Rotary drilling rig noise reduction method and system based on active noise control

Abstract

The invention discloses a rotary drilling rig noise reduction method and system based on active noise control, and belongs to the technical field of engineering machinery noise control. Noise reduction is achieved by obtaining noise signals and working condition parameters, determining a target noise reduction mode, adopting a variable step size self-adaptive filtering algorithm to calculate anti-noise signals, and controlling a loudspeaker to play the anti-noise signals. The system comprises a microphone array, a working condition monitoring module, a mode recognition module, a self-adaptive filtering module, a loudspeaker array, a noise reduction evaluation module and a control module, can adapt to the multi-working-condition dynamic characteristics of the rotary drilling rig, has the rapid convergence capability and good space noise reduction uniformity, and is high in noise reduction accuracy. And the noise reduction effect is obviously improved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of noise control of engineering machinery, and particularly relates to a rotary drilling rig noise reduction method and system based on active noise control, and more particularly to a technical scheme for effectively suppressing low-frequency noise of a rotary drilling rig by using an adaptive filtering algorithm and a multi-working-condition cooperative control strategy. BACKGROUND

[0002] As an important engineering machinery device indispensable in modern infrastructure construction, a rotary drilling rig is widely used in fields such as building engineering, bridge construction, subway construction, and highway construction. With the acceleration of urbanization, the construction site of the rotary drilling rig is increasingly close to residential areas and commercial areas, and the problem of strong noise pollution generated by the rotary drilling rig is increasingly prominent. During the working process of the rotary drilling rig, the power head rotates, the drill rod drills, the hydraulic system operates, and other links generate noise of 70-100 dB or even higher, among which low-frequency noise is dominant, and the frequency range is concentrated in 20-500 Hz. This kind of low-frequency noise has the characteristics of long propagation distance, strong penetration, and slow attenuation. Traditional passive noise reduction measures such as soundproof covers and mufflers have limited effect on the suppression of low-frequency noise, and usually only achieve a noise reduction of 5-15 dB, which is difficult to meet the increasingly stringent environmental noise standard requirements.

[0003] In the prior art, active noise control technology is gradually attracting attention as an effective means of low-frequency noise suppression. For example, a Chinese patent application with the application number 202210636792.6 discloses a method and device for determining the parameters of an active noise reduction filter. The technical scheme determines the transfer functions of the primary path and the secondary path in the active noise reduction system, and determines the parameters of the active noise reduction filter according to the transfer functions, thereby improving the active noise reduction capability. The technical scheme uses a fixed filter parameter determination method to obtain the filter parameters by calculating the product of the generalized inverse matrix of the secondary path matrix and the primary path matrix. However, the technical scheme has the following deficiencies: first, the fixed filter parameter method cannot adapt to the dynamic changes of noise characteristics of the rotary drilling rig under different working conditions. When the rotary drilling rig switches from the drilling working condition to the drill lifting working condition or standby working condition, the frequency spectrum characteristics and amplitude distribution of the noise change significantly, and the fixed parameters cannot maintain the optimal noise reduction effect under multiple working conditions; second, the scheme uses the traditional least mean square algorithm for filter weight updating, and uses a fixed step size, which makes it difficult to balance the convergence speed and steady-state error of the algorithm. When the step size is too large, the convergence speed is fast but the steady-state error is large, and when the step size is too small, the steady-state error is small but the convergence speed is slow; third, the scheme only focuses on the noise reduction effect at the error microphone position and does not consider the spatial distribution characteristics of the sound field in the noise reduction area, which may result in spatial unevenness phenomenon that some positions have good noise reduction effect while other positions have poor noise reduction effect.

[0004] In addition, according to the international front research in 2024-2025, advanced technologies such as variable step-size filtering X least mean square algorithm, neural network accelerated active noise control, and adaptive step-size control based on particle swarm optimization have shown significant advantages in the field of active noise reduction. These technologies can effectively improve the convergence speed and noise reduction performance of the active noise reduction system by dynamically adjusting algorithm parameters and introducing intelligent optimization strategies. However, existing technologies have not combined these advanced algorithms with the multi-working condition characteristics of rotary drilling rigs, and lack specialized optimization solutions for complex working scenarios of rotary drilling rigs.

[0005] Therefore, there is an urgent need to develop an active noise control method and system that can adapt to the dynamic characteristics of rotary drilling rigs in multiple working conditions, has fast convergence ability and good spatial noise reduction uniformity, to meet the strict requirements of modern engineering construction on environmental noise control. SUMMARY

[0006] The purpose of the present application is to overcome the shortcomings of the prior art, provide a rotary drilling rig noise reduction method and system based on active noise control, introduce a multi-working condition adaptive identification mechanism, a variable step-size adaptive filtering algorithm, and a spatial sound field reconstruction evaluation method, realize efficient noise reduction and noise reduction spatial uniformity optimization of rotary drilling rigs in different working conditions, and significantly improve the practicality and noise reduction effect of the active noise reduction system.

[0007] To achieve the above-mentioned purposes, the following technical solutions are adopted in the present application: The present application provides a rotary drilling rig noise reduction method based on active noise control, which acquires rotary drilling rig noise signals and working condition parameters collected by multiple microphones, determines a target noise reduction mode based on the working condition parameters, and acquires corresponding filter initial parameters and step-size adjustment coefficients from a pre-set mode parameter library. An adaptive filtering algorithm is used to calculate an anti-noise signal, which uses a variable step-size mechanism to dynamically adjust the step size according to the energy of the error signal. The anti-noise signal is played by multiple loudspeakers, the residual noise in the noise reduction area is monitored in real time, the noise reduction effect is evaluated using a spatial sound field reconstruction method, and if the noise reduction effect does not reach a pre-set threshold, the update strategy of the filter weight is adjusted.

[0008] The present application also provides a rotary drilling rig noise reduction system based on active noise control, which includes a microphone array, a working condition monitoring module, a mode identification module, an adaptive filtering module, a loudspeaker array, a noise reduction evaluation module, and a control module. The microphone array is used to collect noise signals, the working condition monitoring module acquires working condition parameters, the mode identification module determines a target noise reduction mode and acquires corresponding parameters, the adaptive filtering module calculates an anti-noise signal, the loudspeaker array plays the anti-noise signal, the noise reduction evaluation module evaluates the noise reduction effect using a spatial sound field reconstruction method, and the control module coordinates the operation of each module.

[0009] The present application has the following advantages: Firstly, the application can automatically identify and switch to the corresponding noise reduction mode according to the real-time working condition parameters of the rotary drilling rig by introducing a multi-working condition adaptive identification mechanism. Each mode is equipped with pre-optimized filter initial parameters and step adjustment coefficients, so that the noise reduction system can quickly adapt to the noise characteristic changes of the rotary drilling rig under different working conditions such as drilling, drilling, standby and the like. Compared with the fixed parameter scheme, the fluctuation range of the noise reduction performance of the application during working condition switching is reduced by more than 60%, and the noise reduction effect is improved by 15-25dB.

[0010] Secondly, the application adopts a variable step adaptive filtering algorithm based on error signal energy. By calculating the power change rate of the error signal in real time, the filter weight update step is dynamically adjusted using a nonlinear mapping function. A larger step is used to accelerate convergence when the noise characteristics change rapidly, and a smaller step is used to reduce the steady-state error when it is close to steady state. Compared with the fixed step algorithm, the convergence speed of the application is improved by 40-60%, the steady-state error is reduced by more than 35%, and effective noise reduction can be achieved within 0.5-1.0s.

[0011] Thirdly, the application introduces a spatial sound field reconstruction evaluation method. The sound field of the noise reduction area is represented as a linear combination of multiple sound field bases through sound field base function decomposition. Not only can the overall noise reduction effect be evaluated, but also the frequency components and spatial positions with poor noise reduction effect can be identified. The filter weight update strategy is adjusted accordingly. Compared with the single-point evaluation method, the spatial uniformity in the noise reduction area is improved by more than 50%, and the coverage range of the noise reduction area is expanded by 30-40%.

[0012] Fourthly, the application realizes smooth adjustment of filter parameters during working condition switching using a gradual transition method, avoiding the problem of increased transient noise caused by parameter mutation. The noise fluctuation during working condition switching is controlled within 3dB, significantly improving the stability and user experience of the system.

[0013] Fifthly, the system structure of the application is reasonable, the functions of each module are clear, and it is easy to integrate and modify on the existing rotary drilling rig without affecting the normal working performance of the rotary drilling rig. It has good engineering practical value and application prospect. BRIEF DESCRIPTION OF DRAWINGS

[0014] Figure 1 Figure 1 is a schematic diagram of the overall structure of the rotary drilling rig noise reduction system based on active noise control of the application.

[0015] Figure 2 Figure 2 is a schematic diagram of the overall process of the rotary drilling rig noise reduction method of the application.

[0016] Figure 3 Figure 3 is a schematic diagram of the working process of the mode identification module of the application.

[0017] Figure 4This is a schematic diagram of the adaptive filtering module of the present invention.

[0018] Figure 5 This is a schematic diagram of the variable step size calculation process of the present invention.

[0019] Figure 6 This is a schematic diagram of the spatial sound field reconstruction evaluation process of the present invention.

[0020] Figure 7 This is a schematic diagram illustrating the smooth transition of filter parameters during the switching of operating conditions in this invention. Detailed Implementation

[0021] Please refer to the attached document. Figures 1-7 The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.

[0022] like Figure 1 As shown, the noise reduction system for rotary drilling rigs based on active noise control of the present invention includes a microphone array 1, a working condition monitoring module 2, a pattern recognition module 3, an adaptive filtering module 4, a speaker array 5, a noise reduction evaluation module 6, and a control module 7.

[0023] Microphone array 1 is positioned at different locations on the rotary drilling rig to collect noise signals generated during its operation. Microphone array 1 includes multiple reference microphones and multiple error microphones. Reference microphones are positioned near the main noise sources of the rotary drilling rig, specifically near the power head, drill rod connection, and hydraulic pump station, to collect reference noise signals. Error microphones are positioned within the noise reduction target area, typically a 10-20m radius around the rotary drilling rig, to collect residual noise signals after noise reduction. Preferably, there are 4-8 reference microphones and 8-16 error microphones, with a frequency response range of 20-2000Hz and a sensitivity of not less than -40dB to ensure accurate acquisition of low-frequency noise signals. The noise signals collected by microphone array 1 are converted from analog to digital and transmitted at a sampling rate of 8000-16000Hz to the adaptive filtering module 4 and the noise reduction evaluation module 6.

[0024] The operating condition monitoring module 2 is connected to the control system of the rotary drilling rig to acquire its operating condition parameters in real time. The module connects to the rig control system via a CAN bus or other communication interface, acquiring parameters including power head speed, drill rod depth, hydraulic system pressure, and drilling rate. Preferably, the sampling frequency of the operating condition parameters is 10-50Hz, which can reflect changes in the working status of the rotary drilling rig in real time. The operating condition monitoring module 2 also includes a data preprocessing unit to filter and normalize the acquired operating condition parameters, eliminating measurement noise and dimensional differences, and providing standardized input for subsequent pattern recognition.

[0025] The mode recognition module 3 is connected with the working condition monitoring module 2, and is used for determining a target noise reduction mode based on the working condition parameters, and obtaining corresponding filter initial parameters and step adjustment coefficients from a preset mode parameter library. Figure 3 As shown in the figure, the working process of the mode recognition module 3 includes the following steps: first, the speed signal, the drill pipe depth signal and the hydraulic pressure signal provided by the working condition monitoring module 2 are obtained, and the working condition feature vector is calculated based on the signals. The calculation of the working condition feature vector adopts normalization processing, the speed is normalized to the range of 0-1, the drill pipe depth is normalized according to the maximum drilling depth, and the hydraulic pressure is normalized according to the rated pressure. Then, the working condition feature vector is matched with the preset mode feature vector in terms of similarity, and the Euclidean distance method is adopted for similarity calculation. The noise reduction mode corresponding to the mode feature vector with the highest similarity is selected as the target noise reduction mode.

[0026] The present application presets three noise reduction modes: drilling mode, pipe lifting mode and standby mode. The drilling mode corresponds to the working condition of the rotary drilling rig drilling the drill pipe into the soil layer, at this time the speed is high (usually 10-30 rpm), the hydraulic pressure is large (usually 15-25 MPa), and the noise is mainly low-frequency periodic pulse, with a frequency range concentrated in 50-200 Hz. The pipe lifting mode corresponds to the working condition of the rotary drilling rig lifting the drill pipe, at this time the speed is low or zero, the hydraulic pressure is medium (usually 8-15 MPa), and the noise is mainly hydraulic system noise and mechanical friction noise, with a wide frequency range, concentrated in 100-500 Hz. The standby mode corresponds to the working condition of the rotary drilling rig running under no load or short pause, at this time the speed is zero or low, the hydraulic pressure is low (usually 5-10 MPa), and the noise is mainly from the engine and hydraulic pump, with a frequency concentrated in 50-150 Hz.

[0027] Each noise reduction mode is equipped with pre-optimized filter initial parameters and step adjustment coefficients. These parameters are obtained through offline training, and the specific method is as follows: in an experimental environment, record the noise data of the rotary drilling rig under various working conditions, adopt batch processing, run the traditional filter X least mean square algorithm for each working condition to the convergence state, and take the converged filter weight as the filter initial parameter of the working condition. The step adjustment coefficient is determined through multiple tests, under the condition of ensuring the best balance between convergence speed and steady-state error. Preferably, the step adjustment coefficient of the drilling mode is 0.8-1.2, the step adjustment coefficient of the pipe lifting mode is 0.6-1.0, and the step adjustment coefficient of the standby mode is 0.4-0.8.

[0028] The adaptive filtering module 4 is connected with the microphone array 1 and the mode recognition module 3, and is used for calculating an anti-noise signal based on the noise signal by using an adaptive filtering algorithm. Figure 4As shown, the adaptive filter module 4 includes a filter initialization unit, a secondary path compensation unit, a step size calculation unit, a weight update unit, and an anti-noise generation unit.

[0029] The filter initialization unit initializes the filter weights based on the filter initial parameters provided by the mode recognition module 3. The filter adopts a finite impulse response filter structure, and the filter order is preferably 128-512, which can effectively represent the frequency spectrum characteristics of the rotary drilling rig noise. When the working condition changes and the target noise reduction mode switches, the filter initialization unit does not directly update the filter weights to the new initial parameters, but uses a gradual transition method to gradually adjust the current filter weights to the new filter initial parameters. The transition time is preferably 0.5-2.0s, and the filter weights change smoothly during the transition process according to linear interpolation or cosine interpolation, avoiding the increase of transient noise caused by parameter mutation.

[0030] The secondary path compensation unit is used to filter the reference noise signal based on the secondary path transfer function. The secondary path refers to the acoustic propagation path from the loudspeaker to the error microphone, including the electro-acoustic conversion characteristics of the loudspeaker, the spatial sound propagation characteristics, and the acoustoelectric conversion characteristics of the error microphone. The secondary path transfer function is obtained by an offline identification method. The specific method is: during the system installation and debugging phase, input a known test signal (such as white noise or a sweep signal) to the loudspeaker, and record the response signal collected by the error microphone. The least squares method or recursive least squares method is used to estimate the secondary path transfer function. Preferably, the secondary path transfer function is represented by a finite impulse response filter with an order of 64-256. The secondary path compensation unit filters the reference noise signal to generate a filtered reference signal, which is used for subsequent weight update calculation.

[0031] The step size calculation unit is used to dynamically calculate the current step size value based on the step size adjustment coefficient and the energy of the error signal. As shown, Figure 5 The specific process of step size calculation includes: first, calculate the instantaneous power of the error signal, i.e. the square value of the error signal. Then, the exponential smoothing method is used to calculate the smoothed power of the error signal. The exponential smoothing method can effectively suppress the influence of instantaneous fluctuations, and the calculation formula of the smoothed power is: , wherein, is the smoothed power at the nth moment, is the instantaneous power at the nth moment, is the smoothing coefficient, is the time index. Preferably, the value is 0.05-0.2, which can achieve a good balance between response speed and smoothing effect.

[0032] Next, based on the smoothed power and historical smoothed power, the rate of change of power is calculated: , in, Let n be the rate of change of power at time n. For the length of the history window, Small positive numbers are used to avoid division by zero errors. Preferably, The value is taken from 10 to 50 sampling points. Values The power change rate reflects the dynamic change of the error signal energy. When the error signal energy changes rapidly, the power change rate is large, indicating that the noise characteristics are changing or the system has not yet converged. In this case, a larger step size is needed to accelerate the weight adjustment. When the error signal energy tends to be stable, the power change rate is small, indicating that the system is close to steady state. In this case, a smaller step size is needed to reduce steady-state error.

[0033] The current step size value is determined using a nonlinear mapping function based on the power change rate and step size adjustment coefficient. , in, This represents the current step size at time n. Based on step size, The step size adjustment coefficient provided for pattern recognition module 3 This is the step size variation coefficient. For nonlinear mapping gain, It is the hyperbolic tangent function. Preferably, The value ranges from 0.01 to 0.05. The value ranges from 0.5 to 1.5. The value ranges from 5 to 20. The hyperbolic tangent function has good nonlinear mapping characteristics, capable of mapping the power change rate to the step size adjustment range. When the power change rate is large, the hyperbolic tangent function output is close to 1, indicating a large current step size; when the power change rate is small, the hyperbolic tangent function output is close to 0, indicating a small current step size. This nonlinear mapping mechanism can effectively balance convergence speed and steady-state error, enabling the adaptive filtering algorithm to automatically adjust the learning rate at different stages.

[0034] The weight update unit is connected to the step size calculation unit and is used to update the filter weights based on the current step size value. This invention employs an improved filter-X minimum mean square algorithm for weight update. The filter-X minimum mean square algorithm is a classic algorithm widely used in the field of active noise control. Its core idea is to adaptively adjust the filter weights by minimizing the mean square value of the error signal. The specific formula for weight update is: , in, Let n be the filter weight vector at time n. Let n be the filtered reference signal vector. Let be the error signal at time n. Time index. Filter weight vector. The dimension is the same as the filter order, preferably 128-512 dimensions. (Referencing the filtered reference signal vector.) Provided by the secondary path compensation unit, its calculation method is as follows: , in, This is the filter coefficient vector of the secondary path transfer function. Let n be the reference noise signal vector at time n. This represents the convolution operation.

[0035] Error signal Acquired by the error microphone, its physical meaning is the superposition result of the primary noise signal and the secondary noise-reduced signal at the error microphone's location: , in, Let be the sound pressure signal generated by the primary noise source at the position of the error microphone at time n. Let be the sound pressure signal generated by the secondary noise reduction source at the error microphone position at time n. The goal of active noise control is to make... Approaching zero, meaning the secondary noise reduction signal completely cancels out the primary noise signal.

[0036] The weight update unit performs a weight update operation once at each sampling time, with the update frequency being the same as the sampling frequency of the microphone array 1, preferably 8000-16000Hz. To prevent numerical overflow or instability, the weight update unit also includes a weight limiting mechanism, which limits each component of the filter weight to a preset range, preferably -2 to 2.

[0037] The anti-noise generation unit is connected to the weight update unit and is used to generate an anti-noise signal based on the updated filter weights. The formula for calculating the anti-noise signal is: , in, Let n be the anti-noise signal at time n. This is the transpose of the filter weight vector. The noise signal vector serves as the reference. After digital-to-analog conversion and power amplification, the anti-noise signal drives the speaker array 5 for playback.

[0038] The speaker array 5 is connected to the adaptive filtering module 4 and is used to play anti-noise signals. The speaker array 5 includes multiple speakers arranged around the boundary of the noise reduction target area around the rotary drilling rig, forming an acoustic barrier. Preferably, the number of speakers is 4-8, the frequency response range of the speakers is 20-2000Hz, and the rated power is 50-200W, capable of generating sufficient sound pressure level at a distance of 10-20m to cancel the noise of the rotary drilling rig. The speaker placement needs to be optimized through acoustic simulation to ensure a uniform noise reduction effect within the noise reduction target area. The speaker array 5 is driven by a multi-channel power amplifier, with each speaker corresponding to an independent amplification channel to avoid mutual interference between channels.

[0039] The noise reduction evaluation module 6 is connected to the microphone array 1 and is used to monitor the residual noise in the noise reduction area in real time, and evaluate the noise reduction effect based on the residual noise using a spatial sound field reconstruction method. Figure 6 As shown, the workflow of the noise reduction evaluation module 6 includes: first, acquiring residual noise signals from multiple monitoring points within the noise reduction area, which are provided by error microphones in the microphone array 1; then, based on the residual noise signals, using the sound field basis function decomposition method, decomposing the sound field of the noise reduction area into a linear combination of multiple sound field bases.

[0040] Sound field basis function decomposition is a classic method in acoustics. Its basic principle is to represent a complex sound field as a linear combination of a set of orthogonal or approximately orthogonal basis functions. For the rotary drilling rig noise reduction application in this invention, the sound field basis functions can be spherical harmonic functions or plane wave expansion functions. Preferably, spherical harmonic functions are used to represent the sound field because they possess orthogonality and completeness, accurately characterizing the sound field distribution in three-dimensional space. The spherical harmonic expansion form of the sound field is: , in, For spatial location The sound pressure at that location, For spherical harmonic coefficients, For order is And modulus is spherical harmonic functions, and These are the azimuth and elevation angles in a spherical coordinate system. For a sphere of the first kind, Bessel function For wave number, Radial distance, This is the maximum order. Preferably, A value of 3-6 can achieve a good balance between computational complexity and reconstruction accuracy.

[0041] Spherical harmonic coefficient The solution is obtained using the least squares method. Specifically, it is assumed that [the following is a list of elements] are deployed within the noise reduction region. The error microphone, the first The microphone is located at position The measured residual noise is Then we have: , Write the equations for all measurement points in matrix form: , in, This is the residual noise vector. The sound field basis function matrix, Let be the spherical harmonic coefficient vector. Solve using the least squares method: , in, For matrix transpose, For matrix The inverse matrix.

[0042] After calculating the spherical harmonic coefficients, the complete sound pressure distribution of the noise reduction region can be reconstructed. For any location within the noise reduction region... The sound pressure level can be calculated by substituting the spherical harmonic coefficients into the expansion formula. Based on the reconstructed sound pressure distribution, the average sound pressure level and spatial uniformity of the noise reduction area are calculated. The average sound pressure level is defined as: , in, The average sound pressure level, The volume of the noise reduction region. For position The sound pressure at that location, The volume is represented by a small element. In actual calculations, the integral is approximated using numerical methods. The noise reduction region is discretized into several small grids, the sound pressure at the center point of each grid is calculated, and then the average value is obtained.

[0043] Spatial uniformity is defined as the ratio of the standard deviation to the mean of the sound pressure distribution. , in, For spatial uniformity, The standard deviation of the sound pressure distribution. This represents the average sound pressure distribution. Spatial uniformity reflects the degree of evenness of the noise reduction effect in space; the smaller the spatial uniformity, the more uniform the noise reduction effect in space. Preferably, the spatial uniformity should be less than 0.3, indicating that the difference in noise reduction effect at different locations within the noise reduction area is small.

[0044] The noise reduction effect is evaluated based on the average sound pressure level and spatial uniformity. If the average sound pressure level is lower than the preset target sound pressure level and the spatial uniformity is less than the preset uniformity threshold, the noise reduction effect is considered to have met the preset requirements. Otherwise, the noise reduction effect is considered to have failed to meet the preset requirements, and the filter weight update strategy needs to be adjusted.

[0045] If the noise reduction effect does not reach the preset threshold, the noise reduction evaluation module 6 also performs an adjustment of the filter weight update strategy. Specifically, based on the spherical harmonic coefficients, frequency components with poor noise reduction effects are identified. By analyzing the amplitude of the spherical harmonic coefficients at different frequencies, it can be determined which frequency components have larger residual noise. For frequency components with poor noise reduction effects, the filter weight update step size of the corresponding frequency band is increased to accelerate the weight adjustment of that frequency band, thereby improving the noise reduction effect of that frequency band. For frequency components with good noise reduction effects, the filter weight update step size of the corresponding frequency band is decreased to avoid performance degradation caused by over-adjustment. The specific method for step size adjustment is as follows: the filter weights are divided into several sub-bands according to frequency, each sub-band corresponds to an independent step size, and the step size value is dynamically adjusted according to the noise reduction effect of that sub-band. Preferably, the frequency range of 20-500Hz is divided into 5-10 sub-bands, and the step size of each sub-band is adjusted independently, with an adjustment range of 0.5-2 times the base step size.

[0046] The control module 7 is connected to the pattern recognition module 3, the adaptive filtering module 4, and the noise reduction evaluation module 6, respectively. It coordinates the operation of each module and adjusts the filter weight update strategy when the noise reduction effect fails to reach a preset threshold. The control module 7 is implemented using an embedded processor, preferably an ARM Cortex-A series or a digital signal processor. The main functions of the control module 7 include: system initialization, inter-module data communication management, operating condition switching detection and response, and noise reduction effect monitoring and adjustment.

[0047] During system initialization, control module 7 loads the preset mode parameter library, establishes data communication links between modules, and initializes the hardware parameters of microphone array 1 and speaker array 5. During system operation, control module 7 monitors the operating condition parameters provided by operating condition monitoring module 2 in real time to detect changes in operating conditions. If the change in operating condition parameters exceeds a preset switching threshold, control module 7 triggers the mode switching process, notifies mode recognition module 3 to re-execute the determination of the target noise reduction mode, and controls adaptive filtering module 4 to adjust filter parameters using a gradual transition method.

[0048] like Figure 7 As shown, the smooth transition process of filter parameters during operating condition switching is as follows: Assume the current filter weights are... The initial parameters of the new target filter are The transition time is The sampling frequency is Then, during the transition process, the first The filter weights for each sampling point are: , in, This represents the total number of sampling points during the transition process. This is the index of the current sampling point. The range of values ​​is arrive This linear interpolation method enables a smooth transition in filter weights, avoiding abrupt parameter changes. Preferably, cosine interpolation can also be used to further smooth the transition. , Cosine interpolation changes slowly at the beginning and end of the transient process, but changes rapidly in the middle stage, which can further reduce transient noise fluctuations.

[0049] The control module 7 also monitors the noise reduction effect evaluation results provided by the noise reduction evaluation module 6 in real time. If the noise reduction effect does not reach the preset threshold, the control module 7 transmits the frequency components that need adjustment and the corresponding step size adjustment strategy identified by the noise reduction evaluation module 6 to the adaptive filtering module 4, guiding the adaptive filtering module 4 to optimize the filter weight update strategy. The control module 7 also records system operating data, including noise reduction performance, algorithm convergence time, steady-state error, etc. under various operating conditions, for system performance evaluation and parameter optimization.

[0050] The workflow of this invention is as follows: Figure 2 As shown, the specific steps include: Step S1: The system starts up, the control module 7 performs system initialization, loads the mode parameter library, establishes communication links between modules, and initializes the microphone array 1 and the speaker array 5.

[0051] Step S2: Microphone array 1 begins to collect noise signals from the rotary drilling rig, and operating condition monitoring module 2 begins to acquire operating condition parameters of the rotary drilling rig.

[0052] Step S3: The pattern recognition module 3 determines the target noise reduction mode based on the operating parameters and obtains the corresponding filter initial parameters and step size adjustment coefficients from the pattern parameter library.

[0053] Step S4: Adaptive filtering module 4 initializes the filter weights based on the filter initial parameters and begins to execute the adaptive filtering algorithm.

[0054] Step S5: The secondary path compensation unit filters the reference noise signal to generate a filtered reference signal.

[0055] Step S6: The step size calculation unit calculates the instantaneous power of the error signal, calculates the smoothed power using the exponential smoothing method, calculates the power change rate based on the smoothed power, and determines the current step size value using a nonlinear mapping function.

[0056] Step S7: The weight update unit updates the filter weights based on the current step size, the filtered reference signal, and the error signal.

[0057] Step S8: The anti-noise generation unit generates an anti-noise signal based on the updated filter weights.

[0058] Step S9: The speaker array 5 plays an anti-noise signal to interfere with and cancel out the noise signal of the rotary drilling rig.

[0059] Step S10: The noise reduction evaluation module 6 acquires the residual noise signals of multiple monitoring points within the noise reduction area, reconstructs the complete sound pressure distribution of the noise reduction area using the sound field basis function decomposition method, calculates the average sound pressure level and spatial uniformity, and evaluates the noise reduction effect.

[0060] Step S11: Determine whether the noise reduction effect has reached the preset threshold. If yes, continue monitoring; otherwise, proceed to step S12.

[0061] Step S12: The noise reduction evaluation module 6 identifies frequency components with poor noise reduction effect based on the spherical harmonic coefficients, increases the filter weight update step size of the corresponding frequency band for these frequency components, and decreases the step size of the corresponding frequency band for frequency components with good noise reduction effect, and passes the adjustment strategy to the adaptive filtering module 4.

[0062] Step S13: The adaptive filtering module 4 re-executes the filter weight update according to the adjustment strategy and the adjusted step size, and returns to step S6.

[0063] Step S14: The control module 7 continuously monitors the operating parameters and determines whether the operating conditions have changed. If the change in the operating parameters exceeds the preset switching threshold, then step S15 is executed; otherwise, the process returns to step S6 to continue noise reduction control.

[0064] Step S15: Pattern recognition module 3 redetermines the target noise reduction mode and obtains new filter initial parameters and step size adjustment coefficients.

[0065] Step S16: The adaptive filtering module 4 adopts a gradual transition method to gradually adjust the current filter weights to the new initial filter parameters. The transition time is preferably 0.5-2.0s. Then, return to step S6 to continue executing the noise reduction control.

[0066] In one embodiment of the present invention, a rotary drilling rig is performing pile foundation construction at a construction site, initially in drilling mode. Microphone array 1 includes 6 reference microphones and 12 error microphones. The reference microphones are respectively arranged on the left and right sides of the power head, the upper part of the drill rod, the middle part of the drill rod, the left side of the hydraulic pump station, and the right side of the hydraulic pump station, at a distance of 1-2 meters from the noise source. The error microphones are evenly distributed on a circumference with a radius of 15 meters around the rotary drilling rig, at a height of 1.5 meters above the ground. Speaker array 5 includes 6 speakers, arranged alternately with the error microphones on the same circumference.

[0067] The operating parameters collected by the operating condition monitoring module 2 are: power head speed 18 rpm, drill pipe depth 12 m, and hydraulic system pressure 20 MPa. The pattern recognition module 3 calculates the operating condition feature vector based on these parameters and performs similarity matching with the preset drilling mode feature vector. The similarity reaches 0.92, far higher than the drill lifting mode (similarity 0.15) and standby mode (similarity 0.08). Therefore, the target noise reduction mode is determined to be the drilling mode. The initial filter parameters and step size adjustment coefficients corresponding to the drilling mode are obtained from the mode parameter library. It is 1.0.

[0068] The adaptive filtering module 4 employs a 256th-order finite impulse response (FIR) filter, initializing the filter weights based on the filter's initial parameters. The secondary path transfer function is represented using a 128th-order FIR filter and obtained through an offline identification method. The step size calculation unit sets the smoothing coefficient. The historical window length is 0.1. For 20 sampling points, the basic step size is... The step size variation coefficient is 0.03. The nonlinear mapping gain is 1.0. It is 10.

[0069] After system startup, the adaptive filtering algorithm begins operation. Initially, the error signal has high power and a high rate of power change; the current step size calculated by the step size calculation unit is close to twice the base step size, accelerating filter weight convergence. After approximately 0.8 seconds (about 12,800 sampling points at a sampling frequency of 16,000 Hz), the error signal power decreases significantly, the rate of power change decreases, and the current step size gradually decreases to 1.2 times the base step size. Once the system enters steady state, the current step size stabilizes between 1.0 and 1.2 times the base step size, and the error signal power remains at a low level.

[0070] Noise reduction evaluation module 6 uses the spherical harmonic function decomposition method, with a maximum order of The sound field of the noise reduction area was reconstructed using a setting of 4. The calculated average sound pressure level was 62 dB, representing a noise reduction of 23 dB compared to the original 85 dB. The spatial uniformity was 0.25, indicating that the noise reduction effect was evenly distributed spatially. The noise reduction effect reached the preset threshold, and the system continued monitoring and operation.

[0071] After approximately 10 minutes, the rotary drilling rig completes drilling to the current depth and begins lifting the drill rod, switching the operating mode to drill lifting mode. The operating parameters collected by the operating condition monitoring module 2 show the following changes: the power head speed drops to 0 rpm, the drill rod depth begins to decrease, and the hydraulic system pressure drops to 12 MPa. The control module 7 detects that the change in operating parameters exceeds the switching threshold, triggering the mode switching process. The mode recognition module 3 re-determines the target noise reduction mode as drill lifting mode and obtains the initial parameters and step size adjustment coefficients of the new filter corresponding to drill lifting mode. It is 0.8.

[0072] The adaptive filtering module 4 initiates a smooth transition process for the filter parameters, with a transition time set to 1.5 seconds. Using cosine interpolation, the current filter weights are gradually adjusted to the new initial filter parameters within 1.5 seconds. During the transition, noise fluctuations are controlled within 2.5 dB, and no significant transient noise increase is observed. After the transition is complete, the system continues noise reduction control based on the parameters of the drilling mode, with the average sound pressure level stabilizing at around 65 dB, achieving a noise reduction of 18 dB (the original noise in drilling mode is 83 dB).

[0073] After about 5 minutes, the rotary drilling rig completed the drill rod lifting and entered a brief standby state, switching the operating condition to standby mode. The system then executed the mode switching procedure again, adjusting the filter parameters using the same smooth transition method. In standby mode, the average sound pressure level stabilized at around 58dB, with a noise reduction of 14dB (the original noise in standby mode was 72dB).

[0074] As can be seen from the above embodiments, the rotary drilling rig noise reduction system of the present invention can automatically switch noise reduction modes according to the real-time working conditions of the rotary drilling rig, maintaining good noise reduction effect under different working conditions. Employing a variable step-size adaptive filtering algorithm, the system can achieve effective noise reduction within 1 second, with a convergence speed significantly better than the fixed step-size algorithm (which typically requires 2-3 seconds to converge). Using a spatial sound field reconstruction evaluation method, the system can not only evaluate the overall noise reduction effect but also identify frequency components and spatial locations where the noise reduction effect is poor, and specifically optimize the filter weight update strategy to keep the spatial uniformity within the noise reduction area below 0.3, significantly better than the single-point evaluation method (the spatial uniformity of the single-point evaluation method is typically above 0.5).

[0075] It should be noted that the embodiments described in this invention are merely illustrative and are not intended to limit the scope of protection of this invention. Those skilled in the art can make various modifications and improvements to this invention without departing from its spirit and essence, and these modifications and improvements should also be considered within the scope of protection of this invention.

[0076] For example, in another embodiment of the present invention, multiple adaptive filtering algorithms can be combined to employ different filtering algorithms for different frequency bands. For low-frequency noise (20-100Hz), a variant based on the normalized minimum mean square algorithm can be used to improve the convergence speed in the low-frequency band; for mid-frequency noise (100-300Hz), the improved filtering X minimum mean square algorithm described in this invention can be used; for high-frequency noise (300-500Hz), a frequency domain adaptive filtering algorithm can be used to reduce computational complexity.

[0077] In another embodiment of the present invention, machine learning methods, such as support vector machines or neural networks, can be introduced into the pattern recognition module 3 to classify operating parameters and improve the accuracy and robustness of pattern recognition. Training data can be obtained by recording operating parameters and corresponding optimal filter parameters under different operating conditions in the actual working environment. After offline training, the trained classifier is embedded into the pattern recognition module 3.

[0078] In another embodiment of the present invention, a predictive evaluation mechanism can be introduced into the noise reduction evaluation module 6. Based on historical noise reduction effect data and the changing trends of operating parameters, a time series prediction method (such as a recurrent neural network or a long short-term memory network) is used to predict the noise reduction effect at future times. If the predicted noise reduction effect may be lower than a preset threshold, the filter weight update strategy is adjusted in advance to achieve active rather than passive noise reduction performance optimization.

[0079] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. All equivalent structural transformations made under the inventive concept of the present invention using the contents of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. A method for reducing noise of a rotary drilling rig based on active noise control, characterized in that, The method comprises the following steps: acquiring noise signals collected by multiple microphones of a rotary drilling rig and acquiring working condition parameters of the rotary drilling rig; determining a target noise reduction mode based on the working condition parameters, wherein the target noise reduction mode comprises at least one of a drilling mode, a tripping mode and a standby mode; acquiring corresponding filter initial parameters and step adjustment coefficients from a preset mode parameter library based on the target noise reduction mode; calculating an anti-noise signal based on the noise signals by using an adaptive filtering algorithm, wherein the adaptive filtering algorithm adopts a variable step mechanism, and the variable step is dynamically adjusted according to an energy of an error signal; initializing filter weights of the adaptive filtering algorithm based on the filter initial parameters; determining a current step value based on the step adjustment coefficients and the energy of the error signal in a filter weight updating process; updating the filter weights by using the current step value to generate the anti-noise signal; controlling multiple loudspeakers to play the anti-noise signal to interfere and cancel the noise signals; monitoring residual noise in a noise reduction area in real time, and evaluating a noise reduction effect by using a spatial sound field reconstruction method based on the residual noise; and adjusting an updating strategy of the filter weights based on the residual noise if the noise reduction effect does not reach a preset threshold.

2. The method of claim 1, wherein, The method further comprises the following steps of determining the target noise reduction mode based on the working condition parameters: acquiring a rotating speed signal, a drill rod depth signal and a hydraulic pressure signal of the rotary drilling rig; calculating a working condition feature vector based on the rotating speed signal, the drill rod depth signal and the hydraulic pressure signal; performing similarity matching on the working condition feature vector and a preset mode feature vector; and selecting a noise reduction mode corresponding to a mode feature vector with the highest similarity as the target noise reduction mode.

3. The method of claim 1, wherein, The adaptive filtering algorithm is an improved filter X least mean square algorithm, and the method of calculating the anti-noise signal based on the noise signals by using the adaptive filtering algorithm comprises the following steps: acquiring a reference noise signal collected by a reference microphone and an error signal collected by an error microphone; performing filter processing on the reference noise signal based on a secondary path transfer function to generate a filtered reference signal; calculating a filter output signal based on the filtered reference signal and the filter weights; calculating a weight updating gradient based on the error signal and the filtered reference signal; updating the filter weights based on the current step value and the weight updating gradient; and generating the anti-noise signal based on the updated filter weights and the reference noise signal.

4. The method of claim 1, wherein, The method of determining the current step value based on the step adjustment coefficients and the energy of the error signal comprises the following steps: calculating an instantaneous power of the error signal; calculating a smoothed power of the error signal by using an exponential smoothing method based on the instantaneous power; calculating a power change rate based on the smoothed power and a historical smoothed power; and determining the current step value by using a nonlinear mapping function based on the power change rate and the step adjustment coefficients, wherein the current step value is larger when the power change rate is larger, and the current step value is smaller when the power change rate is smaller.

5. The method of claim 1, wherein, The method of evaluating the noise reduction effect by using the spatial sound field reconstruction method comprises the following steps: The residual noise signals of a plurality of monitoring points in the noise reduction area are acquired; based on the residual noise signals, a sound field basis function decomposition method is used to decompose the sound field of the noise reduction area into a linear combination of a plurality of sound field bases; coefficients corresponding to each sound field base are calculated; based on the coefficients, the complete sound pressure distribution of the noise reduction area is reconstructed; based on the sound pressure distribution, the average sound pressure level and the spatial uniformity of the noise reduction area are calculated; based on the average sound pressure level and the spatial uniformity, the noise reduction effect is evaluated.

6. The method of claim 5, wherein, The update strategy of the filter weight based on the residual noise includes: Based on the coefficients, frequency components with poor noise reduction effect are identified; for the frequency components, the filter weight update step size of the corresponding frequency band is increased; for frequency components with good noise reduction effect, the filter weight update step size of the corresponding frequency band is reduced; based on the adjusted step size, the filter weight update is re-executed.

7. The method of claim 1, wherein, The method further includes: When the working condition of the rotary drilling rig changes, the change amount of the working condition parameter is detected; if the change amount exceeds a preset switching threshold, the step of determining the target noise reduction mode is re-executed; based on the new target noise reduction mode, new filter initial parameters and step size adjustment coefficients are acquired from the mode parameter library; using a gradual transition method, the current filter weight is gradually adjusted to the new filter initial parameters.

8. A rotary drilling rig noise reduction system based on active noise control for implementing the method of any one of claims 1-7, characterized in that, It includes: A microphone array is used to collect noise signals generated during the operation of the rotary drilling rig; a working condition monitoring module is connected to the control system of the rotary drilling rig to acquire working condition parameters of the rotary drilling rig; a mode recognition module is connected to the working condition monitoring module to determine a target noise reduction mode based on the working condition parameters and acquire corresponding filter initial parameters and step size adjustment coefficients from a preset mode parameter library; an adaptive filtering module is connected to the microphone array and the mode recognition module to calculate an anti-noise signal based on the noise signal using an adaptive filtering algorithm, wherein the adaptive filtering module uses a variable step size mechanism to dynamically adjust the step size according to the energy of the error signal; a loudspeaker array is connected to the adaptive filtering module to play the anti-noise signal; a noise reduction evaluation module is connected to the microphone array to monitor residual noise in the noise reduction area in real time and evaluate the noise reduction effect based on the residual noise using a spatial sound field reconstruction method; a control module is connected to the mode recognition module, the adaptive filtering module, and the noise reduction evaluation module to coordinate the operation of each module and adjust the update strategy of the filter weight when the noise reduction effect does not meet a preset threshold.

9. The system of claim 8, wherein, The microphone array includes a plurality of reference microphones and a plurality of error microphones, the reference microphones are arranged near the main noise source of the rotary drilling rig to collect reference noise signals, and the error microphones are arranged in the noise reduction target area to collect residual noise signals after noise reduction.

10. The system of claim 8, wherein, The adaptive filtering module includes: The filter initialization unit is configured to initialize filter weights based on filter initial parameters provided by the pattern recognition module; the secondary path compensation unit is configured to perform filter processing on the reference noise signal based on a secondary path transfer function; the step size calculation unit is configured to dynamically calculate a current step size value based on the step size adjustment coefficient and the energy of the error signal; the weight updating unit, connected to the step size calculation unit, is configured to update the filter weights based on the current step size value; and the anti-noise generation unit, connected to the weight updating unit, is configured to generate the anti-noise signal based on the updated filter weights.

Images