Binding force monitoring method and system of banding machine

Abstract

The invention relates to the technical field of data processing, and discloses a binding force monitoring method and system of a binding machine. The method comprises the following steps: acquiring stress-deformation data through a high-frequency sampling array to form a time sequence characteristic matrix; reconstructing the signal by using adaptive filtering to obtain a key feature point set; constructing a prediction engine by using a deep tensor network based on the feature point set; using a prediction engine to identify waveform abnormity to obtain a deviation characteristic spectrum; generating a compensation strategy graph through a hybrid decision-making system according to the characteristic spectrum; and obtaining a steady-state control scheme through a parameter self-correction algorithm according to the strategy graph. According to the high-frequency binding force monitoring method based on multi-sensor fusion and the self-adaptive control technology, accurate binding force control over different materials is achieved, and the working efficiency and binding quality of the binding machine are improved.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a method and system for monitoring the tightening force of a strapping machine. Background Art

[0002] With the continuous development of automated packaging technology, strapping machines, as important packaging equipment, are widely used in logistics, warehousing, manufacturing and other fields. The traditional strapping machine's tightening force control mainly relies on mechanical force devices or simple closed-loop electrical control systems, and the strapping of different materials is achieved by setting fixed tightening force parameters. This type of control method usually uses a tension sensor to monitor the strap tension in real time, and compares the measured value with the preset value through a comparator. When the preset tension value is reached, the cutting and heat sealing process is triggered to complete the strapping operation. In recent years, some advanced strapping machines have begun to introduce PLC control systems and touch screen human-machine interfaces, allowing operators to manually adjust the tightening force setting value according to different materials, thereby improving the adaptability of the equipment.

[0003] However, there are still many shortcomings in the tightening force monitoring methods of strapping machines in the prior art. First, the traditional single sensor monitoring method cannot fully capture the complex interaction between the strap and the tied object, resulting in low control accuracy of the tightening force; second, the control strategy with fixed parameters is difficult to adapt to the changing characteristics of different materials, which can easily cause problems of over-tightening or over-loose bundling; third, there is a lack of intelligent fault diagnosis mechanism. When an abnormality occurs in the tightening process, it is difficult to quickly locate the cause of the fault and automatically adjust it; fourth, the response lag and oscillation problems in the control process have not been effectively solved, affecting the bundling efficiency and quality. These problems are particularly evident when dealing with high-value, easily deformed or specially shaped materials, which seriously limits the application of strapping machines in high-end manufacturing and precision packaging. Summary of the invention

[0004] The present application provides a method and system for monitoring the tightening force of a strapping machine, which are used to solve the technical problems in the prior art of low tightening force control accuracy, poor material adaptability and lack of intelligent fault diagnosis and automatic adjustment capabilities of strapping machines. By constructing a high-frequency tightening force monitoring method and adaptive control technology based on multi-sensor fusion, accurate tightening force control of different materials can be achieved, thereby improving the working efficiency and strapping quality of the strapping machine.

[0005] In a first aspect, the present application provides a method for monitoring the tightening force of a strapping machine, the method comprising: collecting stress-deformation coupling data during the operation of the strapping machine through a high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series feature matrix; based on the multi-dimensional tightening force time series feature matrix, performing signal reconstruction and dynamic threshold segmentation through an adaptive filtering algorithm to obtain a set of key feature points of the tightening process; based on the set of key feature points of the tightening process, constructing a tightening force dynamic response model through a cross-validated deep tensor network to obtain an accurate prediction engine with material adaptive capability; based on the accurate prediction engine, performing multi-modal abnormal pattern recognition on the tightening force waveform collected in real time to obtain a hierarchical tightening deviation characteristic spectrum; based on the hierarchical tightening deviation characteristic spectrum, performing intelligent fault diagnosis through a hybrid decision tree-fuzzy inference system to obtain a tightening force deviation compensation strategy diagram; based on the tightening force deviation compensation strategy diagram, generating a closed-loop control instruction for the strapping machine through a parameter self-correction iterative algorithm to obtain a tightening force steady-state control scheme.

[0006] In a second aspect, the present application provides a tightening force monitoring system for a strapping machine, the tightening force monitoring system for the strapping machine comprising: A coupling module is used to collect stress-deformation coupling data during the operation of the strapping machine through a high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series characteristic matrix; A reconstruction module, used to perform signal reconstruction and dynamic threshold segmentation through an adaptive filtering algorithm according to the multi-dimensional tightening force time series feature matrix, so as to obtain a set of key feature points of the tightening process; A construction module is used to construct a dynamic response model of the tightening force through a cross-validated deep tensor network based on the key feature point set of the tightening process, so as to obtain an accurate prediction engine with material self-adaptation capability; An identification module, for performing multi-modal abnormal pattern recognition on the real-time collected tightening force waveform based on the precise prediction engine, to obtain a hierarchical tightening deviation characteristic spectrum; A diagnosis module, used for performing intelligent fault diagnosis through a hybrid decision tree-fuzzy inference system according to the hierarchical tightening deviation characteristic spectrum to obtain a tightening force deviation compensation strategy diagram; A generation module is used to generate a closed-loop control instruction for a strapping machine through a parameter self-correction iterative algorithm according to the tightening force deviation compensation strategy diagram to obtain a tightening force steady-state control scheme.

[0007] The third aspect of the present invention provides a computer device, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the computer device executes the above-mentioned method for monitoring the tightening force of the strapping machine.

[0008] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned method for monitoring the tightening force of a strapping machine.

[0009] In the technical solution provided in the present application, the stress-deformation coupling data of the strapping machine during operation is collected through a high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series feature matrix, which not only realizes the comprehensive monitoring of the physical state of the tightening process, but also provides a high-quality original data basis for subsequent analysis, overcoming the limitations of the traditional single sensor monitoring method; an adaptive filtering algorithm is used for signal reconstruction and dynamic threshold segmentation to obtain a set of key feature points in the tightening process, effectively filtering out environmental interference and mechanical vibration noise, greatly improving signal quality and feature extraction accuracy, and making the identification of key operating points more accurate and reliable; a tightening force dynamic response model is constructed through a cross-validated deep tensor network to obtain an accurate prediction engine with material adaptive capabilities. The deep learning model can automatically learn the mechanical properties and response laws of different materials, so that the strapping machine can intelligently adapt to various materials. The tightening requirements of materials have significantly improved the versatility and adaptability of the equipment; based on the precise prediction engine, the multi-modal abnormal pattern recognition of the tightening force waveform collected in real time is carried out to obtain the hierarchical tightening deviation characteristic spectrum, which realizes the early detection and accurate classification of abnormal situations and provides a reliable basis for fault diagnosis; intelligent fault diagnosis is carried out through the hybrid decision tree-fuzzy reasoning system to obtain the tightening force deviation compensation strategy diagram. This hybrid intelligent algorithm fully combines the clear logical reasoning ability of the decision tree and the uncertainty processing ability of the fuzzy reasoning system, and can accurately identify the cause of the fault and generate targeted compensation strategies, which greatly improves the accuracy and efficiency of fault handling; finally, the closed-loop control instructions of the strapping machine are generated through the parameter self-correction iterative algorithm, and the steady-state control scheme of the tightening force is obtained, which realizes the automatic optimization and real-time adjustment of the control parameters and ensures the stability and consistency of the tightening force under various working conditions. It is particularly worth emphasizing that the deep tensor network algorithm used in this scheme innovatively integrates tensor decomposition and deep learning technology. It can not only efficiently process multi-dimensional data structures and capture the complex coupling relationship between different physical quantities, but also significantly enhance the generalization ability of the model through K-fold cross-validation technology, so that the prediction engine can accurately respond to new materials and working conditions that have never been seen; and the introduction of the genetic-simulated annealing hybrid algorithm cleverly combines the complementary advantages of the two optimization algorithms, achieving a good balance between global search and local fine-tuning, and providing an efficient and stable calculation method for the optimization of control parameters, thereby improving the intelligence level and work efficiency of the strapping machine's tightening force monitoring and control, and realizing precise tightening force control of different materials. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0011] Figure 1 A schematic diagram of an embodiment of a method for monitoring the tightening force of a strapping machine in an embodiment of the present application; Figure 2 A schematic diagram of an embodiment of a tightening force monitoring system of a strapping machine in an embodiment of the present application; Figure 3 It is a schematic block diagram of the structure of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0012] Embodiments of the present application provide a method and system for monitoring the tightening force of a strapping machine. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0013] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 In the embodiment of the present application, an embodiment of the method for monitoring the tightening force of the strapping machine includes: Step S101, collecting stress-deformation coupling data during the operation of the strapping machine through a high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series characteristic matrix; Step S102, based on the multi-dimensional tightening force time series feature matrix, signal reconstruction and dynamic threshold segmentation are performed through an adaptive filtering algorithm to obtain a set of key feature points of the tightening process; Step S103, based on the key feature point set of the tightening process, a tightening force dynamic response model is constructed through a cross-validated deep tensor network to obtain an accurate prediction engine with material self-adaptation capability; Step S104: Based on the precise prediction engine, multi-modal abnormal pattern recognition is performed on the real-time collected tightening force waveform to obtain a hierarchical tightening deviation characteristic spectrum; Step S105, based on the hierarchical tightening deviation characteristic spectrum, intelligent fault diagnosis is performed through a hybrid decision tree-fuzzy inference system to obtain a tightening force deviation compensation strategy diagram; Step S106: Based on the tightening force deviation compensation strategy diagram, a closed-loop control instruction for the strapping machine is generated through a parameter self-correction iterative algorithm to obtain a tightening force steady-state control solution.

[0014] It is understandable that the execution subject of the present application may be a tightening force monitoring system of a strapping machine, or a terminal or a server, which is not limited here. The present application embodiment is described by taking a server as the execution subject as an example.

[0015] Specifically, high-frequency sampling sensor array data acquisition is carried out. By arranging pressure sensors, displacement sensors and angle sensors on the strapping machine's tightening device to form a high-frequency sampling sensor array, these sensors synchronously collect contact stress and deformation data between the strap and the tied object at a frequency of 1000Hz. The collected raw data will be added with unified timestamp information to form a multi-source tightening force parameter stream associated with the time series, and then the analog-to-digital conversion process will be performed to generate a digital stress-deformation sampling lattice. These lattices are tensor arranged and matrix reconstructed to construct a tightening state cube containing three-dimensional information of pressure, displacement and angle. Finally, the tightening state cube is reduced in dimension and data enhanced by the feature extraction method to obtain a multi-dimensional tightening force time series feature matrix. Based on the multi-dimensional tightening force time series feature matrix, signal reconstruction and dynamic threshold segmentation are performed by an adaptive filtering algorithm. In this stage, the tightening force time series feature matrix is ​​subjected to wavelet transform multi-scale decomposition to obtain the frequency domain component spectrum of the tightening force signal. The noise energy distribution diagram is calculated based on the frequency domain component spectrum and the signal-to-noise ratio parameters of each frequency band are determined. An adaptive bandpass filter group is constructed based on the signal-to-noise ratio parameters, and a multi-band filter template is generated and applied to the tightening force time series feature matrix to form a denoised tightening force smooth curve. Subsequently, a sliding window analysis is performed on the smooth curve to calculate the slope change rate of the curve and mark the tightening force mutation point. Finally, the key working points in the tightening process are extracted through the dynamic threshold segmentation algorithm to form a set of key feature points in the tightening process.

[0016] The dynamic response model of the tightening force is constructed through a cross-validated deep tensor network. The key feature point set of the tightening process is converted into a tensor representation form, and the tightening force data training matrix is ​​constructed. The matrix is ​​subjected to data amplification and standardization to generate a model training sample library. The deep tensor network structure is designed according to the model training sample library to form a tightening force neural hierarchical architecture. The tightening force neural hierarchical architecture is trained and verified by the K-fold cross-validation method to obtain a parameter-optimized tightening force model. The material property compensation function is constructed based on the parameter-optimized tightening force model, and a multi-type material adaptation coefficient table is generated. The multi-type material adaptation coefficient table is integrated into the parameter-optimized tightening force model to obtain an accurate prediction engine with material adaptive capabilities. Multi-modal abnormal pattern recognition is performed on the tightening force waveform collected in real time. The standard tightening force prediction value generated by the accurate prediction engine is calculated by difference with the tightening force waveform collected in real time to obtain the original tightening force deviation sequence. The waveform features of this sequence are extracted by time-frequency analysis to form a tightening force deviation feature vector. A multidimensional abnormal index system is constructed based on the tightening force deviation feature vector to generate a tightening deviation evaluation matrix. The matrix is ​​subjected to abnormal pattern classification through cluster analysis to obtain a tightening abnormal type library. The tightening abnormal type library is matched with the current tightening force deviation for similarity, the current abnormal pattern category is identified, and the severity of the tightening force deviation is graded and the spatial distribution is analyzed according to the abnormal pattern category to obtain a hierarchical tightening deviation feature spectrum. Based on the hierarchical tightening deviation feature spectrum, intelligent fault diagnosis is performed through a hybrid decision tree-fuzzy inference system. The hierarchical tightening deviation feature spectrum is input into the fault feature parser to extract the tightening fault symptom feature set. A decision tree structure is constructed based on the tightening fault symptom feature set to form a preliminary diagnosis path map for tightening faults. The uncertain nodes in the preliminary diagnosis path diagram of the tightening fault are quantified by the fuzzy membership function to obtain the tightening fault possibility distribution table. Bayesian probability reasoning is performed based on the tightening fault possibility distribution table to calculate the probability ranking of each fault type, from which the main fault source is identified and the tightening force fault diagnosis result is generated. The compensation parameter calculation and strategy mapping are performed on the tightening force fault diagnosis result to obtain the tightening force deviation compensation strategy diagram.

[0017] According to the tightening force deviation compensation strategy diagram, the closed-loop control instructions of the strapping machine are generated through the parameter self-correction iterative algorithm. The torque transfer curve and response delay parameters are extracted from the tightening force deviation compensation strategy diagram, and a tension-relaxation two-state control mapping table is constructed. The mapping table is subjected to physical constraint boundary inspection to form a parameter adjustment range. The control variables within the parameter adjustment range are globally searched and locally fine-tuned through the genetic-simulated annealing hybrid algorithm to obtain a multi-scale tightening force correction. The correction is dynamically smoothed through a nonlinear response compensation function to generate a progressive adjustment curve for vibration resistance. According to the progressive adjustment curve, a feedback gain adaptive adjustment mechanism is designed to output a parameter correction control quantity with time-varying characteristics. The parameter correction control quantity and the historical tightening force data are jointly analyzed in the time domain and frequency domain and then optimized and integrated to obtain a steady-state control scheme for the tightening force.

[0018] For example, in the operation of a strapping machine bundling cartons in a packaging factory, the raw data collected by the high-frequency sampling sensor array includes the pressure change curve between the strap and the carton recorded by the pressure sensor, the strap tightening displacement recorded by the displacement sensor, and the angle change data of the strapping machine clamp recorded by the angle sensor. These data are timestamped and converted into a three-dimensional data matrix. Wavelet decomposition is used to analyze the noise, and the signals below 40Hz are identified as valid signals. A 40-60Hz bandpass filter group is constructed to filter out the vibration noise of the working environment, making the tightening force curve smoother. The smoothed curve is subjected to sliding window analysis, and when the rate of change of the tightening force exceeds 0.5N / ms, it is marked as a key feature point. These feature points are input into a deep tensor network trained with K=5-fold cross validation to construct a dynamic response model, which can automatically adjust the tightening parameters for different materials such as corrugated cartons and plastic boxes. When the real-time tightening force deviates from the predicted value, the deviation characteristics are extracted through time-frequency analysis, and the abnormal increase in the friction force of the strap is found through matching with the abnormality library. The intelligent diagnosis system determines that the strap tension controller is worn through the decision tree and fuzzy reasoning, generates a corresponding compensation strategy, and adjusts the motor output torque through the parameter self-correction algorithm to achieve stable tightening force control.

[0019] In the embodiment of the present application, the stress-deformation coupling data of the strapping machine during operation is collected by a high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series feature matrix, which not only realizes the comprehensive monitoring of the physical state of the tightening process, but also provides a high-quality original data basis for subsequent analysis, overcoming the limitations of the traditional single sensor monitoring method; an adaptive filtering algorithm is used for signal reconstruction and dynamic threshold segmentation to obtain a set of key feature points in the tightening process, effectively filtering out environmental interference and mechanical vibration noise, greatly improving signal quality and feature extraction accuracy, and making the identification of key operating points more accurate and reliable; a tightening force dynamic response model is constructed through a cross-validated deep tensor network to obtain an accurate prediction engine with material adaptive capabilities. The deep learning model can automatically learn the mechanical properties and response laws of different materials, so that the strapping machine can intelligently adapt to various materials. The system can meet the tightening requirements and significantly improve the versatility and adaptability of the equipment; based on the precise prediction engine, the multi-modal abnormal pattern recognition of the tightening force waveform collected in real time is carried out to obtain the hierarchical tightening deviation characteristic spectrum, which realizes the early detection and accurate classification of abnormal situations and provides a reliable basis for fault diagnosis; intelligent fault diagnosis is carried out through the hybrid decision tree-fuzzy reasoning system to obtain the tightening force deviation compensation strategy diagram. This hybrid intelligent algorithm fully combines the clear logical reasoning ability of the decision tree and the uncertainty processing ability of the fuzzy reasoning system, and can accurately identify the cause of the fault and generate targeted compensation strategies, which greatly improves the accuracy and efficiency of fault handling; finally, the closed-loop control instructions of the strapping machine are generated through the parameter self-correction iterative algorithm to obtain the steady-state control scheme of the tightening force, which realizes the automatic optimization and real-time adjustment of the control parameters and ensures the stability and consistency of the tightening force under various working conditions. It is particularly worth emphasizing that the deep tensor network algorithm used in this scheme innovatively integrates tensor decomposition and deep learning technology. It can not only efficiently process multi-dimensional data structures and capture the complex coupling relationship between different physical quantities, but also significantly enhance the generalization ability of the model through K-fold cross-validation technology, so that the prediction engine can accurately respond to new materials and working conditions that have never been seen; and the introduction of the genetic-simulated annealing hybrid algorithm cleverly combines the complementary advantages of the two optimization algorithms, achieving a good balance between global search and local fine-tuning, and providing an efficient and stable calculation method for the optimization of control parameters, thereby improving the intelligence level and work efficiency of the strapping machine's tightening force monitoring and control, and realizing precise tightening force control of different materials.

[0020] In a specific embodiment, the process of executing step S101 may specifically include the following steps: (1) A pressure sensor, a displacement sensor, and an angle sensor are arranged on the tightening device of the strapping machine to form a high-frequency sampling sensor array to collect the contact stress and deformation data between the strap and the tied object; (2) The output signal of the high-frequency sampling sensor array is synchronously sampled at a frequency of 1000 Hz through the signal acquisition circuit to obtain the original stress-deformation coupling data; (3) Add unified timestamp information to the original stress-deformation coupling data to form a multi-source tightening force parameter flow with time series correlation; (4) The multi-source tightening force parameter stream is processed by analog-to-digital conversion to generate a digital stress-deformation sampling matrix; (5) Perform tensor arrangement and matrix reconstruction on the digital stress-deformation sampling point array to construct a taut cube containing three-dimensional information of pressure, displacement, and angle; (6) The dimension reduction and data enhancement of the tightening state cube are performed through feature extraction method to obtain a multi-dimensional tightening force time series feature matrix.

[0021] Specifically, pressure sensors, displacement sensors and angle sensors are reasonably arranged on the tightening device of the strapping machine to form a high-frequency sampling sensor array. The pressure sensor is mainly used to measure the contact stress between the strap and the tied object, the displacement sensor is used to measure the displacement change of the strap during the tightening process, and the angle sensor is used to measure the angle change of the strapping machine chuck. These three sensors work together to fully reflect the changes in the physical state during the tightening process. The output signal of the high-frequency sampling sensor array is synchronously sampled at a frequency of 1000Hz through the signal acquisition circuit. This high-frequency synchronous sampling can ensure that the fast-changing details of the strapping machine's tightening process are captured and accurate stress-deformation original coupling data are obtained. The sampling frequency of 1000Hz means that 1000 data points are collected per second, which can effectively capture transient changes during the high-speed operation of the strapping machine.

[0022] Add unified timestamp information to the original stress-deformation coupling data. By adding precise time stamps to each sampling point, the data from different sensors can be accurately aligned in the time dimension to form a multi-source tightening force parameter stream with time series association. It ensures that the relationship between the changes of various physical quantities over time can be accurately restored during the data analysis process. The multi-source tightening force parameter stream is processed by analog-to-digital conversion, and the analog signal is converted into a digital signal to generate a digital stress-deformation sampling lattice. During the analog-to-digital conversion process, a high-precision ADC converter is used to ensure the accuracy of the data, and anti-aliasing filtering technology is applied to avoid high-frequency interference. The digital stress-deformation sampling lattice is tensor-arranged and matrix-reconstructed. The discrete data points are rearranged according to the spatial and temporal dimensions to construct a tightening state cube containing three-dimensional information of pressure, displacement and angle. This three-dimensional data structure can fully reflect the state changes during the tightening process. The tightening state cube is reduced in dimension and data enhanced by feature extraction methods to obtain a multi-dimensional tightening force time series feature matrix. In the feature extraction process, dimensionality reduction techniques such as principal component analysis (PCA) are applied to reduce data redundancy and extract key features. At the same time, data enhancement technology is used to enrich training data and enhance the generalization ability of the model. The generated multi-dimensional tightening force time series feature matrix will contain the key feature information of the tightening process of the strapping machine.

[0023] In a specific embodiment, the process of executing step S102 may specifically include the following steps: (1) The multi-dimensional tightening force time series feature matrix is ​​decomposed at multiple scales by wavelet transform to obtain the frequency domain component spectrum of the tightening force signal; (2) Calculate the noise energy distribution diagram based on the frequency domain component spectrum and determine the signal-to-noise ratio parameters of each frequency band; (3) Construct an adaptive bandpass filter group based on the signal-to-noise ratio parameters to generate a multi-band filtering template; (4) Applying the multi-band filtering template to the multi-dimensional tightening force time series feature matrix to form a denoised tightening force smooth curve; (5) Perform sliding window analysis on the tightening force smooth curve, calculate the rate of change of the curve slope, and mark the tightening force mutation point; (6) According to the mutation point of the tightening force, the key working points in the tightening process are extracted through the dynamic threshold segmentation algorithm to obtain the key feature point set of the tightening process.

[0024] Specifically, the multi-dimensional tightening force time series feature matrix is ​​decomposed at multiple scales by wavelet transform. Wavelet transform is a time-frequency analysis tool that can analyze signal characteristics at different scales. In the tightening force monitoring of the strapping machine, discrete wavelet transform is used to decompose the tightening force time series feature matrix into sub-signals of different frequency bands, thereby obtaining the frequency domain component spectrum of the tightening force signal. The specific operation is to select a wavelet basis function suitable for the tightening force signal characteristics, such as Daubechies wavelet or Symlet wavelet, decompose the tightening force time series data into high-frequency detail coefficients and low-frequency approximate coefficients through high-pass filters and low-pass filters, recursively decompose the low-frequency approximate coefficients, form a multi-layer decomposition structure, and obtain the frequency domain component spectrum. The noise energy distribution diagram is calculated according to the frequency domain component spectrum, and the effective components and noise components in the signal are identified by analyzing the energy distribution characteristics of each frequency band. In the tightening force monitoring of the strapping machine, the high-frequency part usually contains noise such as environmental interference and mechanical vibration, while the low-frequency part contains effective information about the tightening force. During the calculation process, the energy of each frequency band is counted and compared with the known tightening force signal pattern to calculate the signal-to-noise ratio parameters of each frequency band. The signal-to-noise ratio represents the ratio of effective signal energy to noise energy. The frequency band with a high signal-to-noise ratio contains more effective information. According to the signal-to-noise ratio parameters of each frequency band, a set of bandpass filters is designed, and each filter corresponds to a specific frequency band. In the tightening force monitoring of the strapping machine, the center frequency and bandwidth of the filter will be dynamically adjusted according to the signal-to-noise ratio. The frequency band with a high signal-to-noise ratio uses a narrower passband to retain more details, and the frequency band with a low signal-to-noise ratio uses a wider passband for smoothing. This set of filters constitutes a multi-band filter template, which can be adaptively processed according to the noise characteristics under different environments and working conditions. The multi-band filter template is applied to the multi-dimensional tightening force time series feature matrix. By convolving each frequency band of the time series feature matrix with the corresponding filter template, the irrelevant noise components are filtered out and the effective tightening force signal is retained. The filtered signals of each frequency band are reconstructed and synthesized into a denoised tightening force smooth curve. This smooth curve retains the main characteristics of the original tightening force change, while significantly reducing the impact of noise interference, providing a high-quality data basis for subsequent analysis.

[0025] The fifth step is to perform sliding window analysis on the smooth curve of the tightening force. A fixed-size window is set, and it slides point by point along the time axis to calculate the slope of the tightening force curve in the window and the rate of change of the slope. The slope indicates the speed of change of the tightening force over time, and the rate of change of the slope reflects the acceleration of the change of the tightening force. During the tightening process of the strapping machine, when the slope change rate exceeds a certain threshold, it means that the tightening force has undergone a sudden change. These mutation points usually correspond to key state changes in the tightening process, such as starting to tighten, reaching the preset tension, and completing the tightening. Finally, according to the tightening force mutation points, the key working condition points in the tightening process are extracted by the dynamic threshold segmentation algorithm. The dynamic threshold segmentation algorithm sets different threshold standards for different stages according to the statistical characteristics of the tightening force data and the historical working condition characteristics. The algorithm performs cluster analysis on the marked tightening force mutation points, identifies mutation point groups with similar characteristics, and then sets corresponding threshold conditions according to the characteristics of each group of mutation points, screens out the key points that truly represent the working condition changes, and forms a set of key feature points in the tightening process. These key feature point sets accurately characterize the key stages of the strapping process of the strapping machine.

[0026] For example, when the packaging box is strapped, the multi-dimensional tightening force time series feature matrix obtained contains 30,000 data points collected by three sensors at a sampling rate of 1000Hz during the 10-second tightening process. Through wavelet transform, the data are decomposed into five frequency bands: 0-10Hz, 10-20Hz, 20-40Hz, 40-80Hz and 80-160Hz using a 5-level Daubechies-4 wavelet decomposition. The analysis shows that the 0-40Hz band mainly contains the effective signal of the tightening force, with a signal-to-noise ratio of about 15dB, while the 40-160Hz band mainly contains mechanical vibration and electrical interference, with a signal-to-noise ratio of about 3dB. Based on this, an adaptive bandpass filter group is constructed, which uses a narrower passband to retain details in the 0-40Hz band and suppresses the intensity of the 40-160Hz band. After applying the filter group, the smooth curve of the tightening force clearly shows the three stages of the tightening process: initial tightening, maintaining tension, and completing locking. The sliding window analysis with a window size of 200ms was used to calculate the slope change rate. When the change rate exceeded 0.5N / ms, it was marked as a mutation point. A total of 7 mutation points were detected. By setting a dynamic threshold, 4 key working points were extracted: the strap contacted the packaging box (t=1.2s), reached the initial tension (t=3.5s), reached the target tension (t=6.8s) and completed the locking (t=8.7s). These four points accurately described the key state changes of the entire tightening process.

[0027] In a specific embodiment, the process of executing step S103 may specifically include the following steps: (1) Convert the key feature point set of the tightening process into a tensor representation form and construct the tightening force data training matrix; (2) Perform data amplification and standardization on the tightening force data training matrix to generate a model training sample library; (3) Design a deep tensor network structure based on the model training sample library to form a binding force neural hierarchical architecture; (4) The binding force neural hierarchy architecture is trained and verified through the K-fold cross-validation method to obtain a parameter-optimized binding force model; (5) Construct material property compensation function based on parameter optimization tightening force model and generate multi-type material adaptation coefficient table; (6) The adaptation coefficient table of multiple types of materials is integrated into the parameter optimization tightening force model to obtain an accurate prediction engine with material adaptability.

[0028] Specifically, the key feature point set of the tightening process is converted into a tensor representation form to construct a tightening force data training matrix. Tensor representation is a multidimensional data structure that can simultaneously express information relationships in multiple dimensions. In the tightening force monitoring of the strapping machine, the key feature point set is converted from a time series form to a third-order tensor structure, where the first dimension represents the sample index, the second dimension represents the time window, and the third dimension represents the feature type (such as pressure, displacement, angle, etc.). Through this conversion, each tightening process is encoded as a multidimensional feature block, capturing the coupling relationship between different physical quantities to form a tightening force data training matrix.

[0029] The tightening force data training matrix is ​​amplified and standardized to generate a model training sample library. Data amplification refers to a technology that generates more training samples through a series of transformations to enhance the generalization ability of the model. In the tightening force monitoring of the strapping machine, amplification methods such as time window sliding, random noise addition, and signal scale scaling are used to expand the original tightening force data training matrix into a larger data set. Standardization is to convert features of different dimensions into a unified scale to eliminate the influence of dimensions. Commonly used standardization methods include Z-score standardization (minus the mean divided by the standard deviation) and Min-Max standardization (linear mapping to the 0-1 interval). The amplified and standardized data constitute the model training sample library, providing rich training materials for subsequent deep learning models. According to the model training sample library, a deep tensor network structure is designed to form a tightening force neural hierarchical architecture. The deep tensor network is a special neural network designed specifically for processing data in tensor form and can retain the multi-dimensional structural characteristics of the data. In the tightening force monitoring of the strapping machine, the designed network structure includes a tensor decomposition layer, a tensor convolution layer, and a fully connected layer. The tensor decomposition layer is used to reduce the complexity of the model and extract key features; the tensor convolution layer is used to capture local patterns in different dimensions; and the fully connected layer is used to integrate features of each dimension to achieve prediction. These layers are organized in a hierarchical manner to form a tight neural hierarchical architecture with powerful feature extraction and pattern recognition capabilities.

[0030] K-fold cross validation is a model evaluation method that divides the data set into K subsets, uses K-1 subsets as training sets each time, and uses the remaining 1 subset as a validation set, and repeats this process K times so that each subset is used as an over-validation set. In the strapping machine tightening force monitoring, 5-fold or 10-fold cross validation is used to comprehensively evaluate the performance of the model under different data partitions to avoid overfitting. During the training process, the back propagation algorithm is used to adjust the network parameters to minimize the mean square error between the predicted value and the actual tightening force. After multiple rounds of iterations, a parameter-optimized tightening force model is obtained, which has good generalization ability. Based on the parameter-optimized tightening force model, a material property compensation function is constructed to generate a multi-type material adaptation coefficient table. The material property compensation function is used to adjust the output of the tightening force model to adapt it to the characteristics of different materials. The compensation function can be expressed as:

[0031] in, Indicates the adjusted tightening force value. represents the original predicted binding force value, represents the material characteristic parameter vector, represents the weight coefficient of the i-th compensation term, represents the function of the original predicted value, Represents a function about material characteristics, and n represents the number of compensation items. By analyzing the response characteristics of different materials during the bundling process, the characteristic parameters of each material type are determined, and a multi-type material adaptation coefficient table is generated, which includes parameters such as elastic coefficient, friction coefficient, hardness, etc. of each material and its corresponding compensation coefficient. The multi-type material adaptation coefficient table is integrated into the parameter optimization bundling force model to obtain an accurate prediction engine with material adaptive capabilities. During the integration process, a material identification mechanism and an adaptive adjustment mechanism are established. The material identification mechanism identifies the type of the current bundled material by analyzing the force-deformation curve characteristics at the initial stage of bundling; the adaptive adjustment mechanism searches for the corresponding compensation parameters from the material adaptation coefficient table based on the identification results, and applies them to the prediction model to achieve adaptive adjustment of different materials. The accurate prediction engine can accurately predict the mechanical response of different materials during the bundling process based on real-time monitoring data, providing an accurate control basis for the strapping machine.

[0032] For example, in the carton strapping process of a packaging production line, 100 sets of key feature point sets of the strapping process of different carton sizes and materials are converted into tensor representation forms. Each sample contains 20 time points × 3 types of sensor data (pressure, displacement, angle), forming a 100×20×3 third-order tensor structure. Data amplification is performed by sliding the time window and adding 5% Gaussian noise, expanding the sample size to 500 groups, and using the Z-score method for standardization, so that all features have a mean of 0 and a standard deviation of 1. Based on the processed data, a deep tensor network structure is designed, which includes 2 layers of tensor decomposition layers (kernel size is 3×2), 3 layers of tensor convolution layers (sliding windows are 3, 2, and 1 respectively), and 2 layers of fully connected layers (the number of hidden units is 64 and 32 respectively). The network is trained using the 10-fold cross-validation method, and the average relative error on the validation set is reduced to 3.2%. The analysis found that the tightening force model needs to be adjusted to different degrees for different materials such as corrugated boxes, plastic boxes, and wooden boxes. Therefore, a material property compensation function was constructed, taking into account the elasticity, friction and other characteristics of the material, and generating an adaptation coefficient table containing 8 common packaging materials. These coefficients were integrated into the prediction model to form an accurate prediction engine, which can automatically apply the corresponding flexible compensation coefficient when a corrugated box is detected, and switch to the rigid compensation coefficient when a wooden box is detected, realizing accurate prediction and control of the tightening process of different materials.

[0033] In a specific embodiment, the process of executing step S104 may specifically include the following steps: (1) The standard tightening force prediction value generated by the precise prediction engine is calculated by difference with the tightening force waveform collected in real time to obtain the original tightening force deviation sequence; (2) extracting waveform features of the original tightening force deviation sequence by time-frequency analysis to form a tightening force deviation feature vector; (3) constructing a multidimensional abnormal index system based on the tightening force deviation characteristic vector and generating a tightening deviation evaluation matrix; (4) performing abnormal pattern classification on the tightening deviation assessment matrix through cluster analysis to obtain a tightening abnormality type library; (5) performing similarity matching between the tightening abnormality type library and the current tightening force deviation to identify the current abnormality pattern category; (6) According to the abnormal pattern category, the severity of the tightening force deviation is graded and the spatial distribution is analyzed to obtain a hierarchical tightening deviation characteristic spectrum.

[0034] Specifically, the standard tightening force prediction value generated by the precise prediction engine is calculated differentially from the tightening force waveform collected in real time to obtain the original tightening force deviation sequence. The standard tightening force prediction value is the ideal tightening force curve generated by the precise prediction engine based on historical data and material characteristics, and the tightening force waveform collected in real time is the tightening force data actually measured during the operation of the strapping machine. The difference calculation adopts a point-to-point method to perform a subtraction operation on the predicted value and the measured value at each time point to form a deviation sequence that changes over time. This original tightening force deviation sequence reflects the difference between the actual tightening process and the ideal state, and contains various deviation information during the operation of the strapping machine. The waveform features of the original tightening force deviation sequence are extracted by time-frequency analysis to form a tightening force deviation feature vector. Time-frequency analysis is a method that simultaneously analyzes the characteristics of a signal in the time domain and frequency domain, and can reveal the frequency characteristics of a signal that changes over time. In the monitoring of the tightening force of the strapping machine, time-frequency analysis technologies such as short-time Fourier transform (STFT) and wavelet transform are used to extract multiple characteristic parameters from the original tightening force deviation sequence, including average deviation, maximum deviation, deviation standard deviation, spectrum energy distribution, main frequency component, etc. These characteristic parameters are combined into a high-dimensional feature vector, which comprehensively describes the various characteristics of the tightening force deviation.

[0035] According to the characteristic vector of the tightening force deviation, a multidimensional abnormal index system is constructed to generate a tightening deviation evaluation matrix. The multidimensional abnormal index system is a set of indicators used to evaluate the degree of system abnormality, covering abnormal performance at different angles and levels. In the monitoring of the tightening force of the strapping machine, the constructed index system includes time domain indicators (such as deviation amplitude, duration), frequency domain indicators (such as high-frequency component ratio, spectrum distribution) and time-frequency joint indicators (such as energy concentration, modal distribution), etc. These indicators are arranged according to a certain organizational structure to form a tightening deviation evaluation matrix. Each row represents a time point or time window, each column represents an evaluation indicator, and the value of the matrix element reflects the evaluation result of the corresponding indicator at a specific time. The fourth step is to classify the abnormal patterns of the tightening deviation evaluation matrix through cluster analysis to obtain the tightening abnormality type library. Cluster analysis is an unsupervised learning method that classifies similar samples into the same category by calculating the similarity between samples. In the monitoring of the tightening force of the strapping machine, algorithms such as K-means clustering, hierarchical clustering or density clustering are used to classify the row vectors of the tightening deviation evaluation matrix, and the deviation patterns with similar characteristics are classified into the same category. The classification results form a tightening anomaly type library, each type represents a specific anomaly mode, such as strap slippage, material deformation, mechanical failure, etc.

[0036] The tightening abnormal type library is matched with the current tightening force deviation to identify the current abnormal pattern category. Similarity matching is a pattern recognition method that determines the category to which the sample to be identified belongs by calculating the similarity between the sample to be identified and the known category sample. In the tightening force monitoring of the strapping machine, the similarity is calculated between the feature vector of the current tightening force deviation and the representative vector of each type in the abnormal type library. Commonly used similarity metrics include Euclidean distance, cosine similarity, Mahalanobis distance, etc. According to the similarity sorting, the most likely pattern category of the current abnormality is identified to provide a basis for subsequent fault diagnosis and processing. According to the abnormal pattern category, the severity of the tightening force deviation is graded and the spatial distribution is analyzed to obtain the hierarchical tightening deviation feature spectrum. The severity grading is to divide the abnormality into different levels, such as slight, moderate, severe, etc., according to factors such as the amplitude, duration and scope of influence of the deviation. Spatial distribution analysis studies the distribution characteristics of the deviation at different locations and stages, and identifies the concentrated area and propagation law of the deviation. Taking into account the category, severity and spatial distribution of abnormal patterns, a hierarchical tightening deviation characteristic spectrum is formed. It is a hierarchical abnormal description structure that contains multi-level information such as abnormal type, degree and distribution, providing a comprehensive basis for subsequent fault diagnosis and processing. In the tightening work of the strapping machine of a packaging production line, the precise prediction engine generates a standard tightening force curve based on the size and material of the current tightened carton. In theory, the curve should show a smooth rise and then stabilize from the beginning of contact to the completion of locking. However, the real-time collected tightening force waveform shows obvious fluctuations and declines in the middle section. The original tightening force deviation sequence is obtained by point-to-point subtraction, and it is found that there is a maximum negative deviation of 15N between 3.5 seconds and 5.2 seconds in the tightening process. Wavelet transform is used to perform time-frequency analysis on this deviation sequence, and the average deviation value, maximum deviation value, deviation duration and spectrum energy distribution are extracted to form a 25-dimensional deviation feature vector. According to this feature vector, a multidimensional abnormal index system including time domain stability index, frequency domain purity index and time-frequency consistency index is constructed to generate a tightening deviation evaluation matrix. The K-means clustering algorithm is used to analyze the deviation assessment matrix collected historically. K=8 is set to obtain 8 typical types of tightening anomalies, including strap slippage, material compression deformation, and mechanical jamming. The cosine similarity between the feature vector of the current deviation and the center vector of these 8 anomaly types is calculated, and it is found that the similarity with the "strap slippage" type is the highest, reaching 0.92, confirming that the current anomaly is a strap slippage type. Based on the amplitude and duration of the deviation, it is determined that the slippage is of medium severity, and by analyzing the distribution of deviations at different locations, it is determined that the slippage occurs in the contact area between the strap and the bottom of the carton, forming a hierarchical tightening deviation feature spectrum containing anomaly type, severity, and location information.

[0037] In a specific embodiment, the process of executing step S105 may specifically include the following steps: (1) Inputting the hierarchical tightening deviation characteristic spectrum into a fault feature analyzer to extract a tightening fault symptom feature set; (2) constructing a decision tree structure based on the symptom feature set of the tightening fault to form a preliminary diagnosis path diagram of the tightening fault; (3) quantifying the uncertain nodes in the preliminary diagnosis path diagram of the tightening fault by using a fuzzy membership function to obtain a tightening fault possibility distribution table; (4) performing Bayesian probability reasoning based on the tightening fault probability distribution table to calculate the probability ranking of each fault type; (5) identifying the main fault source from the occurrence probability ranking of the fault types and generating a tightening force fault diagnosis result; (6) Compensation parameter calculation and strategy mapping are performed on the tightening force fault diagnosis result to obtain a tightening force deviation compensation strategy diagram.

[0038] Specifically, the hierarchical tightening deviation characteristic spectrum is input into the fault feature parser to extract the tightening fault symptom feature set. The fault feature parser is a data processing unit specially used for analyzing fault characteristics, which can extract key fault symptoms from complex characteristic spectra. In the strapping machine tightening force monitoring method, the fault feature parser extracts symptom parameters directly related to fault diagnosis from the hierarchical tightening deviation characteristic spectrum through operations such as feature separation, dimensionality reduction and reorganization, including the time characteristics (such as duration, occurrence frequency), amplitude characteristics (such as maximum deviation, average deviation) and spectrum characteristics (such as main frequency components, harmonic distribution) of the deviation. These parameters constitute the tightening fault symptom feature set, which directly reflects the abnormal performance in the tightening process. A decision tree structure is constructed according to the tightening fault symptom feature set to form a preliminary diagnosis path map for the tightening fault. The decision tree is a classification model with a tree structure, which gradually narrows the fault range through a series of conditional judgments. In the strapping machine tightening force monitoring, the root node of the decision tree represents the initial fault judgment state, the internal node represents the test condition for a certain fault feature, the branch represents the different results of the test, and the leaf node represents the fault diagnosis result. The construction process determines the splitting criteria of the decision tree, commonly used indicators include information gain, Gini coefficient, etc., and then selects the optimal features as split nodes based on these criteria, and recursively constructs subtrees. The resulting preliminary diagnosis path diagram of the bundle fault intuitively represents the reasoning path from symptoms to faults, but this path has the limitation of deterministic judgment.

[0039] The uncertain nodes in the preliminary diagnosis path diagram of the tightening fault are quantified by the fuzzy membership function to obtain the tightening fault possibility distribution table. The fuzzy membership function is a function used in fuzzy set theory to represent the degree of membership of an element to a fuzzy set, and its value is between 0 and 1. In the monitoring of the tightening force of the strapping machine, the fuzzy membership function is introduced for quantification of the nodes with uncertainty in the decision tree, such as the relatively vague concept of "high vibration frequency". Commonly used membership functions include triangular function, trapezoidal function, Gaussian function, etc. The appropriate function type is selected according to the specific characteristic distribution. By calculating the membership of each characteristic value to each fault type, the tightening fault possibility distribution table is obtained. Each element in the table represents the probability of a specific fault caused by a specific characteristic value. Bayesian probability reasoning is performed based on the tightening fault possibility distribution table to calculate the probability ranking of each fault type. Bayesian probability reasoning is a statistical inference method based on Bayes' theorem, which can combine prior knowledge and observation data to calculate the posterior probability. In the monitoring of the tightening force of the strapping machine, the Bayesian probability reasoning process can be expressed as:

[0040] in, Represents the fault type under a given symptom set S The posterior probability of Indicates the fault type The likelihood probability of causing the symptom set S to appear, Indicates the fault type The prior probability of Represents the total number of fault types. Likelihood probability It can be obtained from the distribution table of the possibility of bundling failure, the prior probability It comes from historical fault statistics. The posterior probability of each fault type is calculated by the above formula, and sorted according to the probability size to obtain the probability ranking of each fault type. The main fault source refers to one or several fault types with the highest probability, which are the most likely causes of the current abnormal tightening force. In the tightening force monitoring of the strapping machine, a probability threshold (such as 0.6) is set to identify the fault type with a probability exceeding the threshold as the main fault source. If the probability of no fault type exceeds the threshold, the first few fault types with the highest probability are selected as the composite fault source. The identified main fault source, together with the corresponding symptoms, location, severity and other information, constitutes the tightening force fault diagnosis result. The compensation parameter calculation and strategy mapping are performed on the tightening force fault diagnosis result to obtain the tightening force deviation compensation strategy diagram. The compensation parameter calculation is to determine the control parameter adjustment required to correct the tightening force deviation according to the fault type and severity. Strategy mapping is to convert the compensation parameter into a specific control operation instruction sequence. In the tightening force monitoring of the strapping machine, different compensation strategies are adopted according to different fault types, such as adjusting the motor output torque, modifying the strap preload, adjusting the tightening speed, etc. These compensation strategies are arranged in chronological order and spatial position to form a tightening force deviation compensation strategy diagram, providing intuitive operation guidance for subsequent closed-loop control.

[0041] For example, in the process of tightening the strapping machine of an industrial packaging line, the hierarchical tightening deviation characteristic spectrum shows that there is a periodic force drop in the middle of the tightening, with a drop range of 10-15N and a frequency of about 2Hz. These features are input into the fault feature analyzer to extract 10 key symptom parameters including periodicity (2Hz), amplitude (10-15N), duration (2.5 seconds), and occurrence stage (middle of the tightening), forming a tightening fault symptom feature set. Based on these features, a decision tree is constructed to determine whether the deviation is periodic, then the frequency range, then the amplitude, and finally the occurrence stage, forming a preliminary diagnostic path diagram containing 15 nodes. For uncertain nodes such as "frequency is about 2Hz", Gaussian fuzzy membership function is used for quantification, and the membership of the actual frequency 1.8Hz to "low-frequency vibration (1-3Hz)" and "medium-frequency vibration (3-10Hz)" is calculated to be 0.85 and 0.15 respectively. The other uncertain nodes are processed in the same way to obtain a possibility distribution table containing 8 possible fault types. Based on the prior probability of each fault type in the historical data (such as motor vibration 0.25, strap slippage 0.20, mechanical looseness 0.15, etc.) and the likelihood probability of the current symptoms, the Bayesian formula is used to calculate the posterior probability of motor vibration to be 0.68, strap slippage to be 0.22, and other types of probabilities to be lower. Motor vibration is determined to be the main source of faults, and detailed fault diagnosis results are generated based on its location (drive motor) and severity (medium). For motor vibration faults, the compensation parameters that need to be adjusted are calculated, including reducing the motor speed by 10%, increasing the damping coefficient of the PID controller, adjusting the initial tension of the strap, etc., to form a tightening force deviation compensation strategy diagram containing timing operation instructions.

[0042] In a specific embodiment, the process of executing step S106 may specifically include the following steps: (1) extracting the torque transfer curve and response delay parameters from the tightening force deviation compensation strategy diagram, and constructing a tension-relaxation dual-state control mapping table; (2) performing a physical constraint boundary check on the tension-relaxation dual-state control mapping table to form a parameter adjustment range; (3) A genetic-simulated annealing hybrid algorithm is used to perform global search and local fine-tuning of the control variables within the parameter adjustment range to obtain a multi-scale tightening force correction value; (4) dynamically smoothing the multi-scale tightening force correction amount through a nonlinear response compensation function to generate a progressive anti-vibration adjustment curve; (5) designing a feedback gain adaptive adjustment mechanism according to the progressive adjustment curve, and outputting a parameter correction control quantity with time-varying characteristics; (6) The parameter correction control quantity and the historical tightening force data are subjected to a time-domain-frequency domain joint analysis and then optimized and integrated to obtain a steady-state control scheme for the tightening force.

[0043] Specifically, the torque transfer curve and response delay parameters are extracted from the tightening force deviation compensation strategy diagram to construct a tension-relaxation dual-state control mapping table. The torque transfer curve is a function curve that describes the relationship between the output torque of the driving motor of the strapping machine and the actual tightening force, reflecting how the torque is converted into the actual tightening effect. The response delay parameter represents the time required for the system to generate an actual response from receiving the control signal, including motor start-up delay, mechanical transmission delay, and strap deformation delay. The tension-relaxation dual-state control is the two basic states of the strapping machine's tightening process. The tension state is responsible for increasing the tightening force, and the relaxation state is used to release excessive tightening force. By analyzing the adjustment instructions in the tightening force deviation compensation strategy diagram, the torque input value and the corresponding tightening force output value under different working conditions, as well as the time delay of the state conversion, are extracted to construct a tension-relaxation dual-state control mapping table, providing a basic data structure for subsequent precise control. The physical constraint boundary check is performed on the tension-relaxation dual-state control mapping table to form a parameter adjustment range. Physical constraint boundaries refer to the physical limitations that must be followed in parameter adjustment in actual systems, including the performance limits of the equipment itself, the safe operating range, and the material bearing capacity. In the monitoring of the tightening force of the strapping machine, the physical constraints that need to be tested include the maximum motor output torque, the maximum strap tension, the maximum tightening speed, the minimum response time, etc. During the inspection process, the physical limit values ​​of each parameter are determined, and then the parameters in the tension-relaxation dual-state control mapping table are compared with these limit values ​​to ensure that all parameters are within a safe and reasonable range. For parameters that exceed the range, truncation or scaling is performed to form a parameter adjustment range that meets the physical constraints to ensure that the subsequent optimization process does not produce unrealistic control parameters.

[0044] The genetic-simulated annealing hybrid algorithm is used to perform global search and local fine-tuning of the control variables within the parameter adjustment range to obtain the multi-scale tightening force correction. The genetic-simulated annealing hybrid algorithm is an optimization method that combines the global search capability of the genetic algorithm and the local fine-tuning capability of the simulated annealing algorithm. In the tightening force monitoring of the strapping machine, the optimization process of the hybrid algorithm can be expressed as:

[0045] in, represents the optimization objective function of the hybrid algorithm, represents the parameter set of the genetic algorithm, represents the parameter set of the simulated annealing algorithm, represents the weight coefficient of the ith control variable, represents the mixed ratio of genetic algorithm and simulated annealing algorithm, Indicates the temperature The optimization result of the genetic algorithm for the i-th control variable is, Indicates the temperature The optimization result of the simulated annealing algorithm for the i-th control variable, n represents the total number of control variables. The genetic algorithm is used to perform a global search within the parameter adjustment range, and the possible optimal solution area is found through selection, crossover and mutation operations. Then the simulated annealing algorithm is used to perform local fine-tuning in these areas, and the new solution is accepted or rejected according to the energy function and temperature parameters, and the temperature is gradually reduced to improve the accuracy. The multi-scale tightening force correction is obtained, including large-scale overall adjustment and small-scale fine correction, covering the tightening force control requirements at different time and space scales. The multi-scale tightening force correction is dynamically smoothed by the nonlinear response compensation function to generate a progressive adjustment curve for vibration resistance, which is the fourth step. The nonlinear response compensation function is a mathematical function used to adjust the response characteristics of the system, which can process input signals of different sizes and frequencies to different degrees. In the monitoring of the tightening force of the strapping machine, the nonlinear response compensation function is mainly used to smooth the mutation components in the multi-scale correction amount and reduce the oscillation and impact in the control process. Commonly used nonlinear response compensation functions include S-type functions, exponential smoothing functions, and high-order polynomial functions. Through these functions, the multi-scale tightening force correction is dynamically smoothed to make the change of the control quantity more gentle and continuous, forming a vibration-resistant progressive adjustment curve, avoiding mutations and oscillations in the control process. According to the progressive adjustment curve, the feedback gain adaptive adjustment mechanism is designed to output the parameter correction control quantity with time-varying characteristics. The feedback gain adaptive adjustment mechanism is a mechanism that can dynamically adjust the control parameters according to the system state and external environment, so that the control system has better adaptability and robustness. In the tightening force monitoring of the strapping machine, the feedback gain mainly refers to the proportional, integral and differential coefficients in the PID controller, which directly affect the response characteristics of the control system. The design process analyzes the characteristics of the progressive adjustment curve, determines the response speed and stability requirements required at different stages, and then designs the gain adjustment function so that the parameters of the PID controller can be automatically adjusted according to the current state. The adjusted controller outputs the parameter correction control quantity with time-varying characteristics, which can provide the most suitable control signal at different stages of the tightening process.

[0046] The parameter correction control quantity and the historical tightening force data are analyzed in the time domain and frequency domain, and then optimized and integrated to obtain the steady-state control scheme for the tightening force. The time domain and frequency domain joint analysis is an analysis method that considers both the time characteristics and frequency characteristics of the signal, and can comprehensively evaluate the performance of the control strategy. In the tightening force monitoring of the strapping machine, the parameter correction control quantity is compared with the historical tightening force data in the time domain to analyze the time characteristics of the control response; then the data is converted to the frequency domain through Fourier transform to analyze the frequency response characteristics; finally, the control strategy is optimized and adjusted based on the analysis results in the time domain and frequency domain. During the optimization and integration process, multiple factors such as control accuracy, stability, response speed and anti-interference ability are comprehensively considered to form a steady-state control scheme for the tightening force.

[0047] The above describes the method for monitoring the tightening force of the strapping machine in the embodiment of the present application. The following describes the system for monitoring the tightening force of the strapping machine in the embodiment of the present application. Figure 2 In the embodiment of the present application, an embodiment of the tightening force monitoring system of the strapping machine includes: A coupling module is used to collect stress-deformation coupling data during the operation of the strapping machine through a high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series characteristic matrix; A reconstruction module, used to perform signal reconstruction and dynamic threshold segmentation through an adaptive filtering algorithm according to the multi-dimensional tightening force time series feature matrix, so as to obtain a set of key feature points of the tightening process; A construction module is used to construct a dynamic response model of the tightening force through a cross-validated deep tensor network based on the key feature point set of the tightening process, so as to obtain an accurate prediction engine with material self-adaptation capability; An identification module, for performing multi-modal abnormal pattern recognition on the real-time collected tightening force waveform based on the precise prediction engine, to obtain a hierarchical tightening deviation characteristic spectrum; A diagnosis module, used for performing intelligent fault diagnosis through a hybrid decision tree-fuzzy inference system according to the hierarchical tightening deviation characteristic spectrum to obtain a tightening force deviation compensation strategy diagram; A generation module is used to generate a closed-loop control instruction for a strapping machine through a parameter self-correction iterative algorithm according to the tightening force deviation compensation strategy diagram to obtain a tightening force steady-state control scheme.

[0048] Through the coordinated cooperation of the above-mentioned components, the stress-deformation coupling data of the strapping machine during operation is collected through a high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series feature matrix, which not only realizes the comprehensive monitoring of the physical state of the tightening process, but also provides a high-quality original data basis for subsequent analysis, overcoming the limitations of the traditional single sensor monitoring method; an adaptive filtering algorithm is used for signal reconstruction and dynamic threshold segmentation to obtain a set of key feature points in the tightening process, effectively filtering out environmental interference and mechanical vibration noise, greatly improving signal quality and feature extraction accuracy, and making the identification of key operating points more accurate and reliable; a tightening force dynamic response model is constructed through a cross-validated deep tensor network to obtain an accurate prediction engine with material adaptive capabilities. This deep learning model can automatically learn the mechanical properties and response laws of different materials, enabling the strapping machine to intelligently adapt to various The tightening requirements of various materials have significantly improved the versatility and adaptability of the equipment; based on the precise prediction engine, the multi-modal abnormal pattern recognition of the tightening force waveform collected in real time is carried out to obtain the hierarchical tightening deviation characteristic spectrum, which realizes the early detection and accurate classification of abnormal situations and provides a reliable basis for fault diagnosis; intelligent fault diagnosis is carried out through the hybrid decision tree-fuzzy reasoning system to obtain the tightening force deviation compensation strategy diagram. This hybrid intelligent algorithm fully combines the clear logical reasoning ability of the decision tree and the uncertainty processing ability of the fuzzy reasoning system, and can accurately identify the cause of the fault and generate targeted compensation strategies, which greatly improves the accuracy and efficiency of fault handling; finally, the closed-loop control instructions of the strapping machine are generated through the parameter self-correction iterative algorithm, and the steady-state control scheme of the tightening force is obtained, which realizes the automatic optimization and real-time adjustment of the control parameters and ensures the stability and consistency of the tightening force under various working conditions. It is particularly worth emphasizing that the deep tensor network algorithm used in this scheme innovatively integrates tensor decomposition and deep learning technology. It can not only efficiently process multi-dimensional data structures and capture the complex coupling relationship between different physical quantities, but also significantly enhance the generalization ability of the model through K-fold cross-validation technology, so that the prediction engine can accurately respond to new materials and working conditions that have never been seen; and the introduction of the genetic-simulated annealing hybrid algorithm cleverly combines the complementary advantages of the two optimization algorithms, achieving a good balance between global search and local fine-tuning, and providing an efficient and stable calculation method for the optimization of control parameters, thereby improving the intelligence level and work efficiency of the strapping machine's tightening force monitoring and control, and realizing precise tightening force control of different materials.

[0049] Reference Figure 3 In an embodiment of the present invention, a computer device is also provided. The computer device may be a server, and its internal structure may be as follows: Figure 3As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0050] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0051] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0052] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0053] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0054] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program codes.

[0055] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for monitoring the tightening force of a strapping machine, characterized in that: The method for monitoring the tightening force of the strapping machine comprises: The stress-deformation coupling data of the strapping machine during operation is collected by a high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series characteristic matrix. According to the multi-dimensional tightening force time series feature matrix, signal reconstruction and dynamic threshold segmentation are performed through an adaptive filtering algorithm to obtain a set of key feature points of the tightening process; According to the key feature point set of the tightening process, a tightening force dynamic response model is constructed through a cross-validated deep tensor network to obtain an accurate prediction engine with material adaptive capability; Based on the precise prediction engine, multi-modal abnormal pattern recognition is performed on the real-time collected tightening force waveform to obtain a hierarchical tightening deviation characteristic spectrum; According to the hierarchical tightening deviation characteristic spectrum, intelligent fault diagnosis is performed through a hybrid decision tree-fuzzy inference system to obtain a tightening force deviation compensation strategy diagram; According to the tightening force deviation compensation strategy diagram, a closed-loop control instruction of the strapping machine is generated through a parameter self-correction iterative algorithm to obtain a tightening force steady-state control scheme.

2. The method for monitoring the tightening force of a strapping machine according to claim 1, characterized in that: The stress-deformation coupling data during the operation of the strapping machine is collected by the high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series characteristic matrix, including: A pressure sensor, a displacement sensor and an angle sensor are arranged on the tightening device of the strapping machine to form a high-frequency sampling sensor array to collect contact stress and deformation data between the strap and the tied object; The output signal of the high-frequency sampling sensor array is synchronously sampled at a frequency of 1000 Hz by a signal acquisition circuit to obtain original stress-deformation coupling data; Adding unified timestamp information to the stress-deformation original coupling data to form a multi-source tightening force parameter flow with time series association; The multi-source tightening force parameter stream is subjected to analog-to-digital conversion processing to generate a digital stress-deformation sampling point matrix; Performing tensor arrangement and matrix reconstruction on the digital stress-deformation sampling lattice to construct a tightened state cube containing three-dimensional information of pressure, displacement and angle; The dimension reduction and data enhancement of the tightening state cube are performed by a feature extraction method to obtain a multi-dimensional tightening force time series feature matrix.

3. The method for monitoring the tightening force of a strapping machine according to claim 1, characterized in that: According to the multi-dimensional tightening force time series feature matrix, signal reconstruction and dynamic threshold segmentation are performed through an adaptive filtering algorithm to obtain a set of key feature points of the tightening process, including: Performing multi-scale decomposition on the multi-dimensional tightening force time series characteristic matrix through wavelet transform to obtain a frequency domain component spectrum of the tightening force signal; Calculating a noise energy distribution diagram according to the frequency domain component spectrum to determine a signal-to-noise ratio parameter for each frequency band; constructing an adaptive bandpass filter group according to the signal-to-noise ratio parameter to generate a multi-band filtering template; Applying the multi-band filtering template to the multi-dimensional tightening force time series feature matrix to form a denoised tightening force smooth curve; Performing sliding window analysis on the tightening force smooth curve, calculating the curve slope change rate, and marking the tightening force mutation point; According to the tightening force mutation point, the key operating points in the tightening process are extracted by a dynamic threshold segmentation algorithm to obtain a set of key feature points in the tightening process.

4. The method for monitoring the tightening force of a strapping machine according to claim 1, characterized in that: According to the key feature point set of the tightening process, a tightening force dynamic response model is constructed through a cross-validated deep tensor network to obtain an accurate prediction engine with material adaptive capability, including: The key feature point set of the tightening process is converted into a tensor representation form to construct a tightening force data training matrix; Performing data amplification and standardization processing on the tightening force data training matrix to generate a model training sample library; Designing a deep tensor network structure based on the model training sample library to form a binding force neural hierarchical architecture; The tightening force neural hierarchy architecture is trained and verified by a K-fold cross-validation method to obtain a parameter-optimized tightening force model; Optimizing the tightening force model based on the parameters to construct a material property compensation function and generate a multi-type material adaptation coefficient table; The multi-type material adaptation coefficient table is integrated into the parameter optimization tightening force model to obtain an accurate prediction engine with material self-adaptation capability.

5. The method for monitoring the tightening force of a strapping machine according to claim 1, characterized in that: Based on the precise prediction engine, multi-modal abnormal pattern recognition is performed on the real-time collected tightening force waveform to obtain a hierarchical tightening deviation characteristic spectrum, including: Performing difference calculation between the standard tightening force prediction value generated by the precise prediction engine and the tightening force waveform collected in real time to obtain an original tightening force deviation sequence; Extracting waveform features of the original tightening force deviation sequence by time-frequency analysis to form a tightening force deviation feature vector; Constructing a multi-dimensional abnormal index system according to the tightening force deviation characteristic vector to generate a tightening deviation evaluation matrix; The tightening deviation assessment matrix is ​​subjected to abnormal pattern classification by cluster analysis to obtain a tightening abnormality type library; Perform similarity matching between the tightening abnormality type library and the current tightening force deviation to identify the current abnormal pattern category; The severity of the tightening force deviation is graded and the spatial distribution is analyzed according to the abnormal pattern category to obtain a hierarchical tightening deviation characteristic spectrum.

6. The method for monitoring the tightening force of a strapping machine according to claim 1, characterized in that: According to the hierarchical tightening deviation characteristic spectrum, intelligent fault diagnosis is performed through a hybrid decision tree-fuzzy inference system to obtain a tightening force deviation compensation strategy diagram, including: Inputting the hierarchical tightening deviation characteristic spectrum into a fault feature analyzer to extract a tightening fault symptom feature set; Constructing a decision tree structure according to the tightening fault symptom feature set to form a preliminary diagnosis path diagram for the tightening fault; Quantifying the uncertain nodes in the preliminary diagnosis path diagram of the tightening fault by using a fuzzy membership function to obtain a tightening fault possibility distribution table; Perform Bayesian probability reasoning based on the tightening fault probability distribution table to calculate the occurrence probability ranking of each fault type; Identify the main fault source from the occurrence probability ranking of the fault types, and generate a tightening force fault diagnosis result; Compensation parameter calculation and strategy mapping are performed on the tightening force fault diagnosis result to obtain a tightening force deviation compensation strategy diagram.

7. The method for monitoring the tightening force of a strapping machine according to claim 1, characterized in that: The method of generating a closed-loop control instruction for the strapping machine by a parameter self-correction iterative algorithm based on the tightening force deviation compensation strategy diagram to obtain a tightening force steady-state control scheme includes: Extract the torque transfer curve and response delay parameters from the tightening force deviation compensation strategy diagram to construct a tension-relaxation dual-state control mapping table; Performing a physical constraint boundary check on the tension-relaxation dual-state control mapping table to form a parameter adjustment interval; A genetic-simulated annealing hybrid algorithm is used to perform global search and local fine adjustment on the control variables within the parameter adjustment interval to obtain a multi-scale tightening force correction amount; Dynamically smoothing the multi-scale tightening force correction amount through a nonlinear response compensation function to generate a progressive anti-vibration adjustment curve; According to the progressive adjustment curve, a feedback gain adaptive adjustment mechanism is designed to output a parameter correction control quantity with time-varying characteristics; The parameter correction control quantity and the historical tightening force data are subjected to a time-domain-frequency-domain joint analysis and then optimized and integrated to obtain a tightening force steady-state control scheme.

8. A system for monitoring the tightening force of a strapping machine, used to implement the method for monitoring the tightening force of a strapping machine as claimed in any one of claims 1 to 7, characterized in that: The tightening force monitoring system of the strapping machine comprises: A coupling module is used to collect stress-deformation coupling data during the operation of the strapping machine through a high-frequency sampling sensor array to obtain a multi-dimensional tightening force time series characteristic matrix; A reconstruction module, used to perform signal reconstruction and dynamic threshold segmentation through an adaptive filtering algorithm according to the multi-dimensional tightening force time series feature matrix, so as to obtain a set of key feature points of the tightening process; A construction module is used to construct a dynamic response model of the tightening force through a cross-validated deep tensor network based on the key feature point set of the tightening process, so as to obtain an accurate prediction engine with material adaptive capability; An identification module, for performing multi-modal abnormal pattern recognition on the real-time collected tightening force waveform based on the precise prediction engine, to obtain a hierarchical tightening deviation characteristic spectrum; A diagnosis module, used for performing intelligent fault diagnosis through a hybrid decision tree-fuzzy inference system according to the hierarchical tightening deviation characteristic spectrum to obtain a tightening force deviation compensation strategy diagram; A generation module is used to generate a closed-loop control instruction for a strapping machine through a parameter self-correction iterative algorithm according to the tightening force deviation compensation strategy diagram to obtain a tightening force steady-state control scheme.

9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and is characterized in that when the processor executes the computer program, the method for monitoring the tightening force of the strapping machine described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the processor is enabled to execute the method for monitoring the tightening force of a strapping machine according to any one of claims 1 to 7.

Images