Method for designing TBM tunneling parameter noise reduction filter based on Fourier transform

Abstract

The invention discloses a TBM tunneling parameter noise reduction filter design method based on Fourier transform, and the method comprises the steps: analyzing the frequency domain distribution characteristics of a cutterhead torque idling stage and a complete tunneling stage based on a fast Fourier transform method; a noise reduction transition section parameter selection method which takes an idle frequency as a noise reduction cut-off frequency and retains main information of a tunneling section frequency domain is provided; thirdly, in order to eliminate tunneling parameter high-frequency vibration noise, different low-pass filter parameter design and comparison and selection methods are provided; and finally, on the basis of the optimal filter, performing engineering application on noise reduction of field monitoring data and extraction of rock mass dig-ability indexes in the tunnel construction process. According to the method, a frequency domain guidance basis is provided for TBM tunneling data noise reduction target parameter selection, a standardized filter parameter design and comparison and selection process is established, meanwhile, it is guaranteed that vibration noise elimination of tunneling data is achieved through an optimal filter, and TBM monitoring data processing standardization and field monitoring data quality are improved.

Description

Technical Field

[0001] The invention relates to the field of tunnel boring machine construction monitoring data informatization, and in particular to a TBM excavation parameter noise reduction filter design method based on Fourier transform. Background Art

[0002] The improvement of the informatization level of the tunnel construction process makes the real-time monitoring of on-site engineering data more convenient and faster. The excavation process of the tunnel boring machine is a dynamic impact process in which the roller cutter installed on its cutter head penetrates the rock mass of the excavation face and cuts the rock mass under the thrust provided by the hydraulic system. In this process, due to the influence of the surrounding rock resistance, there is a large amount of impact vibration noise information in the original data of the machine rock breaking obtained by on-site monitoring. The existence of high-frequency noise data in the monitoring data will, on the one hand, conceal the real data change law of the roller cutter rock breaking process, and on the other hand, it will also cause large errors in the calculation of the statistical characteristics of the data. These errors will seriously affect the accuracy and credibility of the excavability index reflecting the difficulty of on-site rock excavation, thereby affecting the accuracy of rock quality assessment and reducing construction efficiency.

[0003] The traditional method of removing vibration noise from monitoring data of tunnel construction sites mainly relies on some empirical parameters to smooth the excavation parameters. The setting of these parameters does not take into account the inherent vibration characteristics of the tunnel boring machine itself, and lacks the physical process guidance of the machine's on-site idling test. On the other hand, the vibration noise levels of different excavation parameters in the TBM construction process are not exactly the same due to the influence of rock breaking process excavation. The workload of analyzing the characteristics of different monitoring data separately is large and the efficiency is low. Therefore, how to select reasonable noise reduction parameters according to the frequency domain characteristics and inherent rotation properties of different monitoring data of the tunnel boring machine site, and realize batch automatic noise reduction processing of different monitoring data can effectively improve the processing efficiency of on-site monitoring data.

[0004] Eliminating high-frequency noise of TBM field excavation parameters and accurately calculating rock digging indexes depends on the setting of reasonable noise reduction parameters. Unreasonable noise reduction results will result in high-frequency vibration signals not completely suppressing or eliminating low-frequency trend information of the data. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a TBM excavation parameter noise reduction filter design method based on Fourier transform in view of the deficiencies of the above-mentioned prior art. The TBM excavation parameter noise reduction filter design method based on Fourier transform is based on the excavation data under two working conditions: normal excavation on site and idling test on site. The frequency domain characteristics of different parameters are analyzed by Fourier transform method to guide the selection of noise reduction parameters. By comparing the frequency domain characteristics of cutter head torque, hydraulic thrust, and penetration parameters, a standardized noise reduction process based on torque is established, which can greatly reduce the noise reduction parameter analysis process. In addition, the parameters after noise reduction can effectively eliminate the vibration phenomenon of the original data, improve the calculation accuracy of the on-site excavability index, and provide more accurate results for the evaluation of the difficulty of on-site rock excavation.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A method for designing a TBM excavation parameter denoising filter based on Fourier transform includes the following steps.

[0008] Step 1: TBM idling test: The TBM performs a cutterhead idling test at the excavation site and monitors the cutterhead torque data of the TBM over time.

[0009] Step 2, obtain the natural frequency of the TBM cutter head idling rotation: perform Fourier frequency domain transformation on the TBM cutter head torque data obtained in step 1 that varies with time, and form a curve of the idling cutter head torque amplitude spectrum varying with frequency; the frequency corresponding to the maximum peak value in the curve of the idling cutter head torque amplitude spectrum varying with frequency is taken as the natural frequency of the TBM idling rotation.

[0010] Step 3: TBM excavation test: The TBM excavates normally and monitors the excavation operation data; the excavation operation data includes the cutter head torque and penetration.

[0011] Step 4, TBM excavation operation data transformation: Each excavation operation parameter is subjected to Fourier frequency domain transformation; for example, the cutter head torque in the excavation operation data is transformed into a curve of the excavation cutter head torque amplitude spectrum versus frequency.

[0012] Step 5: Obtain the noise reduction target parameters: The noise reduction target parameters include the cutoff frequency f c , passband frequency f pass and stopband frequency f stop ; Take the TBM rotation natural frequency obtained in step 2 as the cutoff frequency f c ; In the curve of torque amplitude spectrum of the cutting head in step 4, the frequency variation curve is greater than the cutoff frequency f c The starting point of the first peak segment is taken as the stop band frequency f stop ; then the passband frequency f pass The calculation formula is:

[0013] f pass =2·f c -f stop

[0014] Step 6: Construct filters: Based on the noise reduction target parameters obtained in step 5, establish N different types of filters; where N≥2.

[0015] Step 7, filter selection: For the N types of filters constructed in step 6, draw the response characteristic curves in the frequency domain respectively, and select the optimal filter type by comparing the frequency response characteristics of different filters.

[0016] In step 1, during the cutter head idling test, the cutter head speed value needs to be set to the minimum value of its rated speed range.

[0017] In step 3, the penetration data obtained through monitoring must cover the entire working range of the rated penetration of the TBM machine.

[0018] In step 3, the excavation operation data also includes the cutterhead thrust.

[0019] In step 4, the penetration in the tunneling operation data is transformed into a penetration amplitude spectrum versus frequency curve after Fourier transformation; the hydraulic thrust in the tunneling operation data is formed into a cutterhead thrust amplitude spectrum versus frequency curve; when the tunneling cutterhead torque amplitude spectrum versus frequency curve, the penetration amplitude spectrum versus frequency curve and the hydraulic thrust amplitude spectrum versus frequency curve have the same distribution, enter step 5 and use the frequency domain analysis noise reduction parameter f of the cutterhead torque c 、f pass and f stop The same noise reduction parameters are designed as hydraulic thrust and penetration filters; otherwise, go to step 4, perform Fourier transform on the hydraulic thrust and penetration TBM excavation section data to obtain the respective amplitude spectrum versus frequency curves, and then refer to step 5 to obtain the noise reduction parameters f for hydraulic thrust and penetration, respectively. c 、f pass and f stop .

[0020] In step 6, N=4, and the four filters are respectively FIR filter, window function FIR filter, Butterworth IIR filter and Chebyshev II filter.

[0021] In step 7, the method for selecting the optimal filter type is: a filter with a frequency response characteristic curve without oscillation and a transition band that matches the noise reduction target parameter.

[0022] The method also includes step 8, denoising the excavation parameters: using the optimal filter selected in step 8 to perform denoising on the excavation operation data of the on-site excavation monitoring during the TBM construction process.

[0023] It also includes step 9, calculating the excavation index: the excavation operation data monitored during on-site excavation during the TBM construction process includes the average thrust of a single blade, the average torque value of a single blade and the penetration rate.

[0024] The excavability indexes include the linear fitting slope a of the average thrust and penetration of a single knife, the thrust intercept b, and the linear fitting slope TPI of the torque and penetration of a single knife;

[0025] The average thrust of a single knife after noise reduction and the penetration after noise reduction are linearly fitted to obtain the thrust-penetration linear fitting slope a and thrust intercept b after noise reduction; the average torque value of a single knife after noise reduction and the penetration after noise reduction are fitted to obtain the torque-penetration linear fitting slope TPI after noise reduction; the three indicators of the thrust-penetration linear fitting slope a, thrust intercept b and torque-penetration linear fitting slope TPI after noise reduction are used to reflect the difficulty of rock excavation during construction.

[0026] The single-blade average thrust and single-blade average torque values ​​are collected using corresponding sensors, and the penetration is collected using a displacement sensor.

[0027] The present invention has the following beneficial effects:

[0028] (1) The present invention provides a physical basis for the design of digital filter noise reduction parameters through a frequency domain feature analysis method based on Fourier transform. The noise reduction cutoff frequency is designed based on the peak frequency of the machine's own rotation inherent frequency domain energy spectrum obtained from the TBM on-site cutterhead idling test, and the passband and stopband parameters are designed based on the complete excavation section of the TBM effective construction, which can ensure that the effective information in the monitoring data is retained while removing high-frequency parameters. At the same time, the parameter selection is more reasonable, solving the problem that there is no basis for setting the cutoff frequency of the filter noise reduction work at different engineering sites.

[0029] (2) The present invention establishes a parameter design and frequency domain response comparison process for different digital filters, and obtains the filter that can achieve the best effect through comparison and selection to achieve the work of removing the vibration noise of the excavation parameters. The established standardized filter design and noise reduction processing method is suitable for the noise reduction processing of excavation data monitored on-site in different tunnel projects, which can effectively eliminate the invalid noise information in the recorded data and improve the quality of the monitoring data.

[0030] (3) The method for denoising the excavation parameters during TBM construction established by the present invention is suitable for real-time data processing of on-site monitoring data. The data processed based on the low-pass digital filter can better reflect the rock-machine interaction process, and is verified in the application comparison of on-site excavation indicators before and after denoising. This method can significantly improve the goodness of fit of on-site excavation indicators, thereby better reflecting the difficulty of rock excavation and realizing the evaluation and judgment of rock quality during construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a principle flow chart of the TBM excavation parameter noise reduction filter design method based on Fourier transform of the present invention.

[0032] Figure 2 It is the frequency domain characteristic curve of the low-pass digital filter.

[0033] Figure 3 The process of selecting transition section parameters for denoising TBM excavation data.

[0034] Figure 4 Schematic diagram of the frequency domain amplitude spectrum of penetration, thrust and torque in the excavation section of Yinsong Tunnel; among them, (a) is the frequency domain amplitude spectrum of penetration; (b) is the frequency domain amplitude spectrum of cutterhead thrust; (c) is the frequency domain amplitude spectrum of cutterhead torque.

[0035] Figure 5 Schematic diagram of frequency response characteristic curve results for different filter types.

[0036] Figure 6 These are example results before and after noise reduction in the excavation section near pile number 71138 of the Yinsong Project.

[0037] Figure 7 This is the time domain diagram of the average thrust, average torque value and penetration of a single blade before noise reduction during TBM on-site construction of the Yinsong Project.

[0038] Figure 8 This is the fitting diagram of the average thrust and average torque value of a single blade and the penetration rate before noise reduction during TBM on-site construction of the Yinsong Project.

[0039] Fig. 9 This is the time domain diagram of the average thrust, average torque value and penetration of a single blade after noise reduction during TBM on-site construction of the Yinsong Project.

[0040] Fig.10 This is the fitting diagram of the average thrust and average torque value of a single tool and the penetration rate after noise reduction during TBM on-site construction of the Yinsong Project. DETAILED DESCRIPTION

[0041] The specific implementation method of the present invention is described in detail with a representative example of monitoring data of the effective rock breaking tunneling section of the TBM on-site normal excavation and the on-site cutterhead idling test within the fourth section of the Yinsong Project in Jilin Province.

[0042] like Figure 1 and Figure 3 As shown, the design method of TBM excavation parameter denoising filter based on Fourier transform includes the following steps.

[0043] Step 1: TBM idling test: The TBM performs a cutterhead idling test at the excavation site and monitors the cutterhead torque data over time. During the cutterhead idling test, the cutterhead speed value needs to be set to the minimum value of its rated speed range.

[0044] In this embodiment, the TBM field cutterhead idling test experiment collected a total of 100 cutterhead idling test monitoring data, which included the cutterhead torque data at the minimum speed value within the rated working speed range of 3.97 rpm to 7.6 rpm of the tunnel boring machine. Therefore, it meets the data requirements for noise reduction parameter design.

[0045] The above idling test adopts the minimum speed value within the working range of the rated working speed of 3.97r / min to 7.6r / min.

[0046] Step 2, obtain the natural frequency of TBM rotation: perform Fourier frequency domain transformation on the time-varying cutter head torque data of TBM obtained in step 1 to form a curve of idling cutter head torque amplitude spectrum versus frequency; take the frequency corresponding to the maximum peak value in the curve of idling cutter head torque amplitude spectrum versus frequency as the natural frequency of TBM rotation f, and the specific calculation formula is:

[0047] f = argmax(Amplitude)

[0048] Where Amplitude is the amplitude of the idling torque in the frequency domain spectrum, and the frequency corresponding to the maximum amplitude is the idling noise frequency of the TBM machine itself. The frequency of the cutter head torque monitored on site greater than this value can be considered as high-frequency noise information.

[0049] The cutterhead idling test can measure the natural frequency of the machine's rotation when it is not cutting rock. This frequency represents the rotation noise frequency of the machine itself. The cutterhead speed value during the TBM on-site cutterhead idling test needs to be set to the minimum value of the rated speed range. This is because the minimum speed corresponds to the minimum frequency of the machine's idling noise, which can provide a noise frequency reference for the setting of the noise reduction frequency.

[0050] Taking the data collected from the Yinsong Water Conveyance Tunnel Project as an example, some examples of the data collected on site during the TBM cutterhead torque idling test phase and the frequency domain amplitude spectrum distribution obtained by Fourier transform are shown in the figure below. Figure 3 Shown on the left.

[0051] exist Figure 3In the left diagram, the frequency corresponding to the maximum amplitude in the frequency domain amplitude spectrum is at 0.08; the second peak corresponds to a position higher than the idling frequency, and the corresponding frequency value is 0.15. By analyzing the distribution characteristics of the frequency domain energy spectrum density, since the idling of the tunnel boring machine itself can be regarded as a noise signal, it can be considered that the interval exceeding the idling frequency of the tunnel boring machine is also a noise signal. Therefore, as the natural frequency value of TBM rotation f c =0.08.

[0052] Step 3, TBM excavation test: The TBM excavates normally and monitors the excavation operation data; wherein the excavation operation data preferably includes cutter head thrust, cutter head torque and penetration, etc., and the excavation parameter sampling frequency is preferably 1 Hz / s.

[0053] The above penetration data must cover the entire working range of the rated penetration of the TBM machine. The reason is that the penetration parameter is the depth of the cutter penetrating into the tunnel excavation face. Different penetrations represent the intensity of the interaction between the machine cutter and the rock mass. Only when the collected excavation data includes the rated working range of the machine penetration can it be guaranteed that the collected data covers the vibration characteristics under all working conditions of the current machine.

[0054] In this embodiment, field monitoring data of the effective rock breaking stage of normal TBM excavation were collected from Section 4 of the Yinsong Water Diversion Tunnel in Jilin Province, totaling 13,200 rock breaking excavation sections. Each excavation section contained data on cutter disc torque, total cutter disc thrust and penetration. These effective on-site excavation data included data within the range of 30 mm / rev, the maximum penetration of the machine, and the data covered the high-frequency and low-frequency vibration characteristics of the current machine under rated working conditions.

[0055] Step 4, excavation operation data transformation: each excavation operation data is subjected to Fourier frequency domain transformation; among them, the cutter head torque in the excavation operation data is transformed into a curve of the excavation cutter head torque amplitude spectrum changing with frequency; the penetration in the excavation operation data is transformed into a curve of the penetration amplitude spectrum changing with frequency; the hydraulic thrust in the excavation operation data is formed into a curve of the cutter head thrust amplitude spectrum changing with frequency.

[0056] When the distribution of the cutter head torque amplitude spectrum with frequency, the penetration amplitude spectrum with frequency, and the hydraulic thrust amplitude spectrum with frequency is consistent, go to step 5 and use the frequency domain analysis noise reduction parameter f of the cutter head torque to reduce the noise. c 、f pass and f stop The same noise reduction parameters are designed as hydraulic thrust and penetration filters; otherwise, go to step 4, perform Fourier transform on the hydraulic thrust and penetration TBM excavation section data to obtain the respective amplitude spectrum versus frequency curves, and then refer to step 5 to obtain the noise reduction parameters f for hydraulic thrust and penetration, respectively. c 、fpass and f stop .

[0057] In this embodiment, some examples of the cutter head torque data collected on site during normal excavation of a TBM and the frequency domain amplitude spectrum distribution obtained by Fourier transform are shown in the following figure: Figure 3 Shown on the right.

[0058] The frequency domain amplitude spectra of penetration, cutterhead thrust and cutterhead torque collected in the normal effective excavation section of the Yinsong water diversion tunnel are as follows: Figure 4 As shown in the figure, it can be seen that the frequency domain spectrum distribution between different tunneling construction parameters is basically consistent with the frequency domain amplitude spectrum distribution of the cutter head torque, and the cutter head penetration and thrust also show peaks at the local peak of the torque high-frequency amplitude. The idling inherent noise frequency obtained based on the torque also affects the penetration and thrust data, so the cutter head torque noise reduction parameters are also applicable to the thrust and penetration parameters.

[0059] Step 5: Get the noise reduction target parameters

[0060] like Figure 2 As shown, the noise reduction target parameters include the cutoff frequency f c , passband frequency f pass and stopband frequency f stop In addition, δ p and δ s They are the passband and stopband ripple parameters respectively. The two parameters represent the fluctuation amplitude of the filter frequency response in the passband and stopband respectively. When the value is smaller, the filter response is better.

[0061] The TBM rotational natural frequency obtained in step 2 is used as the cutoff frequency f c In this embodiment, f c =0.08, that is, the center value of the noise reduction transition section parameter is set to 0.08.

[0062] In the curve of torque amplitude spectrum of the cutting head in step 4, the frequency is greater than the cutoff frequency f c The starting point of the first peak segment is taken as the stop band frequency f stop According to Figure 3 The frequency domain energy spectrum distribution of the cutter head torque in the normal rock breaking stage on the right shows that the starting value of the first peak section greater than the idling cutoff frequency domain of 0.08 is 0.10, so f stop Set it to 0.10. The advantage of this setting is that it can remove large high-frequency data in the peak range.

[0063] The above passband frequency f pass The calculation formula is:

[0064] f pass =2·fc -f stop

[0065] In this embodiment:

[0066] f pass =2·f c -f stop =0.06

[0067] Therefore, the passband frequency f of the low-pass filter transition section parameter is pass =0.06.

[0068] In summary, based on the above analysis process, the noise reduction target transition zone parameter of the Songshui water diversion tunnel engineering data can be obtained as f pass =0.06Hz,f c =0.08Hz, f stop =0.10Hz.

[0069] Step 6, constructing filters: Based on the noise reduction target parameters obtained in step 5, establish N different types of filters; wherein N ≥ 2. In this embodiment, preferably N = 4, and the four filters are FIR filter (Firls), window function FIR filter (Firwin), Butterworth IIR filter (Butterworth) and Chebyshev II filter (Cheby2).

[0070] Firls filter parameter design frequency parameters and target parameters: based on the field monitoring data acquisition frequency f s =1Hz / sCalculate the maximum frequency parameter f of the filter s / 2=0.5Hz. Divide the maximum frequency value into n parts and calculate the frequency step value Then we get the initial frequency parameter sequence [0,Δx,2Δx,...,nΔx], and then we sum the sequence and f pass , f c , f stop , insert the initial sequence and arrange them according to the numerical size. For example,

[0071] Based on the maximum frequency parameter of the TBM on site of Yinsong Tunnel, the initial frequency parameter sequence [0, 0.1, 0.2, 0.3, 0.4, 0.5] is obtained, and then f pass =0.06Hz,f stop =0.10Hz, f c=0.08Hz is inserted into the sequence and sorted from small to large to obtain the frequency parameter sequence [0, 0.06, 0.08, 0.1, 0.2, 0.3, 0.4, 0.5]. The target parameter is set to set the amplitude attenuation coefficient of the frequency parameter less than or equal to the passband frequency to 1, and other values ​​to zero. Therefore, the target parameter with a frequency less than or equal to 0.06 is set to 1, and the target parameter sequence is [1, 1, 0, 0, 0, 0, 0].

[0072] For the noise reduction parameter design of Firwin low-pass filter, Butterworth filter low-pass filter and Chebyshev II filter, the noise reduction design parameter f pass =0.06Hz,f stop =0.10Hz, f c =0.08Hz Existing mature digital filter design libraries such as MATLAB and SciPy can realize the design of low-pass filters for noise reduction.

[0073] Step 7, filter selection: For the N types of filters constructed in step 6, draw the response characteristic curves in the frequency domain respectively, and select the optimal filter type by comparing the frequency response characteristics of different filters.

[0074] The above-mentioned method for selecting the optimal filter type is preferably: a filter with a frequency response characteristic curve without oscillation and a transition band that matches the noise reduction target parameters.

[0075] The frequency domain response curves of different types of low-pass filters are drawn, and the frequency response characteristics of four types of low-pass filters are obtained as follows: Figure 5 As shown. Figure 5 In the figure, the ordinate is the normalized amplitude, the maximum value of the normalized amplitude is 1, and the minimum value is 0, which correspond to the maximum and minimum values ​​of the noise reduction parameter data. The frequency response characteristic curve indicates the attenuation degree of the parameter value at different frequencies of the noise reduction parameter. For example, the high-frequency noise elimination means reducing the data amplitude at a larger frequency to 0 to achieve the noise reduction effect.

[0076] It can be seen that Firls and Cheby2 filters have some oscillations in the stopband, which will result in the noise in the stopband not being completely removed. Overall, the frequency response curves of Firwin and Butter filters are more reasonable, but the transition band of the Butterworth filter is more consistent with the design target range of 0.06-0.10, so it is the optimal filter.

[0077] Step 8: Denoising of tunneling parameters: Use the optimal filter selected in step 8 to perform noise reduction on the tunneling operation data of on-site excavation monitoring during the TBM construction process.

[0078] The above optimal filter and the design parameters of the Butterworth filter are used to reduce the noise of the effective excavation parameters of the TBM in the Yinsong Project.

[0079] In this embodiment, the penetration of the TBM machine in the effective rock breaking stage near pile number 71138.00 is subjected to high-frequency noise elimination using a filter with optimal frequency response, and the excavation section data before and after the noise reduction processing using the optimal filter are obtained as follows: Figure 6 As shown, it can be seen that the designed low-pass filter can effectively eliminate the high-frequency vibration noise in the original data while retaining the local vibration information in the original data.

[0080] Step 9, calculate the excavation index: The excavation operation data monitored during the on-site excavation of the TBM construction process include the average thrust of a single blade, the average torque value of a single blade, and the penetration rate. Among them, the average thrust of a single blade and the average torque value of a single blade are collected by stress sensors, and the penetration rate is collected by displacement sensors. Then, the measured data is recorded and transmitted to the computer in the TBM main control room through the programmable logic control unit, and then the excavation data is input into the designed low-pass filter to eliminate the high-frequency noise in the original data.

[0081] The excavability indexes include the linear fitting slope a of the average thrust and penetration of a single knife, the thrust intercept b, and the linear fitting slope TPI of the torque and penetration of a single knife;

[0082] The average thrust of a single knife after noise reduction and the penetration after noise reduction are linearly fitted to obtain the thrust-penetration linear fitting slope a and thrust intercept b after noise reduction; the average torque value of a single knife after noise reduction and the penetration after noise reduction are fitted to obtain the torque-penetration linear fitting slope TPI after noise reduction; the three indicators of the thrust-penetration linear fitting slope a, thrust intercept b and torque-penetration linear fitting slope TPI after noise reduction are used to reflect the difficulty of rock excavation during construction.

[0083] The linear fitting slope a of the above-mentioned single-blade average thrust and penetration indicates the thrust increment required to increase the unit penetration, and the intercept b indicates the thrust value required for initial penetration into the rock mass. The linear fitting slope TPI of the single-blade torque and penetration indicates the torque increment required to increase the unit penetration. These three indicators can all reflect the difficulty of the rock mass to be excavated. The larger the value, the more difficult the rock mass is to excavate.

[0084] In this embodiment, the machine excavation parameters such as cutter head thrust, cutter head torque and penetration parameters in the new section of Yinsong Tunnel Project with pile numbers ranging from 71083.27 to 71082.83 during tunnel excavation are monitored and collected using force and displacement sensors. Then, the single-blade average thrust and single-blade average torque values ​​in the TBM construction monitoring excavation parameters after noise reduction are linearly fitted with the penetration parameters, and the original data before and after noise reduction and the noise reduction data are plotted respectively. Figure 7 to Figure 10 As shown in the figure, it can be seen that the denoised data can effectively remove the vibration phenomenon of the original data, and the linear fitting goodness of fit is improved from 0.87 (before denoising) and 0.93 (before denoising) to 0.93 and 0.97 respectively. This shows that denoising can significantly improve the quality of on-site excavability indicators, thereby better reflecting the difficulty of rock excavation during construction.

[0085] The present invention provides a frequency domain guidance basis for the selection of target parameters for TBM excavation data noise reduction, solving the current problems of unclear noise reduction targets and unreasonable noise reduction parameter selection. The related technology provides a standardized filter parameter design and comparison process, while ensuring the use of the optimal filter to eliminate vibration noise in excavation data, providing technical support for further improving the standardization of TBM monitoring data processing and the quality of field monitoring data.

[0086] The preferred embodiments of the present invention are described in detail above; however, the present invention is not limited to the specific details in the above embodiments. Within the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all belong to the protection scope of the present invention.

Claims

1. A method for designing a TBM excavation parameter noise reduction filter based on Fourier transform, characterized in that: The steps include: Step 1: TBM idling test: The TBM cutterhead idling test is carried out at the excavation site, and the cutterhead torque data of the TBM changing over time is monitored; Step 2, obtaining the natural frequency of the TBM cutterhead idling rotation: the TBM cutterhead torque data obtained in step 1 that changes with time is subjected to Fourier frequency domain transformation to form a curve of the idling cutterhead torque amplitude spectrum changing with frequency; the frequency corresponding to the maximum peak value in the curve of the idling cutterhead torque amplitude spectrum changing with frequency is taken as the natural frequency of the TBM idling rotation; Step 3: TBM excavation test: The TBM excavates normally and monitors the excavation operation data; the excavation operation data includes the cutter head torque and penetration; Step 4, TBM excavation operation data transformation: each excavation operation parameter is subjected to Fourier frequency domain transformation; wherein the cutter head torque in the excavation operation data is transformed into a curve of the excavation cutter head torque amplitude spectrum versus frequency; Step 5: Obtain the noise reduction target parameters: The noise reduction target parameters include the cutoff frequency f c , passband frequency f pass and stopband frequency f stop ; Take the TBM rotation natural frequency obtained in step 2 as the cutoff frequency f c ; In the curve of torque amplitude spectrum of the cutting head in step 4, the frequency variation curve is greater than the cutoff frequency f c The starting point of the first peak segment is taken as the stop band frequency f stop ; then the passband frequency f pass The calculation formula is: f pass =2·f c -f stop Step 6, constructing filters: based on the noise reduction target parameters obtained in step 5, establishing N different types of filters; wherein N ≥ 2; Step 7, filter selection: For the N types of filters constructed in step 6, draw the response characteristic curves in the frequency domain respectively, and select the optimal filter type by comparing the frequency response characteristics of different filters.

2. The method for designing a TBM excavation parameter noise reduction filter based on Fourier transform according to claim 1, characterized in that: In step 1, during the cutter head idling test, the cutter head speed value needs to be set to the minimum value of its rated speed range.

3. The method for designing a TBM excavation parameter noise reduction filter based on Fourier transform according to claim 1, characterized in that: In step 3, the penetration data obtained through monitoring must cover the entire working range of the rated penetration of the TBM machine.

4. The method for designing a TBM excavation parameter noise reduction filter based on Fourier transform according to claim 1, characterized in that: In step 3, the excavation operation data also includes the cutterhead thrust.

5. The method for designing a TBM excavation parameter noise reduction filter based on Fourier transform according to claim 4 is characterized in that: In step 4, the penetration in the tunneling operation data is transformed into a penetration amplitude spectrum versus frequency curve after Fourier transformation; the hydraulic thrust in the tunneling operation data is formed into a cutterhead thrust amplitude spectrum versus frequency curve; when the tunneling cutterhead torque amplitude spectrum versus frequency curve, the penetration amplitude spectrum versus frequency curve and the hydraulic thrust amplitude spectrum versus frequency curve have the same distribution, enter step 5 and use the frequency domain analysis noise reduction parameter f of the cutterhead torque c 、f pass and f stop The same noise reduction parameters are designed as hydraulic thrust and penetration filters; otherwise, go to step 4, perform Fourier transform on the hydraulic thrust and penetration TBM excavation section data to obtain the respective amplitude spectrum versus frequency curves, and then refer to step 5 to obtain the noise reduction parameters f for hydraulic thrust and penetration, respectively. c 、f pass and f stop .

6. The method for designing a TBM excavation parameter noise reduction filter based on Fourier transform according to claim 1, characterized in that: In step 6, N=4, and the four filters are respectively FIR filter, window function FIR filter, Butterworth IIR filter and Chebyshev II filter.

7. The method for designing a TBM excavation parameter noise reduction filter based on Fourier transform according to claim 1, characterized in that: In step 7, the method for selecting the optimal filter type is: a filter with a frequency response characteristic curve without oscillation and a transition band that matches the noise reduction target parameter.

8. The method for designing a TBM excavation parameter noise reduction filter based on Fourier transform according to claim 1, characterized in that: The method also includes step 8, denoising the excavation parameters: using the optimal filter selected in step 8 to perform denoising on the excavation operation data of the on-site excavation monitoring during the TBM construction process.

9. The method for designing a TBM excavation parameter noise reduction filter based on Fourier transform according to claim 8, characterized in that: It also includes step 9, calculating the rock digging index: the excavation operation data monitored during the on-site excavation of the TBM construction process includes the average thrust of a single blade, the average torque value of a single blade and the penetration rate; The excavability indexes include the linear fitting slope a of the average thrust and penetration of a single knife, the thrust intercept b, and the linear fitting slope TPI of the torque and penetration of a single knife; The average thrust of a single knife after noise reduction and the penetration after noise reduction are linearly fitted to obtain the thrust-penetration linear fitting slope a and thrust intercept b after noise reduction; the average torque value of a single knife after noise reduction and the penetration after noise reduction are fitted to obtain the torque-penetration linear fitting slope TPI after noise reduction; the three indicators of the thrust-penetration linear fitting slope a, thrust intercept b and torque-penetration linear fitting slope TPI after noise reduction are used to reflect the difficulty of rock excavation during construction.

10. The method for designing a TBM excavation parameter noise reduction filter based on Fourier transform according to claim 9, characterized in that: The single-blade average thrust and single-blade average torque values ​​are collected using corresponding force sensors, and the penetration is collected using displacement sensors.

Images