CN110564905A - Signal processing method and system for blast furnace lining impact echo detection - Google Patents
Signal processing method and system for blast furnace lining impact echo detection Download PDFInfo
- Publication number
- CN110564905A CN110564905A CN201910950250.4A CN201910950250A CN110564905A CN 110564905 A CN110564905 A CN 110564905A CN 201910950250 A CN201910950250 A CN 201910950250A CN 110564905 A CN110564905 A CN 110564905A
- Authority
- CN
- China
- Prior art keywords
- signal
- blast furnace
- furnace lining
- echo
- obtaining
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- C—CHEMISTRY; METALLURGY
- C21—METALLURGY OF IRON
- C21B—MANUFACTURE OF IRON OR STEEL
- C21B7/00—Blast furnaces
- C21B7/24—Test rods or other checking devices
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Materials Engineering (AREA)
- Metallurgy (AREA)
- Organic Chemistry (AREA)
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
- Length Measuring Devices Characterised By Use Of Acoustic Means (AREA)
Abstract
The invention discloses a signal processing method and a system for detecting blast furnace lining impact echo, which are characterized in that a time difference ultrasonic method is utilized to calibrate the wave velocity of a material of a measured object to obtain calibrated wave velocity information, the impact echo signal is subjected to empirical mode decomposition to obtain an intrinsic mode component, the impact echo signal is filtered based on the intrinsic mode component to obtain a filtering signal and an energy operator based on the filtering signal, the filtering signal is classified, the thickness information of the innermost layer of the blast furnace lining is obtained based on the classification result and the wave velocity information, the technical problem that the detection result of the erosion state of the blast furnace lining is greatly deviated due to the aliasing effect generated by the propagation of elastic waves in a non-uniform layered medium in the prior art is solved, the acquired impact echo signal is denoised and the filtering signal is further classified by adopting the energy operator, various types of echoes can be effectively distinguished, therefore, the corresponding characteristics of the interface are accurately extracted, and finally, the thickness information of the innermost layer of the blast furnace lining is accurately extracted.
Description
Technical Field
The invention relates to the technical field of blast furnace smelting, in particular to a signal processing method and a signal processing system for blast furnace lining impact echo detection.
background
Blast furnace smelting is a continuous production process, the operation condition in the furnace is complex, and various physical and chemical reactions are numerous. In recent years, with the increase of the demand of steel in various fields, the output of various iron works is gradually increased, the service lives of the hearth and the bottom of the blast furnace face great challenges in the face of the improvement of smelting strength, and the outer wall of the blast furnace hearth can be summarized into three types of materials, namely, an innermost carbon layer, a middle tamping material and a tamping material, and an outermost steel plate furnace shell. In the iron making process, molten iron in a molten state at an inner hearth of a blast furnace is produced by chemical reactions of over 1000 kinds of gases and materials such as coke and iron ore. In such a severe environment of high temperature and high pressure, the innermost carbon layer of the furnace wall can be subjected to high-temperature molten corrosion and molten iron circulation scouring to generate oxidation reaction and metal corrosion, and the magnesia carbon bricks in the carbon layer are easily corroded at the high temperature of the molten iron, so that detection of the thickness of the furnace wall has great significance for detection of the corrosion state of the furnace lining of the blast furnace in operation production, and safe and efficient production.
The existing blast furnace lining detection mainly achieves the monitoring purpose by embedding monitoring equipment, carrying out echo nondestructive detection and establishing a furnace lining furnace wall heat transfer model. The embedded monitoring equipment is mainly characterized in that optical fibers and thermocouples are embedded in a furnace lining, and the erosion state of the furnace lining is judged according to the existence of a transmission signal of the equipment, but the method has requirements on a blast furnace and is not strong in adaptability. The echo nondestructive test uses an elastic wave time-of-flight method to calculate the local thickness and cracks of the furnace lining, but has the disadvantages of low precision, easy numerical value shift, and the like. The heat transfer model method is used for establishing a heat transfer model aiming at the furnace wall and the furnace bottom through mechanism analysis and predicting the erosion state, but the method needs to do many hypothesis processing, so the model precision is not high.
The patent publication No. CN 1436861A blast furnace lining ablation monitoring device and the monitoring method thereof, the device pre-embeds optical fibers, transceivers and the like in brick joints when a blast furnace is built, when molten iron slowly erodes to a certain U-shaped detection element, the optical fiber protection tube and the optical fibers inside the optical fiber protection tube are fused by the high temperature of the molten iron, and then the ablation condition of the blast furnace lining is judged by the existence of optical fiber end reflected light signals. However, the patent requires that the light and the matching device are embedded in the furnace lining, and although the cost of the whole set of equipment is not high, the method has requirements on the blast furnace, is not suitable for the constructed blast furnace or the operated blast furnace, and has certain limitations.
System, method and apparatus for non-disruptive and non-destructive inspection of metallurgical furnaces and the like according to patent publication No. CN 1825056 a, which proposes a method for determining the condition of the refractory lining using instantaneously propagating stress waves, taking into account the effect of temperature on the velocity of the compression waves through the heated refractory material, slag. However, the patent does not consider the aliasing effect generated by the reflection of each interface when the wave propagates in the combined medium, and severe deviation is easily generated in the calculation result when the physical property difference of the materials is large.
The patent publication No. CN 101812559A blast furnace lining erosion analysis monitoring method provides a method for constructing a heat transfer model according to the temperature of a blast furnace lining thermocouple, the material of a refractory material and the parameters of a cooling system and calculating the position of an isotherm, thereby displaying the position change and the shape of an erosion boundary. However, the thermocouple needs to be embedded in advance, and the thermocouple is prone to failure in a long-term high-temperature state, so that local data failure can be caused, further, the model is inaccurate, and adverse effects can be caused on blast furnace guidance.
Disclosure of Invention
The signal processing method and the signal processing system for detecting the blast furnace lining impact echo solve the technical problem that in the prior art, the detection result deviation of the blast furnace lining erosion state is large due to the fact that the aliasing effect generated by the propagation of elastic waves in a non-uniform layered medium is not considered.
In order to solve the technical problem, the signal processing method for detecting the blast furnace lining impact echo provided by the invention comprises the following steps:
Carrying out wave velocity calibration on the material of the measured object by using a time difference ultrasonic method to obtain calibrated wave velocity information;
performing empirical mode decomposition on the impact echo signal to obtain an intrinsic mode component;
filtering the impulse echo signal based on the intrinsic mode component to obtain a filtered signal;
And classifying the filtering signals based on the energy operator of the filtering signals, and obtaining the thickness information of the innermost layer of the blast furnace lining based on the classification result and the wave velocity information.
further, performing empirical mode decomposition on the impulse echo signal, and obtaining an intrinsic mode component includes:
Obtaining all maximum values and minimum value points of the impact echo signal, thereby obtaining an upper envelope line corresponding to the maximum values and a lower envelope line corresponding to the minimum value points;
Based on the upper envelope line, the lower envelope line and a preset threshold condition, performing empirical mode decomposition on the impact echo signal to obtain an intrinsic mode component, wherein the preset threshold condition specifically comprises:
Wherein s isi(k-1)(n) represents si(n) the k-1 th decomposition quantity, s, not satisfying the threshold conditionik(n) represents si(n) difference of the k-1 th decomposition amount with the average value m (n), andWherein s ish(n) is the upper envelope, slAnd (n) is a lower envelope line, alpha is a screening threshold, and alpha is 0.2, and when the result meets the threshold condition, the decomposition is completed.
Further, filtering the impulse echo based on the eigenmode component, and obtaining a filtered signal includes:
Acquiring an intrinsic mode component threshold value of sudden change of noise energy distribution based on the intrinsic mode component;
And carrying out wavelet soft threshold denoising on the intrinsic mode component starting from the intrinsic mode component threshold value, thereby obtaining a filtering signal.
Further, based on the eigenmode component, a calculation formula for obtaining the eigenmode component threshold value of the sudden change of the noise energy distribution specifically includes:
When a local minimum precedes the global minimum:
Otherwise:
Where N denotes the data length of a single imf component,to representandthe continuous mean square error of two adjacent imf components,The value of k corresponding to the minimum value of the local function is shown,and representing the value of k corresponding to the minimum value of the global function.
further, classifying the filtered signals based on energy operators of the filtered signals, and obtaining the thickness information of the innermost layer of the blast furnace lining based on the classification result and the wave velocity information comprises the following steps:
Based on an energy operator of the filtering signal, obtaining arrival time information from the impact echo to the detection surface;
And obtaining the thickness information of the innermost layer of the blast furnace lining based on the arrival time information and the wave velocity information.
Further, obtaining the time-of-arrival information of the impulse echo to the detection surface based on an energy operator of the filtered signal comprises:
Obtaining the amplitude envelope of the filtered signal according to the filtered signal, wherein the specific calculation formula of the amplitude envelope of the filtered signal is as follows:
Wherein, | a (n) | represents an amplitude envelope of the filtered signal, s ' (n) represents an nth signal reconstructed after the de-noising by the EMD based on the CMSE criterion, s ' (n +1) represents an nth +1 signal reconstructed after the de-noising by the EMD based on the CMSE criterion, s ' (n-1) represents an nth-1 signal reconstructed after the de-noising by the EMD based on the CMSE criterion, ψ [ s ' (n) ] represents an energy operator function of s ' (n), ψ [ s ' (n +1) -s ' (n-1) ] represents an energy operator function of [ s ' (n +1) -s ' (n-1) ];
And obtaining the arrival time information of the impact echo to the detection surface based on the local minimum value of the amplitude envelope of the filtering signal.
Further, the wave velocity calibration of the material of the measured object by using the time difference ultrasonic method to obtain the calibrated wave velocity information comprises the following steps:
Smearing the medical ultrasonic coupling agent on the transmitting probe and the receiving probe, and directly contacting the surface of the transmitting probe with the surface of the receiving probe;
Calculating the propagation time occupied by a transmitting crystal in the transmitting probe to the surface of a transmitting source and the propagation time occupied by a receiving crystal in the receiving probe to the surface of a receiving source to obtain first propagation time;
Carrying out wave velocity calibration on a material of a measured object to obtain the propagation time of a wave in the measured object, wherein the propagation time is the sum of the propagation time of the wave in the material and the first propagation time;
and obtaining the calibrated wave velocity information according to the propagation time and the thickness of the measured object.
the invention provides a signal processing system for detecting blast furnace lining impact echo, which comprises:
The device comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, and further realizes the steps of the signal processing method for detecting the blast furnace lining impact echo when the processor executes the computer program.
Compared with the prior art, the invention has the advantages that:
the invention provides a signal processing method and a system for detecting blast furnace lining impact echo, which calibrate the wave velocity of a material of a measured object by utilizing a time difference ultrasonic method to obtain calibrated wave velocity information, carry out empirical mode decomposition on the impact echo signal to obtain an intrinsic mode component, filter the impact echo signal based on the intrinsic mode component to obtain a filter signal and an energy operator based on the filter signal, classify the filter signal, and obtain the thickness information of the innermost layer of the blast furnace lining based on the classification result and the wave velocity information, thereby solving the technical problem that the detection result deviation of the blast furnace lining erosion state is large due to the aliasing effect generated by the propagation of elastic waves in a non-uniform layered medium in the prior art, denoising the acquired impact echo signal and further classifying the filter signal by adopting the energy operator, and effectively distinguishing various types of echo, therefore, the corresponding characteristics of the interface are accurately extracted, and finally, the thickness information of the innermost layer of the blast furnace lining is accurately extracted.
drawings
FIG. 1 is a flow chart of a signal processing method for detecting blast furnace lining impact echo according to a first embodiment of the present invention;
FIG. 2 is a flowchart of a signal processing method for detecting blast furnace lining shock echoes according to a second embodiment of the present invention;
fig. 3 is a schematic diagram of light rays propagating in the medium 1 and the medium 2 of the stress wave of the second embodiment of the present invention;
FIG. 4 is a schematic diagram of calibrating the wave velocity of the material of the object to be measured according to the second embodiment of the present invention;
FIG. 5 is a schematic view of a refractory clay brick according to a second embodiment of the present invention;
FIG. 6 is a diagram illustrating the envelope result of the energy operator according to the second embodiment of the present invention;
FIG. 7 is a block diagram of a signal processing system for detecting blast furnace lining shock echoes according to an embodiment of the invention.
Reference numerals:
10. A memory; 20. a processor; 100. a transmitting element; 200. and a collecting element.
Detailed Description
in order to facilitate an understanding of the invention, the invention will be described more fully and in detail below with reference to the accompanying drawings and preferred embodiments, but the scope of the invention is not limited to the specific embodiments below.
The embodiments of the invention will be described in detail below with reference to the drawings, but the invention can be implemented in many different ways as defined and covered by the claims.
Example one
referring to fig. 1, a signal processing method for detecting blast furnace lining impact echo according to an embodiment of the present invention includes:
s101, calibrating the wave velocity of a material of a measured object by using a time difference ultrasonic method to obtain calibrated wave velocity information;
step S102, carrying out empirical mode decomposition on the impact echo signal to obtain an intrinsic mode component;
step S103, filtering the impact echo signal based on the intrinsic mode component to obtain a filtered signal;
And S104, classifying the filtering signals based on the energy operator of the filtering signals, and obtaining the thickness information of the innermost layer of the blast furnace lining based on the classification result and the wave velocity information.
the signal processing method for detecting blast furnace lining impact echo provided by the embodiment of the invention comprises the steps of calibrating the wave velocity of a material of a measured object by utilizing a time difference ultrasonic method to obtain calibrated wave velocity information, carrying out empirical mode decomposition on the impact echo signal to obtain an intrinsic mode component, filtering the impact echo signal based on the intrinsic mode component to obtain a filtering signal and an energy operator based on the filtering signal, classifying the filtering signal, and obtaining the thickness information of the innermost layer of the blast furnace lining based on the classification result and the wave velocity information, so that the technical problem that the detection result deviation of the blast furnace lining erosion state is large due to the aliasing effect generated by the propagation of elastic waves in a non-uniform layered medium in the prior art is solved, and various types of echoes can be effectively distinguished by denoising the acquired impact echo signal and further adopting the energy operator to classify the filtering signal, therefore, the corresponding characteristics of the interface are accurately extracted, and finally, the thickness information of the innermost layer of the blast furnace lining is accurately extracted.
specifically, the embodiment denoises the acquired impact echo signal and further classifies the filtering signal by adopting an energy operator, so that the time and the frequency of the wave which is generated by different interfaces and reaches a detection surface for the first time under the conditions of no interference and no echo overlap can be effectively distinguished, various types of echoes are distinguished, the corresponding characteristics of the interface are further accurately extracted, finally, the thickness information of the innermost layer of the blast furnace lining is accurately extracted, the aliasing effect generated by the propagation of elastic waves in a non-uniform layered medium can be eliminated, the detection error generated by the aliasing effect is overcome, the accuracy of the extracted thickness information of the innermost layer of the blast furnace lining is high, the erosion state of the lining is further accurately reflected through the thickness information, and the method has important significance on the safe production of the blast furnace.
example two
Referring to fig. 2, a signal processing method for detecting blast furnace lining impact echo according to a second embodiment of the present invention includes:
step S201, the material of the object to be measured is subjected to wave velocity calibration by using a time difference ultrasonic method, and calibrated wave velocity information is obtained.
specifically, in the present embodiment, when processing the signal of blast furnace lining impact echo detection, it is first necessary to study the propagation mode of the wave in the solid. The nature of the thickness measurement by using the shock echo is calculated by using the characteristic frequency corresponding to the interface of the measured object and the environment and the propagation speed of the shock wave in the measured object. And because the shock wave is reflected and transmitted at the heterogeneous interface of the heterogeneous medium such as a steel-concrete structure or a carbon-concrete structure, as shown in fig. 3.
The stress wave transmitting end excites an impact, the frequency is different according to the different contact time of the transmitting device and the object to be measured, and the corresponding main frequency can be different. In FIG. 3, S1For a longitudinal wave of the excited stress wave to propagate downwards in the medium 1, S2is S1Reflected wave, S, at the interface of the media 1, 23is S1Transmitted wave of S3The reflected wave after contacting the innermost interface of the measured object is S4the reflected wave S4When reaching the interface between the media 1, 2, the same reflection and refraction, S, occur6Is S4Transmitted wave of S5Is S4The reflected wave of (1) is analogized in turn, S7Bit S5Reflected wave of (2), S8Is S7the transmitted wave of (1).
Because the thickness value of the lining of the blast furnace is far smaller than the propagation velocity value of the elastic wave in the lining of the blast furnace, the time difference between the reflection and refraction of the elastic wave at each heterogeneous interface and the reflection of the elastic wave continuing to generate at the next heterogeneous interface reaches the detection surface is very small, so that the characteristics containing the thickness information are interfered by aliasing of the waveform.
When the wave velocity calibration is carried out on the material of the tested object, the time difference ultrasonic method is utilized to detect 3 opposite surfaces of the refractory material, and the medical ultrasonic coupling agent is firstly coated on the transmitting probe and the receiving probe, so that the influence on the experimental result caused by the air layer between the probe and the tested object is avoided. Before testing the refractory material, it is necessary to bring the transmitting probe face and the receiving probe face into direct contact, as shown in fig. 4.
In fig. 4, the dashed box is a crystal, which includes a transmitting element 100 and an acquisition element 200, and the method is used to acquire signals and obtain the head wave receiving time t by the instrument1At this time, there is TOF (time of flight) calculation formula:
In the formula (1), t represents the propagation time of the wave in the measured object, s represents the thickness of the measured medium in the longitudinal wave direction, and v represents the wave velocity of the longitudinal wave in the test direction.
As shown in formula (1), t1Representing the time it takes for the actual source of emission and source of reception to be in contact at no distance within the probe. And then calibrating the material to obtain the time t of the propagation of the wave in the material2plus the propagation time t taken by the emitting crystal to the emitting source surface and the receiving crystal to the receiving source surface1I.e. by
t＝t1+t2 (2)
order towherein f issIs the sampling rate of the device, fmthe test wave velocity is the main frequency peak of the test data, when m is larger than 2, the data has integrity, otherwise, the sampling is incomplete, and the test wave velocity is smaller.
finally, when the wave velocity of the refractory material is calibrated, the wave velocities obtained by testing different opposite surfaces are different, and the refractory clay brick shown in the figure 5 is taken as an example.
The embodiment introduces the concept of the shape coefficient, and when the object to be measured is infinite, the shape coefficient is 1; when the object is a brick as shown in fig. 5, the influence of the boundary is considered, and when an excitation point on any one surface is selected, the wave propagates in all directions in the object, so that the reflected wave generated by the boundary interferes with the calculation.
TABLE 1
| shape of | plate-like shape | cube | Cylinder | Ring |
| Coefficient beta | 0.96 | 0.87 | 0.92 | 0.96 |
After introducing the shape factor, the calculation formula of the wave velocity is:
the average value result of 5 times of ultrasonic measurement of the surface 1 and the opposite surface thereof by using time difference is 4230.8m/s, the average value result of 5 times of ultrasonic measurement of the surface 2 and the opposite surface thereof by using time difference is 4318.2m/s, in a Cartesian coordinate system, the physical properties of parts of the X axis perpendicular to the surface 1, the Y axis perpendicular to the surface 2 and the X direction and the Y direction are different, so that the wave speeds are different, and the refractory brick is known to be an anisotropic medium.
Step S202, empirical mode decomposition is carried out on the impact echo signal, and intrinsic mode components are obtained.
Because the production environment of the blast furnace is full of a large amount of noise, including instantaneous vibration caused by overhaul operation of workers, collision between materials and mechanical devices, electronic interference of instruments and the like, the traditional method for converting time domain signals into frequency domains for denoising by Fourier transform is not suitable for instantaneous non-stationary signals such as impact echoes, and has no precedent for specific signals of a measured object, and a proper wavelet basis cannot be accurately selected for denoising the measured object consisting of complex non-uniform media, so that the embodiment adopts an EMD denoising and similar wavelet soft threshold method for signal processing. EMD (Empirical Mode Decomposition), also called Empirical Mode Decomposition, is an Empirical signal Decomposition method proposed by Huang et al, NASA, and decomposes a complex wideband signal into a plurality of narrowband signals, each component becomes an eigenmode function (imf), which has a good effect when processing unknown target signals, theoretically all signals can be decomposed by EMD, and signal components from high amplitude to low amplitude and low frequency and residual signals are obtained.
Specifically, the present embodiment first searches all the maximum and minimum points of the signal s (n) and interpolates them with cubic splinesfitting all the maximum value points and minimum value points by using a value algorithm to obtain an upper envelope line s corresponding to the maximum value pointsh(n) and a lower envelope sl(n) then there is the formula of the mean value
Set variable s1(n) is the first decomposition quantity, let s1(n) s (n) -m (n), if s is calculated1If (n) satisfies the condition (eigenmode component) imf, imf1 is s1(n), if not, then s1(n) as raw data, let s11(n)＝s1(n) -m (n), repeating the above steps until s1(k-1)(n) satisfying the condition that s is allowed to stand after k-times of screening1(k-1)(n) is imf1 component, imf1 component is stripped off, and residual signal r (1) is s (n) -s1(n), repeating the above steps for r (1) to obtain the remaining imf component, considering the practical situation, the average value of the upper and lower envelopes is almost impossible to be 0, so a threshold value needs to be set when s is1(k-1)(n) satisfies the condition:
In the formula (5), s1(k-1)(n) represents s1(n) the k-1 th decomposition quantity, s, not satisfying the threshold condition1k(n) represents s1And (n) the difference between the k-1 th decomposition amount and the average value m (n), wherein alpha is a screening threshold and is 0.2.
the method finally comprises the following steps:
in the formula (6), r (n) is the residual quantity of the intrinsic mode function of s (n) after empirical mode decomposition, and is a monotonous signal or a constant. The original signal is decomposed into a sum of imf components and the remaining components through empirical mode decomposition.
Step S203, acquiring an eigenmode component threshold value of noise energy distribution mutation based on the eigenmode component;
Specifically, after the signal in this embodiment is subjected to EMD decomposition, imf components with different frequencies, high-frequency components corresponding to small orders, and high-frequency portions of the signal corresponding to low-order components are obtained, and it is generally considered that the high-frequency portions contain high-frequency noise, which may be a noise signal generated by worker maintenance or a collision with a sharp object. The low-frequency component corresponding to the high-order number concentrates on the part of the effective signal of the impulse echo detection, so that a screening component k needs to be determined, and imf components after the k-th order are the dominant signals of the detection signal, so as to achieve the aim of filtering noise. In this embodiment, an EMD denoising method based on a continuous mean square error criterion is selected, and a continuous mean square error is obtained by continuously calculating the mean square error of two adjacent features:
Where N represents the data length of a single imf component,Representing the kth reconstructed imf component,Representing the (k +1) th reconstructed imf component,Represents the amplitude of the k imf th component at the i-th time point of the time domain of length N,Imf representing the amplitude of the (k +1) th imf-th component at the i-th point in time in a time domain of length N in timek(ti) Representing the difference of the amplitudes of two adjacent imf components at the ith time point of the time domain with the time length of N,to representandThe continuous mean square error of two adjacent imf components.
in this embodiment, the kth imf component corresponding to the abrupt change of the noise energy distribution in the formula is first found, and because some imf energies are drowned by noise modal energy when the test signal is dominant, and then become the global minimum, in this case, k needs to be defined in cases:
When a local minimum precedes the global minimum:
otherwise:
Where N denotes the data length of a single imf component,To representandThe continuous mean square error of two adjacent imf components,The value of k corresponding to the minimum value of the local function is shown,and k corresponding to the minimum value of the global function is taken.
Step S204, carrying out wavelet soft threshold denoising on the eigenmode component starting from the eigenmode component threshold value, thereby obtaining a filtering signal.
the embodiment obtains the determined k value according to step S203, and then imfkand (3) further denoising all the initial components by using a wavelet soft threshold function:
applying it modified on IMF, there are
Wherein, imf'kRepresents the k imf component after soft threshold de-noising, t in equation (11)jThe threshold value for the jth imf component is given by equation (12):
T in formula (12)jMeaning taking at the jth componentN is the length of the signal, σjFor the standard deviation of the interference signal over j components, byjEstimated as y/0.6745, y is the absolute median of the wavelet coefficients on the k-th component.
after the above steps, the signal is reconstructed according to equations (11) and (12), and a signal s' (n) with the high-frequency interference signal and part of the low-frequency noise removed is obtained.
after the steps, the signal obtained by the embodiment basically only has the test signal and the interference similar to the test signal, including the low-frequency interference caused by the low-frequency impact with similar frequency or the friction between materials. The signal also includes reflected and refracted waves generated by the media of the layers forming the furnace wall at the furnace hearth in the figure 3, and the elastic waves reach the detection surface with similar time difference and are difficult to distinguish. Except that the echo of the first layer of medium can be clearly distinguished in a time-frequency domain, the thickness information of the remaining second layer and the thickness information of the remaining third layer are difficult to identify, so that signals are classified, and time-frequency information corresponding to the reflected wave of the innermost layer is extracted.
and S205, acquiring the arrival time information from the impact echo to the detection surface based on the energy operator of the filtering signal.
And step S206, obtaining the thickness information of the innermost layer of the blast furnace lining based on the arrival time information and the wave velocity information.
Specifically, echo data s' (n) after noise reduction in this embodiment is basically composed of wavelets generated by the same excitation, as shown in fig. 3, the final result has high correlation, various wavelets overlap to cause aliasing, and because the difference between media constituting the furnace lining is not particularly large, the arrival time of 3 reflections generated at the interface between the second layer and the third layer and the arrival time of the last layer reflected to the detection surface are infinitely close to each other, so that the corresponding thickness information of the two layers is difficult to distinguish, and corrosion detection is interfered.
in order to distinguish the time and frequency of arrival of the primary reflection-free echoes generated by different interfaces at the detection surface, and extract the corresponding characteristics of the interfaces, the embodiment further processes the signal s' (n) by using an energy operator sum.
Energy operator ψ for s' (n)d[s'(n)]The following definitions are provided:
ψd[s'(n+1)]＝[s'(n)]2-s'(n+1)s'(n-1) (14)
Using 3 point symmetric difference formula, the first derivative of s' (n) is dividedAnd second orderRespectively convert into
The energy function of the signal s' (n) is converted into
ψc[s'(t)]＝(ψd[s'(n+1)]+2ψc[s'(n)]+ψc[s'(n-1)]/4T2 (17)
Considering equation (17) as a single signal system, the transfer function can be expressed as
H(z)＝z·(1+2z-1+z-2)/4 (18)
in equation (18), when the input signal is ψc[s'(n)]then, the energy operator function psi [ s' (n) after integral calculation can be output]At this time, the amplitude envelope | a (n) | of the noise-reduced signal s' (n) is calculated according to equation (19).
Wherein, | a (n) | represents an amplitude envelope of the filtered signal, s ' (n) represents an nth signal reconstructed after EMD denoising based on a CMSE (continuous Mean Square Error) criterion, s ' (n +1) represents an n +1 th signal reconstructed after EMD denoising based on the CMSE criterion, s ' (n-1) represents an n-1 th signal reconstructed after EMD denoising based on the CMSE criterion, ψ [ s ' (n) ] represents an energy operator function of s ' (n), and ψ [ s ' (n +1) -s ' (n-1) ] represents an energy operator function of [ s ' (n +1) -s ' (n-1) ];
after the amplitude envelope is obtained, the embodiment uses | a (n) | as a detection function to detect the arrival time of each echo to the detection surface, and further obtains the specific time of the trough in the envelope curve by using the local minimum value and records the specific time by using a sign function.
sign(|a(n+1)|-|a(n)|-sign(|a(n)|-|a(n-1)|)＝2 (20)
In the method, the starting time of the envelope result just corresponds to the arrival time of the signal, in addition, for the detection of the impact echo of the multilayer medium, after the corresponding thickness information of the surface wave is eliminated, the wave can be clearly shown by transmitting the corresponding head wave signal on the first layer medium, and the various types of echoes can be effectively distinguished by combining the arrival time corresponding to the energy operator.
further, in this embodiment, a combination of a magnesium-carbon refractory brick and a clay refractory brick is used for performing an experiment, the thicknesses in the testing directions are 114mm and 115mm, respectively, the wave velocities in the testing directions are 4318.3m/s and 5997m/s, respectively, an electronic instrument of a Hunan core instrument is used as a monitoring device, the frequency of a shock wave is set to 500KHz, the gain is 50db, and a signal of the previous 130 μ s is intercepted, as shown in FIG. 6, in the structure of the magnesium-carbon-clay refractory brick, through calculation of an energy operator, amplitude abrupt changes occur at 52, 92, 106, and 130 μ s, and the corresponding time point when an echo reaches a receiving point is the corresponding one, and according to prior knowledge, the following judgments can be made: the wave arriving at 52 mus is only reflected in the magnesia carbon brick; the wave arriving at 92 mus passes through the clay brick and the magnesia carbon brick and is reflected once respectively; the wave arriving at 106 mus is reflected twice in the clay brick and once in the magnesia carbon brick; the wave arriving at 130 mus is reflected 2 times in the clay brick and 2 times in the magnesia carbon brick.
According to the propagation rule, the echo time of 52 mu s arrival corresponds to the propagation time of the wave passing through the first layer of medium, so that the thickness of the clay brick is calculated by the formula (1) to be 112.2mm, and the error of the calculation result and the actual numerical value is 1.4%. The echo time of 92 mus arrival corresponds to the time for the wave to pass through the first and second layers of media, and the total thickness calculated by equation (1) is 232.1mm, so the thickness of the coal brick is 119.9 mm. The error between the calculated result and the actual value is 5%.
According to the method, the time and the frequency of the primary non-reflection echo generated by different interfaces reaching the detection surface can be effectively distinguished by denoising the collected impact echo signal and further classifying the filtering signal by adopting the energy operator, so that various types of echoes are distinguished, the corresponding characteristics of the interfaces are accurately extracted, the thickness information of the innermost layer of the blast furnace lining is finally and accurately extracted, the aliasing effect generated by the propagation of elastic waves in a non-uniform layered medium is eliminated, the detection error generated by the aliasing effect is overcome, the accuracy of the extracted thickness information of the innermost layer of the blast furnace lining is high, the erosion state of the lining is further accurately reflected through the thickness information, and the method has important significance for the safe production of the blast furnace.
Referring to fig. 7, a signal processing system for detecting blast furnace lining impact echo according to an embodiment of the present invention includes:
The device comprises a memory 10, a processor 20 and a computer program stored on the memory 10 and capable of running on the processor 20, wherein the processor 20 implements the steps of the signal processing method for detecting blast furnace lining impact echo provided by the embodiment when executing the computer program.
The specific working process and working principle of the signal processing system for detecting blast furnace lining impact echo in this embodiment can refer to the working process and working principle of the signal processing method for detecting blast furnace lining impact echo in this embodiment.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (8)
1. A signal processing method for detecting blast furnace lining impact echo is characterized by comprising the following steps:
carrying out wave velocity calibration on the material of the measured object by using a time difference ultrasonic method to obtain calibrated wave velocity information;
Performing empirical mode decomposition on the impact echo signal to obtain an intrinsic mode component;
Based on the intrinsic mode component, filtering the impact echo signal to obtain a filtered signal;
And classifying the filtering signals based on the energy operator of the filtering signals, and obtaining the thickness information of the innermost layer of the blast furnace lining based on the classification result and the wave velocity information.
2. the signal processing method for detecting the blast furnace lining shock echo according to claim 1, wherein the empirical mode decomposition of the shock echo signal to obtain the intrinsic mode component comprises:
obtaining all maximum values and minimum value points of the impact echo signal, so as to obtain an upper envelope line corresponding to the maximum values and a lower envelope line corresponding to the minimum value points;
Performing empirical mode decomposition on the impulse echo signal based on the upper envelope line, the lower envelope line and a preset threshold condition to obtain an intrinsic mode component, wherein the preset threshold condition specifically comprises:
Wherein s isi(k-1)(n) represents si(n) the k-1 th decomposition quantity, s, not satisfying the threshold conditionik(n) represents si(n) difference of the k-1 th decomposition amount with the average value m (n), andWherein s ish(n) is the upper envelope, slAnd (n) is a lower envelope line, alpha is a screening threshold, and alpha is 0.2, and when the result meets the threshold condition, the decomposition is completed.
3. The signal processing method for detecting the blast furnace lining shock echo according to claim 2, wherein the step of filtering the shock echo based on the intrinsic mode component to obtain a filtered signal comprises:
Acquiring an intrinsic mode component threshold value of sudden change of noise energy distribution based on the intrinsic mode component;
And carrying out wavelet soft threshold denoising on the intrinsic mode component starting from the intrinsic mode component threshold value, thereby obtaining a filtering signal.
4. the signal processing method for detecting the blast furnace lining impact echo according to claim 3, wherein the calculation formula for obtaining the threshold value of the intrinsic modal component of the sudden change of the noise energy distribution based on the intrinsic modal component is specifically as follows:
When a local minimum precedes the global minimum:
otherwise:
where N denotes the data length of a single imf component,To representAndThe continuous mean square error of two adjacent imf components,the value of k corresponding to the minimum value of the local function is shown,and representing the value of k corresponding to the minimum value of the global function.
5. The signal processing method for detecting the blast furnace lining impact echo according to any one of claims 1 to 4, wherein classifying the filtered signal based on an energy operator of the filtered signal, and obtaining the thickness information of the innermost layer of the blast furnace lining based on the classification result and the wave velocity information comprises:
Obtaining the arrival time information of the impact echo to the detection surface based on the energy operator of the filtering signal;
and obtaining the thickness information of the innermost layer of the blast furnace lining based on the arrival time information and the wave velocity information.
6. the signal processing method for detecting the blast furnace lining shock echo according to claim 5, wherein the obtaining the arrival time information of the shock echo to the detection surface based on the energy operator of the filtered signal comprises:
Obtaining the amplitude envelope of the filtering signal according to the filtering signal, wherein a specific calculation formula of the amplitude envelope of the filtering signal is as follows:
wherein, | a (n) | represents an amplitude envelope of the filtered signal, s ' (n) represents an nth signal reconstructed after being denoised by EMD based on the CMSE criterion, s ' (n +1) represents an nth +1 signal reconstructed after being denoised by EMD based on the CMSE criterion, s ' (n-1) represents an nth-1 signal reconstructed after being denoised by EMD based on the CMSE criterion, ψ [ s ' (n) ] represents an energy operator function of s ' (n), and ψ [ s ' (n +1) -s ' (n-1) ] represents an energy operator function of [ s ' (n +1) -s ' (n-1) ];
And obtaining the arrival time information of the shock echo to the detection surface based on the local minimum value of the amplitude envelope of the filtering signal.
7. The signal processing method for detecting the blast furnace lining impact echo according to claim 6, wherein the step of calibrating the wave velocity of the material to be measured by using the time difference ultrasonic method comprises the following steps:
smearing the medical ultrasonic coupling agent on the transmitting probe and the receiving probe, and directly contacting the surface of the transmitting probe with the surface of the receiving probe;
Calculating the propagation time occupied by the transmitting crystal in the transmitting probe to the surface of an emitting source and the propagation time occupied by the receiving crystal in the receiving probe to the surface of a receiving source to obtain first propagation time;
carrying out wave velocity calibration on a material of a measured object to obtain the propagation time of a wave in the measured object, wherein the propagation time is the sum of the propagation time of the wave in the material and the first propagation time;
and obtaining calibration wave velocity information according to the propagation time and the thickness of the measured object.
8. a signal processing system for blast furnace lining shock echo detection, the system comprising:
memory (10), processor (20) and computer program stored on the memory (10) and executable on the processor (20), characterized in that the steps of the method according to any of the preceding claims 1 to 7 are implemented when the computer program is executed by the processor (20).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910950250.4A CN110564905B (en) | 2019-10-08 | 2019-10-08 | Signal processing method and system for blast furnace lining impact echo detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910950250.4A CN110564905B (en) | 2019-10-08 | 2019-10-08 | Signal processing method and system for blast furnace lining impact echo detection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110564905A true CN110564905A (en) | 2019-12-13 |
| CN110564905B CN110564905B (en) | 2020-10-09 |
Family
ID=68784167
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910950250.4A Active CN110564905B (en) | 2019-10-08 | 2019-10-08 | Signal processing method and system for blast furnace lining impact echo detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110564905B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112226561A (en) * | 2020-10-10 | 2021-01-15 | 鞍钢股份有限公司 | Blast furnace lining monitoring method based on impact echo method |
| CN112326408A (en) * | 2020-10-09 | 2021-02-05 | 南京理工大学 | System and method for measuring wave velocity of solid medium under confining pressure state |
| CN115323092A (en) * | 2022-08-04 | 2022-11-11 | 华北理工大学 | Blast furnace lining ablation monitoring device and monitoring method thereof |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1087428A (en) * | 1992-11-25 | 1994-06-01 | 鞍山钢铁公司 | The method and the device of the residual lining Ultrasonic Detection of blast furnace |
| EP0841758A1 (en) * | 1996-11-06 | 1998-05-13 | Telediffusion De France Sa | Method of echo cancelling for a radio signal with constant envelope and associated devices |
| CN1825056A (en) * | 2005-02-22 | 2006-08-30 | 哈茨有限公司 | Systems, methods and apparatus for non-disruptive and non-destructive inspection of metallurgical furnaces |
| CN101822548A (en) * | 2010-03-19 | 2010-09-08 | 哈尔滨工业大学（威海） | Ultrasound signal de-noising method based on correlation analysis and empirical mode decomposition |
| CN103615996A (en) * | 2013-11-14 | 2014-03-05 | 大连理工大学 | Method for measuring thickness of coatings through ultrasonic signal spectrum filter technology in nondestructive mode |
| CN106556647A (en) * | 2016-11-18 | 2017-04-05 | 湘潭大学 | A kind of impact echo data processing method |
| CN107144829A (en) * | 2017-06-29 | 2017-09-08 | 南京信息工程大学 | A kind of efficient laser radar echo signal antinoise method |
| CN110108240A (en) * | 2019-04-23 | 2019-08-09 | 北京理工大学 | A kind of thin layer thickness measurement method based on adaptive-filtering |
-
2019
- 2019-10-08 CN CN201910950250.4A patent/CN110564905B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1087428A (en) * | 1992-11-25 | 1994-06-01 | 鞍山钢铁公司 | The method and the device of the residual lining Ultrasonic Detection of blast furnace |
| EP0841758A1 (en) * | 1996-11-06 | 1998-05-13 | Telediffusion De France Sa | Method of echo cancelling for a radio signal with constant envelope and associated devices |
| CN1825056A (en) * | 2005-02-22 | 2006-08-30 | 哈茨有限公司 | Systems, methods and apparatus for non-disruptive and non-destructive inspection of metallurgical furnaces |
| CN101822548A (en) * | 2010-03-19 | 2010-09-08 | 哈尔滨工业大学（威海） | Ultrasound signal de-noising method based on correlation analysis and empirical mode decomposition |
| CN103615996A (en) * | 2013-11-14 | 2014-03-05 | 大连理工大学 | Method for measuring thickness of coatings through ultrasonic signal spectrum filter technology in nondestructive mode |
| CN106556647A (en) * | 2016-11-18 | 2017-04-05 | 湘潭大学 | A kind of impact echo data processing method |
| CN107144829A (en) * | 2017-06-29 | 2017-09-08 | 南京信息工程大学 | A kind of efficient laser radar echo signal antinoise method |
| CN110108240A (en) * | 2019-04-23 | 2019-08-09 | 北京理工大学 | A kind of thin layer thickness measurement method based on adaptive-filtering |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112326408A (en) * | 2020-10-09 | 2021-02-05 | 南京理工大学 | System and method for measuring wave velocity of solid medium under confining pressure state |
| CN112226561A (en) * | 2020-10-10 | 2021-01-15 | 鞍钢股份有限公司 | Blast furnace lining monitoring method based on impact echo method |
| CN115323092A (en) * | 2022-08-04 | 2022-11-11 | 华北理工大学 | Blast furnace lining ablation monitoring device and monitoring method thereof |
| CN115323092B (en) * | 2022-08-04 | 2023-09-12 | 华北理工大学 | Blast furnace lining ablation monitoring device and monitoring method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110564905B (en) | 2020-10-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110564905B (en) | Signal processing method and system for blast furnace lining impact echo detection | |
| KR102417558B1 (en) | Detection and Monitoring of Changes in Metal Structures Using Multimode Acoustic Signals | |
| Demma et al. | The reflection of guided waves from notches in pipes: a guide for interpreting corrosion measurements | |
| Verma et al. | Interaction of low-frequency axisymmetric ultrasonic guided waves with bends in pipes of arbitrary bend angle and general bend radius | |
| JP2003207463A (en) | Nondestructive inspection method for concrete structure and structure other than the same | |
| US11092573B2 (en) | Apparatus, systems, and methods for determining nonlinear properties of a material to detect early fatigue or damage | |
| CN101509604A (en) | Method and device for detecting and assessing deposit in metal pipe | |
| US10996203B2 (en) | Detection, monitoring, and determination of location of changes in metallic structures using multimode acoustic signals | |
| CN105758499A (en) | Non-contact type liquid level detection system and method based on ultrasonic pulse echo method | |
| CN112154324B (en) | Using multimode acoustic signals to detect, monitor and determine the location of changes in metal structures | |
| WO2007132509A1 (en) | Ultrasonic inspection method utilizing resonant phenomena | |
| Kaphle et al. | Acoustic emission technique-Opportunities, challenges and current work at QUT | |
| CN113983976B (en) | Ultrasonic pipeline thickness measuring method based on FPGA | |
| Elwalwal et al. | Crack inspection using guided waves (GWs)/structural health monitoring (SHM) | |
| Saifullin et al. | Methods of Leak Search from Pipeline for Acoustic Signal Analysis | |
| CN212179841U (en) | Instrument device for detecting thickness of carbon deposition layer of smoke exhaust pipe | |
| CN111256630A (en) | Method for rapidly measuring thickness of metal plate by utilizing electromagnetic ultrasonic guided wave frequency dispersion characteristic | |
| Gao et al. | A novel NaCl concentration detection method based on ultrasonic impedance method | |
| Gunarathna et al. | Challenges in Monitoring Metallic Pipeline Corrosion Using Ultrasonic Waves—A Review Article | |
| Takimoto et al. | An echo analysis method for the ultrasonic measurement of micrometric wall-thickness loss inside pipes | |
| Butkus et al. | An acoustic method of determination of leakage coordinates in gas pipelines | |
| CN117268299B (en) | Method and system for detecting wall thickness of desulfurization pipeline based on electromagnetic ultrasound | |
| Ziganshin et al. | Technique for search of pipeline leakage according to acoustic signals analysis | |
| JP2001337077A (en) | Device for non-destructively inspecting exfoliation in concrete structure | |
| Bayoumi et al. | New approach for reliability assessment of guided wave-based structure health monitoring system on a pipe application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |