the environmental error rate Ec k Is a fuzzy subset of the frequency h of the frequency converter.
(5) The application if-then conditional statement expression fuzzy statement is specifically as follows:
1)If E is NB and Ec is NB then h is PB；
2)If E is NB and Ec is NS then h is PS；
3)If E is NB and Ec is ZO then h is ZO；
4)If E is NB and Ec is PS then h is NS；
5)If E is NB and Ec is PB then h is NS；
6)If E is NS and Ec is NB then h is PS；
7)If E is NS and Ec is NS then h is ZO；
8)If E is NS and Ec is ZO then h is NS；
9)If E is NS and Ec is PS then h is NS；
10)If E is NS and Ec is PB then h is NS；
11)If E is ZO and Ec is NB then h is ZO；
12)If E is ZO and Ec is NS then h is ZO；
13)If E is ZO and Ec is ZO then h is ZO；
14)If E is ZO and Ec is PS then h is NS；
15)If E is ZO and Ec is PB then h is NB；
16)If E is PS and Ec is NB then h is ZO；
17)If E is PS and Ec is NS then h is NS；
18)If E is PS and Ec is ZO then h is NS；
19)If E is PS and Ec is PS then h is NB；
20)If E is PS and Ec is PB then h is NB；
21)If E is PB and Ec is NB then h is ZO；
22)If E is PB and Ec is NS then h is PS；
23)If E is PB and Ec is ZO then h is PB；
24)If E is PB and Ec is PS then h is PS；
25)If E is PB and Ec is PB then h is ZO。
according to the 25 control rules, a fuzzy control rule table is obtained, and is shown in table 1.
TABLE 1 fuzzy rule control Table
(6) The Mamdni minimum algorithm is adopted as an reasoning mode, the trigonometric function is a membership function of the input quantity E, ec, and a membership function diagram of E, ec is obtained through MATLAB simulation, and is shown in fig. 4 and 5.
(7) Input E, ec and an h fuzzy inference graph are obtained through MATLAB simulation, and fig. 6 is shown.
(8) And obtaining an input quantity E, ec and an h fuzzy inference graph through MATLAB simulation. And adopting Sugeno reasoning mode and trigonometric function as membership functions of each input quantity, and obtaining a control curved surface diagram of E, ec and h through MATLAB simulation, wherein the control curved surface diagram is shown in figure 7.
Step 8, determining a fuzzy input value; the actual environment condition of the tunneling roadway determines expected thresholds of all environment parameters, the highest ratio of the predicted value to the expected value is selected as a given value of the fuzzy controller, the air quantity fuzzy controller is operated, the fan frequency converter is intelligently controlled, and intelligent regulation and control of the air quantity are realized.
(1) Respectively establishing corresponding upper limit of environmental parameters according to the actual condition of the tunneling roadway and the related regulations of the coal mine, wherein the carbon monoxide concentration threshold value is a * The gas concentration threshold value is b * The dust concentration threshold value is c * A temperature threshold value of d * A humidity threshold value of e * 。
(2) And selecting the maximum value of the ratio of the environment predicted value to the corresponding threshold value as the fuzzy control input quantity, and operating the air quantity fuzzy controller.
Drawings
FIG. 1 is a schematic diagram of a local intelligent ventilation control model based on multi-sensor data fusion;
FIG. 2 is a flowchart of an LSTM neural network algorithm prediction model in the environmental parameter prediction system of the present invention;
FIG. 3 is a block diagram of a stroke volume fuzzy controller according to the present invention;
FIG. 4 is a membership diagram of the input E using trigonometric function according to the present invention;
FIG. 5 is a membership diagram of the input Ec using trigonometric function according to the present invention;
FIG. 6 is a fuzzy inference chart of the input E, ec and output h according to the present invention;
fig. 7 is a graph showing the control curves of the input amount E, ec and the output amount h according to the present invention.
Claims (3)
1. A local intelligent ventilation method based on multi-sensor data fusion is characterized in that: the method comprises the following steps:
step 1, establishing a tunneling working face environment parameter detection system; arranging carbon monoxide, gas, dust, wind speed and temperature and humidity sensors in each l m of a tunneling roadway, arranging N parts in total, acquiring carbon monoxide, gas, dust, wind speed and temperature and humidity in the roadway as environmental parameters in real time every minute, and acquiring a roadway internal environmental parameter set at the moment
Wherein a, b, c, d, e, f respectively represents carbon monoxide concentration, gas concentration, dust concentration, temperature, humidity and wind speed, s is the number of the sensor, s epsilon N, i represents the ith minute,
the carbon monoxide concentration detected by the carbon monoxide sensor at the ith minute s is represented to obtain a total set P of environmental parameters in the roadway:
step 2, converting the parameter set P into a time sequence more suitable for analysis by using a python program, and cleaning data of the generated data set to reduce errors; invoking a pandas library to convert the environmental parameter set into a time sequence set A of the same type of sensor according to the sensor type s ,B s ,C s ,D s ,E s ,F s Wherein A is s ,B s ,C s ,D s ,E s ,F s Respectively representing a time series set of carbon monoxide concentration, gas concentration, dust concentration, wind speed, temperature and humidity detected by a No. s sensor, supplementing the missing values existing in the time series set of corresponding environmental parameters by adopting a linear interpolation method, and adopting the following formula:
wherein x is p And x p+r The environmental parameter values at the p-th time and the p+r time, x p+q The environmental parameter value missing at the p+q time;
step 3, data fusion of data level, feature level and decision level is sequentially carried out on the data set; the method comprises the steps of completing data level fusion by adopting an arithmetic mean value recursive estimation method and a self-adaptive weighting algorithm, performing feature engineering processing by adopting a feature set to complete feature level fusion, training data after feature level fusion by adopting an LSTM algorithm, and establishing a multi-sensor data fusion-based environment parameter monitoring system to perform data fusion on a single-category environment parameter time sequence after supplementation by using the arithmetic mean value recursive estimation method, wherein the method comprises the following specific steps of:
(1) The specific steps of completing data level fusion by adopting an arithmetic average value recursive estimation method and an adaptive weighting algorithm are as follows:
1) The method for performing arithmetic average recursion estimation comprises the following specific steps:
a. the arithmetic mean of the environmental parameter set is calculated as follows:
wherein the method comprises the steps of
For the arithmetic mean of the collection of such sensor parameters, n is the number of elements in the collection, x i The parameter value is the parameter value at the i-th moment;
b. will be
As a recursion estimation initial value, consistency test is carried out on subsequent detection parameters, and the formula is as follows:
wherein the method comprises the steps of
Is an error requirement;
c. if the above formula is satisfied, consider x i+1 Compliance with consistency check, recalculate x i And x i+1 Is a recursive estimate of x' i+1 And x' i+1 Consistency test is carried out on the sensor detection value at the next moment, and recursive estimation is carried out; if not, calculate x i+2 And x i+3 Is a recursive estimate of x' i+2 、x′ i+3 For x 'respectively' i+1 And x' i+2 ，x′ i+2 And x' i+3 Consistency checking, if both are satisfied, reserving x i+1 The method comprises the steps of carrying out a first treatment on the surface of the If not, reject x i+1 Supplementing the removed data by adopting a linear interpolation method, and
still as a recurrence initial;
for the estimated parameter x' i+1 The calculation formula is as follows:
wherein delta i 2 、
Is x i 、x i+1 Is a variance of (2);
for the estimated parameter x' i+1 Variance of (2)
The calculation formula is as follows:
2) The self-adaptive weighting algorithm is carried out, and the specific steps are as follows:
a. calculating the weighting factor theta of the s-th sensor data after j times of measurement s The calculation formula is as follows:
b. the fusion estimation value Y is calculated, and the calculation formula is as follows:
3) Repeating the steps to respectively obtain data fusion values of carbon monoxide concentration, gas concentration, dust concentration, wind speed, temperature and humidity;
(2) And carrying out characteristic engineering treatment on the data set subjected to data level fusion, wherein the specific steps are as follows:
performing feature coupling on the original environmental parameter features by a feature polynomial fusion method, removing repeated items, and adding 23 coupling feature combinations ab, ac, ad, ae, bc, bd, be, cd, ce, abc, abd, abe, acd, ade, bcd, bce, bde, cde, abcd, acde, bcde, abce, abcde;
(3) Performing LSTM algorithm processing on the data set subjected to feature level fusion to complete decision level fusion, wherein the method comprises the following specific steps of:
setting a lag phase step LS, a neutral phase step GS and a time step TS of a data set by taking an environmental parameter characteristic sample as a supervision sample, taking the difference value of the lag phase step and the neutral phase step as the number of environmental parameter lag phase characteristics as new time sequence characteristics, constructing an LSTM multi-parameter data fusion environmental parameter prediction model, and carrying out normalization processing on a characteristic parameter fusion value, wherein the formula is as follows:
wherein T is n Normalized value of the feature, T is the original feature value, T max T is the maximum characteristic value in the data set min Feature minima in the data set;
step 4, predicting environmental parameters by using an environmental parameter detection system model; extracting space-time characteristics affecting environmental changes from sensor monitoring data by adopting an LSTM multi-parameter data fusion environment prediction system; dividing the normalized data set into a non-test set 5% according to a time sequence, dividing the non-test set 95% into a training set and dividing the non-test set 5% into a verification set, performing super-parameter optimization by a grid search method, and determining corresponding parameters of an LSTM model;
step 5, comparing and analyzing according to the environmental parameters and the predicted parameters in the tunneling roadway, and correcting the model; inputting the normalized characteristic parameter training set, verification set and test set into a model, and training and testing the model to realize the prediction of environmental parameters;
step 6, setting an air quantity fuzzy controller; comprising the following steps: determining the structure of a fuzzy controller, formulating a univariate definition fuzzy input and output set, defining a membership function, establishing a fuzzy control rule and determining an actual control quantity;
step 7, determining a given value of the fuzzy controller; according to the actual environment condition of the tunneling roadway, determining expected thresholds of all environment parameters, selecting the highest ratio of the predicted value to the expected value as a given value of the fuzzy controller, operating the air quantity fuzzy controller, performing intelligent control on the fan frequency converter, and realizing intelligent regulation and control of the air quantity.
2. The local intelligent ventilation method based on multi-sensor data fusion according to claim 1, wherein: the wind speed sensor in the step 1 is used for monitoring the average wind speed of a tunneling roadway; the gas concentration sensor is used for monitoring the gas concentration of the tunneling roadway; the carbon monoxide sensor is used for monitoring the carbon monoxide concentration of the tunneling roadway; the dust concentration sensor is used for monitoring the dust concentration in the tunneling roadway; the temperature and humidity sensor is used for monitoring the temperature and humidity of the tunneling tunnel and realizing the real-time acquisition of the environmental parameters such as wind speed, gas concentration, carbon monoxide concentration, dust concentration, temperature and humidity in the tunneling tunnel.
3. The local intelligent ventilation method based on multi-sensor data fusion according to claim 1, wherein: in the step 2, the python program is used to call the pandas library and the corresponding programming code to perform format conversion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310305069.4A CN116305985A (en) | 2023-03-27 | 2023-03-27 | Local intelligent ventilation method based on multi-sensor data fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310305069.4A CN116305985A (en) | 2023-03-27 | 2023-03-27 | Local intelligent ventilation method based on multi-sensor data fusion |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116305985A true CN116305985A (en) | 2023-06-23 |
Family
ID=86837701
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310305069.4A Pending CN116305985A (en) | 2023-03-27 | 2023-03-27 | Local intelligent ventilation method based on multi-sensor data fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116305985A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117115758A (en) * | 2023-10-25 | 2023-11-24 | 山西榕行智能科技有限公司 | Control method based on intelligent main coal flow transport AI monitoring system of coal mine |
-
2023
- 2023-03-27 CN CN202310305069.4A patent/CN116305985A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117115758A (en) * | 2023-10-25 | 2023-11-24 | 山西榕行智能科技有限公司 | Control method based on intelligent main coal flow transport AI monitoring system of coal mine |
| CN117115758B (en) * | 2023-10-25 | 2023-12-26 | 山西榕行智能科技有限公司 | Control method based on intelligent main coal flow transport AI monitoring system of coal mine |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lindemann et al. | Anomaly detection and prediction in discrete manufacturing based on cooperative LSTM networks | |
| CN112990556A (en) | User power consumption prediction method based on Prophet-LSTM model | |
| CN116757534A (en) | Intelligent refrigerator reliability analysis method based on neural training network | |
| CN112036075A (en) | Abnormal data judgment method based on environmental monitoring data association relation | |
| CN114282443B (en) | Residual service life prediction method based on MLP-LSTM supervised joint model | |
| CN113325721B (en) | Model-free adaptive control method and system for industrial system | |
| CN116050665B (en) | Heat supply equipment fault prediction method | |
| CN110874616A (en) | Transformer operation prediction method based on LSTM network and Markov chain correction error | |
| CN112329990A (en) | User power load prediction method based on LSTM-BP neural network | |
| CN116776745B (en) | Method and system for predicting pollutant concentration and track movement based on edge calculation | |
| CN113219871B (en) | Curing room environmental parameter detecting system | |
| CN116305985A (en) | Local intelligent ventilation method based on multi-sensor data fusion | |
| CN114169254A (en) | Abnormal energy consumption diagnosis method and system based on short-term building energy consumption prediction model | |
| CN110782067A (en) | Sintering end point fluctuation range prediction method based on fuzzy information granulation | |
| Lv et al. | Non-iterative T–S fuzzy modeling with random hidden-layer structure for BFG pipeline pressure prediction | |
| CN110222825B (en) | Cement product specific surface area prediction method and system | |
| CN112580254B (en) | Multi-dimensional analysis and early warning method for converter transformer based on multi-level decision fusion | |
| CN114970745B (en) | Intelligent security and environment big data system of Internet of things | |
| CN115062764B (en) | Intelligent illuminance adjustment and environmental parameter Internet of things big data system | |
| CN115688865A (en) | Industrial soft measurement method for long and short term memory network for flue gas of desulfurization process | |
| CN114486262B (en) | CNN-AT-LSTM-based bearing residual service life prediction method | |
| CN114386672B (en) | Environment big data Internet of things intelligent detection system | |
| CN114390376B (en) | Fire big data remote detection and early warning system | |
| CN114399024B (en) | Oil gas concentration big data intelligent detection system | |
| CN113218598B (en) | Air tightness detection system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |