Abrasive belt full life cycle wear prediction and compensation method and system based on acoustic signals

Abstract

The application discloses a kind of based on acoustic signal's abrasive belt full life cycle wear prediction and compensation method and system, including the matching data set of establishing abrasive belt wear parameter and acoustic signal characteristic parameter on time;Based on the matching data set, through neural network, establish simulation abrasive belt wear parameter prediction model;According to the real-time acoustic signal characteristic parameter of abrasive belt polishing operation process, obtain abrasive belt wear parameter prediction value, and then compensate the instantaneous polishing speed of abrasive belt, to realize the stability of control the instantaneous polishing amount of abrasive belt.This application predicts abrasive belt wear parameter through neural network model, compensates the instantaneous polishing speed of abrasive belt based on abrasive belt wear parameter prediction value in the full life cycle of abrasive belt, avoids the problem of underpolishing or overpolishing caused by abrasive belt wear, improves the stability of abrasive belt polishing operation and the profile accuracy of workpiece machining surface.In addition, the application can accurately grasp the timing of abrasive belt replacement by real-time monitoring of abrasive belt wear parameter prediction value.

Description

Technical Field

[0001] This invention relates to the field of grinding technology, and in particular to a method and system for predicting and compensating for the wear of abrasive belts throughout their entire life cycle based on acoustic signals. Background Technology

[0002] Belt polishing primarily utilizes fixtures and auxiliary equipment to create pressure between the workpiece and the surface of the abrasive belt. The relative movement of the abrasive grains and the workpiece surface causes interference, resulting in a cutting process. Belt polishing technology is characterized by high efficiency, flexible processing methods, and high adaptability, and is widely used in the grinding and polishing of various parts.

[0003] Because the abrasive grains of the abrasive belt are also damaged during the polishing process, its removal capacity gradually changes with the extension of polishing time. During polishing operations, the wear level and removal capacity of the abrasive belt directly affect its polishing ability, thus impacting the surface finish and contour accuracy of the workpiece. Existing technologies employ various methods for assessing abrasive belt wear and predicting its lifespan. For example, relying on experienced workers to judge the wear level and replacement timing is highly subjective and its accuracy is limited by the worker's experience. Another method is to judge the abrasive belt wear based on the workpiece's material removal rate and then predict its lifespan. This method requires frequent shutdowns to weigh the workpiece and cannot monitor the abrasive belt wear during the processing.

[0004] Existing technologies often fail to accurately determine the wear level and service life of abrasive belts, and are even less capable of online prediction and compensation for dynamic changes in the abrasive belt's polishing capabilities. This makes it easy for over- or under-wearing to occur during the abrasive belt polishing process, and it is also difficult to determine when to replace the abrasive belt. Summary of the Invention

[0005] In view of the above-mentioned defects or improvement needs of the existing technology, the purpose of this invention is to provide a method and system for predicting and compensating the wear of abrasive belts throughout their entire life cycle based on acoustic signals. By using a Bayesian neural network model, the wear parameters of the abrasive belt are predicted based on the characteristic parameters of acoustic signals during the abrasive belt grinding and polishing process. The dynamic changes in the abrasive belt grinding and polishing capacity are compensated by controlling the abrasive belt grinding and polishing speed, so as to ensure the stability of the abrasive belt grinding and polishing operation.

[0006] To achieve this objective, the present invention adopts the following technical solution:

[0007] This invention provides a method for predicting and compensating for the wear of abrasive belts throughout their entire life cycle based on acoustic signals, comprising:

[0008] S100. Establish a time-matching dataset of abrasive belt wear parameters and acoustic signal characteristic parameters;

[0009] S200. Using the acoustic signal feature parameters in the matching dataset as the input of the neural network and the sand belt wear parameters in the matching dataset as the output of the neural network, the neural network is trained to establish a simulated sand belt wear parameter prediction model.

[0010] S300 During the sand belt grinding and polishing operation, real-time acoustic signal characteristic parameters are acquired and input into the simulated sand belt wear parameter prediction model to obtain the predicted value of sand belt wear parameters.

[0011] S400. Based on the predicted values ​​of the abrasive belt wear parameters, the instantaneous abrasive belt polishing speed is compensated to achieve stable control of the instantaneous abrasive belt polishing amount.

[0012] Further, step S100 includes:

[0013] Acquire the acoustic signal characteristic parameters and polishing amount of the abrasive belt throughout its entire life cycle under various polishing parameters; wherein, the polishing parameters include polishing length, feed speed, polishing speed and polishing pressure, and the acoustic signal characteristic parameters include acoustic frequency, gravity frequency and cumulative power;

[0014] Based on the polishing parameters and the polishing amount, the abrasive belt wear parameters for each sampling time are determined; wherein, the abrasive belt wear parameters include the abrasive belt wear coefficient and the abrasive belt polishing removal coefficient;

[0015] A matching dataset is established based on the acoustic signal characteristic parameters and the sand belt wear parameters at each sampling time.

[0016] Furthermore, obtaining the acoustic signal characteristic parameters of the abrasive belt throughout its entire lifecycle under various grinding and polishing parameters includes:

[0017] During the belt grinding and polishing process, information on the sound frequency and sound power under various grinding and polishing parameters is obtained. The amplitude-frequency relationship is obtained through Fourier transform, and then the gravity frequency and cumulative power are calculated.

[0018] Further, the step of determining the abrasive belt wear parameters at each sampling time based on the abrasion parameters and the abrasion amount includes:

[0019] For abrasive belts with the same grinding and polishing parameters, throughout their entire lifespan, based on the abrasion coefficient of 1 during the stable grinding and polishing period, the removal coefficient of the abrasive belt grinding and polishing system is calculated using the following formula:

[0020]

[0021] Based on the polishing amount at each sampling time and the polishing amount during the polishing stabilization period, calculate the abrasion coefficient of the abrasive belt at each sampling time throughout its entire life cycle:

[0022]

[0023] Among them, P i Let V be the polishing pressure of the i-th grinding point. m ) j For the j-th grinding and polishing speed, V f For the feed rate, k p h(Sta) represents the belt abrasion and polishing removal coefficient. ij k(Sta) represents the polishing amount during the stable period of belt polishing. w t represents the belt wear coefficient during the stable period of belt grinding and polishing. n For the nth sampling time, h(t) n ) ij Let k(t) be the polishing amount corresponding to the nth sampling time in belt polishing at the i-th polishing pressure and j-th polishing speed. n ) w The wear coefficient of the sand belt corresponding to the nth sampling time.

[0024] Furthermore, step S200 includes:

[0025] A three-layer Bayesian neural network model is constructed, including an input layer, an output layer, and a hidden layer. The sound frequency, gravity frequency, and cumulative power in the matching dataset are used as the input vectors of the Bayesian neural network model, and the sanding belt polishing removal coefficient and sanding belt wear coefficient are used as the output vectors. The Bayesian neural network is trained to establish a model for predicting sanding belt wear parameters.

[0026] Specifically, the weights of each input vector are determined based on the feature parameters of each acoustic signal in the matching dataset.

[0027] Further, step S400 includes:

[0028] Based on the predicted wear coefficient of the abrasive belt, compensation is made for the instantaneous abrasion and polishing speed of the abrasive belt, including:

[0029] Real-time control of instantaneous grinding and polishing speed ensures that the product of the predicted abrasion coefficient of the abrasive belt and the instantaneous grinding and polishing speed remains constant.

[0030] Furthermore, the method also includes:

[0031] When the predicted value of the sanding belt wear coefficient meets the sanding belt wear coefficient threshold, a sanding belt replacement alarm is issued.

[0032] This invention also provides a system for predicting and compensating for the wear of abrasive belts throughout their entire life cycle, comprising:

[0033] The training data module is used to create a time-matched dataset of abrasive belt wear parameters and acoustic signal characteristic parameters.

[0034] The neural network training and model determination module is used to train the neural network by using the acoustic signal feature parameters in the matching dataset as the input of the neural network and the sand belt wear parameters in the matching dataset as the output of the neural network, and to establish a simulated sand belt wear parameter prediction model.

[0035] The belt abrasion prediction and compensation module is used to acquire real-time acoustic signal characteristic parameters during belt abrasion and polishing operations and input them into the simulated belt abrasion parameter prediction model to obtain belt abrasion parameter prediction values. It is also used to compensate for the instantaneous abrasion speed of the belt based on the predicted belt abrasion parameter values, so as to achieve stable control of the instantaneous abrasion amount of the belt.

[0036] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned method for predicting and compensating wear throughout the life cycle of abrasive belts.

[0037] The present invention also provides a computer-readable storage medium storing computer instructions that cause the computer to execute the aforementioned method for predicting and compensating wear throughout the life cycle of abrasive belts.

[0038] The beneficial effects of this invention are:

[0039] This invention provides a method and system for predicting and compensating for the wear of abrasive belts throughout their entire lifecycle based on acoustic signals. The method includes: establishing a time-matching dataset of abrasive belt wear parameters and acoustic signal feature parameters; training the neural network using the acoustic signal feature parameters in the matching dataset as input and the abrasive belt wear parameters in the matching dataset as output to establish a simulated abrasive belt wear parameter prediction model; during the abrasive belt grinding and polishing operation, acquiring real-time acoustic signal feature parameters and inputting them into the simulated abrasive belt wear parameter prediction model to obtain predicted abrasive belt wear parameter values; and compensating for the instantaneous grinding and polishing speed of the abrasive belt based on the predicted abrasive belt wear parameter values ​​to achieve stable control of the instantaneous grinding and polishing amount of the abrasive belt.

[0040] This invention uses a neural network model to predict abrasive belt wear parameters. Throughout the abrasive belt's lifespan, the instantaneous grinding and polishing speed is compensated based on these predicted wear parameters, preventing under- or over-grinding caused by belt wear and improving the stability of the abrasive belt polishing operation and the contour accuracy of the workpiece surface. Furthermore, by monitoring the predicted abrasive belt wear parameters, this invention can accurately determine the timing of belt replacement, avoiding under- or over-use of the abrasive belt.

[0041] Additional aspects and advantages of this application will be set forth in part in the description which follows, and will become apparent from the description or may be learned by practice of this application. Attached Figure Description

[0042] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0043] Figure 1 This is a flowchart of the method for predicting and compensating for the wear of abrasive belts throughout their entire life cycle, as described in this invention.

[0044] Figure 2 This is a schematic diagram illustrating the changes in the grinding and polishing capabilities of the abrasive belt throughout its entire lifecycle, as described in this embodiment of the invention.

[0045] Figure 3 This is a schematic diagram of a Bayesian neural network in an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of the abrasive belt full life cycle wear prediction and compensation system in an embodiment of the present invention. Detailed Implementation

[0047] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, the accompanying drawings show only the parts relevant to the present invention, and not all of the structures.

[0048] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined as in the embodiments of this application.

[0049] This application provides a method, system, electronic device, and medium for predicting and compensating for the wear of sanding belts throughout their entire life cycle based on acoustic signals. It utilizes a neural network algorithm to predict the wear parameters of the sanding belt based on acoustic signals, and then compensates for the sanding belt polishing speed to maintain a constant sanding belt polishing amount and ensure the stability of the sanding belt polishing operation.

[0050] In belt polishing operations, the belt wear coefficient and belt polishing removal coefficient are usually used to reflect the belt's polishing ability. Analysis of the polishing process can provide formulas for the belt polishing amount in relation to the belt wear coefficient and the belt polishing removal coefficient:

[0051] dh = k p ·k w ·P·V m·dt;

[0052]

[0053] Where h is the abrasive belt polishing amount, k p k is the belt polishing removal coefficient. w V is the abrasion coefficient of the abrasive belt, P is the abrasion pressure, and V is the abrasion pressure. m V is the polishing speed, t is the polishing time, l is the polishing length, and V is the polishing velocity. f This refers to the feed rate.

[0054] As can be seen from the above formulas, the abrasion belt wear coefficient and the abrasion belt polishing removal coefficient directly affect the polishing amount of the abrasion belt. The abrasion belt polishing removal coefficient is an inherent property of the abrasion belt and is a constant, independent of the degree of abrasion. However, due to differences in the material and arrangement of abrasive grains within the abrasion belt, the abrasion belt polishing coefficient varies. The abrasion belt wear coefficient reflects the degree of abrasion and exhibits a non-linear change during the polishing process. During on-site polishing operations, the polishing sound varies significantly at different stages of the abrasion belt's lifespan. In this application, the acoustic signal is used as a strongly correlated quantity of the abrasion belt wear degree in the algorithm design. Based on a neural network algorithm, the acoustic signal is used to accurately obtain information about the abrasion belt wear degree.

[0055] This application provides a method for predicting and compensating for the full lifecycle wear of abrasive belts based on acoustic signals. See [link to relevant documentation]. Figure 1 The method includes steps S100-S400.

[0056] S100. Establish a time-matching dataset of abrasive belt wear parameters and acoustic signal characteristic parameters.

[0057] Define the abrasive belt's polishing capacity K = k p ·k w During belt abrasion, by controlling various abrasion parameters and monitoring the abrasion volume in real time, the abrasion capacity of the belt can be calculated using a formula, thus determining the changes in abrasion capacity throughout the belt's lifespan. (See also...) Figure 2 It can be seen that the polishing capability of the abrasive belt throughout its entire life cycle can be roughly divided into three stages: In the initial stage, due to the sharpness of the abrasive grains, the polishing capability is relatively strong; in the middle stage, the polishing capability stabilizes; and in the later stage, the abrasive grains gradually become dulled and worn during continuous polishing, leading to a decrease in polishing capability. Since the abrasive belt's polishing removal coefficient is constant, the changing trend of its polishing capability throughout its life cycle also reflects the changing trend of its wear coefficient. (Continue to see...) Figure 2 The value of the polishing capability during the polishing stabilization period is K. sta The abrasion coefficient of the abrasive belt during this period is defined as 1. In this embodiment, the abrasion belt abrasion removal coefficient and the abrasion belt abrasion coefficient throughout the entire life cycle are determined based on the abrasion belt abrasion coefficient during the stable period of abrasion.

[0058] In this embodiment, the polishing amount of the abrasive belt throughout its entire lifecycle under various polishing parameters is first obtained. Then, based on the polishing parameters and polishing amount, the abrasive belt wear parameters for each sampling time are determined. The polishing parameters include polishing length, feed speed, polishing velocity, and polishing pressure. The abrasive belt wear parameters include the abrasive belt wear coefficient and the abrasive belt polishing system removal coefficient. Specifically, for the abrasive belt with the same polishing parameters throughout its entire lifecycle, based on the abrasive belt wear coefficient being 1 during the polishing stability period, the value of the abrasive belt polishing system removal coefficient is calculated using the following formula:

[0059]

[0060] Furthermore, based on the polishing amount during each sampling time and the polishing amount during the polishing stabilization period, the wear coefficient of the abrasive belt corresponding to each sampling time throughout its entire life cycle is calculated:

[0061]

[0062] Among them, P i Let V be the polishing pressure of the i-th grinding point. m ) j For the j-th grinding and polishing speed, V f For the feed rate, k p h(Sta) represents the belt abrasion and polishing removal coefficient. ij k(Sta) represents the polishing amount during the stable period of belt polishing. w t represents the belt wear coefficient during the stable period of belt grinding and polishing. n For the nth sampling time, h(t) n ) ij Let k(t) be the polishing amount corresponding to the nth sampling time in belt polishing at the i-th polishing pressure and j-th polishing speed. n ) w The wear coefficient of the sand belt corresponding to the nth sampling time.

[0063] Therefore, the wear coefficient and polishing removal coefficient of the abrasive belt throughout its entire life cycle can be determined through measurement. These values ​​are then matched with the characteristic parameters of the synchronously measured acoustic signal to establish a matching dataset, which can be used to train a Bayesian neural network model. The acoustic signal characteristic parameters include acoustic frequency, gravity frequency, and cumulative power. During the abrasive belt polishing process, information on acoustic frequency and acoustic power under various polishing parameters is obtained. The amplitude-frequency relationship is obtained through Fourier transform, and then the gravity frequency and cumulative power are calculated.

[0064]

[0065]

[0066] Where, f(t)n Let be the sound frequency corresponding to the nth sampling time, p(f) be the power of the sound frequency, and S(t) be the power of the sound frequency. n ) ap S(t) represents the cumulative power corresponding to the nth sampling time. n ) gf The gravity frequency corresponds to the nth sampling time.

[0067] S200. Using the acoustic signal feature parameters in the matching dataset as the input of the neural network and the sand belt wear parameters in the matching dataset as the output of the neural network, the neural network is trained to establish a simulated sand belt wear parameter prediction model.

[0068] Construct a Bayesian neural network model using the parameters in step S100, such as Figure 3 As shown, a three-layer Bayesian neural network model is constructed, including an input layer, an output layer, and a hidden layer. The sound frequency, gravity frequency, and cumulative power from the matching dataset are used as input vectors to the Bayesian neural network model, while the belt polishing removal coefficient and belt wear coefficient are used as output vectors. The Bayesian neural network is trained to establish a model for predicting simulated belt wear parameters. The hidden layer represents the results obtained under different weighting coefficients for different operational relationships between the three input vectors. In this embodiment, the weights of each input vector are determined based on the sound signal feature parameters in the matching dataset: the weight of sound frequency is 0.5, the weight of cumulative power is 0.25, and the weight of gravity frequency is 0.25. It can be understood that the larger the amount of data in the matching dataset, the more accurate the prediction model obtained through Bayesian neural network deep learning.

[0069] k w =g(S gf f, S aq );

[0070] k p =const;

[0071] During the S300 belt grinding and polishing operation, real-time acoustic signal characteristic parameters are acquired and input into the simulated belt wear parameter prediction model to obtain the predicted values ​​of belt wear parameters.

[0072] S400: Based on the predicted values ​​of abrasive belt wear parameters, the instantaneous abrasive belt grinding speed is compensated to achieve stable control of the instantaneous abrasive belt grinding amount.

[0073] In this embodiment, the change in the sand belt polishing coefficient is compensated by controlling the instantaneous polishing speed, which is relatively easy to control in real time, so that the product of the predicted sand belt wear coefficient and the instantaneous polishing speed remains constant, thereby ensuring the stability of the instantaneous polishing amount.

[0074] Furthermore, in the method of this embodiment, when the predicted value of the sanding belt wear coefficient meets the sanding belt wear coefficient threshold, a sanding belt replacement alarm is issued. The sanding belt wear coefficient threshold is set by relevant technical personnel. Generally speaking, the sanding belt wear coefficient threshold is 0.5, which also corresponds to the limit value of the sanding belt's polishing capacity for scrapping.

[0075] In summary, this application uses a neural network model to predict abrasive belt wear parameters and compensates for the instantaneous grinding and polishing speed of the abrasive belt throughout its entire lifespan based on the predicted wear parameters. This avoids under- or over-grinding caused by abrasive belt wear, improving the stability of abrasive belt grinding and polishing operations and the contour accuracy of the workpiece surface. Furthermore, by monitoring the predicted abrasive belt wear parameters, this application can accurately determine the timing of abrasive belt replacement, preventing under- or over-use of the abrasive belt.

[0076] This application also provides a system for predicting and compensating for the wear of abrasive belts throughout their entire life cycle. A schematic diagram of this system is shown below. Figure 4 As shown, it includes a training data module, a neural network training and model determination module, and a belt wear prediction and compensation module.

[0077] The training data module is used to establish a time-matching dataset of abrasive belt wear parameters and acoustic signal characteristic parameters.

[0078] The neural network training and model determination module is used to train the neural network by using the acoustic signal feature parameters in the matching dataset as the input of the neural network and the sand belt wear parameters in the matching dataset as the output of the neural network, and to establish a simulated sand belt wear parameter prediction model.

[0079] The belt abrasion prediction and compensation module is used to acquire real-time acoustic signal characteristic parameters during belt abrasion and polishing operations and input them into a simulated belt abrasion parameter prediction model to obtain predicted belt abrasion parameter values. It is also used to compensate for the instantaneous abrasion speed of the belt based on the predicted belt abrasion parameter values, so as to achieve stable control of the instantaneous abrasion amount of the belt.

[0080] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for predicting and compensating for the wear of a sanding belt throughout its entire life cycle.

[0081] This application also provides a computer-readable storage medium that stores computer instructions that enable a computer to execute a method for predicting and compensating wear throughout the life cycle of abrasive belts, which will not be described in detail here.

[0082] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0083] The above description is only a partial embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for predicting and compensating for the full life-cycle wear of abrasive belts based on acoustic signals, characterized in that, include: S100. Establish a time-matching dataset of abrasive belt wear parameters and acoustic signal characteristic parameters; S200. Using the acoustic signal feature parameters in the matching dataset as the input of the neural network and the sand belt wear parameters in the matching dataset as the output of the neural network, the neural network is trained to establish a simulated sand belt wear parameter prediction model. S300 During the sand belt grinding and polishing operation, real-time acoustic signal characteristic parameters are acquired and input into the simulated sand belt wear parameter prediction model to obtain the predicted value of sand belt wear parameters. S400. Based on the predicted values ​​of the abrasive belt wear parameters, the instantaneous abrasive belt polishing speed is compensated to achieve stable control of the instantaneous abrasive belt polishing amount. Step S100 includes: Acquire the acoustic signal characteristic parameters and polishing amount of the abrasive belt throughout its entire life cycle under various polishing parameters; wherein, the polishing parameters include polishing length, feed speed, polishing speed and polishing pressure, and the acoustic signal characteristic parameters include acoustic frequency, gravity frequency and cumulative power; Based on the polishing parameters and the polishing amount, the abrasive belt wear parameters for each sampling time are determined; wherein, the abrasive belt wear parameters include the abrasive belt wear coefficient and the abrasive belt polishing removal coefficient; A matching dataset is established based on the acoustic signal characteristic parameters and the sand belt wear parameters at each sampling time. The step of determining the abrasive belt wear parameters for each sampling time period based on the abrasion parameters and the abrasion amount includes: For abrasive belts with the same polishing parameters, throughout their entire lifespan, based on the wear coefficient of 1 during the stable polishing period, the polishing removal coefficient of the abrasive belt is calculated using the following formula: ； Based on the polishing amount at each sampling time and the polishing amount during the polishing stabilization period, calculate the abrasion coefficient of the abrasive belt at each sampling time throughout its entire life cycle: ； in, For the first i The grinding and polishing pressure. For the first j A grinding and polishing speed For feed rate, This represents the removal coefficient from belt abrasion and polishing. This refers to the polishing amount during the stable period of belt polishing. The abrasion coefficient of the abrasive belt during the stable period of abrasive belt grinding and polishing. For the first n Each sampling time, For the first i The grinding and polishing pressure, the first j In belt polishing at the first polishing speed, the first... n The amount of polishing corresponding to each sampling time. For the first n The abrasion coefficient of the sand belt corresponding to each sampling time; Step S400 includes: Based on the predicted wear coefficient of the abrasive belt, compensation is made for the instantaneous abrasion and polishing speed of the abrasive belt, including: Real-time control of instantaneous grinding and polishing speed ensures that the product of the predicted abrasion coefficient of the abrasive belt and the instantaneous grinding and polishing speed remains constant.

2. The method for predicting and compensating for the full life cycle wear of abrasive belts according to claim 1, characterized in that, The acquisition of acoustic signal characteristic parameters of the abrasive belt throughout its entire life cycle under various grinding and polishing parameters includes: During the belt grinding and polishing process, information on the sound frequency and sound power under various grinding and polishing parameters is obtained. The amplitude-frequency relationship is obtained through Fourier transform, and then the gravity frequency and cumulative power are calculated.

3. The method for predicting and compensating for the full life cycle wear of abrasive belts according to claim 2, characterized in that, Step S200 includes: A three-layer Bayesian neural network model is constructed, including an input layer, an output layer, and a hidden layer. The sound frequency, gravity frequency, and cumulative power in the matching dataset are used as the input vectors of the Bayesian neural network model, and the sanding belt polishing removal coefficient and sanding belt wear coefficient are used as the output vectors. The Bayesian neural network is trained to establish a model for predicting sanding belt wear parameters. Specifically, the weights of each input vector are determined based on the feature parameters of each acoustic signal in the matching dataset.

4. The method for predicting and compensating for the full life cycle wear of abrasive belts according to claim 1, characterized in that, The method further includes: When the predicted value of the sanding belt wear coefficient meets the sanding belt wear coefficient threshold, a sanding belt replacement alarm is issued.

5. A system for predicting and compensating for the wear of abrasive belts throughout their entire life cycle, used to implement the method for predicting and compensating for the wear of abrasive belts throughout their entire life cycle as described in any one of claims 1-4, characterized in that, include: The training data module is used to create a time-matched dataset of abrasive belt wear parameters and acoustic signal characteristic parameters. The neural network training and model determination module is used to train the neural network by using the acoustic signal feature parameters in the matching dataset as the input of the neural network and the sand belt wear parameters in the matching dataset as the output of the neural network, and to establish a simulated sand belt wear parameter prediction model. The belt abrasion prediction and compensation module is used to acquire real-time acoustic signal characteristic parameters during belt abrasion and polishing operations and input them into the simulated belt abrasion parameter prediction model to obtain belt abrasion parameter prediction values. It is also used to compensate for the instantaneous abrasion speed of the belt based on the predicted belt abrasion parameter values, so as to achieve stable control of the instantaneous abrasion amount of the belt.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the sanding belt full life cycle wear prediction and compensation method as described in any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause the computer to perform the sanding belt full life cycle wear prediction and compensation method as described in any one of claims 1 to 4.

Images