Intelligent packaging production line control system and method

Abstract

The invention belongs to the technical field of monitoring and fault early warning of an intelligent packaging production line, discloses an intelligent packaging production line control system and method, and aims to realize comprehensive monitoring and fault early warning of the intelligent packaging production line. Comprising the steps of collecting electric power parameter data of an electrical component; the power parameter data comprises current, voltage, energy utilization ratio data and a vibration signal of the electrical component; preprocessing the power parameter data to form an electrical characteristic data set; training according to the electrical characteristic data set to obtain a fault state prediction model, and predicting based on the fault state prediction model to obtain a fault probability value of the electrical component; judging whether the intelligent packaging production line has a fault according to the fault probability value of the electrical component; when it is detected that the intelligent packaging production line has a fault, the electrical component having the fault is analyzed, and targeted regulation and control measures are taken, so that efficient fault identification and regulation and control in various special environments are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of monitoring and fault warning of intelligent packaging production lines. More specifically, the present invention relates to a control system and method for an intelligent packaging production line. Background Art

[0002] The control system of an intelligent packaging production line is an indispensable part of modern industrial production. It not only improves the automation level of production but also ensures the consistency and reliability of product quality. With the development of technology, traditional packaging production lines can no longer meet the growing demands for high efficiency, precision, and intelligence. Therefore, the development of more advanced and efficient intelligent control systems has become an important direction for the industry's development.

[0003] The existing control systems of intelligent packaging production lines perform poorly in fault detection under complex environments. The main problems include: data is easily distorted by electromagnetic interference and temperature fluctuations; it is difficult to integrate multi-source sensor data and unable to comprehensively reflect the equipment status; relying on static threshold judgment leads to false alarms and missed alarms; lacking a real-time monitoring mechanism delays problem handling. Therefore, there is an urgent need for a more intelligent and comprehensive solution to ensure data accuracy, integrate multi-source information, and provide real-time and dynamic fault warning and regulation to improve the reliability and efficiency of the production line. Summary of the Invention

[0004] To overcome the above-mentioned defects of the prior art and to achieve the above object, the present invention provides the following technical solutions: A control system and method for an intelligent packaging production line, including:

[0005] A power data acquisition module: used to collect power parameter data of electrical components; the power parameter data includes current, voltage, energy utilization ratio data, and vibration signals of electrical components;

[0006] A data processing module: used to preprocess the power parameter data to form an electrical feature data set;

[0007] An anomaly analysis module: trains a fault state prediction model based on the electrical feature data set, and predicts the fault probability value of electrical components based on the fault state prediction model;

[0008] An anomaly judgment module: used to judge whether there is a fault in the intelligent packaging production line according to the fault probability value of electrical components;

[0009] A fault control module, used to analyze the electrical components with faults and take targeted regulation measures when a fault is detected in the intelligent packaging production line.

[0010] Further, the method of preprocessing the power parameter data to form an electrical feature data set includes:

[0011] Adjust the current and voltage data using an improved temperature compensation algorithm, and calculate the energy utilization ratio based on the adjusted current and voltage data; convert the vibration signal into a frequency spectrum using the Fourier transform method, perform time series decomposition on the frequency spectrum, and extract the signal characteristics in different frequency ranges;

[0012] Perform noise suppression, time alignment, normalization, and data fusion processing on the adjusted current, voltage, energy utilization ratio data, and signal characteristics in different frequency ranges to obtain an electrical feature dataset;

[0013] The method of time alignment is: for each time period wt, calculate the mean values of the current, voltage, and energy utilization ratio at all times respectively to obtain the power characteristics of the wt time period;

[0014] The method of data fusion is: splice the normalized power characteristics and signal characteristics within the time period wt column by column into a performance feature matrix as the electrical feature dataset within the time period wt.

[0015] Furthermore, the method of adjusting the current and voltage data using the improved temperature compensation algorithm includes:

[0016] Use a thermistor sensor to measure the ambient temperature at any time t within the time period wt, and calculate the actual resistance value of the current sensor or voltage sensor at time t through the improved temperature compensation algorithm formula

[0017] where, T t represents the temperature of the current sensor or voltage sensor at time t, R(T t ) is the actual resistance value of the current sensor or voltage sensor at time t, R 0 is the resistance value of the current sensor or voltage sensor at the reference temperature T 0 , and α is the temperature coefficient;

[0018] Define the room temperature as the reference temperature T 0 , measure the resistance values of the current sensor or voltage sensor at different temperatures, and record the corresponding temperature-resistance values (T rr , R rr ), where, T rr represents the rr-th temperature value, and R rr represents the corresponding rr-th resistance value;

[0019] Perform logarithmic conversion on the temperature-resistance values (T rr , R rr ) to linearize the exponential relationship ln(R rr ) = ln(R 0 ) + α(T rr - T0 );

[0020] Use the least squares method to fit the logarithmic conversion formula to obtain the best estimates of ln(R 0 ) and α, and obtain 0 from ln(R

[0021] Adjust the current and voltage according to the actual resistance value; and calculate the energy utilization ratio through the adjusted current and voltage data.

[0022] Furthermore, the method of adjusting the current and voltage according to the actual resistance value includes:

[0023] For a shunt-based current sensor, use the actual resistance value of the current sensor to calculate the current correction coefficient of the current sensor at time t where DI represents the current sensor, RB DI represents the nominal resistance of the current sensor DI, XS DI represents the current correction coefficient of the current sensor, R(T t ) DI represents the actual resistance value of the current sensor;

[0024] Calculate the ratio of the actual measured current on the electrical component of the current sensor at time t to the correction coefficient to obtain the adjusted current data;

[0025] The electrical components include the main power distribution cabinet, motor frequency converter, motor drive, central controller, and heating element in the intelligent packaging production line;

[0026] For a voltage divider-based voltage sensor, use the actual resistance value of the voltage sensor to calculate the voltage correction coefficient of the voltage sensor at time t where DV represents the voltage sensor, XS DV is the voltage correction coefficient of the voltage sensor, are the actual resistance values of the two voltage-dividing resistors R 1 and R 2 of the voltage sensor respectively, are the nominal resistance values of the two voltage-dividing resistors R 1 and R 2 of the voltage sensor respectively;

[0027] Calculate the product of the actual measured voltage on the electrical component of the voltage sensor at time t and the voltage correction coefficient to obtain the adjusted voltage data.

[0028] Furthermore, the method of calculating the energy utilization ratio through the adjusted current and voltage data includes:

[0029] Based on the adjusted current and voltage data, calculate the product of the current and voltage to obtain the apparent power SG;

[0030] Use a power analyzer to measure the power factor of the electrical component at time t Calculate the product of SG and to obtain the active power wherein, represents the phase difference angle between the current and the voltage, represents the power factor;

[0031] Calculate the product of the apparent power and to obtain the reactive power wherein, represents the sine phase difference angle;

[0032] Take the ratio of the active power to the reactive power to obtain the energy utilization ratio.

[0033] Further, the method for converting the vibration signal into a spectrum by using the Fourier transform method includes:

[0034] Set the sampling rate according to the Nyquist sampling theorem, and use a displacement sensor to collect the vibration signal of the electrical component of the intelligent packaging production line within the wt period at the set sampling rate;

[0035] Use band-pass filtering and mean value taking method to filter and remove the mean value of the collected vibration signal;

[0036] Set a time window with a window width of wk;

[0037] Initialize the window index m to start from 0, and extract the signal segment ZD of the m'th window from the processed vibration signal ZD[n], m [n] = ZD[mL + n], where n represents the sample index within the window, used to traverse the sample points within each window, n = 0, 1, 2,..., wk - 1, and L represents the preset sliding step, that is, the overlapping degree between adjacent windows;

[0038] Introduce a window function for the signal segment within the window, use the window function as the weight of the signal segment, define the window function as the Hamming window function, and apply it to the signal segment ZD m [n], to obtain the weighted signal segment ZD hm [m][n] = ZD[mL + n] × hm[n], where hm[n] represents the Hamming window function, representing the weighted signal segment within the m'th time window;

[0039] For the weighted signal segment ZD within each time window w[m][n]Apply the fast Fourier transform to obtain the spectrum within the time window where XP[m, k] represents the spectrum of the m-th time window at the k-th discrete frequency, k = 0, 1, 2,..., wk - 1, j is the imaginary unit satisfying j 2 = -1, e is the natural constant, e -j2πkn / wk is the complex exponential function used to transform the time-domain signal to the frequency domain;

[0040] Integrate the spectra within all time windows, with time as the horizontal axis and frequency as the vertical axis, to form a time-frequency spectrogram containing the complete spectrum within the wt period.

[0041] Furthermore, the method of performing time series decomposition on the spectrum and extracting signal features within different frequency ranges includes:

[0042] Based on the time-frequency spectrogram, mark all peak positions in the time-frequency spectrogram. For each peak, select the frequency band containing the peak and a preset ratio as the frequency range;

[0043] For each frequency range, calculate the sum of the squares of the corresponding spectrum values on the time-frequency spectrogram to obtain the energy value of the frequency range, and select the Ne frequency ranges with the largest energy values among all frequency ranges as the frequency components;

[0044] Use the Morlet wavelet as the basis function, and perform multi-scale continuous wavelet transform for each time point in the time-frequency spectrogram to obtain the wavelet coefficient matrix at different scales

[0045] where f c represents the c-th frequency component, c = 1, 2,..., Ne, tp represents the continuous time variable, a i represents the i-th scale, b j represents the j-th discrete position on the time axis of the time-frequency spectrogram, j = 1, 2,..., J, x(tp, f c ) represents the spectrum component in the time-frequency domain spectrogram, ψ * represents the complex conjugate of the Morlet wavelet function, represents at f c , a i and b j the wavelet coefficient;

[0046] For each frequency component f c , construct a wavelet coefficient matrix with scale a i as rows and b j as columns Take all the calculated wavelet coefficients as the values at each position in the matrix;

[0047] According to the wavelet coefficient matrix under each frequency component, each scale a i is mapped to the corresponding frequency where f cy represents the sampling rate, and ω 0 represents the central angular frequency of the Morlet wavelet function;

[0048] Set the frequency bands as low-frequency band, medium-frequency band and high-frequency band. According to the frequency f i , all scales a i are assigned to the corresponding frequency bands;

[0049] For each frequency band, calculate the total energy sum of the wavelet coefficients of all relevant scales within it where E pdz represents the total energy under the frequency band, a i ∈pdz means all scales belonging to the frequency band, and pdz represents the frequency band;

[0050] Combine the energies of all frequency bands in the wavelet coefficient matrix into a vector to obtain the frequency band energy vector under each frequency component, which is used as the signal feature within each frequency range.

[0051] Furthermore, the training method of the fault state prediction model includes:

[0052] Collect a sample set, including the YB group performance data set and the corresponding true labels, and divide the sample set into a training set and a validation set according to a ratio;

[0053] Construct a fault state prediction model, introduce L1 and L2 regularization in the model, use the performance data set as the input, and use the fault state probability value of the electrical component as the output label. The fault state prediction model is a gradient boosting decision tree model;

[0054] Initialize the hyperparameters of the model, as well as the L1 regularization strength α and the L2 regularization strength λ;

[0055] Define a parameter network containing multiple α and λ values, and use the k-fold cross-validation method to calculate the average F1 score of each parameter combination, and select the parameter combination with the highest average F1 score as the optimal parameter combination;

[0056] Define the logarithmic loss as the loss function and introduce a regularization term, where TL represents the loss function after introducing the regularization term, α∑ j |w j | represents the L1 regularization term, and w jDenote the weight of the j-th feature. Take the number of samples covered when the j-th feature is used as a splitting point in all trees as the weight of the feature. Denote the L2 regularization term, and Lo denote the logarithmic loss function.

[0057] Use the optimal parameter combination and the initialized model hyperparameters as model parameters, and adjust the model parameters according to the gradient using the training set and the loss function in each iteration of the model.

[0058] For each iteration, update the model parameters using the training set, use AUC as the evaluation metric, and calculate the AUC value on the validation set.

[0059] Set a performance improvement threshold xnt. If the difference between the AUC value after iteration and the AUC value of the previous iteration is greater than xnt, it is determined that the performance of the model has improved. If the difference between the AUC value after iteration and the AUC value of the previous iteration is less than or equal to xnt, it is determined that the performance of the model has not improved. If the performance of the model on the validation set has not improved in consecutive pat iterations, stop training; obtain the trained gradient boosting decision tree model.

[0060] Further, the method for judging whether there is a fault in the intelligent packaging production line according to the fault probability value of the electrical component includes:

[0061] Collect historical fault scores, sort them by time, and form a fault score sequence.

[0062] Set a sliding window HD and the step size hd of the sliding window.

[0063] Apply the sliding window HD to the fault score sequence. For all fault scores within the initial sliding window, use the ROC curve method to calculate the true positive rate TPR and false positive rate FPR corresponding to each possible fault score threshold.

[0064] According to the TPR and FPR within the initial sliding window, use the Youden index to calculate the optimal threshold yz.

[0065] Move the window forward by one step size nd, recalculate the TPR and FPR for all fault scores within the new sliding window, and calculate the optimal threshold yz.

[0066] Take the mean of all optimal thresholds within the fault score sequence as the current optimal threshold.

[0067] When the fault score is greater than or equal to the optimal threshold, it indicates that there is a fault in the electrical component of the intelligent packaging production line. When the fault score is less than the optimal threshold, it indicates that the electrical component of the intelligent packaging production line is normal.

[0068] If any electrical component in the intelligent packaging production line fails, it is determined that the intelligent packaging production line has a fault.

[0069] Further, the method of analyzing the faulty electrical component and taking targeted control measures when it is detected that the intelligent packaging production line has a fault includes:

[0070] When it is determined that the intelligent packaging production line has a fault, analyze the category of the faulty electrical component and take targeted control measures for different categories of electrical components;

[0071] Main distribution cabinet, motor frequency converter, motor driver, central controller and heating element

[0072] For the main distribution cabinet, check and adjust the input voltage and current to the preset normal range, optimize the load distribution in the main distribution cabinet to make the load of each circuit balanced, and check the setting value of the circuit breaker to enable it to correctly respond to abnormal situations;

[0073] For the motor frequency converter, obtain the actual operating state of the motor, adjust the operating frequency of the motor to make the motor operate in the best state, and optimize the start and stop procedures of the motor frequency converter to reduce the impact during motor startup and the shutdown time;

[0074] For the motor driver, adjust the torque output of the motor driver, reduce the load of the motor driver, and optimize the speed setting of the motor to make it work in coordination with other parts of the production line;

[0075] For the central controller, recalibrate the production parameters in the central controller to make the set production parameters meet the current production requirements;

[0076] For the heating element, check and adjust the temperature setting value of the heating element to keep it within the preset safe range, and at the same time calibrate the temperature controller to make the accuracy of temperature measurement greater than the preset accuracy threshold.

[0077] The technical effects and advantages of the control system and method for an intelligent packaging production line of the present invention:

[0078] The present invention integrates modules such as power data acquisition and adjustment, signal feature extraction, data processing, anomaly analysis, and anomaly judgment, aiming to achieve comprehensive monitoring and fault early warning of an intelligent packaging production line; uses an improved temperature compensation algorithm to adjust current and voltage data, and calculates the energy utilization ratio based on the adjusted data, improving the accuracy of current and voltage data and reducing the influence of ambient temperature changes on measurement results; collects vibration signals of electrical components, uses the Fourier transform method to convert the vibration signals into spectra, and performs time series decomposition to extract signal features in different frequency ranges, realizing an effective conversion from the time domain to the frequency domain, being able to capture vibration characteristics at different frequencies, providing detailed frequency domain information, and enhancing the understanding and analysis ability of complex vibration modes; preprocesses power data and signal features, including noise suppression, time alignment, normalization, and data fusion, to form a performance dataset, clearing noise interference, achieving time synchronization of multi-source data, and facilitating comprehensive analysis; uses an improved Gradient Boosting Decision Tree (GBDT) algorithm combined with L1 and L2 regularization to perform anomaly analysis on the performance dataset, outputs the probability value of the fault state as the fault score, effectively preventing model overfitting, improving the generalization ability and stability of the model, providing reliable fault prediction, and reducing false alarms and missed detections; according to the fault score, dynamically adjusts the threshold through the sliding window method, uses the ROC curve and Youden index to determine the optimal threshold, and judges whether there is a fault, realizing the dynamic adjustment of the threshold, improving the sensitivity and specificity of fault detection, reducing the false alarm rate and missed detection rate, enabling the system to not only timely detect and early warn potential faults, but also perform targeted regulation on the faulty electrical components, helping maintenance personnel quickly locate and solve problems, thereby ensuring the efficient operation and stable production of the intelligent packaging production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 is a schematic diagram of a control system for an intelligent packaging production line of the present invention;

[0080] Figure 2 is a schematic diagram of a control method for an intelligent packaging production line of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0081] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0082] Embodiment 1:

[0083] Please refer to Figure 1As shown, a control system and method for an intelligent packaging production line according to this embodiment include:

[0084] Power data acquisition module: used to collect power parameter data of electrical components; the power parameter data includes current, voltage, energy utilization ratio data, and vibration signals of electrical components;

[0085] Data processing module: used to preprocess the power parameter data to form an electrical feature data set;

[0086] Abnormal analysis module: trains and obtains a fault state prediction model based on the electrical feature data set, and predicts the fault probability value of electrical components based on the fault state prediction model;

[0087] Abnormal judgment module: used to judge whether there is a fault in the intelligent packaging production line according to the fault probability value of electrical components;

[0088] Fault control module, used to analyze the electrical components with faults and take targeted control measures when a fault is detected in the intelligent packaging production line;

[0089] The method of preprocessing the power parameter data to form an electrical feature data set includes:

[0090] Using an improved temperature compensation algorithm to adjust the current and voltage data, calculating the energy utilization ratio based on the adjusted current and voltage data; using the Fourier transform method to convert the vibration signal into a frequency spectrum, performing time series decomposition on the frequency spectrum, and extracting signal features in different frequency ranges;

[0091] Performing noise suppression, time alignment, normalization, and data fusion processing on the adjusted current, voltage, energy utilization ratio data, and signal features in different frequency ranges to obtain an electrical feature data set;

[0092] The method of time alignment is: for each time period wt, calculate the mean values of current, voltage, and energy utilization ratio at all times respectively to obtain the power characteristics of the wt time period;

[0093] The method of data fusion is: splicing the normalized power characteristics and signal characteristics within the time period wt into a performance feature matrix column by column as the electrical feature data set within the time period wt

[0094] The method of using the improved temperature compensation algorithm to adjust the current and voltage data includes:

[0095] Using a thermistor sensor to measure the ambient temperature at any t moment within the time period wt, and calculating the actual resistance value of the current sensor (such as a shunt) or voltage sensor (such as a voltage divider) at the t moment through the improved temperature compensation algorithm formula Where, Tt represents the temperature of the current sensor or voltage sensor at time t, R(T t ) is the actual resistance value of the current sensor or voltage sensor at time t, R 0 is the resistance value of the current sensor or voltage sensor at the reference temperature T 0 , and α is the temperature coefficient;

[0096] Define the room temperature as the reference temperature T 0 , measure the resistance values of the current sensor or voltage sensor at different temperatures, and record the corresponding temperature-resistance values (T rr , R rr ), where T rr represents the rr-th temperature value, and R rr represents the corresponding rr-th resistance value;

[0097] Perform a logarithmic transformation on the temperature-resistance values (T rr , R rr ) to linearize the exponential relationship ln(R rr ) = ln(R 0 ) + α(T rr - T 0 );

[0098] Use the least squares method to fit the logarithmic transformation formula to obtain the best estimated values of ln(R 0 ) and α, and obtain 0 according to ln(R

[0099] Adjust the current and voltage according to the actual resistance value, and calculate the power factor through the adjusted current and voltage data;

[0100] It should be noted that in a highly automated packaging workshop, there are a large number of motors, inverters, and other electronic devices. The electromagnetic interference generated by these devices may affect the accuracy of power data. The working principle of most current sensors (such as Hall effect sensors or shunts) depends on the change of resistance. The increase in the sensor temperature will cause the resistance of the sensor to increase, thereby affecting the measured current value. Similarly, voltage sensors usually measure voltage through voltage dividers or other forms of resistance networks. Temperature changes will also affect the resistance values of these resistance elements, thereby changing the measurement results. Therefore, after calculating the actual resistance value through the corrected temperature compensation algorithm, the output signals of the current and voltage sensors can be corrected by this actual resistance value;

[0101] An improved temperature compensation algorithm ensures accurate current and voltage data can be obtained even in a complex electromagnetic environment. The real-time temperature sensor readings are combined with temperature compensation to dynamically adjust the current and voltage data, eliminating the influence of temperature fluctuations and ensuring data consistency and accuracy;

[0102] The method of adjusting current and voltage according to the actual resistance value includes:

[0103] For a shunt-based current sensor, use the actual resistance value of the current sensor to calculate the current correction factor of the current sensor at time t where DI represents the current sensor, and RB DI represents the nominal resistance of the current sensor DI, and XS DI represents the current correction factor of the current sensor, and R(T t ) DI represents the actual resistance value of the current sensor;

[0104] Calculate the ratio of the actual measured current on the electrical component of the current sensor at time t to the correction factor to obtain the adjusted current data;

[0105] The electrical components include the main distribution cabinet (main power supply of the production line) in the intelligent packaging production line, the motor frequency converter (frequency converter for controlling the speed and torque of the motor), the motor driver (motor and its controller for driving moving components such as conveyor belts and robotic arms), the central controller (central controller responsible for coordinating the operations of each subsystem), and the heating element (device for heating sealing materials);

[0106] For a voltage divider-based voltage sensor, use the actual resistance value of the voltage sensor to calculate the voltage correction factor of the voltage sensor at time t where DV represents the voltage sensor, and XS DV is the voltage correction factor of the voltage sensor, are the actual resistance values of the two voltage dividing resistors R 1 and R 2 of the voltage sensor respectively, are the nominal resistance values of the two voltage dividing resistors R 1 and R 2 of the voltage sensor respectively;

[0107] Calculate the product of the actual measured voltage on the electrical component of the voltage sensor at time t and the voltage correction factor to obtain the adjusted voltage data;

[0108] It should be noted that the reason for the two resistors of the voltage sensor is usually related to the design of the voltage divider circuit. The voltage divider is a commonly used component in the voltage sensor and is used to convert high voltage into low voltage suitable for measurement. The voltage divider consists of two series resistors, and these two resistors together determine how the input voltage is distributed to the output. The voltage divider is a simple circuit that uses two resistors R 1 and R 2 to proportionally reduce the input voltage V in to the output voltage V out (i.e., the voltage actually measured by the sensor). Its working principle is based on Ohm's law and Kirchhoff's voltage law, and the formula is

[0109] The two resistors in the voltage sensor are mainly used to achieve the voltage division function, ensure that the output voltage is suitable for measurement, and can realize proportional adjustment, impedance matching, temperature compensation, and protection functions by appropriately selecting the resistance values. By performing temperature compensation on these two resistors, the accuracy of the measurement data can be further improved, ensuring the stable operation of the intelligent packaging production line;

[0110] The method for calculating the energy utilization ratio from the adjusted current and voltage data includes:

[0111] Based on the adjusted current and voltage data, calculate the product of the current and voltage to obtain the apparent power SG;

[0112] Use a power analyzer to measure the power factor of the electrical component at time t Calculate the product of SG and to obtain the active power where, represents the phase difference angle between the current and voltage, represents the power factor;

[0113] Calculate the product of the apparent power and to obtain the reactive power where, represents the sine phase difference angle;

[0114] Take the ratio of the active power to the reactive power to obtain the energy utilization ratio PQ;

[0115] It should be noted that when the PQ value is high, it means that the active power accounts for a relatively large proportion relative to the reactive power. This usually indicates that the energy utilization efficiency of the system is high because more electrical energy is used for useful work rather than being stored or released;

[0116] When the PQ value is low, it means that the reactive power is relatively large. This situation may indicate that there are more reactive loads in the system, such as induction motors or other inductive loads, which will cause increased current, increase line losses, and may require larger cable and transformer capacity;

[0117] The method of converting the vibration signal into a frequency spectrum using the Fourier transform method comprises:

[0118] The sampling rate is set according to the Nyquist sampling theorem, and the displacement sensor is used to collect the vibration signals of the electrical components of the intelligent packaging production line within the wt period according to the set sampling rate;

[0119] The collected vibration signal is filtered and the mean value is removed using bandpass filtering and averaging method;

[0120] Set a time window with a window width of wk;

[0121] Initialize the window index m to start from 0, and extract the signal segment ZD of the mth window from the processed vibration signal ZD[n] m [n] = ZD[mL+n], where n represents the sample index in the window, which is used to traverse the sample points in each window, n = 0, 1, 2, ..., wk-1, and L represents the preset sliding step size, that is, the degree of overlap between adjacent windows, which is usually selected as a part of the window width (such as L = wt / 2 or L = wt / 4);

[0122] Introduce a window function for the signal fragment within the window, use the window function as the weight of the signal fragment, define the window function as a Hamming window function, and apply it to the signal fragment within the window ZD m [n], get the weighted signal segment ZD hm [m][n]=ZD[mL+n]×hm[n], where hm[n] represents the Hamming window function, ZD hm [m][n] represents the weighted signal segment in the mth time window;

[0123] It should be noted that the window function is a mathematical function used to weight the signal to reduce spectrum leakage and improve the quality of frequency domain analysis. It is applied to the signal fragment within each window to generate a weighted signal;

[0124] For each weighted signal segment ZD in the time window w [m][n] Apply fast Fourier transform to get the spectrum within the time window Where XP[m,k] represents the spectrum of the mth time window at the Kth discrete frequency, K = 0, 1, 2, ..., wk-1, j is an imaginary unit, and j satisfies 2 =-1, e is a natural constant, e-j2πkn / wk is a complex exponential function used to transform a time-domain signal into the frequency domain;

[0125] Integrate the spectra within all time windows, with time on the horizontal axis and frequency on the vertical axis, to form a time-frequency spectrogram containing the complete spectrum within the wt period;

[0126] It should be noted that in some environments with multiple variable vibration sources, mechanical components in the packaging production line (such as motors and conveyor belt drives) will generate complex vibration patterns, which may mask potential fault signals. By converting the time-domain vibration signal into the frequency domain, abnormal frequency components can be effectively identified;

[0127] The methods for performing time series decomposition on the spectrum and extracting signal features within different frequency ranges include:

[0128] Based on the time-frequency spectrogram, mark all significant peak positions in the time-frequency spectrogram. For each significant peak, select the frequency band containing the peak and a preset ratio (such as ±10% or ±20%) as the frequency range;

[0129] For each frequency range, calculate the sum of the squares of the corresponding spectrum values on the time-frequency spectrogram to obtain the energy value of the frequency range, and select the Ne frequency ranges with the largest energy values among all frequency ranges as the frequency components;

[0130] Use the Morlet wavelet as the basis function. For each time point in the time-frequency spectrogram, perform multi-scale continuous wavelet transform to obtain the wavelet coefficient matrix at different scales

[0131] where, f c represents the c-th frequency component, c = 1, 2,..., Ne, tp represents the continuous time variable used to traverse the entire time axis, and an integration operation is performed on the entire time axis to calculate the inner product between the wavelet basis function and the signal. During the integration process, represents all possible time points from negative infinity to positive infinity, a i represents the i-th scale, obtained based on the frequency component, that is f cy represents the sampling rate, ω 0 represents the central angular frequency of the Morlet wavelet function (usually taking the value of 5 or 6), b j is the displacement parameter, representing the j-th discrete position on the time axis of the time-frequency spectrogram, j = 1, 2,..., J, representing the specific time point of interest when calculating the wavelet coefficients, x(tp, f c ) represents a spectrum component in the time-frequency spectrogram, expressed as a function of time t, and the signal intensity or amplitude at f c , ψ *denotes the complex conjugate of the Morlet wavelet function, denotes at f c , a i and b j under the wavelet coefficients;

[0132] For each frequency component f c , construct a wavelet coefficient matrix JW i with rows of scale a j and columns of b fc , and take all the calculated wavelet coefficients as the values at each position in the matrix;

[0133] According to the wavelet coefficient matrix under each frequency component, map each scale a i to the corresponding frequency where, f cy represents the sampling rate, and ω 0 represents the central angular frequency of the Morlet wavelet function;

[0134] Set the frequency bands as low-frequency band, medium-frequency band and high-frequency band (for example, set the frequency bands as low-frequency band 0 - 30Hz, medium-frequency band 30 - 100Hz and high-frequency band 100 - 300Hz). According to the frequency f i , allocate all scales a i to the corresponding frequency bands (for example, if the frequency corresponding to a certain scale falls into the low-frequency band, then classify it into the low-frequency band);

[0135] For each frequency band, calculate the total energy sum of the wavelet coefficients of all relevant scales within it where, E pdz represents the total energy under the frequency band, a i ∈pdz means all scales belonging to the frequency band, and pdz represents the frequency band;

[0136] Combine the energies of all frequency bands in the wavelet coefficient matrix into a vector to obtain the frequency band energy vector under each frequency component, as the signal feature within each frequency range;

[0137] The training method of the fault state prediction model includes:

[0138] Collect a sample set, including the YB group performance data set and the corresponding true labels, and divide the sample set into a training set and a validation set according to a ratio;

[0139] Construct a fault state prediction model, introduce L1 and L2 regularization in the model, use the performance data set as the input, and use the fault state probability value of the electrical component as the output label. The fault state prediction model is a gradient boosting decision tree model;

[0140] Initialize the hyperparameters of the model, as well as the L1 regularization strength α and the L2 regularization strength λ;

[0141] Define a parameter network containing multiple values of α and λ, and use k-fold cross-validation to calculate the average F1 score for each parameter combination. Select the parameter combination with the highest average F1 score as the optimal parameter combination;

[0142] Define the logarithmic loss as the loss function and introduce the regularization term, where TL represents the loss function after introducing the regularization term, α∑ j |w j | represents the L1 regularization term, α represents the regularization strength, controlling the degree of regularization. A larger α will increase sparsity and may remove more features. w j represents the weight of the j-th feature. Take the number of samples covered when the j-th feature is used as a split point in all trees as the weight of the feature, represents the L2 regularization term, λ represents the L2 regularization strength, controlling the degree of regularization. A larger λ will make the model smoother and reduce the risk of overfitting. Lo represents the logarithmic loss function, where YB represents the number of samples, y i represents the true label of the i-th sample, represents the predicted probability that the i-th sample is in the fault state;

[0143] Use the optimal parameter combination and the initialized model hyperparameters as the model parameters. In each iteration of the model, adjust the model parameters according to the gradient using the training set and the loss function (the regularization term will directly affect the gradient calculation, thereby controlling the update of the parameters);

[0144] For each iteration, update the model parameters using the training set, and use AUC as the evaluation metric to calculate the AUC value on the validation set;

[0145] Set a performance improvement threshold xnt. If the difference between the AUC value after iteration and the AUC value of the previous iteration is greater than xnt, it is determined that the performance of the model has improved. If the difference between the AUC value after iteration and the AUC value of the previous iteration is less than or equal to xnt, it is determined that the performance of the model has not improved. If the performance of the model on the validation set has not improved in consecutive pat iterations, stop training; obtain the trained gradient boosting decision tree model;

[0146] It should be noted that introducing a regularization term into the gradient boosting decision tree can significantly improve the model performance and technical effects. By adding a regularization term to the loss function, regularization effectively prevents overfitting, limits the magnitude or number of model parameters, reduces the model complexity, thereby making the model more general and performing better on unseen data. Regularization improves the generalization ability, enabling the model to not only perform well on the training data but also maintain excellent performance on the validation set and test set. Especially when using L1 regularization, it encourages sparse solutions, reduces the influence of unimportant features, and enhances the model interpretability. L2 regularization, on the other hand, makes the model smoother and more stable by preventing the parameters from being too large, which helps to stabilize the model training process and reduce the problems of gradient explosion or vanishing, especially in high-dimensional datasets. In addition, regularization can also accelerate the model convergence, simplify feature selection, and improve the reliability and stability of the model in practical applications, reducing the risks of false positives and false negatives. In summary, regularization makes the model more robust and efficient in processing complex data, improving the overall performance and reliability.

[0147] In some environments that require long-term continuous operation, intelligent packaging production lines usually need to run continuously for a long time, which increases the risk of electrical system failures. By improving the Gradient Boosting Decision Tree (GBDT), it can efficiently process a large amount of historical data, capture long-term dependencies, and predict future fault trends.

[0148] The method for determining whether there is a fault in the intelligent packaging production line according to the fault probability value of electrical components includes:

[0149] Collect historical fault scores, sort them by time, and form a fault score sequence.

[0150] Set a sliding window HD and the step size hd of the sliding window (usually set to half of the time window size).

[0151] Apply the sliding window HD to the fault score sequence. For all the fault scores within the initial sliding window, use the ROC curve method to calculate the true positive rate TPR and false positive rate FPR corresponding to each possible fault score threshold.

[0152] According to the TPR and FPR within the initial sliding window, use the Youden index to calculate the optimal threshold yz.

[0153] Move the window forward by one step size nd. For all the fault scores within the new sliding window, recalculate the TPR and FPR, and calculate the optimal threshold yz.

[0154] Take the mean of all the optimal thresholds within the fault score sequence as the current optimal threshold.

[0155] When the fault score is greater than or equal to the optimal threshold, it indicates that there is a fault in the electrical components of the intelligent packaging production line. When the fault score is less than the optimal threshold, it indicates that the electrical components of the intelligent packaging production line are normal;

[0156] If any electrical component in the intelligent packaging production line fails, it is determined that there is a fault in the intelligent packaging production line;

[0157] The method of analyzing the faulty electrical components and taking targeted control measures when it is detected that there is a fault in the intelligent packaging production line includes:

[0158] When it is determined that there is a fault in the intelligent packaging production line, analyze the category of the faulty electrical components and take targeted control measures for different categories of electrical components;

[0159] Main power distribution cabinet, motor frequency converter, motor driver, central controller and heating element

[0160] For the main power distribution cabinet, check and adjust the input voltage and current to the preset normal range, optimize the load distribution in the main power distribution cabinet to make the load of each circuit balanced, and check the setting value of the circuit breaker to enable it to correctly respond to abnormal situations;

[0161] For the motor frequency converter, obtain the actual operating state of the motor, adjust the operating frequency of the motor to make the motor operate in the best state, and optimize the start and stop procedures of the motor frequency converter to reduce the impact during motor startup and the shutdown time;

[0162] For the motor driver, adjust the torque output of the motor driver, reduce the load of the motor driver, and optimize the speed setting of the motor to make it work in coordination with other parts of the production line;

[0163] For the central controller, recalibrate the production parameters in the central controller to make the set production parameters meet the current production requirements;

[0164] For the heating element, check and adjust the temperature setting value of the heating element to keep it within the preset safe range, and at the same time calibrate the temperature controller to make the accuracy of temperature measurement greater than the preset accuracy threshold.

[0165] In this embodiment, by integrating modules such as power data acquisition and adjustment, signal feature extraction, data processing, anomaly analysis, and anomaly judgment, comprehensive monitoring and fault warning of the intelligent packaging production line are realized; the improved temperature compensation algorithm is used to adjust the current and voltage data, and the energy utilization ratio is calculated based on the adjusted data, improving the accuracy of the current and voltage data and reducing the influence of ambient temperature changes on the measurement results; the vibration signals of electrical components are collected, the Fourier transform method is used to convert the vibration signals into spectra, and time series decomposition is performed to extract signal features in different frequency ranges, realizing an effective conversion from the time domain to the frequency domain, being able to capture the vibration characteristics at different frequencies, providing detailed frequency domain information, and enhancing the understanding and analysis ability of complex vibration modes; the power data and signal features are preprocessed, including noise suppression, time alignment, normalization, and data fusion, to form a performance data set, removing noise interference, realizing the time synchronization of multi-source data, and facilitating comprehensive analysis; the improved gradient boosting decision tree (GBDT) algorithm combined with L1 and L2 regularization is used to perform anomaly analysis on the performance data set, and the probability value of the fault state is output as the fault score, effectively preventing model overfitting, improving the generalization ability and stability of the model, providing reliable fault prediction, and reducing false alarms and missed alarms; according to the fault score, the threshold is dynamically adjusted by the sliding window method, and the ROC curve and Youden index are used to determine the optimal threshold to judge whether there is a fault, realizing the dynamic adjustment of the threshold, improving the sensitivity and specificity of fault detection, reducing the false alarm rate and missed alarm rate, not only being able to detect and warn potential faults in time, but also providing a detailed fault analysis report to help maintenance personnel quickly locate and solve problems, thus ensuring the efficient operation and stable production of the intelligent packaging production line; this method is particularly suitable for intelligent packaging production lines with high requirements for stability, complex environmental conditions, and the need for continuous and efficient operation. Through targeted technical means and innovative methods, efficient fault identification and preventive maintenance in various special environments are ensured.

[0166] Embodiment 2:

[0167] Please refer to Figure 2 As shown, for the parts not described in detail in this embodiment, refer to the description content of Embodiment 1. Provide xxxx, including:

[0168] S1. Collect the power parameter data of electrical components; the power parameter data includes current, voltage, energy utilization ratio data, and the vibration signals of electrical components;

[0169] S2. Preprocess the power parameter data to form an electrical feature data set;

[0170] S3. Train and obtain a fault state prediction model based on the electrical feature data set, and predict the fault probability value of the electrical components based on the fault state prediction model;

[0171] S4. Determine whether there is a fault in the intelligent packaging production line according to the fault probability value of the electrical components;

[0172] S5. When it is detected that there is a fault in the intelligent packaging production line, analyze the electrical components with faults and take targeted control measures.

[0173] Embodiment 3:

[0174] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the above-provided control method for an intelligent packaging production line.

[0175] Since the electronic device introduced in this embodiment is the electronic device adopted for implementing the control method for an intelligent packaging production line in an embodiment of the present application, based on the control method for an intelligent packaging production line introduced in an embodiment of the present application, those skilled in the art can understand the specific implementation manners and various forms of variation of the electronic device in this embodiment. Therefore, the specific implementation of how this electronic device implements the method in an embodiment of the present application will not be described in detail here. As long as those skilled in the art implement the electronic device adopted for the control method for an intelligent packaging production line in an embodiment of the present application, it falls within the scope of protection of the present application.

[0176] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula that is closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0177] The above are only the preferred embodiments of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for ordinary technical users in the technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as within the protection scope of the present invention.

Claims

1. An intelligent packaging production line control system and method, characterized in that: include: Power data acquisition module: used to collect power parameter data of electrical components; Power parameter data include current, voltage, energy utilization rate data and vibration signals of electrical components; Data processing module: used to pre-process the power parameter data to form an electrical characteristic data set; Abnormal analysis module: a fault state prediction model is obtained based on the training of the electrical characteristic data set, and the fault probability value of the electrical component is predicted based on the fault state prediction model; Abnormal judgment module: used to judge whether there is a fault in the intelligent packaging production line according to the fault probability value of the electrical components; The fault control module is used to analyze the faulty electrical components and take targeted control measures when a fault is detected in the intelligent packaging production line.

2. According to claim 1, a smart packaging production line control system and method, characterized in that: The method of preprocessing the power parameter data to form an electrical characteristic data set includes: The improved temperature compensation algorithm is used to adjust the current and voltage data, and the energy utilization ratio is calculated based on the adjusted current and voltage data; the vibration signal is converted into a spectrum using the Fourier transform method, and the spectrum is decomposed into a time series to extract signal features in different frequency ranges; The adjusted current, voltage, energy utilization ratio data and signal characteristics in different frequency ranges are subjected to noise suppression, time alignment, standardization and data fusion processing to obtain an electrical characteristic data set; The time alignment method is as follows: for each time period wt, the mean values ​​of current, voltage and energy utilization ratio at all times are calculated to obtain the power characteristics of the time period wt; The method of data fusion is: the standardized power characteristics and signal characteristics within the time period wt are spliced ​​into a performance characteristic matrix by column as the electrical characteristic data set within the time period wt.

3. The intelligent packaging production line control system and method according to claim 2 is characterized in that: The method for adjusting current and voltage data using an improved temperature compensation algorithm comprises: Use thermistor sensor to measure the ambient temperature at any time t within the wt period, and use the improved temperature compensation algorithm formula to calculate the actual resistance value of the current sensor or voltage sensor at time t according to the ambient temperature. Among them, T t represents the temperature of the current sensor or voltage sensor at time t, R(T t ) is the actual resistance value of the current sensor or voltage sensor at time t, R0 is the resistance value of the current sensor or voltage sensor at the reference temperature T0, and α is the temperature coefficient; Define room temperature as reference temperature T0, measure the resistance value of current sensor or voltage sensor at different temperatures, and record the corresponding temperature-resistance value (T rr , R rr ), where T rr Indicates the rrth temperature value, R rr Indicates the corresponding rrth resistance value; The temperature-resistance value (T rr , R rr ) is logarithmically transformed to linearize the exponential relationship ln(R rr )=ln(R0)+α(T rr -T0); The logarithmic transformation formula is fitted using the least squares method to obtain the best estimate of ln(R0) and α. According to ln(R0), According to the actual resistance value, the current and voltage are adjusted; and the energy utilization ratio is calculated through the adjusted current and voltage data.

4. The intelligent packaging production line control system and method according to claim 3, characterized in that: The method of adjusting the current and voltage according to the actual resistance value includes: For shunt-based current sensors, the current correction factor of the current sensor at time t is calculated using the actual resistance value of the current sensor. Among them, DI represents the current sensor, RB DI Indicates the nominal resistance of the current sensor DI, XS DI Represents the current correction factor of the current sensor, R(T t ) DI Indicates the actual resistance value of the current sensor; Calculate the ratio of the current actually measured by the current sensor on the electrical component at time t to the correction coefficient to obtain adjusted current data; The electrical components include the main power distribution cabinet, motor inverter, motor driver, central controller and heating element in the intelligent packaging production line; For a voltage sensor based on a voltage divider, the voltage correction factor of the voltage sensor at time t is calculated using the actual resistance value of the voltage sensor. Where DV represents the voltage sensor, XS DV is the voltage correction factor of the voltage sensor, and are the actual resistance values ​​of the two voltage divider resistors R1 and R2 of the voltage sensor, and are the nominal resistance values ​​of the two voltage-dividing resistors R1 and R2 of the voltage sensor; The product of the voltage actually measured by the voltage sensor on the electrical component at time t and the voltage correction coefficient is calculated to obtain the adjusted voltage data.

5. The intelligent packaging production line control system and method according to claim 4, characterized in that: The method for calculating the energy utilization ratio by using the adjusted current and voltage data includes: Based on the adjusted current and voltage data, the product of current and voltage is calculated to obtain the apparent power SG; Use a power analyzer to measure the power factor of electrical components at time t Calculate SG and The product of , we get the active power in, represents the phase difference angle between current and voltage, Indicates power factor; Calculate the apparent power and The product of in, represents the sinusoidal phase difference angle; Take the ratio of active power to reactive power to get the energy utilization ratio.

6. The intelligent packaging production line control system and method according to claim 5, characterized in that: The method of converting the vibration signal into a frequency spectrum using the Fourier transform method comprises: The sampling rate is set according to the Nyquist sampling theorem, and the displacement sensor is used to collect the vibration signals of the electrical components of the intelligent packaging production line within the wt period according to the set sampling rate; The collected vibration signal is filtered and the mean value is removed using bandpass filtering and averaging method; Set a time window with a window width of wk; Initialize the window index m to start from 0, and extract the signal segment ZD of the m'th window from the processed vibration signal ZD[n] m [n] = ZD[mL+n], where n represents the sample index in the window, which is used to traverse the sample points in each window, n = 0, 1, 2, ..., wk-1, and L represents the preset sliding step size, that is, the degree of overlap between adjacent windows; Introduce a window function for the signal fragment within the window, use the window function as the weight of the signal fragment, define the window function as a Hamming window function, and apply it to the signal fragment within the window ZD m [n], get the weighted signal segment ZD hm [m][n]=ZD[mL+n]×hm[n], where hm[n] represents the Hamming window function, ZD hm [m][n] represents the weighted signal segment in the mth time window; For each weighted signal segment ZD in the time window w [m][n] Apply fast Fourier transform to get the spectrum within the time window Where XP[m,k] represents the spectrum of the mth time window at the Kth discrete frequency, K = 0, 1, 2, ..., wk-1, j is an imaginary unit, and j satisfies 2 =-1, e is a natural constant, e -j2πkn / wk It is a complex exponential function, which is used to convert the time domain signal to the frequency domain; The spectra in all time windows are integrated, with time as the horizontal axis and frequency as the vertical axis, to form a time-spectrum diagram containing the complete spectrum within the wt period.

7. The intelligent packaging production line control system and method according to claim 6, characterized in that: The method of performing time series decomposition on the spectrum to extract signal features in different frequency ranges includes: Based on the time-spectrum diagram, all peak positions are marked in the time-spectrum diagram, and for each peak, a frequency band including the peak and a preset ratio is selected as a frequency range; For each frequency range, calculate the square of the corresponding spectrum value on the time-spectrum diagram and sum them to obtain the energy value of the frequency range, and select the Ne frequency ranges with the largest energy values ​​of all frequency ranges as frequency components; Using Morlet wavelet as the basis function, multi-scale continuous wavelet transform is performed for each time point in the time-spectrum diagram to obtain the wavelet coefficient matrix at different scales. Among them, f c represents the cth frequency component, c=1,2,...,Ne, tp represents a continuous time variable, a i represents the i-th scale, b j represents the jth discrete position on the time axis of the time-frequency spectrum, j = 1, 2, ..., J, x(tp,f c represents the spectral components in the time-frequency domain spectrogram, ψ * represents the complex conjugate of the Morlet wavelet function, Indicates that in f c 、a i and b j The wavelet coefficients under ; For each frequency component f c , construct a i For the line, with b j The wavelet coefficient matrix is All calculated wavelet coefficients as the value for each position in the matrix; According to the wavelet coefficient matrix under each frequency component, each scale a i Mapping to the corresponding frequency Among them, f cy represents the sampling rate, ω0 represents the central angular frequency of the Morlet wavelet function; Set the frequency band to low frequency band, middle frequency band and high frequency band, according to the frequency f i , all scales a i allocated to the corresponding frequency bands; For each frequency band, calculate the sum of the energy of the wavelet coefficients of all relevant scales within it Among them, E pdz represents the total energy in the frequency band, a i ∈pdz represents all scales belonging to a frequency band, and pdz represents the frequency band; The energy of all frequency bands in the wavelet coefficient matrix is ​​combined into a vector to obtain the frequency band energy vector under each frequency component as the signal feature in each frequency range.

8. The intelligent packaging production line control system and method according to claim 7, characterized in that: The training method of the fault state prediction model includes: Collect sample sets, including the YB group performance data set and the corresponding true labels, and divide the sample sets into training sets and validation sets in proportion; Constructing a fault state prediction model, and introducing L1 and L2 regularization into the model, taking the performance data set as input, and taking the fault state probability value of the electrical component as the output label, the fault state prediction model is a gradient boosting decision tree model; Initialize the model's hyperparameters as well as the L1 regularization strength α and the L2 regularization strength λ; Define a parameter network containing multiple α and λ values, and use the k-fold cross-validation method to calculate the average F1 score of each parameter combination, and select the parameter combination with the highest average F1 score as the optimal parameter combination; Define logarithmic loss as the loss function and introduce regularization terms, Among them, TL represents the loss function after the regularization term is introduced, α∑ j |w j | represents the L1 regularization term, w j represents the weight of the jth feature, and the number of samples covered when the jth feature is used as a split point in all trees is taken as the weight of the feature. represents the L2 regularization term, Lo represents the logarithmic loss function; The optimal parameter combination and the initialized model hyperparameters are used as model parameters, and the model parameters are adjusted according to the gradient in each iteration of the model using the training set and the loss function; For each iteration, the model parameters are updated using the training set, and the AUC value is calculated on the validation set using the AUC as the evaluation metric; A performance improvement threshold xnt is set. If the difference between the AUC value after the iteration and the AUC value of the previous iteration is greater than xnt, the model performance is judged to be improved. If the difference between the AUC value after the iteration and the AUC value of the previous iteration is less than or equal to xnt, the model performance is judged to be not improved. If the model performance on the validation set does not improve in consecutive pat iterations, the training is stopped. The trained gradient boosting decision tree model is obtained.

9. The intelligent packaging production line control system and method according to claim 8, characterized in that: The method for judging whether the intelligent packaging production line has a fault according to the fault probability value of the electrical component includes: Collect historical fault scores, sort them by time, and form a fault score sequence; Set the sliding window HD and the sliding window step hd; Apply the sliding window HD to the fault score sequence, and for all fault scores in the initial sliding window, use the ROC curve method to calculate the true positive rate TPR and false positive rate FPR corresponding to each possible fault score threshold; According to the TPR and FPR in the initial sliding window, the optimal threshold yz is calculated using the Youden index; Move the window forward by a step length nd, recalculate TPR and FPR for all fault scores in the new sliding window, and calculate the optimal threshold yz; Take the average of all the best thresholds in the fault score sequence as the current best threshold; When the fault score is greater than or equal to the optimal threshold, it means that the electrical components of the intelligent packaging production line are faulty; when the fault score is less than the optimal threshold, it means that the electrical components of the intelligent packaging production line are normal; If any electrical component in the intelligent packaging production line fails, it is determined that the intelligent packaging production line has a fault.

10. The intelligent packaging production line control system and method according to claim 9, characterized in that: When a fault is detected in the intelligent packaging production line, the method of analyzing the faulty electrical components and taking targeted control measures includes: When it is determined that there is a fault in the intelligent packaging production line, analyze the type of electrical components that have failed, and take targeted control measures for different types of electrical components; Main switchboard, motor inverter, motor drive, central controller and heating elements For the main distribution cabinet, check and adjust the input voltage and current to the preset normal range, optimize the load distribution in the main distribution cabinet to balance the load of each circuit, and check the setting value of the circuit breaker so that it can respond correctly to abnormal situations; For the motor inverter, obtain the actual state of the motor operation, adjust the motor's operating frequency, make the motor run in the best state, optimize the motor inverter's start and stop procedures, and reduce the impact and downtime when the motor starts; For motor drives, adjust the torque output of the motor drive, reduce the load on the motor drive, and optimize the speed setting of the motor to coordinate with other parts of the production line; For the central controller, recalibrate the production parameters in the central controller to make the set production parameters meet the current production requirements; For the heating element, check and adjust the temperature setting value of the heating element to make it within the preset safety range, and calibrate the temperature controller so that the accuracy of temperature measurement is greater than the preset accuracy threshold.

Images