Ultrasonic notching system and method for notching machine

Abstract

The invention relates to the technical field of cutting, in particular to an ultrasonic notching system and method for a notching machine, the system comprises an ultrasonic notching module, a force sense sensing unit, an acoustic sensing unit and a controller, and the controller is electrically connected with an ultrasonic transducer, the force sense sensing unit and the acoustic sensing unit. A processor configured to: receive and process the force signal and the acoustic signal; on the basis of a preset process control model, the real-time cutting state is jointly calculated through the force signal and the acoustic signal; and according to the deviation between the real-time cutting state and the target state, working parameters of the ultrasonic transducer are dynamically adjusted in a closed-loop mode. The invention provides an ultrasonic notching system and method for a notching machine. According to the system, through combination of the ultrasonic incision module, the force sense sensing unit and the acoustic sensing unit, cutting force and acoustic signals in the cutting process can be monitored in real time, and then accurate analysis and dynamic adjustment of the cutting state are achieved.

Description

Technical Field

[0001] This invention relates to the field of cutting technology, and more specifically to an ultrasonic cutting system and method for a cutting machine. Background Technology

[0002] In modern medical and industrial applications, incision machines serve as crucial cutting tools, widely used for tissue cutting and material processing. However, existing incision machines suffer from several significant problems during the cutting process. First, the cutting force and cutting status of traditional incision machines are often difficult to monitor in real time, leading to unstable cutting results and potentially causing tissue damage or material waste. Second, current technologies typically lack effective utilization of acoustic signals during the cutting process, failing to fully analyze the cutting status and thus affecting cutting accuracy and efficiency. Furthermore, traditional incision machines rely heavily on manual experience for adjusting cutting parameters, lacking intelligent feedback mechanisms, resulting in slow response times and difficulty in achieving precise control during the cutting process. Summary of the Invention

[0003] To overcome the shortcomings and deficiencies of existing technologies, this invention provides an ultrasonic cutting system and method for a cutting machine. This system, through the combination of an ultrasonic cutting module, a force sensing unit, and an acoustic sensing unit, can monitor the cutting force and acoustic signals during the cutting process in real time, thereby achieving precise analysis and dynamic adjustment of the cutting state.

[0004] This invention is achieved through the following technical solution: In a first aspect, the present invention discloses an ultrasonic incision system for an incision machine, comprising: An ultrasonic cutting module, comprising an ultrasonic transducer and a cutting head connected to the ultrasonic transducer. A force sensing unit is configured on the force transmission path of the ultrasonic cutting module to monitor the cutting force on the cutting head during the cutting process in real time and generate a force signal. An acoustic sensing unit is configured near the cutting head to collect high-frequency acoustic vibrations generated during the cutting process in real time and generate acoustic signals. and a controller, which is electrically connected to the ultrasonic transducer, the force sensing unit, and the acoustic sensing unit, and is configured to: Receive and process the force signal and the acoustic signal; Based on a preset process control model, the real-time cutting state is calculated by jointly solving the force signal and the acoustic signal. The operating parameters of the ultrasonic transducer are dynamically and in a closed-loop manner adjusted based on the deviation between the real-time cutting state and the target state.

[0005] In conjunction with the first aspect, the force sensing unit is further described as a multi-axis piezoelectric force sensor, used to acquire at least the components of axial cutting force and lateral cutting force.

[0006] In conjunction with the first aspect, the acoustic sensing unit is further described as a microelectromechanical system microphone or a contact acoustic emission sensor covering a frequency range of 20kHz to 100kHz.

[0007] In conjunction with the first aspect, the process control model stored internally by the controller further includes a probability density model for defining the optimal process window and a dynamic control law model for calculating parameter adjustment amounts.

[0008] Secondly, the present invention discloses an ultrasonic incision control method for an incision machine, which includes the following steps: S100. By performing multiple calibration cuts on standard samples, simultaneously collecting cutting force data and acoustic data, a multi-dimensional feature space defining the optimal cutting quality is established, and an optimal process window model is constructed. S200. Real-time acquisition of force and acoustic signals of the target product generated by the force sensing unit and the acoustic sensing unit; S300. Extract features from the real-time acquired force and acoustic signals, map them to a multi-dimensional feature space, and calculate the deviation between the current working point and the optimal process window. S400. Based on the degree of deviation, a nonlinear dynamic control algorithm is applied to adjust the operating parameters of the ultrasonic transducer in real time. The operating parameters include output power, operating frequency, or vibration amplitude, so as to drive the current operating point into the optimal process window and realize closed-loop control.

[0009] In conjunction with the second aspect, furthermore, the workflow for constructing the optimal process window model in step S100 includes: Fourier transform and feature extraction were performed on the force and acoustic signals from the N sets of collected calibration data to obtain... There are 1 data point, each data point is characterized by cutting force. Harmony Acoustic Spectrum constitute; Using Gaussian mixture model to analyze this Probability density estimation is performed on each data point to establish the probability density function for the optimal process window. The mathematical formula is:

[0010]

[0011] in, For data points in the feature space; The number of Gaussian distributions in the mixture model; For this is the first A Gaussian-distributed mixture of weights, satisfying ; For the first One Gaussian component; For the first The mean vector of Gaussian components; For the first The covariance matrix of Gaussian components; is the dimension of the feature space; This is the set of model parameters that need to be learned using the EM algorithm.

[0012] In conjunction with the second aspect, further, in step S400, the ultrasonic output power is adjusted... The workflow includes: Real-time calculation of current cutting force With the center of the optimal process window error ; The power adjustment is calculated using a nonlinear gain controller incorporating an sigmoid saturation function. ; Ultrasonic output power The update formula is:

[0013]

[0014] in, yes The final output power at any given moment; This is the base power setting value; for The cutting force error at any moment, It is a real-time force feedback value. This represents the expected value of the optimal force characteristic in the optimal process window model; These are the proportional and integral gain coefficients, respectively. It is the hyperbolic tangent function; For adjustment The scaling factor of the function shape; This is the integral term over historical error.

[0015] In conjunction with the second aspect, step S300 further includes diagnosing tool wear conditions and compensatoryly adjusting the operating frequency. The workflow includes: For real-time acoustic signals Perform a short-time Fourier transform to obtain its time-frequency spectrum; Extracting the higher harmonic energy ratio from the time-frequency spectrum. Spectral Entropy As a characteristic indicator of quality degradation; Construct a tool health index Its value decreases over time depending on the degree of deterioration of the characteristic indicators; According to health index For basic operating frequency Make fine-tuning compensation; Health Index The iterative update algorithm is as follows:

[0016] Operating frequency The compensatory adjustment formula is:

[0017] in, for The health index of the cutting tool at any given time; This is the learning rate or decay rate constant; This represents the ratio of higher harmonic energy to fundamental frequency energy. The preset failure threshold; This is the spectral entropy. The preset failure threshold; for Operating frequency after time-compensation; Basic operating frequency; This is the maximum frequency compensation coefficient; It is a natural exponential function; is the time constant.

[0018] Thirdly, the present invention discloses an electronic device comprising 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 ultrasonic incision control method as described above.

[0019] Fourthly, the present invention discloses a computer-readable storage medium having a computer program stored thereon, the program being executed by a processor to implement the ultrasonic incision control method as described above.

[0020] The beneficial effects of this invention are: This invention discloses an ultrasonic cutting system, method, electronic device, and storage medium for a cutting machine. By monitoring cutting force and acoustic signals in real time, the system can accurately assess the cutting state, ensuring the stability and safety of the cutting process. Secondly, based on a preset process control model, the system can dynamically adjust the operating parameters of the ultrasonic transducer to achieve closed-loop control, thereby improving cutting accuracy and efficiency. Furthermore, the intelligent design of the system reduces reliance on human experience, making the cutting process more automated and intelligent, reducing operational risks, and enhancing the user experience. In summary, the ultrasonic cutting system of this invention has significant application value and broad market prospects in the field of cutting machine technology. Attached Figure Description

[0021] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0022] Figure 1 This is a schematic diagram of the modules of an ultrasonic incision system according to an embodiment of the present invention.

[0023] Figure 2 This is a flowchart of an ultrasonic incision control method according to an embodiment of the present invention.

[0024] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0025] Figure 4 This is a schematic diagram of a storage medium provided in an embodiment of the present invention. Detailed Implementation

[0026] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0027] Example 1 Please see Figure 1 The figure shows a structural block diagram of an ultrasonic incision system according to an embodiment of the present invention. As shown, the system includes an ultrasonic incision module, a force sensing unit, an acoustic sensing unit, and a controller.

[0028] Specifically, the ultrasonic cutting module is the core component that performs the cutting action. In this embodiment, it includes an ultrasonic transducer for generating high-frequency electrical signals and converting them into mechanical vibrations, and a cutting head connected to the output end of the transducer. The tip of the cutting head generates high-frequency vibrations under the drive of the ultrasonic transducer, thereby cutting the blow-molded product (not shown in the figure) placed at the workstation.

[0029] The force sensing unit functions to sense the physical load during the cutting process in real time and quantitatively. To achieve this, the force sensing unit is positioned along the force transmission path of the ultrasonic cutting module. In a preferred embodiment, the force sensing unit can be installed between the ultrasonic transducer and the mounting bracket (not shown in the figure), or below the worktable used to fix the blow-molded product. When the cutting head cuts the product, the reaction force of the product on the cutting head is transmitted to the force sensing unit without attenuation through the cutting head, transducer, and other components. This unit monitors this cutting force in real time and converts it into an electrical signal (i.e., a force signal), which is then sent to the controller.

[0030] To support the claims, in one specific embodiment of the invention, the force sensing unit is specifically a multi-axis piezoelectric force sensor. Compared to a single-axis sensor, a multi-axis sensor can simultaneously measure force components in at least three dimensions, namely the axial cutting force along the tool feed direction. and two mutually perpendicular lateral cutting forces. and This multi-dimensional force information is extremely valuable because of the axial force. It mainly reflects the thickness and hardness of the material, while lateral force... , Abnormal fluctuations may indicate more complex quality issues such as unilateral tool wear, product misalignment, or burrs on the cut. Therefore, using a multi-axis force sensor can provide the controller with richer and more comprehensive information for judging the process status.

[0031] The acoustic sensing unit is designed to monitor subtle changes in the cutting process. It is positioned near the cutting head to collect high-frequency acoustic vibrations generated during the cutting process, such as friction between the blade and the plastic, and the resulting breakage. These vibrations are then converted into electrical signals (i.e., acoustic signals) and sent to the controller.

[0032] In one specific embodiment of the present invention, the acoustic sensing unit can be a MEMS (Micro-Electro-Mechanical Systems) microphone covering a frequency range of kHz to kHz. MEMS microphones have advantages such as small size, high sensitivity, and cost-effectiveness. They can be installed non-contactly very close to the cutting head, effectively picking up high-frequency sound waves generated by cutting that are inaudible to the human ear but contain rich state information. In another alternative embodiment, the acoustic sensing unit can also be a contact acoustic emission sensor, which can be directly mounted on a fixture or workbench to fix the product. Its advantage lies in its ability to directly pick up structural vibrations propagating through a solid medium, exhibiting an extremely high signal-to-noise ratio and being virtually unaffected by ambient noise in the workshop environment.

[0033] Specifically, in this embodiment, the controller interacts with the ultrasonic transducer, force sensing unit, and acoustic sensing unit via cable or wireless means to exchange data and commands. The controller's input end receives real-time sensing signals from the units, and its output end sends control commands for operating parameters to the ultrasonic transducer.

[0034] Its core functions are configured as follows: First, it performs high-speed processing and fusion analysis on the input force and acoustic signals; then, based on the analysis results, it jointly calculates the current real-time cutting state (e.g., whether the cutting is smooth, whether the cut quality is excellent, and whether the tool is worn); finally, it compares the real-time state with a preset target state that represents the standard cutting, and dynamically and in a closed loop adjusts the working parameters of the ultrasonic transducer (such as output power and working frequency) according to the deviation between them, so as to keep the cutting process in the optimal state at all times.

[0035] Specifically, the controller typically consists of hardware such as a microprocessor, memory, and input / output interfaces. Its core technology lies in the process control model embedded in memory. In a preferred embodiment of the invention, this process control model includes at least two key parts: A probability density model a is used to define the optimal process window: This model is not a simple threshold, but a multidimensional probability distribution model built on big data and machine learning (e.g., using a Gaussian mixture model, GMM). It mathematically depicts a region or probability cloud representing the optimal cutting state by learning from force-acoustic data of a large number of high-quality cut samples. The higher the probability that any real-time working point falls into this region, the better its cutting quality.

[0036] A dynamic control law model b for calculating parameter adjustment amounts: This model is one or a set of advanced nonlinear control algorithms. When the real-time operating point deviates from the central region of the probability density model a, the control law model will accurately calculate, based on the direction and distance of the deviation, how much adjustment is needed to parameters such as power and frequency in order to pull it back to the optimal region in the fastest and most stable way.

[0037] In summary, the embodiments of the present invention creatively integrate sensor information from two different modes, force and acoustic, and construct a closed-loop control system capable of real-time perception, deep understanding, and dynamic adaptive adjustment by means of an advanced mathematical model stored in the controller. This fundamentally solves the problems of poor stability and reliance on human experience in the existing ultrasonic cutting process.

[0038] Example 2 like Figure 2 As shown, this embodiment discloses an ultrasonic incision control method for an incision machine, which includes the following steps: S100. By performing multiple calibration cuts on the standard sample, cutting force data and acoustic data are collected simultaneously to establish a multi-dimensional feature space that defines the optimal cutting quality and construct an optimal process window model. This step is an offline calibration or learning process performed before the start of mass production. Its core purpose is to establish a mathematical benchmark for the controller that can accurately judge the standard cut, namely the optimal process window (OPW) model.

[0039] When implementing step S100 in detail: First, select several standard blow-molded products, or prepare a batch of test samples with uniform size and material; Experienced process engineers manually set a set of initial ultrasonic parameters (power, frequency, amplitude, etc.) considered optimal, and then used the ultrasonic cutting system of this invention to cut these samples multiple times. During the cutting process, the quality of each cut (such as smoothness, burr-free, and no thermal damage) was ensured to meet the highest standards by manual inspection or high-precision inspection equipment (such as microscopes and profilometers).

[0040] While performing standard cuts, the controller synchronously and at high frequency records the complete, unprocessed raw data stream acquired by the force sensing unit and the acoustic sensing unit. For example, it records the complete curve of the cutting force changing over time during each cut, as well as the time-domain waveform of the acoustic signal.

[0041] After collecting a sufficient number of high-quality data samples, the controller performs feature extraction on these data. For example, it extracts features such as average cutting force, peak cutting force, and force fluctuation variance from the force signal; simultaneously, it performs a Fast Fourier Transform (FFT) on the acoustic signal to extract key features such as fundamental frequency energy, harmonic energy ratio, and spectral entropy from its spectrum. In this way, each standard cut is transformed into a data point in the "force-sound" multidimensional feature space.

[0042] Finally, the controller applies its built-in machine learning algorithm to train on all these high-quality data points, ultimately generating a probability density function that describes the spatial distribution of these points. The high-probability-density region defined by this function is the optimal process window model we need, and it is stored in memory.

[0043] S200. Real-time acquisition of force and acoustic signals of the target product generated by the force sensing unit and the acoustic sensing unit.

[0044] Once the system enters normal online production mode, this method enters a real-time monitoring cycle. In step S200, when a target product to be processed is sent to the workstation and cutting begins, the force sensing unit and acoustic sensing unit in the system immediately start working, continuously collecting real-time force and acoustic signals during the cutting process, and transmitting these raw analog signals to the controller in real-time as digital signal streams after A / D conversion.

[0045] S300. Extract features from the real-time acquired force and acoustic signals, map them into a multi-dimensional feature space, and calculate the deviation between the current operating point and the optimal process window.

[0046] In step S300, after the controller receives the real-time signal stream from the monitoring stage, it will immediately use the same algorithm as the modeling stage to extract features in real time, thereby calculating the "force-sound" feature vector at the current moment, that is, the real-time coordinates of the current working point in the multi-dimensional feature space.

[0047] The controller then substitutes these real-time workpoint coordinates into a pre-stored OPW model for evaluation. This model allows the controller to accurately calculate the deviation of the current workpoint from the center of the optimal region. This deviation can be a probability value or a distance metric (e.g., the Mahalanobis distance from the workpoint to the center of the probability distribution). This deviation value provides a quantitative assessment of the current cutting quality.

[0048] S400. Based on the degree of deviation, a nonlinear dynamic control algorithm is applied to adjust the operating parameters of the ultrasonic transducer in real time, including but not limited to output power, operating frequency, or vibration amplitude, so as to drive the current operating point into the optimal process window and realize closed-loop control.

[0049] Step S400 is crucial for the system to achieve closed-loop operation; it is the process by which the controller takes action based on the analysis results.

[0050] Specifically, the controller takes the deviation calculated during the analysis phase as input and passes it to the dynamic control law model stored in memory. This control law model immediately and automatically calculates precise adjustments to one or more ultrasonic parameters based on the magnitude and direction of the input deviation. For example, if the model analyzes that the current cutting force is too high but the sound spectrum is normal (possibly indicating an area with increased product wall thickness), it might calculate a positive power adjustment. If the analysis reveals that the cutting force is normal but the harmonic components of the sound signal are abnormally increased (which may indicate that the tool is starting to wear out), it may calculate a small frequency adjustment. The calculated adjustment amount is immediately converted into a control command, which is sent to the ultrasonic transducer through the controller's output interface. Upon receiving the command, the transducer instantly updates its operating parameters, such as increasing the output power or fine-tuning the operating frequency.

[0051] After parameter adjustments, the cutting state changes, and this change is immediately captured by sensors in the monitoring phase, forming a rapidly evolving closed-loop feedback loop of monitoring, analysis, control, and re-monitoring. This loop continues at an extremely high frequency until the current cutting task is completed, ensuring that throughout the entire cutting process, regardless of any disturbances encountered (such as material inhomogeneity, voltage fluctuations, tool wear, etc.), the actual cutting point can be locked within the optimal process window, achieving a high degree of stability and consistency in cutting quality.

[0052] Example 3 In step S100 of Example 2, the workflow for constructing the optimal process window model includes: Fourier transform and feature extraction were performed on the force and acoustic signals from the N sets of collected calibration data to obtain... There are 1 data point, each data point is characterized by cutting force. Harmony Acoustic Spectrum constitute; Using Gaussian mixture model to analyze this Probability density estimation is performed on each data point to establish the probability density function for the optimal process window. The mathematical formula is:

[0053]

[0054] In Example 2, the goal of the modeling phase S100 is to establish a mathematical model for the system that can accurately define the "optimal cutting state." In this preferred embodiment, this goal is achieved through the following more specific workflow: The first step involves collecting calibration data and extracting features. First, the system operator selects N sets of standard samples for standard cutting operations, which are then synchronously recorded by the controller. Force signal time series corresponding to the group Harmony acoustic signal time series .

[0055] Next, the controller's response to this We perform dimensionality reduction and feature extraction on the original, high-dimensional time series data, transforming it into... A low-dimensional but highly information-rich feature vector. For each group (e.g., the first... The data (groups) are processed as follows: Force signal characteristics Extraction of force signals: Statistical analysis was conducted to extract key indicators that reflect the magnitude and stability of the cutting load, such as the steady-state average cutting force. Cutting force standard deviation .

[0056] Acoustic signal characteristics Extraction of acoustic signals: Applying the Fast Fourier Transform, we transform it from the time domain to the frequency domain to obtain the spectrum. The spectrum contains deeper information about the quality of the cut. Subsequently, key features are extracted from the spectrum, such as: Higher harmonic energy ratio : Calculate the ratio of energy in a specific high-order harmonic frequency band to the fundamental frequency energy. Generally, the smoother and burr-free the cut, the lower this ratio.

[0057] Spectral entropy : Calculate the entropy value of the spectrum. A spectrum with concentrated energy has a low spectral entropy, while a chaotic and disordered spectrum (which may mean undesirable cutting behavior such as tearing and friction) has a high spectral entropy.

[0058] After the above processing, the initial Complex time series data were successfully transformed into multidimensional feature vectors , where each vector ,this The data points together constitute the sample distribution of the "optimal cutting state" in the four-dimensional feature space.

[0059] Step 2: Probability density estimation is performed using a Gaussian mixture model. This yields... After identifying data points representing the standard state, this step aims to find a mathematical model that can accurately describe the inherent distribution patterns of these data points. Those skilled in the art might think of setting a simple rectangular window (i.e., the upper and lower limits of each feature) to define the standard range, but the drawbacks of this approach are obvious: it ignores the correlation between features (e.g., a sample may have a slightly larger cutting force, but it is still good as long as its spectral entropy is extremely low), and the boundaries are too rigid. Therefore, in this preferred embodiment, the controller employs a more advanced Gaussian mixture model algorithm to analyze this... The probability density is estimated for each data point. The core idea of ​​GMM is that it does not assume all standard data points belong to a single cluster, but rather that they are composed of [various data points]. It is formed by the linear superposition of K different, potential Gaussian distributions (which can be understood as K different superior subclasses). The controller trains the model on the collected N feature vectors using a built-in expectation-maximization iterative algorithm, ultimately learning the complete parameter set of the GMM model. At this point, we obtain a continuous probability density function that can describe the optimal process window. .

[0060] The physical meaning of each parameter in the formula is explained in detail here: in, For each data point in the feature space, there is a current working point to be evaluated, consisting of real-time force and acoustic features. The number of Gaussian distributions in the mixture model is a hyperparameter representing the number of subcategories the system expects the standard state to contain. For example, the standard states for cutting different parts of a product (such as the bottle body and the bottle neck) may correspond to different Gaussian components; For this is the first A Gaussian-distributed mixture of weights, satisfying ; For the first The Gaussian component represents the state belonging to the th best among all excellent states. The probability or proportion of each subcategory; For the first The mean vector of the nth Gaussian component, which physically represents the nth Gaussian component. The ideal center point or desired state of a superior sub-category is one of the core objectives for subsequent closed-loop control systems to track. For the first The covariance matrix of the nth Gaussian component describes the... The shape, extension direction, and size of a superior subclass in the multidimensional feature space can tell us, for example, whether the average cutting force and spectral entropy are negatively correlated in that subclass, and how large their respective allowable fluctuation ranges are. is the dimension of the feature space; This is the set of model parameters that need to be learned using the EM algorithm.

[0061] Thus, an optimal process window (OPW) model that is far more accurate, flexible, and robust than the traditional thresholding method has been constructed. This model is stored in the controller's memory. In subsequent online production, the controller only needs to process the feature vectors extracted in real time. Substitute into the probability density function This yields a probability value between 0 and 1, which directly and quantitatively reflects the cutting quality at the current moment, providing a precise and nonlinear decision-making basis for subsequent closed-loop control.

[0062] Example 4 In step S400, the ultrasonic output power is adjusted. The workflow includes: Real-time calculation of current cutting force With the center of the optimal process window error ; The power adjustment is calculated using a nonlinear gain controller incorporating an sigmoid saturation function. ; Ultrasonic output power The update formula is:

[0063]

[0064] in, yes The final output power at any given moment; This is the base power setting value; for The cutting force error at any moment, It is a real-time force feedback value. This represents the expected value of the optimal force characteristic in the optimal process window model; These are the proportional and integral gain coefficients, respectively. It is the hyperbolic tangent function; For adjustment The scaling factor of the function shape; This is the integral term over historical error.

[0065] In the overall closed-loop control process, power adjustment is the most direct and effective means of combating load changes (such as uneven product wall thickness and material hardness fluctuations). This embodiment aims to provide a power regulation method that is far superior to traditional linear control (such as standard PID).

[0066] Specifically, the first step: calculation of real-time cutting force error. The controller continuously performs this step while the system is performing online cutting.

[0067] First, the controller obtains the current real-time cutting force characteristic value from the analysis phase. To enhance stability, here... It can be the moving average of the force output by the force sensing unit within a very short time window (e.g., 5 milliseconds).

[0068] Simultaneously, the controller retrieves the expected value of the optimal cutting force from the optimal process window (OPW) model stored in memory. This value is the force characteristic component at the center of the Gaussian component with the highest probability in the GMM model, and physically represents the most ideal cutting load.

[0069] The controller calculates the difference between the two in real time to obtain the cutting force error signal: This error signal It precisely quantifies the degree and direction of deviation between the current load and the ideal state. Positive values ​​indicate that the load is too large, and negative values ​​indicate that the load is too light.

[0070] Step 2: Calculation of nonlinear power adjustment based on the sigmoid saturation function Error signal obtained Then, the controller calls its dynamic control law model to calculate the power adjustment. . Those skilled in the art know that if a simple linear proportional controller is used ( = • This will present a dilemma: if the proportionality coefficient... Setting the power too high in pursuit of a fast response can lead to a sudden surge in power when there is a large error (such as a tool suddenly encountering a reinforcing rib), potentially causing localized thermal damage to the product or impacting the transducer; if If the settings are too low to ensure safety, the system response will become sluggish and unable to compensate for load changes in a timely manner. To overcome this deficiency, this preferred embodiment employs a nonlinear controller incorporating an sigmoid saturation function (specifically, a hyperbolic tangent function tanh). The core advantage of this controller lies in its variable gain, enabling it to intelligently adapt to errors of varying magnitudes.

[0071] The specific calculation process follows the mathematical formula defined in this embodiment. The components of the formula and their working principle are explained in detail below:

[0072] Nonlinear proportional term : This provides an immediate response to errors. The key lies in the hyperbolic tangent function. The introduction of error. When the absolute value is very small (i.e., the cutting state is close to ideal), At this point, the term approximates a linear high-gain proportional controller (gain approximately...). • It can quickly and accurately correct minute deviations.

[0073] When error When the absolute value is very large (i.e., the load changes drastically), The value of this term will rapidly approach +1 or -1 and saturate. This means that no matter how much the error increases, the output of this term will be smoothly limited to [value missing]. Within this range. This soft saturation characteristic perfectly simulates: when encountering a large disturbance, decisively applying a sufficiently large but not excessive adjustment force, thus achieving a perfect balance between rapid response and prevention of overshoot. Parameters Used to adjust the transition speed in the saturation region.

[0074] Nonlinear integral term : This item aims to eliminate potential static errors in the system. For example, if the material of the entire batch of products is slightly harder than that of the calibration sample, the proportionality item alone may not be able to completely eliminate the error.

[0075] The integral term accumulates historical errors and produces a continuous compensation output until the error... Driven back to zero.

[0076] Used here as well The function is used to prevent integral saturation. In traditional PID control, if the error persists for a long time, the integral term accumulates excessively, causing significant overshoot even after the error disappears. The integral saturation in this algorithm is also... Function constraints make the integration process safer and more stable. Parameters Used to adjust the sensitivity of the integral response.

[0077] Step 3: Final update of output power. The controller will calculate the power adjustment amount. With a base power value set for the current product Add them together to get the final target output power. The command is sent to the ultrasonic transducer via the output interface, and the transducer's power drive circuit will immediately adjust the output power to... .

[0078] In summary, this embodiment introduces a sophisticated nonlinear control law, enabling the system to exhibit intelligent responses similar to those of a human expert when faced with load disturbances of varying degrees. This improves the dynamic stability of the cutting process and the consistency of product quality, while also enhancing equipment safety and adaptability to process changes.

[0079] Example 5 Step S300 also includes diagnosing tool wear and compensatoryly adjusting the operating frequency. The workflow includes: For real-time acoustic signals Perform a short-time Fourier transform to obtain its time-frequency spectrum; Extracting the higher harmonic energy ratio from the time-frequency spectrum. Spectral Entropy As a characteristic indicator of quality degradation; Construct a tool health index Its value decreases over time depending on the degree of deterioration of the characteristic indicators; According to health index For basic operating frequency Make fine-tuning compensation; Health Index The iterative update algorithm is as follows:

[0080] Operating frequency The compensatory adjustment formula is:

[0081] in, for The health index of the cutting tool at any given time; This is the learning rate or decay rate constant; This represents the ratio of higher harmonic energy to fundamental frequency energy. The preset failure threshold; This is the spectral entropy. The preset failure threshold; for Operating frequency after time-compensation; Basic operating frequency; This is the maximum frequency compensation coefficient; It is a natural exponential function; is the time constant.

[0082] The workflow of this embodiment can be divided into the following steps.

[0083] Step 1: Time-Frequency Domain Analysis of Acoustic Signals Specifically, during the cutting process, the controller processes the real-time acoustic signals from the acoustic sensing unit. The signal is then processed. Unlike the modeling phase, which might involve performing an FFT on the entire signal, a short-time Fourier transform (SFT) is preferred here to capture the dynamic evolution of the wear state. The controller divides the continuous acoustic signal stream into a series of small, overlapping time windows and performs an FFT on each window. The result is a time-spectrum plot that clearly shows how the spectral components of the signal dynamically change over time, providing a basis for extracting dynamic features.

[0084] Step 2: Extraction of quality degradation characteristic indicators From the real-time generated time-frequency spectrum, the controller will focus on extracting the following two types of feature indicators that can sensitively reflect the tool wear state: Higher harmonic energy ratio The controller calculates the signal energy within integer harmonics of the fundamental ultrasonic frequency (e.g., around 80kHz for the second harmonic and 120kHz for the third harmonic), and compares this energy with the energy within the fundamental frequency band to obtain the ratio. As mentioned earlier, tool wear exacerbates the nonlinearity of cutting, which leads to a significant increase in this ratio.

[0085] Spectral entropy Spectral entropy is an indicator in information theory used to measure the uncertainty or randomness of a signal. The controller calculates the entropy value of the spectrum within the current time window. In a healthy cutting process, the spectral shape is relatively fixed and ordered, and the spectral entropy is low. As the tool wears, broadband noise generated by friction and random fractures increases, the spectrum becomes flatter and more chaotic, and the spectral entropy increases accordingly.

[0086] Step 3: Tool Health Index Construction and iteration To transform the two indirect characteristic indicators mentioned above into an intuitive tool health metric that can be used for decision-making, this embodiment constructs a tool health index. .

[0087] It was designed as a number that decreases from 1 to 0. This indicates that the cutting tool is in brand new, optimal condition. This indicates that the cutting tool has been completely worn out and is no longer usable.

[0088] The value is iteratively updated based on a damage accumulation model, the specific iterative algorithm of which is given in the mathematical formula of this embodiment. Its physical meaning is explained below: This formula is essentially a discrete-time integrator. At each time step, it calculates an instantaneous damage rate.

[0089] This instantaneous damage rate is derived from the normalized harmonic ratio. Spectral Entropy This is a joint decision. The controller has preset failure thresholds for both metrics. and The closer the current value is to the threshold, the more severe the wear and tear, and the greater the calculated instantaneous damage rate.

[0090] parameter The decay rate constant controls the overall rate of decline in the health index.

[0091] Ultimately, the controller uses the health index from the previous moment. Subtracting the accumulated damage within the current time step yields the health index at the current moment. In this way, the system achieves quantitative tracking of the wear process throughout the entire life cycle of the cutting tool.

[0092] Step 4: Compensatory Frequency Adjustment Based on Health Index Get real-time tool health index Afterwards, the system can not only achieve predictive maintenance (e.g., in...) It can automatically alarm when the value is below 0.2 and prompt the tool to be replaced, and can also perform active compensatory control.

[0093] Studies have found that fine-tuning the working frequency of ultrasonic waves can, to some extent, change the resonance matching state between the cutter head and the material, and can sometimes partially compensate for the decrease in cutting efficiency caused by the dulling of the blade.

[0094] The frequency adjustment algorithm used in this embodiment has been given in the mathematical formula of this embodiment, and its core is a non-linear exponential function mapping relationship: This is the optimal base operating frequency set for the new tool. The core exponential term (1-eτcH(t)-1) constructs a clever compensation strategy: When the tool health When the value approaches 1 (slight wear), the value of this index term approaches 0, the frequency adjustment is minimal, and the system strives for stability.

[0095] along with As wear decreases (moderate wear), this value increases non-linearly and rapidly, and the system begins to perform more significant frequency compensation in an attempt to extract the remaining value of the tool and maintain cut quality.

[0096] parameter The maximum compensation range is limited to ensure that frequency adjustments do not deviate from the transducer's efficient operating range. Parameters The time constant controls the aggressiveness of the compensation curve.

[0097] The controller will calculate the target frequency The data is sent to an ultrasonic transducer, enabling real-time, intelligent process compensation for tool wear without interrupting production.

[0098] Example 6 Please refer to Figure 3 The diagram illustrates a structural schematic of a computer device (equivalent to the controller in embodiments 1-5) provided by some embodiments of this application. The computer device 20 includes a processor 200, a memory 201, a bus 202, and a communication interface 203. The processor 200, the communication interface 203, and the memory 201 are connected via the bus 202. The memory 201 stores a computer program that can run on the processor 200. The processor 200 is used to operate according to the instructions to execute the steps of the method in any one of embodiments 2-5.

[0099] The memory 201 may include high-speed random access memory (RAM) or non-volatile memory, such as at least one disk storage device. Communication between this system network element and at least one other network element is achieved through at least one communication interface 203 (which can be wired or wireless), such as the Internet, wide area network, local area network, or metropolitan area network.

[0100] Bus 202 can be an ISA bus, PCI bus, or EISA bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. The memory 201 is used to store programs. After receiving an execution instruction, the processor 200 executes the program. The ultrasonic incision control method disclosed in any of the foregoing embodiments of this application can be applied to the processor 200, or implemented by the processor 200.

[0101] The processor 200 may be an integrated circuit chip with signal processing capabilities. In implementation, each step of the above method can be completed by the integrated logic circuitry in the hardware of the processor 200 or by instructions in software form. The processor 200 may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), an off-the-shelf programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor may be a microprocessor or any conventional processor. The steps of the methods disclosed in the embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software modules may reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. The storage medium is located in memory 201. The processor 200 reads the information in memory 201 and, in conjunction with its hardware, completes the steps of the above method.

[0102] Example 7 This embodiment also provides a computer-readable storage medium corresponding to the ultrasonic incision control method provided in the foregoing embodiments. Please refer to... Figure 4 The computer-readable storage medium shown is an optical disc containing a computer program (i.e., program product 30) that, when run by a processor, executes the ultrasonic incision control method provided in any of the foregoing embodiments.

[0103] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here.

[0104] The computer-readable storage medium provided in the above embodiments of this application and the ultrasonic incision control method provided in the embodiments of this application are based on the same application concept and have the same beneficial effects as the methods adopted, run or implemented by the application programs stored therein.

[0105] It should be noted that in the above text, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0106] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0107] In summary, the ultrasonic cutting system, method, electronic device, and storage medium for a cutting machine of the present invention, by real-time monitoring of cutting force and acoustic signals, can accurately assess the cutting state, ensuring the stability and safety of the cutting process. Secondly, based on a preset process control model, the system can dynamically adjust the operating parameters of the ultrasonic transducer to achieve closed-loop control, thereby improving cutting accuracy and efficiency. Furthermore, the intelligent design of the system reduces reliance on human experience, making the cutting process more automated and intelligent, reducing operational risks, and enhancing the user experience. Therefore, the ultrasonic cutting system of the present invention has significant application value and broad market prospects in the field of cutting machine technology.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. An ultrasonic incision system for an incision machine, characterized in that, include: An ultrasonic cutting module, comprising an ultrasonic transducer and a cutting head connected to the ultrasonic transducer. A force sensing unit is configured on the force transmission path of the ultrasonic cutting module to monitor the cutting force on the cutting head during the cutting process in real time and generate a force signal. An acoustic sensing unit is configured near the cutting head to collect high-frequency acoustic vibrations generated during the cutting process in real time and generate acoustic signals. and a controller, which is electrically connected to the ultrasonic transducer, the force sensing unit, and the acoustic sensing unit, and is configured to: Receive and process the force signal and the acoustic signal; Based on a preset process control model, the real-time cutting state is calculated by jointly solving the force signal and the acoustic signal. The operating parameters of the ultrasonic transducer are dynamically and in a closed-loop manner adjusted based on the deviation between the real-time cutting state and the target state.

2. The ultrasonic incision system for an incision machine according to claim 1, characterized in that, The force sensing unit is a multi-axis piezoelectric force sensor used to acquire at least the components of axial cutting force and lateral cutting force.

3. The ultrasonic incision system for an incision machine according to claim 1, characterized in that, The acoustic sensing unit is a microelectromechanical system microphone or a contact acoustic emission sensor covering a frequency range of 20kHz to 100kHz.

4. An ultrasonic incision system for an incision machine according to claim 1, characterized in that, The process control model stored inside the controller includes a probability density model for defining the optimal process window and a dynamic control law model for calculating parameter adjustment amounts.

5. An ultrasonic incision control method for an incision machine, characterized in that, Includes the following steps: S100. By performing multiple calibration cuts on standard samples, simultaneously collecting cutting force data and acoustic data, a multi-dimensional feature space defining the optimal cutting quality is established, and an optimal process window model is constructed. S200. Real-time acquisition of force and acoustic signals of the target product generated by the force sensing unit and the acoustic sensing unit; S300. Extract features from the real-time acquired force and acoustic signals, map them to a multi-dimensional feature space, and calculate the deviation between the current working point and the optimal process window. S400. Based on the degree of deviation, a nonlinear dynamic control algorithm is applied to adjust the operating parameters of the ultrasonic transducer in real time to drive the current operating point into the optimal process window.

6. The ultrasonic incision control method for an incision machine according to claim 5, characterized in that, In step S100, the workflow for constructing the optimal process window model includes: Fourier transform and feature extraction were performed on the force and acoustic signals from the N sets of collected calibration data to obtain... There are 1 data point, each data point is characterized by cutting force. Harmony Acoustic Spectrum constitute; Using Gaussian mixture model to analyze this Probability density estimation is performed on each data point to establish the probability density function for the optimal process window. The mathematical formula is: in, For data points in the feature space; The number of Gaussian distributions in the mixture model; For this is the first A Gaussian-distributed mixture of weights, satisfying ; For the first One Gaussian component; For the first The mean vector of Gaussian components; For the first The covariance matrix of Gaussian components; is the dimension of the feature space; This is the set of model parameters that need to be learned using the EM algorithm.

7. The ultrasonic incision control method for an incision machine according to claim 5, characterized in that, In step S400, the ultrasonic output power is adjusted. The workflow includes: Real-time calculation of current cutting force With the center of the optimal process window error ; The power adjustment is calculated using a nonlinear gain controller incorporating an sigmoid saturation function. ; Ultrasonic output power The update formula is: in, yes The final output power at any given moment; This is the base power setting value; for The cutting force error at any moment, It is a real-time force feedback value. This represents the expected value of the optimal force characteristic in the optimal process window model; These are the proportional and integral gain coefficients, respectively. It is the hyperbolic tangent function; For adjustment The scaling factor of the function shape; This is the integral term over historical error.

8. The ultrasonic incision control method for an incision machine according to claim 5, characterized in that, Step S300 also includes diagnosing tool wear and compensatoryly adjusting the operating frequency. The workflow includes: For real-time acoustic signals Perform a short-time Fourier transform to obtain its time-frequency spectrum; Extracting the higher harmonic energy ratio from the time-frequency spectrum. Spectral Entropy As a characteristic indicator of quality degradation; Construct a tool health index It decays over time depending on the degree of deterioration of the characteristic indicators; According to health index For basic operating frequency Make fine-tuning compensation; Health Index The iterative update algorithm is as follows: Operating frequency The compensatory adjustment formula is: in, for The health index of the cutting tool at any given time; This is the learning rate or decay rate constant; This represents the ratio of higher harmonic energy to fundamental frequency energy. The preset failure threshold; This is the spectral entropy. The preset failure threshold; for Operating frequency after time-compensation; Basic operating frequency; This is the maximum frequency compensation coefficient; It is a natural exponential function; is the time constant.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, the processor running the computer program to implement the ultrasonic incision control method as described in any one of claims 5-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by a processor to implement the ultrasonic incision control method as described in any one of claims 5-8.

Images