Method, system, device, medium and program product thereof for controlling the dynamic balancing of a grinding machine

Abstract

The present disclosure provides a control method, system, device, medium and program product for dynamic balancing of a grinding machine. The control method comprises: obtaining an original amplitude of a grinding unit; controlling a dynamic balancing assembly to perform a first correction movement based on a preset initial angle step, and obtaining a first correction amplitude of the grinding unit after the first correction movement; determining an initial phase value of the correction centrifugal force based on the first correction amplitude and the original amplitude; and performing an iterative correction step until the actual correction amplitude meets a preset dynamic balancing condition. Through the adaptive variable step optimization strategy optimized by the fuzzy neural network, the balance accuracy of the rotating mechanical structure (such as the grinding unit in the grinding machine) is dynamically improved through multiple rounds of fine adjustment, overcoming the problems of slow convergence or insufficient accuracy of traditional coordinate wheel exchange, and the defects of influence coefficient method relying on previous calibration values, and having the significant advantages of fast convergence speed and high final accuracy, meeting the high-precision working condition requirements.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of machine tool control, in particular to a control method, system, device, medium and program product thereof for dynamic balancing of a grinding machine. BACKGROUND

[0002] Dynamic balancing technology is a key link in the operation of rotating machinery. The core goal is to reduce the vibration amplitude of the system during operation by adding or removing appropriate unbalance to the rotor or tool, thereby improving the stability of the equipment and prolonging the service life. Existing dynamic balancing methods are mainly divided into coordinate wheel changing method and influence coefficient method. Both methods have their own defects:

[0003] The coordinate wheel changing method, also known as the step-by-step adjustment method, gradually adjusts the phase and amplitude of the unbalance vector by changing the angle position of the trial weight, gradually reducing the vibration. The disadvantage is that the convergence speed is limited, and multiple iterations are required to achieve a better balancing effect. In addition, when the vibration measurement is disturbed by noise, the adjustment accuracy is easily affected. Traditional coordinate wheel changing often uses only a single angle step, making it difficult to balance precision while ensuring efficiency.

[0004] The influence coefficient method establishes a linear relationship matrix between the unbalance and the measured vibration by adding trial weights at different positions, and directly calculates the compensation amount using matrix operations. The disadvantage is that the influence coefficient needs to be calibrated in advance through multiple trial weights, and the calibration value is easily affected by the actual environment. Differences in working conditions of rotating machinery may result in errors in the final balancing result. SUMMARY

[0005] The technical problem to be solved by the present disclosure is to overcome the defects in the prior art that the dynamic balancing control precision of rotating machinery cannot be guaranteed based on the coordinate wheel changing method or the influence coefficient method, and to provide a control method, system, device, medium and program product thereof for dynamic balancing of a grinding machine.

[0006] The present disclosure solves the above technical problems by the following technical solutions:

[0007] In a first aspect, a control method for dynamic balancing of a grinding machine is provided. The grinding machine includes a grinding unit, and the grinding unit includes a dynamic balancing assembly for generating a correction centrifugal force to dynamically balance the grinding unit. The control method comprises:

[0008] Obtaining an original amplitude of the grinding unit;

[0009] Controlling the dynamic balancing assembly to perform a first correction movement based on a preset initial angle step, and obtaining a first correction amplitude of the grinding unit after the first correction movement;

[0010] determining an initial phase value of the correction centrifugal force based on the first correction amplitude and the original amplitude;

[0011] performing an iterative correction step, comprising:

[0012] inputting the amplitude variation trend parameter of the grinding unit as an input into a preset fuzzy neural network to generate a step correction coefficient;

[0013] obtaining a correction parameter of the correction centrifugal force based on the step correction coefficient and an expected correction amplitude of the grinding unit;

[0014] adjusting the correction centrifugal force based on the correction parameter and obtaining an actual correction amplitude of the grinding unit after correction;

[0015] obtaining an amplitude correction error based on the expected correction amplitude and the actual correction amplitude, and feeding back the amplitude correction error to the preset fuzzy neural network to optimize the preset fuzzy neural network;

[0016] repeating the iterative correction step until the actual correction amplitude meets a preset dynamic balance condition.

[0017] Optionally, the step of obtaining the original amplitude of the grinding unit comprises:

[0018] obtaining an original vibration signal of the grinding unit;

[0019] adding a Hanning window to the original vibration signal and performing a fast Fourier transform to obtain a corresponding vibration information spectrum;

[0020] processing the vibration information spectrum to obtain an original spectrum component parameter set corresponding to a real vibration frequency;

[0021] The original spectrum component parameter set includes an original amplitude.

[0022] Optionally, the step of processing the vibration information spectrum to obtain an original spectrum component parameter set corresponding to a real vibration frequency comprises:

[0023] determining a main peak spectrum line with the largest amplitude and its left and right two adjacent auxiliary spectrum lines based on a preset search interval;

[0024] obtaining a first amplitude of the main peak spectrum line, a second amplitude of the left auxiliary spectrum line, and a third amplitude of the right auxiliary spectrum line;

[0025] calculating a target frequency deviation for correcting the main peak spectrum line based on the first amplitude, the second amplitude, and the third amplitude;

[0026] The first amplitude corresponding to the main peak spectrum line is corrected based on the target frequency deviation, and the original amplitude is calculated.

[0027] Optionally, the dynamic balancing assembly comprises a first counterweight and a second counterweight coaxially arranged on a main shaft of the grinding unit, a resultant centrifugal force generated by the first counterweight and the second counterweight being the correction centrifugal force, and the correction centrifugal force comprising a centrifugal force phase and a centrifugal force amplitude.

[0028] The step of adjusting the correction centrifugal force based on the correction parameter comprises:

[0029] Based on the correction parameter, a phase value variable and / or an amplitude variable corresponding to the current centrifugal force phase and / or the centrifugal force amplitude are obtained.

[0030] The first counterweight and the second counterweight are synchronously controlled to perform corresponding phase adjustment based on the phase value variable and / or the amplitude variable.

[0031] Optionally, an included angle between the first counterweight and the second counterweight is a centrifugal force included angle, and the step of synchronously controlling the first counterweight and the second counterweight to perform corresponding phase adjustment based on the phase value variable and / or the amplitude variable comprises:

[0032] First, a phase value variable corresponding to the current centrifugal force phase is obtained based on the correction parameter, and the first counterweight and the second counterweight are synchronously controlled to perform same-direction phase adjustment, and the centrifugal force included angle is maintained.

[0033] In response to the actual correction amplitude meeting a preset phase balance condition, an amplitude variable corresponding to the current centrifugal force phase is obtained based on the correction parameter, and the first counterweight and the second counterweight are synchronously controlled to perform opposite-direction phase adjustment to the centrifugal force included angle.

[0034] Optionally, in response to the actual correction amplitude meeting the preset phase balance condition, the control method further comprises:

[0035] In response to the amplitude variable being lower than a preset change threshold, and the step length correction coefficient output by the preset fuzzy neural network this time representing a first update direction of the current phase adjustment being opposite to a reference direction of the phase adjustment last time, the step length correction coefficient is output twice again, and a second update direction and a third update direction of the phase adjustment are obtained.

[0036] In response to the inconsistency between the second update direction and the third update direction, the first weight counterweight and the second weight counterweight are controlled to adjust the centrifugal force angle in the reference direction based on the step size correction coefficient.

[0037] In a second aspect, a control system for dynamic balancing of a grinding machine is provided. The grinding machine includes a grinding unit, and the grinding unit includes a dynamic balancing component. The dynamic balancing component is used to generate a corrective centrifugal force to dynamically balance the grinding unit. The control system includes an original amplitude acquisition module, an initial movement module, a phase determination module, and an iterative correction module.

[0038] The original amplitude acquisition module is used to acquire the original amplitude of the grinding unit;

[0039] The initial movement module is used to control the dynamic balancing component to perform the initial correction movement based on a preset initial angle step size, and to obtain the initial correction amplitude of the grinding unit after the initial correction movement.

[0040] The phase determination module is used to determine the initial phase value of the corrected centrifugal force based on the first correction amplitude and the original amplitude.

[0041] The iterative correction module is used to perform iterative correction steps, including:

[0042] The amplitude variation trend parameter of the grinding unit is used as input and input into a preset fuzzy neural network to generate step size correction coefficient;

[0043] The correction parameters of the corrected centrifugal force and the expected correction amplitude of the grinding unit are obtained based on the step size correction coefficient.

[0044] The centrifugal force is adjusted based on the correction parameters, and the actual correction amplitude of the grinding unit after correction is obtained;

[0045] The amplitude correction error is obtained based on the expected correction amplitude and the actual correction amplitude, and the amplitude correction error is fed back to the preset fuzzy neural network to optimize the preset fuzzy neural network;

[0046] Repeat the iterative correction steps until the actual correction amplitude meets the preset dynamic balance conditions.

[0047] Thirdly, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and for running on the processor, wherein the processor executes the computer program to implement the control method for dynamic balancing of a grinding machine as described in the first aspect.

[0048] In a fourth aspect, a computer readable storage medium is provided, and a computer program is stored on the computer readable storage medium, and the computer program is executed by a processor to implement the control method for dynamic balancing of a grinding machine according to the first aspect.

[0049] In a fifth aspect, a computer program product is provided, and the computer program product comprises a computer program, and the computer program is executed by a processor to implement the control method for dynamic balancing of a grinding machine according to the first aspect.

[0050] On the basis of common sense in the art, the above-mentioned preferred conditions can be combined arbitrarily, that is, to obtain each preferred example of the present disclosure.

[0051] The positive progress effect of the present disclosure is that through the adaptive variable step size optimization strategy optimized by the fuzzy neural network, the balance accuracy of the rotating mechanical structure (such as the grinding unit in the grinding machine) is dynamically improved through multiple rounds of fine adjustment, which overcomes the problems of slow convergence or insufficient accuracy of traditional coordinate wheel changing, and the defects of the influence coefficient method depending on the previous calibration value, and has the significant advantages of fast convergence speed and high final accuracy, which meets the demand of high-precision working conditions. BRIEF DESCRIPTION OF DRAWINGS

[0052] Figure 1 A flowchart of a control method for dynamic balancing of a grinding machine is provided for an exemplary embodiment of the present disclosure.

[0053] Figure 2 A flowchart of step S101 in a control method for dynamic balancing of a grinding machine is provided for an exemplary embodiment of the present disclosure.

[0054] Figure 3 A flowchart of step S1013 in a control method for dynamic balancing of a grinding machine is provided for an exemplary embodiment of the present disclosure.

[0055] Figure 4 A structural schematic diagram of a dynamic balancing assembly in a control method for dynamic balancing of a grinding machine is provided for an exemplary embodiment of the present disclosure.

[0056] Figure 5 A flowchart of step S1043 in a control method for dynamic balancing of a grinding machine is provided for an exemplary embodiment of the present disclosure.

[0057] Figure 6 A flowchart of step S10432 in a control method for dynamic balancing of a grinding machine is provided for an exemplary embodiment of the present disclosure.

[0058] Figure 7 A module schematic diagram of a control system for dynamic balancing of a grinding machine is provided for an exemplary embodiment of the present disclosure.

[0059] Figure 8 A hardware structure schematic diagram of an electronic device is provided for an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0060] The present disclosure is further illustrated by the following examples without thereby limiting the present disclosure to the described examples.

[0061] The prefix words such as "first", "second" are used in the embodiments of the present disclosure only to distinguish different description objects, and do not have the limited effect on the position, order, priority, quantity or content of the described objects. The use of ordinal words such as ordinal words in the embodiments of the present disclosure does not constitute a limitation on the described objects, and the description of the described objects should be referred to the description of the context in the claims or embodiments, and should not constitute redundant limitation because of the use of such prefix words. In addition, in the description of the embodiments, unless otherwise stated, the meaning of "a plurality of" is two or more.

[0062] Embodiment 1

[0063] Figure 1 A flow chart of a control method of a grinding machine dynamic balancing is provided for an exemplary embodiment of the present disclosure, the grinding machine includes a grinding unit, the grinding unit includes a dynamic balancing assembly, the dynamic balancing assembly is used to generate a correction centrifugal force to dynamically balance the grinding unit, the control method includes:

[0064] S101, obtaining an original amplitude of the grinding unit;

[0065] S102, controlling the dynamic balancing assembly to perform a first correction movement based on a preset initial angle step, and obtaining a first correction amplitude of the grinding unit after the first correction movement;

[0066] Specifically, the preset fuzzy neural network has an initial angle step matched with the grinding unit, and the dynamic balancing assembly performs a preset first correction movement at the initial angle step. For example, a coordinate is established based on the central axis of the grinding assembly, and the initial angle step represents a tentative first correction movement of the dynamic balancing assembly in the downward direction along the longitudinal central axis of the grinding assembly, so as to determine the starting direction of the iterative correction cycle.

[0067] S103, determining an initial phase value of the correction centrifugal force based on the first correction amplitude and the original amplitude;

[0068] Specifically, by comparing the size between the original amplitude before and after the movement and the first correction amplitude, if the first correction amplitude after the first movement is reduced compared to the original amplitude, it indicates that the corresponding correction centrifugal force of the dynamic balancing assembly after the first movement is in the target phase region of the offset force of the grinding assembly, and the current phase value after the first correction can be selected as the initial phase value. Correspondingly, if the first correction amplitude after the first movement is increased compared to the original amplitude, it indicates that the corresponding correction centrifugal force of the dynamic balancing assembly after the first movement is in the opposite region of the target phase region, and the reverse phase value of the current phase value after the first correction can be selected as the initial phase value. At the same time, the amplitude deviation between the original amplitude and the first correction amplitude is also used as part of the amplitude variation trend parameter for subsequent iterative correction steps.

[0069] Through the first tentative movement, the initial phase value is adaptively and quickly and accurately obtained for the subsequent iterative correction cycle.

[0070] S104, an iterative correction cycle step is executed, including:

[0071] S1041, the amplitude variation trend parameter of the grinding unit is input as input, a step correction coefficient is generated by inputting a preset fuzzy neural network;

[0072] Specifically, the non-linear mapping and adaptive learning ability of the fuzzy neural network (Fuzzy Neural Network, FNN) are used to efficiently, accurately and intelligently output the adjustment amplitude of the displacement information of the dynamic balancing assembly. In each iteration stage, an adaptive change step optimization strategy based on fuzzy neural network control is adopted to quickly converge when far away from the balance point and ensure accuracy when close to the balance point, and finally obtain the real-time unbalance and the required correction force.

[0073] The step correction coefficient output by the preset fuzzy neural network is fed back through the understanding of the amplitude variation trend of the grinding unit. For example, the amplitude variation trend of the grinding unit in the current iterative correction step includes the correction amplitude deviation between the updated correction amplitude after the last iteration and the preset target balance amplitude, and the change amount of the amplitude deviation value in the current iteration, wherein the change amount of the amplitude deviation value can be the difference between the correction amplitude deviations of the last two times.

[0074] The preset fuzzy neural network adaptively scales and updates the step, synchronously and accurately corrects the correction centrifugal force of the dynamic balancing assembly, and compared with the direct output of the step, can effectively avoid the shock caused by the too large step, ensure that the final correction action is smooth, stable and impact-free, and improve the convergence stability and correction accuracy.

[0075] In one embodiment, the amplitude variation trend includes the amplitude deviation, and the amplitude deviation calculated after the movement to the position and the cumulative amplitude deviation change amount As the input amount of the preset fuzzy neural network, the calculation formula is as follows:

[0076] ;

[0077] ;

[0078] wherein, is the current extracted updated correction amplitude, is the target balance amplitude, is the current correction amplitude deviation, is the previous correction amplitude deviation, and k is the kth step optimization iteration time.

[0079] The deviation change amount reflects the change trend and speed of the amplitude deviation. When the change amount is negative, under the condition that the correction amplitude of the correction centrifugal force is unchanged, it is indicated that the center of gravity of the grinding unit as a whole is approaching the target position of dynamic balance. If the amplitude deviation is always increasing and the deviation change amount is positive, it is indicated that the center of gravity of the grinding unit as a whole is moving away from the target position of dynamic balance.

[0080] S1042, obtaining the correction parameter of the correction centrifugal force and the expected correction amplitude of the grinding unit based on the step correction coefficient;

[0081] In one embodiment, according to the output of the fuzzy neural network, the step is adjusted to be:

[0082] ;

[0083] wherein, is the previous step, indicates the adaptive correction coefficient of the step; when the error is large and the descending speed is fast, the FNN output is large to speed up the search; when the error converges slowly or oscillates, the FNN output is small to reduce the step and improve the stability.

[0084] S1043, adjusting the correction centrifugal force based on the correction parameter, and obtaining the actual correction amplitude of the grinding unit after correction;

[0085] Specifically, according to the updated step and the phase controlled to the position in the previous step, the theoretical phase value of the double-weight counterweight is determined. At the same time, the dynamic balance assembly is driven to move to the corrected position based on the step correction coefficient, and the actual phase value corresponding to the actual change of the dynamic balance assembly is measured .

[0086] S1044, obtaining an amplitude correction error based on the expected correction amplitude and the actual correction amplitude, and feeding back the amplitude correction error to the preset fuzzy neural network to optimize the preset fuzzy neural network;

[0087] On the basis of the foregoing steps, the actual correction amplitude is compared with the theoretical correction amplitude to obtain a correction amplitude error as a compensation signal fed back to the preset fuzzy neural network, so as to realize adaptive updating of the step length, and the formula is as follows:

[0088] ;

[0089] Specifically, through continuous feedback of the correction amplitude error value in multiple rounds, an iterative loop of the preset fuzzy neural network is constructed, so that the preset fuzzy neural network adaptively optimizes internal parameters, outputs a more reasonable step length correction coefficient, and makes the grinding unit finally reach dynamic balance.

[0090] S105, repeating the iterative correction step until the actual correction amplitude meets the preset dynamic balance condition.

[0091] In the scheme, through the adaptive variable step length optimization strategy optimized by the fuzzy neural network, the balance precision of the rotating mechanical structure (such as the grinding unit in the grinding machine) is dynamically improved through multiple rounds of fine adjustment, which overcomes the problems of slow convergence or insufficient precision of the traditional coordinate wheel exchange and the defects of the influence coefficient method depending on the previous calibration value, and has the significant advantages of fast convergence speed and high final precision, which meets the demand of high-precision working conditions.

[0092] As an implementable way, as shown in Figure 2 , the step S101 includes:

[0093] S1011, obtaining an original vibration signal of a grinding unit;

[0094] Specifically, the grinding unit can adopt a grinding wheel, and a vibration sensor is arranged near the grinding wheel, such as near the grinding wheel spindle box, for collecting real-time vibration signals of the grinding wheel, and when the real-time vibration signals represent that the grinding wheel is in an unbalanced state, such as when the real-time vibration signals are greater than a preset vibration balance threshold, the current real-time vibration signals are taken as the original vibration signals for subsequent dynamic balance calculation.

[0095] In one embodiment, the original vibration signal analyzed is a single frequency complex exponential signal, and the sequence of the signal after discretization is set as , and the specific expression of the signal is:

[0096] ;

[0097] wherein, is an original vibration amplitude; is an original vibration frequency; original vibration phase; sampling frequency; signal length.

[0098] S1012, add a Hanning window to the original vibration signal and perform Fast Fourier Transform (FFT) to obtain the corresponding vibration information spectrum;

[0099] Specifically, FFT can decompose complex vibration signals mixed together and difficult to distinguish on time domain waveform into independent frequency components, providing a quantitative analysis method. The specific amplitude of vibration can be directly read from the vibration information spectrum obtained by transformation, and the severity of the current grinding wheel imbalance state can be judged according to the amplitude.

[0100] By adding a Hanning window, the frequency spectrum leakage caused by signal truncation can be effectively suppressed, that is, only a signal of limited length is analyzed, and the signal energy is more concentrated in the main lobe, laying a foundation for subsequent accurate correction.

[0101] In one embodiment, a Hanning window is added to the vibration signal sequence;

[0102] wherein the Hanning window of signal length N The calculation formula is as follows:

[0103] ;

[0104] wherein, is the N-order rectangular window truncated in the spectrum analysis of Fast Fourier.

[0105] The original vibration signal is added with a Hanning window and Fast Fourier Transform is performed, and the first half of the spectrum obtained after transformation is taken; the amplitude of this part of the spectrum is scaled and compensated, and the scaling coefficient is 4 / N, wherein N is the number of sampling points (i.e. signal length), to eliminate the amplitude attenuation introduced by the Hanning window. The calculation formula of the Fourier transform after windowing is:

[0106] ;

[0107] ;

[0108] wherein k is the frequency index.

[0109] Since the input signal is a real signal, the Fourier transform result has conjugate symmetry, only the first half of the FFT result is extracted, that is, X(1:N / 2) is taken, where N is the length of the signal (consistent with the number of sampling points). The Fourier transform result is normalized to restore its correct amplitude, divided by N for amplitude normalization; due to the signal amplitude attenuation caused by the Hanning window processing, the attenuation factor is 2 (the maximum value of the Hanning window is 0.5), in order to compensate for the amplitude attenuation caused by the window function, an additional coefficient 4 is multiplied. The energy loss caused by the window function is corrected by multiplying twice, so that the final spectral amplitude is restored to the true value.

[0110] At the same time, by adding a Hanning window function to the signal and increasing the length of the window sequence, the amplitude-frequency phase information can be accurately extracted. The obtained frequency spectrum result not only preserves the true amplitude of the signal, but also eliminates the adverse effects of the window function, and can provide more accurate frequency domain information for subsequent signal analysis and processing.

[0111] S1013, obtaining an original spectral component parameter set corresponding to the real vibration frequency from the vibration information spectrum;

[0112] The original spectral component parameter set includes an original frequency, an original phase value, and an original amplitude value.

[0113] Specifically, the vibration information spectrum obtained by FFT transformation is a discrete spectrum that is discontinuous on the frequency axis. The interval between each frequency point in the spectrum is the frequency resolution, which is determined by the sampling rate and the number of FFT points. Since the real vibration frequency of the grinding wheel is usually difficult to coincide with the points in the vibration information spectrum, that is, the real vibration frequency appears between two discrete frequency points in the vibration information spectrum, resulting in a fence effect, that is, the energy of the real frequency leaks to the adjacent discrete frequency points, making the peak value on the spectrum appear lower than the actual peak value and the position deviates. And the spectrum leakage phenomenon, that is, when the sampling time is not an integer multiple of the real frequency period, the energy will leak to a wider frequency band.

[0114] In one embodiment, the target spectral component parameter set of the real vibration frequency is obtained from the vibration information spectrum by a spectrum interpolation method. Specifically, a three-point interpolation spectrum correction algorithm is used, three points refer to selecting three adjacent feature points on the spectrum curve, such as the maximum peak point and its left and right two points, by fitting a function with the known point values, the target spectral component parameter set of the real frequency is calculated to correct the fence effect and spectrum leakage phenomenon mentioned above.

[0115] The original frequency is used to calibrate the vibration frequency causing the imbalance, and the real vibration frequency signal can be separated from various non-synchronous vibration signals, the signal-to-noise ratio is improved, and the accuracy of subsequent amplitude and phase analysis is ensured. The original amplitude value represents the size of the imbalance, and is used to establish the mathematical relationship between the vibration amplitude of the grinding wheel and the dynamic balancing assembly. The original phase value is used to represent the angular position of the imbalance, and provides guidance for adjusting the position of the dynamic balancing assembly.

[0116] The windowed FFT and three-line interpolation method are introduced, which can effectively suppress the FFT fence effect and spectrum energy leakage, and significantly improve the accuracy of amplitude and phase estimation. At the same time, the spectrum correction technology is applied to the online dynamic balancing of rotating machinery, and the phase measurement error caused by non-integer period sampling is solved.

[0117] The original imbalance vector of the grinding unit is obtained based on the original frequency, the original phase value and the original amplitude value;

[0118] Specifically, the imbalance state of the grinding wheel is mathematically described as a complete two-dimensional vector by the original frequency, the original phase value and the original amplitude value, as the original imbalance vector. Among them, the original amplitude value is the modulus of the original imbalance vector, and the original phase value is the argument of the original imbalance vector.

[0119] As an implementable way, as shown in Figure 3 , the step S1013 comprises:

[0120] S10131, based on a preset search interval, determining a main peak spectrum line with the largest amplitude and its left and right two adjacent auxiliary spectrum lines;

[0121] Specifically, according to the upper limit and lower limit interval range of the preset search interval, the vibration information spectrum is converted into the corresponding peak spectrum line serial number .

[0122] ;

[0123] ;

[0124] Among them, , is the upper limit and lower limit of the specified search interval, is the sampling frequency;

[0125] In the set frequency search interval, the maximum spectrum line amplitude and the corresponding index are found as the rough frequency position, and the formula is as follows:

[0126] ;

[0127] Among them, is the main peak spectrum line serial number.

[0128] S10132, acquire a first amplitude value of a main peak spectrum line, a second amplitude value of a left auxiliary spectrum line, and a third amplitude value of a right auxiliary spectrum line;

[0129] wherein the target frequency deviation value is a frequency deviation value between the main peak spectrum line and the true vibration frequency of the grinding unit;

[0130] Specifically, the main spectrum line and two adjacent auxiliary spectrum lines are used to construct a parabolic model, and the frequency offset is calculated through a ratio relationship to achieve accurate frequency estimation; when the interpolation interval is less than 1 / 4 of the main lobe width, the auxiliary spectrum line is located in the main lobe or the first side lobe, which can ensure that the interpolated spectrum line has a high signal-to-noise ratio. Wherein, the point with a larger amplitude in the left and right adjacent points of the peak spectrum line is determined to determine the corresponding side of the true spectrum peak deviation.

[0131] S10133, based on the first amplitude value, the second amplitude value, and the third amplitude value, calculate a target frequency deviation for correcting the main peak spectrum line;

[0132] Specifically, the maximum spectrum line and the spectrum line with a higher amplitude among its adjacent spectrum lines are used to calculate the target frequency deviation δ of the true spectrum peak relative to the maximum spectrum line sequence number , and the formula is as follows:

[0133] ,

[0134] The above formula is derived based on the characteristics of the Hanning window spectrum, and can accurately estimate the frequency deviation.

[0135] S10134, based on the target frequency deviation, correct the first amplitude value corresponding to the main peak spectrum line to obtain an original amplitude value.

[0136] After obtaining the target frequency deviation δ, the corrected amplitude, phase, and frequency are calculated based on the target frequency deviation, and the specific process is as follows:

[0137] Due to the fence effect, the rough amplitude is usually smaller than the true amplitude. The rough amplitude is compensated by the frequency deviation to obtain the corrected accurate amplitude. The formula of the corrected original amplitude A is:

[0138] ;

[0139] wherein N is the signal length, is the peak spectrum line sequence number, is the target frequency deviation. ​

[0140] Similarly, the deviation of frequency will also cause the deviation of phase. The formula uses the frequency deviation to linearly compensate the coarse phase to obtain the accurate phase value, and the corrected original phase The formula is:

[0141] ;

[0142] Wherein, N is the signal length, is the peak line number, is the target frequency deviation.

[0143] With the coarse frequency and the calculated fine frequency deviation, a more accurate frequency value than the frequency spectrum resolution is obtained, and the corrected original frequency The formula is:

[0144] ;

[0145] Wherein, N is the signal length, is the sampling frequency, is the frequency resolution, denotes the normalized frequency, is the peak line number.

[0146] As an implementable way, as shown in Figure 4 The dynamic balancing assembly 200 includes a first counterweight 210 and a second counterweight 220 coaxially arranged on the main shaft of the grinding unit, the centrifugal force generated by the first counterweight 210 and the second counterweight 220 is the correction centrifugal force, and the correction centrifugal force includes a centrifugal force phase and a centrifugal force amplitude;

[0147] As shown in Figure 5 The step of adjusting the correction centrifugal force based on the correction parameter in step S1043 includes:

[0148] S10431, based on the correction parameter, the phase value variable and / or the amplitude variable corresponding to the current centrifugal force phase and / or the centrifugal force amplitude are obtained;

[0149] In one embodiment, based on the correction parameter, the phase value variable is applied on the centrifugal force phase of the last iteration, and based on the phase value variable, the first phase value of the first counterweight 210 and the second phase value of the second counterweight 220 are determined; based on the correction parameter, the amplitude variable is applied on the centrifugal force amplitude of the last iteration, and based on the amplitude variable, the third phase value of the first counterweight 210 and the fourth phase variable of the second counterweight 220 are determined.

[0150] S10432, synchronously control the first counterweight and the second counterweight to make corresponding phase adjustment based on the phase variable and / or the amplitude variable.

[0151] Specifically, the first correction phase value of the first counterweight 210 is determined by the first phase value, and the second correction phase value of the second counterweight 220 is determined by the second phase value; or the first correction phase value of the first counterweight 210 is determined by the third phase value, and the second correction phase value of the second counterweight 220 is determined by the fourth phase value; or the first correction phase value of the first counterweight 210 is determined by the first phase value and the third phase value, and the second correction phase value of the second counterweight 220 is determined by the second phase value and the fourth phase value.

[0152] In this scheme, the correction centrifugal force is mapped as the phase value adjustment of the two counterweights; when the phase is controlled, the two counterweights are adjusted in the same direction as a whole. When the amplitude is controlled, the two counterweights are controlled to adjust in opposite directions. For example, when the two counterweights are adjusted in the same direction, the angle of the centrifugal force becomes smaller, and the corresponding correction centrifugal force increases; when the two counterweights are adjusted in opposite directions, the angle of the centrifugal force becomes larger, and the corresponding correction centrifugal force decreases. The amplitude and phase control of the correction centrifugal force are effectively decoupled, and the complexity of the control algorithm is significantly simplified. The nonlinear mapping relationship between the angle of the centrifugal force and the correction centrifugal force ensures high precision in eliminating small residual imbalance and improves the final dynamic balance effect.

[0153] As an implementable way, as shown in Figure 6 The angle between the first counterweight and the second counterweight is the angle of the centrifugal force, and the step of synchronously controlling the first counterweight and the second counterweight to make corresponding phase adjustment based on the phase variable and / or the amplitude variable includes:

[0154] S10432, synchronously control the first counterweight and the second counterweight to make corresponding phase adjustment based on the phase variable and / or the amplitude variable.

[0155] Specifically, in the phase balance process, there are two cases of phase optimization, and the specific implementation formula is as follows:

[0156] (1) when the counterweight rotates in the direction in which the angle increases,

[0157] ;

[0158] wherein, an angle between the two counterweights at the kth step of optimization, a phase value of the resultant force generated by the two counterweights at the kth step of optimization, a corresponding phase value of the first counterweight at the kth step of optimization, and a corresponding phase value of the second counterweight at the kth step of optimization, a phase value variable of rotation.

[0159] (2) when the counterweights rotate in a direction in which the orientation angle decreases,

[0160] ;

[0161] wherein, an angle between the two counterweights at the (k-1)th step of optimization, a phase value of the resultant force generated by the two counterweights at the (k-1)th step of optimization, a corresponding phase value of the first counterweight at the (k-1)th step of optimization, and a corresponding phase value of the second counterweight at the (k-1)th step of optimization, a phase value variable of rotation.

[0162] Accordingly, a corresponding correction centrifugal force is generated, a dynamic balance correction is completed, and an adjusted actual correction amplitude is obtained.

[0163] The above steps are repeated, the counterweights are continuously driven to move, the vibration amplitude is gradually reduced, and when the amplitude after phase adjustment is no longer reduced, if the amplitude deviation after continuous two times of adjustment both represent that the actual correction amplitude is increasing, i.e., the preset phase dynamic balance condition is met, the amplitude balance phase is entered.

[0164] S104322, in response to the actual correction amplitude meeting the preset phase balance condition, obtaining an amplitude variable corresponding to the centrifugal force phase based on the correction parameter, and synchronously controlling the first counterweight and the second counterweight to perform reverse phase adjustment on the centrifugal force angle.

[0165] After the counterweights reach the expected phase balance position, the amplitude balance phase is entered, and a stability checking mechanism is introduced.

[0166] The amplitude balance process also adopts a fuzzy neural network controller to realize variable step length optimization. After an output corresponding step length value is obtained, the two counterweights are driven to rotate at the same speed in opposite directions to adjust the centrifugal force angle , so as to adjust the size of the resultant force vector, and achieve the purpose of amplitude balance.

[0167] There are two cases of amplitude balance, and the specific implementation formulas are as follows:

[0168] (1) the angle between the weight counterweight is increased,

[0169] ;

[0170] (2) the weight counterweight is rotated in the direction of the angle reduction,

[0171] ;

[0172] Repeat the above optimization steps until the actual correction amplitude after adjustment meets the preset dynamic balance condition, such as the actual correction amplitude is less than the preset first dynamic balance amplitude threshold, then the balancing process is ended.

[0173] In the scheme, by setting an independent phase convergence condition, the two-dimensional vector optimization problem is reduced to one-dimensional phase optimization and amplitude optimization, the correction direction is first aligned, and then the correction amount is accurately adjusted. It ensures that the grinding unit can approach the target balance point with the least number of iterations and the fastest speed, avoids large amplitude adjustment in the case of inaccurate phase, and prevents oscillation or divergence in the iteration process.

[0174] As an implementable way, step S104322 further includes:

[0175] In response to the amplitude variable being lower than the preset change threshold, and the step length correction coefficient output by the preset fuzzy neural network this time representing a first update direction of the current phase adjustment being opposite to the reference direction of the last phase adjustment, the step length correction coefficient is output twice again, and a second update direction and a third update direction of the phase adjustment are obtained.

[0176] In response to the second update direction and the third update direction being inconsistent, the first weight counterweight and the second weight counterweight are controlled to adjust the centrifugal force angle in the reference direction based on the step length correction coefficient.

[0177] Specifically, since the amplitude balance stage has a small amplitude change range, the preset fuzzy neural network is prone to misadjustment and local optimal solution. Therefore, a stability checking mechanism is introduced, that is, when the amplitude deviation change is lower than the preset change threshold, and the amplitude deviation change represents a change in the direction of the counterweight disc rotation, the two-direction optimization is performed, that is, the adjustment of the dynamic balancing assembly is performed first, and the stability of the current iteration correction step is checked through the two output step length adjustment parameters of the preset fuzzy neural network. If the directions of the calculation results of the two-direction optimization are inconsistent, it indicates that the current iteration correction step lacks stability, and the reference direction is maintained to adjust the two weight counterweights to correct the centrifugal force amplitude, thereby improving the adjustment stability and convergence reliability of the entire iteration correction step.

[0178] In the present solution, a stability verification step is provided to accurately identify the zero-crossing point where the iterative process enters the critical flat region that may generate oscillation from the efficient gradient descent phase. Active intervention is made when the direction reversal is detected to improve the stability of the grinding unit.

[0179] The control method for dynamic balancing of a grinding machine provided by the embodiment effectively suppresses the FFT fence effect and spectrum energy leakage through FFT and three-line interpolation method, and significantly improves the accuracy of amplitude and phase estimation. Through the adaptive variable step size optimization strategy optimized by fuzzy neural network, the balance accuracy of rotating mechanical structure (such as grinding unit in grinding machine) is dynamically improved through multiple rounds of fine adjustment, which overcomes the problems of slow convergence or insufficient accuracy of traditional coordinate wheel exchange, and the defects of influence coefficient method depending on previous calibration values, and has the significant advantages of fast convergence speed and high final accuracy, meeting the demand of high-precision working conditions.

[0180] Embodiment 2

[0181] Corresponding to the foregoing embodiment of the control method for dynamic balancing of a grinding machine, the present disclosure also provides an embodiment of a control system for dynamic balancing of a grinding machine.

[0182] Figure 7 A module schematic diagram of a control system for dynamic balancing of a grinding machine is provided for an exemplary embodiment of the present disclosure, the grinding machine comprising a grinding unit, the grinding unit comprising a dynamic balancing assembly for generating a correction centrifugal force for dynamically balancing the grinding unit, the control system for dynamic balancing of the grinding machine 100 comprising an original amplitude acquisition module 101, a first correction module 102, a phase determination module 103 and an iterative correction module 104;

[0183] The original amplitude acquisition module 101 is configured to acquire an original amplitude of the grinding unit.

[0184] The first correction module 102 is configured to control the dynamic balancing assembly to perform a first correction movement based on a preset initial angle step size, and acquire a first correction amplitude of the grinding unit after the first correction movement.

[0185] The phase determination module 103 is configured to determine an initial phase value of the correction centrifugal force based on the first correction amplitude and the original amplitude.

[0186] The iterative correction module 104 is configured to perform an iterative correction step, comprising:

[0187] The amplitude variation trend parameter of the grinding unit is input into a preset fuzzy neural network as an input to generate a step correction coefficient.

[0188] a correction parameter of the correction centrifugal force is obtained based on the step correction coefficient, and a desired correction amplitude of the grinding unit;

[0189] the correction centrifugal force is adjusted based on the correction parameter, and an actual correction amplitude of the grinding unit after correction is obtained;

[0190] an amplitude correction error is obtained based on the desired correction amplitude and the actual correction amplitude, and the amplitude correction error is fed back to the preset fuzzy neural network to optimize the preset fuzzy neural network;

[0191] the iterative correction step is repeated until the actual correction amplitude meets a preset dynamic balance condition.

[0192] As an implementable manner, the original amplitude obtaining module 101 comprises an original vibration obtaining unit, a Fourier transform unit, an interpolation correction unit, and a vector obtaining unit.

[0193] The original vibration obtaining unit is configured to obtain an original vibration signal of the grinding unit.

[0194] The Fourier transform unit is configured to add a Hanning window to the original vibration signal and perform fast Fourier transform to obtain corresponding vibration information spectrum.

[0195] The interpolation correction unit is configured to process the vibration information spectrum to obtain an original spectrum component parameter set corresponding to a true vibration frequency.

[0196] The original spectrum component parameter set comprises an original amplitude.

[0197] The vector obtaining unit is configured to obtain an original unbalance vector of the grinding unit based on the original frequency, an original phase value, and the original amplitude.

[0198] As an implementable manner, the original amplitude obtaining module 101 further comprises a spectrum line determination unit, an amplitude obtaining unit, a frequency deviation calculation unit, and a calculation correction unit.

[0199] The spectrum line determination unit is configured to determine a main peak spectrum line with a maximum amplitude and two adjacent auxiliary spectrum lines thereof on the left and right based on a preset search interval.

[0200] The amplitude obtaining unit is configured to obtain a first amplitude of the main peak spectrum line, a second amplitude of the left auxiliary spectrum line, and a third amplitude of the right auxiliary spectrum line.

[0201] The frequency deviation calculation unit is configured to calculate a target frequency deviation for correcting the main peak spectrum line based on the first amplitude, the second amplitude, and the third amplitude.

[0202] The computing correction unit is configured to correct a first amplitude corresponding to a main peak value of the spectrum based on a target frequency deviation, and to obtain an original amplitude.

[0203] As an implementable manner, the dynamic balancing assembly comprises a first counterweight and a second counterweight coaxially arranged on a spindle of the grinding unit, a resultant centrifugal force generated by the first counterweight and the second counterweight is the correction centrifugal force, and the correction centrifugal force comprises a centrifugal force phase and a centrifugal force amplitude.

[0204] The counterweight moving unit is further configured to obtain a phase variable and / or an amplitude variable corresponding to the centrifugal force phase and / or the centrifugal force amplitude based on the correction parameter;

[0205] The counterweight moving unit is further configured to synchronously control the first counterweight and the second counterweight to perform corresponding phase adjustment based on the phase variable and / or the amplitude variable.

[0206] As an implementable manner, an included angle between the first counterweight and the second counterweight is a centrifugal force included angle.

[0207] The iterative correction module 104 is further configured to obtain a phase variable corresponding to the centrifugal force phase based on the correction parameter, synchronously control the first counterweight and the second counterweight to perform same-direction phase adjustment, and maintain the centrifugal force included angle.

[0208] The iterative correction module 104 is further configured to, in response to the actual correction amplitude meeting a preset phase balance condition, obtain an amplitude variable corresponding to the centrifugal force phase based on the correction parameter, and synchronously control the first counterweight and the second counterweight to perform opposite-direction phase adjustment on the centrifugal force included angle.

[0209] As an implementable manner, the iterative correction module 104 further comprises a stability verification unit.

[0210] The stability verification unit is configured to, in response to the amplitude variable being lower than a preset change threshold, and the step correction coefficient output by the preset fuzzy neural network at this time being opposite to a reference direction of a previous phase adjustment, output the step correction coefficient twice again, and obtain a second update direction and a third update direction of the phase adjustment.

[0211] The counterweight moving unit is further configured to, in response to the second update direction and the third update direction being inconsistent, control the first counterweight and the second counterweight to adjust the centrifugal force included angle in the reference direction based on the step correction coefficient.

[0212] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts are referred to the part of the method embodiments. The system embodiments described above are only illustrative, wherein the units described as separate components can or can not be physically separated, and the components of the units can or can not be physical units, i.e., can be located in one place or distributed on multiple network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the present disclosure.

[0213] Embodiment 3

[0214] Figure 8 A structural schematic diagram of an electronic device is shown for an example embodiment of the present disclosure, which includes a memory, a processor, and a computer program stored on the memory and used to run on the processor, and the processor implements the control method of the dynamic balancing of the grinding machine of any of the above embodiments when executing the computer program. Figure 8 The electronic device 90 shown is only an example and should not limit the functions and use range of the embodiments of the present disclosure.

[0215] As shown in Figure 8 The electronic device 90 can be in the form of a general computing device, for example, it can be a server device. The components of the electronic device 90 can include but are not limited to: the above-mentioned at least one processor 91, the above-mentioned at least one memory 92, a bus 93 connecting different system components including the memory 92 and the processor 91.

[0216] The bus 93 includes a data bus, an address bus, and a control bus.

[0217] The memory 92 can include volatile memory, such as a random access memory (RAM) 921 and / or a cache memory 922, and can further include a read-only memory (ROM) 923.

[0218] The memory 92 can further include a program tool 925 (or utility tool) having a set of (at least one) program modules 924, such as an operating system, one or more application programs, other program modules, and program data, each of which or some combination of which can include implementation of a network environment.

[0219] The processor 91 performs various function applications and data processing by running the computer program stored in the memory 92, such as the control method of the dynamic balancing of the grinding machine provided by any of the above embodiments.

[0220] The electronic device 90 can also communicate with one or more external devices 94 such as a keyboard or a pointing device, among others. This communication can occur via Input / Output (I / O) interface 95. Still yet, the electronic device 90 can communicate with one or more networks, such as a local area network (LAN), a wide area network (WAN), and / or the Internet, through network adapter 96. As depicted, network adapter 96 communicates with the other components of the electronic device 90 through bus 93. It should be appreciated that although not shown, other hardware and / or software modules could be used in conjunction with the electronic device 90. Such as, but not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data archival storage systems, etc.

[0221] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the foregoing detailed description, such division is merely exemplary and not mandatory. Indeed, according to embodiments of the present disclosure, features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, features and functions of one unit / module described above can be further divided into units / modules embodied by multiple units / modules.

[0222] Embodiment 4

[0223] The embodiments of the present disclosure further provide a computer readable storage medium, which has stored thereon a computer program. The program is executed by a processor to implement the control method for dynamic balancing of a grinding machine according to any one of the above embodiments.

[0224] More specifically, the readable storage medium can include, but is not limited to, a portable disc, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0225] Embodiment 5

[0226] The embodiments of the present disclosure further provide a computer program product, which comprises a computer program. The computer program is executed by a processor to implement the control method for dynamic balancing of a grinding machine according to any one of the above embodiments.

[0227] The program code for carrying out the computer program product of the present disclosure can be written in any combination of one or more programming languages, and the program code can be executed entirely on the user device, partly on the user device, as a stand-alone software package, partly on the user device and partly on a remote device, or entirely on a remote device.

[0228] Although the specific embodiments of the present disclosure are described above, those skilled in the art should understand that this is only an example, and the protection scope of the present disclosure is defined by the appended claims. Those skilled in the art can make various changes or modifications to the embodiments without departing from the principles and essence of the present disclosure, and these changes and modifications all fall within the protection scope of the present disclosure.

Claims

1. A control method of dynamic balancing of a grinding machine, characterized by, The grinding machine comprises a grinding unit, the grinding unit comprises a dynamic balancing assembly for generating a correction centrifugal force for dynamically balancing the grinding unit, the control method comprises: obtaining an original amplitude of the grinding unit; controlling the dynamic balancing assembly to perform a first correction movement based on a preset initial angle step, and obtaining a first correction amplitude of the grinding unit after the first correction movement; determining an initial phase value of the correction centrifugal force based on the first correction amplitude and the original amplitude; performing an iterative correction step, comprising: inputting an amplitude change trend parameter of the grinding unit into a preset fuzzy neural network to generate a step correction coefficient; obtaining a correction parameter of the correction centrifugal force based on the step correction coefficient, and an expected correction amplitude of the grinding unit; adjusting the correction centrifugal force based on the correction parameter, and obtaining an actual correction amplitude of the grinding unit after the correction; obtaining an amplitude correction error based on the expected correction amplitude and the actual correction amplitude, and feeding back the amplitude correction error to the preset fuzzy neural network to optimize the preset fuzzy neural network; repeating the iterative correction step until the actual correction amplitude meets a preset dynamic balancing condition.

2. The control method of dynamic balancing of a grinding machine according to claim 1, characterized in that, The step of obtaining the original amplitude of the grinding unit comprises: obtaining an original vibration signal of the grinding unit; adding a Hanning window to the original vibration signal and performing a fast Fourier transform to obtain a corresponding vibration information spectrum; processing an original spectrum component parameter set corresponding to a real vibration frequency from the vibration information spectrum; wherein the original spectrum component parameter set comprises the original amplitude.

3. The control method of claim 2, wherein, The step of processing an original spectrum component parameter set corresponding to a real vibration frequency from the vibration information spectrum comprises: determining a main peak spectrum line with the largest amplitude and its left and right two adjacent auxiliary spectrum lines based on a preset search interval; obtaining a first amplitude of the main peak spectrum line, a second amplitude of the left auxiliary spectrum line, and a third amplitude of the right auxiliary spectrum line; calculating a target frequency deviation for correcting the main peak spectrum line based on the first amplitude, the second amplitude, and the third amplitude; calculating the original amplitude based on the first amplitude corresponding to the main peak spectrum line corrected based on the target frequency deviation.

4. The control method of claim 1, wherein, The dynamic balancing assembly comprises a first counterweight and a second counterweight coaxially arranged on a main shaft of the grinding unit, a resultant centrifugal force generated by the first counterweight and the second counterweight is the correction centrifugal force, and the correction centrifugal force comprises a centrifugal force phase and a centrifugal force amplitude; The step of adjusting the correction centrifugal force based on the correction parameter comprises: obtaining a phase value variable and / or an amplitude variable corresponding to the current centrifugal force phase and / or the centrifugal force amplitude based on the correction parameter; synchronously controlling the first counterweight and the second counterweight to perform corresponding phase adjustment based on the phase value variable and / or the amplitude variable.

5. The control method of claim 4, wherein, An included angle between the first counterweight and the second counterweight is a centrifugal force included angle, and the step of synchronously controlling the first counterweight and the second counterweight to perform corresponding phase adjustment based on the phase value variable and / or the amplitude value variable comprises: Firstly, a phase value variable corresponding to the current centrifugal force phase is obtained based on the correction parameter, synchronous control is performed on the first counterweight and the second counterweight to perform the same direction phase adjustment, and the centrifugal force included angle is maintained; In response to the actual correction amplitude meeting a preset phase balance condition, an amplitude value variable corresponding to the current centrifugal force phase is obtained based on the correction parameter, and the first counterweight and the second counterweight are synchronously controlled to perform reverse phase adjustment on the centrifugal force included angle.

6. The control method of claim 5, wherein, In response to the actual correction amplitude meeting a preset phase balance condition, the control method further comprises: In response to the amplitude value variable being lower than a preset change threshold, and the step length correction coefficient output by the preset fuzzy neural network at this time being opposite to the reference direction of the last phase adjustment, the step length correction coefficient is output twice, and a second update direction and a third update direction of the phase adjustment are obtained; In response to the second update direction and the third update direction being inconsistent, the first counterweight and the second counterweight are controlled to adjust the centrifugal force included angle in the reference direction based on the step length correction coefficient.

7. A control system for dynamic balancing of a grinding machine, characterized in that The grinding machine comprises a grinding unit, the grinding unit comprises a dynamic balancing assembly for generating a correction centrifugal force for dynamically balancing the grinding unit, and a control system of the dynamic balancing of the grinding machine comprises an original amplitude acquisition module, a first movement module, a phase determination module and an iterative correction module; The original amplitude acquisition module is configured to acquire an original amplitude of the grinding unit. The first movement module is configured to control the dynamic balancing assembly to perform a first correction movement based on a preset initial angle step, and acquire a first correction amplitude of the grinding unit after the first correction movement. The phase determination module is configured to determine an initial phase value of the correction centrifugal force based on the first correction amplitude and the original amplitude. The iterative correction module is configured to perform an iterative correction step, comprising: inputting an amplitude change trend parameter of the grinding unit into a preset fuzzy neural network to generate a step length correction coefficient; obtaining a correction parameter of the correction centrifugal force and an expected correction amplitude of the grinding unit based on the step length correction coefficient; adjusting the correction centrifugal force based on the correction parameter, and acquiring an actual correction amplitude of the grinding unit after correction; obtaining an amplitude correction error based on the expected correction amplitude and the actual correction amplitude, and feeding back the amplitude correction error to the preset fuzzy neural network to optimize the preset fuzzy neural network; repeating the iterative correction step until the actual correction amplitude meets a preset dynamic balancing condition.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory for running on the processor, characterized in that, The processor implements the control method of the dynamic balancing of the grinding machine according to any one of claims 1 to 6 when executing the computer program. The processor implements the control method of the dynamic balancing of the grinding machine according to any one of claims 1 to 6 when executing the computer program.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the control method for dynamic balancing of a grinding machine according to any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program, when executed by a processor, implements the control method for dynamic balancing of a grinding machine according to any one of claims 1 to 6.

Images