12-axis PCB drilling machine energy consumption management system based on resource dynamic scheduling

Abstract

The invention relates to the technical field of energy consumption management, and particularly discloses a 12-axis PCB drilling machine energy consumption management system based on resource dynamic scheduling, which collects input and output differential current data and power factor change rate data in real time by deploying a current sensor and a power monitoring module on each motion axis driver; a current anomaly characteristic value and a power stability characteristic value are extracted through a Bayesian point-of-change detection algorithm and a Hilbert-Huang transform method, and a composite energy consumption anomaly characteristic vector is constructed; the random forest model is combined to intelligently score the energy consumption state of each shaft, the working mode of the abnormal shaft is dynamically adjusted according to the scoring result, and self-adaptive scheduling and optimization of energy consumption are achieved; the system continuously optimizes operation parameters through a closed-loop feedback mechanism, the energy consumption of the whole machine is effectively reduced on the premise of guaranteeing the processing efficiency, the intelligence, refinement and greenization levels of equipment operation are improved, and the system has good engineering application value and popularization prospects.

Description

Technical Field

[0001] The present invention relates to the technical field of energy consumption management, and particularly to an energy consumption management system for a 12-axis PCB drilling machine based on resource dynamic scheduling. Background Art

[0002] With the increasing demand of the electronic manufacturing industry for the processing of high-density and high-precision printed circuit boards (PCBs), multi-axis CNC drilling machines, as key equipment in the PCB production process, their operating efficiency and energy consumption levels directly affect the overall manufacturing cost and production efficiency. The 12-axis PCB drilling machine has significant advantages in improving production efficiency due to its multi-station synchronous processing ability. However, due to its complex structure and frequent coordination of moving axes, there are large fluctuations in energy consumption during the operation of each axis. In particular, abnormal energy consumption states are prone to occur in links such as motor drive and power conversion, resulting in energy waste, equipment aging, and even a decline in processing quality. Therefore, constructing a real-time energy consumption monitoring and dynamic scheduling management system for 12-axis PCB drilling machines is of great significance for realizing green manufacturing and improving equipment energy efficiency.

[0003] Deficiencies of the prior art:

[0004] Currently, the energy consumption management of traditional PCB drilling machines mostly relies on single-parameter monitoring (such as voltage, current, or total power consumption) and static energy-saving strategies, lacking the ability to finely model and real-time evaluate the energy consumption behaviors of each moving axis in a multi-axis system. Existing methods usually use Fourier transform or simple statistical analysis means to process current and power data, which are difficult to adapt to the characteristics of non-linear and non-stationary signals, resulting in insensitive identification of early energy consumption abnormalities and prone to missed or misjudged phenomena. In addition, most systems do not establish an effective closed-loop feedback mechanism and cannot dynamically adjust the equipment operation parameters according to the real-time energy consumption state, thus limiting the further exploration of energy-saving potential. To sum up, the current technology still has obvious deficiencies in aspects such as abnormal identification accuracy, response speed, and energy-saving control intelligence, and there is an urgent need for a new energy consumption management system that integrates multi-dimensional feature analysis and intelligent decision-making mechanisms to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide an energy consumption management system for a 12-axis PCB drilling machine based on resource dynamic scheduling to solve the problems in the above background.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] An energy consumption management system for a 12-axis PCB drilling machine based on resource dynamic scheduling, comprising:

[0008] Data acquisition module, which is used to collect the input and output differential current data and power factor change rate data of each axis in real time through current sensors and power monitoring modules deployed on the drive of each moving axis of the drilling machine;

[0009] Differential current analysis module, which is used to normalize the collected differential current data, construct a differential current fluctuation sequence, calculate the current anomaly eigenvalue, and evaluate whether the motor operating state is abnormal;

[0010] Power factor analysis module, which is used to perform sliding window statistical analysis on the power factor change rate data, construct a power factor dynamic sequence, calculate the power stability eigenvalue, and identify abnormal conditions in the energy conversion process;

[0011] Abnormality identification module, which constructs the current anomaly eigenvalue and the power stability eigenvalue into a composite energy consumption anomaly feature vector and inputs it into the machine learning model to determine whether there is an abnormal energy consumption state currently;

[0012] Dynamic scheduling module, which intelligently adjusts the working modes of 12 moving axes according to the abnormality identification result to achieve dynamic optimization of energy consumption and adaptive scheduling of resources.

[0013] As a further solution of the present invention: evaluating whether the motor operating state is abnormal specifically includes:

[0014] Collecting the input and output differential current data in real time, normalizing the collected differential current data, constructing a differential current fluctuation sequence, analyzing the differential current fluctuation sequence, calculating the current anomaly eigenvalue according to the analysis result, and determining whether the current anomaly eigenvalue is greater than or equal to a preset threshold. If so, the motor operating state is abnormal; if not, the motor operating state is normal.

[0015] As a further solution of the present invention: the process of obtaining the current anomaly eigenvalue is as follows:

[0016] Collecting the input and output differential current data in real time through current sensors deployed on the drive of each moving axis, normalizing the collected differential current data, arranging the normalized differential current values in chronological order to obtain a normalized differential current sequence under a unified scale;

[0017] Among them, the normalization process is to calculate the ratio of the difference between the differential current value and the average value of the differential current data to the standard deviation of the differential current data to obtain the normalized differential current value;

[0018] Based on the normalized differential current sequence, the Bayesian change point detection algorithm is used to identify the energy state mutation points and calculate the divergence between the empirical distributions of the two segments of data before and after the mutation; Divergence;

[0019] Calculate among all the change point positions the ratio of the mean value to the standard deviation value of the divergence to obtain the current anomaly degree eigenvalue.

[0020] As a further solution of the present invention: The identification of abnormal conditions in the energy conversion process specifically includes:

[0021] Real-time collect the power factor change rate data of each axis, perform sliding window statistical analysis on the power factor change rate data, construct a power factor dynamic sequence, analyze the power factor dynamic sequence, calculate the power stability eigenvalue according to the analysis result, and determine whether the power stability eigenvalue is greater than or equal to a preset threshold. If so, there is an abnormality in the energy conversion process; if not, the energy conversion process is normal.

[0022] As a further solution of the present invention: The process of obtaining the power stability eigenvalue is as follows:

[0023] Real-time sample the power factor of each axis during the operation of the device, calculate its change rate within a unit time to obtain a power factor change rate sequence, use the sliding window technique to segment the power factor change rate sequence, set the window size as and the step size as and perform statistical analysis on the data in each window, extract the statistical features of local mean, variance, and extreme values, construct a power factor dynamic sequence, arrange the statistical features in all windows in chronological order to form a new time series, denoted as the power factor dynamic sequence, perform empirical mode decomposition on the power factor dynamic sequence, decompose it into several intrinsic mode functions, and apply the Hilbert transform to each intrinsic mode function to obtain its corresponding analytic signal;

[0024] According to the Hilbert transform result, further calculate its instantaneous amplitude and instantaneous frequency;

[0025] According to the instantaneous amplitudes of all the intrinsic mode functions, calculate the total energy distribution of all the intrinsic mode functions, calculate the difference between the maximum instantaneous frequency and the minimum instantaneous frequency, and calculate the ratio of the total energy distribution of all the intrinsic mode functions to the difference between the maximum instantaneous frequency and the minimum instantaneous frequency to obtain the power stability eigenvalue.

[0026] As a further solution of the present invention: The construction of the current anomaly degree eigenvalue and the power stability eigenvalue into a composite energy consumption anomaly feature vector and inputting it into a machine learning model specifically includes:

[0027] Obtain the current abnormality characteristic values and power stability characteristic values of each axis during the operation of the device, construct the current abnormality characteristic values and power stability characteristic values into a composite energy consumption abnormality characteristic vector, and use it as the input of the machine learning model. Minimize the error between the predicted abnormal energy consumption score and the actual abnormal energy consumption score as the training objective of the machine learning model. Output the abnormal energy consumption score according to the trained machine learning model, and the machine learning model is a random forest model.

[0028] As a further solution of the present invention: The training process of the machine learning model is as follows:

[0029] Extract the current abnormality characteristic values and power stability characteristic values of each axis from the historical data of the device operation, construct each set of collected characteristic values into a composite energy consumption abnormality characteristic vector, and use it as the input characteristics of the model; Combine the corresponding actual abnormal energy consumption score labels as the target output to form a complete training sample set. Use the random forest algorithm to model the above training samples. The random forest constructs multiple decision trees and integrates their prediction results. During the training process, continuously adjust the model parameters including: the number of trees and the maximum depth, and evaluate the model performance through cross-validation to minimize the error between the predicted abnormal energy consumption score and the actual score.

[0030] As a further solution of the present invention: The judgment of whether there is an abnormal energy consumption state currently specifically includes:

[0031] Compare the abnormal energy consumption scores of each axis during the current device operation with the preset threshold, and judge whether the abnormal energy consumption score is greater than or equal to the preset threshold. If so, there is an abnormal energy consumption state currently. If not, there is no abnormal energy consumption state currently.

[0032] As a further solution of the present invention: According to the abnormal recognition result, intelligently adjust the working modes of 12 motion axes, specifically including:

[0033] For the motion axes determined to be in an abnormal energy consumption state, the system automatically triggers an energy-saving optimization strategy, including reducing the acceleration, deceleration, and operating speed of the axis, and switching to a low-power standby mode;

[0034] For the motion axes determined to be in a normal energy consumption state, maintain the original working parameters unchanged and continue to execute the current production task;

[0035] After the adjusted working mode runs for a period of time, the system collects relevant parameters again and repeats the abnormal recognition process to form a closed-loop feedback mechanism to prevent performance degradation caused by a single misjudgment.

[0036] The beneficial effects of the present invention:

[0037] (1) By introducing the Bayesian change point detection algorithm to deeply analyze the normalized differential current fluctuation sequence, the present invention can accurately identify the mutation points of the energy state during the operation of the motor, and construct a statistically robust current abnormality eigenvalue by calculating the divergence between the empirical distributions before and after the mutation, so as to achieve high-sensitivity detection of early faults or minor abnormalities of the motor. At the same time, in the power factor analysis link, the Hilbert-Huang transform (HHT) method is innovatively adopted to perform empirical mode decomposition on the power factor dynamic sequence formed after sliding window processing, extract the instantaneous frequency and instantaneous amplitude information of each order of intrinsic mode function, and further calculate the power stability eigenvalue by combining the ratio of the total energy distribution to the frequency fluctuation range, effectively capturing the non-linear and non-stationary characteristics in the energy conversion process. This dual feature extraction mechanism comprehensively characterizes the operation state of the device from two dimensions of current flow and energy conversion, significantly improving the recognition accuracy and adaptability of the system to abnormal energy consumption under complex working conditions, avoiding the problems of missed judgment and misjudgment caused by the failure of the signal stationarity assumption in traditional methods, and providing a solid technical support for realizing high-reliability energy consumption management and intelligent diagnosis.

[0038] (2) The present invention constructs a composite energy consumption evaluation model based on the fusion of current abnormality and power stability characteristics. Through multi-dimensional feature extraction of the differential current fluctuation characteristics and the dynamic change of the power factor, a composite energy consumption abnormality feature vector with strong characterization ability is formed, and it is introduced as an input feature into the machine learning model trained based on the random forest algorithm to realize intelligent scoring and abnormality recognition of the energy consumption state of each motion axis of the 12-axis PCB drilling machine. During the training process of the model, historical data is used for supervised learning, and the accuracy and generalization ability of the scoring prediction are improved by continuously optimizing the model parameters and the cross-validation mechanism, ensuring that the system can stably and reliably identify abnormal energy consumption behaviors under different working conditions. Based on the recognition results, the system further automatically adjusts the working modes of each axis according to the scoring level, including reducing the running speed and acceleration of the abnormal axis or switching to the low-power standby state, while maintaining the normal axis in the high-efficiency operation range, and forming a closed-loop feedback control strategy through the periodic data collection and re-evaluation mechanism, significantly reducing the overall energy consumption level of the machine under the premise of ensuring the processing efficiency and process quality, and improving the intelligent, self-adaptive and green manufacturing capabilities of the equipment operation, providing a practical technical path and system solution for the energy-saving optimization of modern high-precision manufacturing equipment. Brief Description of the Drawings

[0039] The present invention will be further described below with reference to the drawings.

[0040] Figure 1 It is a flow block diagram of the energy consumption management system of the 12-axis PCB drilling machine based on resource dynamic scheduling of the present invention. Detailed Embodiment

[0041] 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 the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts belong to the scope of protection of the present invention.

[0042] Please refer to Figure 1 As shown, the present invention is an energy consumption management system for a 12-axis PCB drilling machine based on dynamic resource scheduling, including:

[0043] A data acquisition module, which is used to collect the input and output differential current data and power factor change rate data of each axis in real time through current sensors and power monitoring modules deployed on the drive of each moving axis of the drilling machine;

[0044] A differential current analysis module, which is used to normalize the collected differential current data, construct a differential current fluctuation sequence, calculate the current anomaly characteristic value, and evaluate whether the motor operation state is abnormal;

[0045] A power factor analysis module, which is used to perform sliding window statistical analysis on the power factor change rate data, construct a power factor dynamic sequence, calculate the power stability characteristic value, and identify abnormal situations in the energy conversion process;

[0046] An anomaly recognition module, which constructs the current anomaly characteristic value and the power stability characteristic value into a composite energy consumption anomaly characteristic vector and inputs it into a machine learning model to determine whether there is an abnormal energy consumption state at present;

[0047] A dynamic scheduling module, which intelligently adjusts the working modes of 12 moving axes according to the anomaly recognition result to achieve dynamic optimization of energy consumption and adaptive scheduling of resources.

[0048] In the data acquisition module, through the current sensors and power monitoring modules deployed on the drives of each moving axis of the drilling machine, the input and output differential current data and power factor change rate data of each axis are collected in real time, specifically including:

[0049] High-precision current sensors and power monitoring modules are respectively installed on each drive of the 12 moving axes of the drilling machine; among them, the current sensor is used to collect the input current and output current signals of the drive motor, and the power monitoring module is used to collect voltage signals and electrical parameter information such as active power and reactive power.

[0050] During the system operation, the current sensor synchronously samples the input current and output current of each axis, with a sampling frequency not lower than 10 kHz, to ensure that the rapid change process of the current can be captured, forming the original input current sequence and the original output current sequence.

[0051] For each moment, calculate the difference between the input current and the output current to obtain the differential current data. The power monitoring module real-time collects the instantaneous power factor of each axis, records the power factor values within a continuous time period, and then calculates the difference in power factor between adjacent two sampling points to obtain the power factor change rate. The power factor change rate time series is obtained by synthesizing the power factor change rates at all time points.

[0052] The collected differential current data and power factor change rate data are transmitted to the central processing unit through wireless communication, and preprocessing operations such as filtering and noise reduction, and outlier removal are performed before entering the subsequent analysis module to improve the data quality and analysis accuracy.

[0053] In the differential current analysis module, normalize the collected differential current data, construct the differential current fluctuation sequence, and calculate the current abnormality characteristic value to evaluate whether the motor operation state is abnormal. Specifically, it includes: real-time collecting the input and output differential current data, normalizing the collected differential current data, constructing the differential current fluctuation sequence, analyzing the differential current fluctuation sequence, according to the analysis result, calculating the current abnormality characteristic value, and judging whether the current abnormality characteristic value is greater than or equal to the preset threshold. If so, the motor operation state is abnormal; if not, the motor operation state is normal.

[0054] The process of obtaining the current abnormality characteristic value is as follows:

[0055] Through the current sensors deployed on each motion axis driver, real-time collect the input and output differential current data, normalize the collected differential current data, and arrange the normalized differential current values in chronological order to obtain the normalized differential current sequence under a unified scale;

[0056] Among them, the normalization process is to calculate the ratio of the difference between the differential current value and the average value of the differential current data to the standard deviation of the differential current data to obtain the normalized differential current value;

[0057] Based on the normalized differential current sequence, use the Bayesian change point detection algorithm to identify the energy state mutation points and calculate the divergence between the empirical distributions of the two segments of data before and after the mutation. The calculation expression is: ;

[0058] In the formula, The divergence between the empirical distributions of two segments of data before and after mutation divergence, representing the empirical distribution before mutation, representing the empirical distribution after mutation, representing the mutation point of the energy state;

[0059] Calculate the ratio of the mean value to the standard deviation of the divergence among all change point positions to obtain the current anomaly characteristic value.

[0060] It should be noted that: the differential current analysis module of the present invention performs normalization processing on the collected differential current data of each axis input and output, constructs a differential current fluctuation sequence under a unified scale, and identifies the mutation point of the energy state based on the Bayesian change point detection algorithm. Furthermore, it calculates the divergence between the empirical distributions before and after mutation, and finally obtains the current anomaly characteristic value through the ratio of the divergence mean value to the standard deviation, thereby realizing the real-time evaluation of the motor operating state. The core innovation of the technical solution of the differential current analysis module lies in: for the first time, introducing the Bayesian change point detection method into the monitoring of the motor operating state of the PCB drilling machine, which can effectively identify the non-stationary mutation behavior in the current signal and improve the sensitivity to early faults or minor anomalies; at the same time, combining the methods of statistical normalization and distribution divergence analysis enhances the robustness and comparability of the feature extraction process, and avoids misjudgment problems caused by equipment differences or working condition changes. This technical effect significantly improves the recognition accuracy and response speed of the system for motor energy consumption anomalies under complex working conditions, provides a reliable decision-making basis for subsequent dynamic scheduling and energy-saving control, and has good engineering application prospects.

[0061] In the power factor analysis module, perform sliding window statistical analysis on the power factor change rate data, construct a power factor dynamic sequence, and calculate the power stability characteristic value to identify anomalies during the energy conversion process, specifically including:

[0062] Real-time collect the power factor change rate data of each axis, perform sliding window statistical analysis on the power factor change rate data, construct a power factor dynamic sequence, analyze the power factor dynamic sequence, and according to the analysis results, calculate the power stability characteristic value, and judge whether the power stability characteristic value is greater than or equal to the preset threshold. If so, there is an anomaly during the energy conversion process; if not, the energy conversion process is normal.

[0063] The process of obtaining the power stability characteristic value is as follows:

[0064] Real-time sample the power factor of each axis during the equipment operation process, calculate its change rate within a unit time to obtain a power factor change rate sequence, and use the sliding window technology to segment the power factor change rate sequence. Set the window size to and the step size to , and perform statistical analysis on the data within each window, extract the statistical features of local mean, variance, and extreme values, construct a dynamic power factor sequence. Arrange the statistical features within all windows in chronological order to form a new time series, denoted as the dynamic power factor sequence. Perform empirical mode decomposition on the dynamic power factor sequence to decompose it into several intrinsic mode functions. Apply the Hilbert transform to each intrinsic mode function to obtain its corresponding analytic signal. The calculation expression is: ; In the formula, is the result of the Hilbert transform of the intrinsic mode function, is the imaginary unit, represents the number of intrinsic mode functions, represents the time series acquisition points, represents the th intrinsic mode function at the time point of the analytic signal, represents at the time point the th intrinsic mode function;

[0065] According to the result of the Hilbert transform, further calculate its instantaneous amplitude and instantaneous frequency. Among them, the instantaneous amplitude is the modulus value of the analytic signal, and the calculation expression of the instantaneous frequency is: ;

[0066] In the formula, represents the th intrinsic mode function at the time point of the instantaneous frequency, represents the angle between the analytic signal in the complex plane and the positive direction of the real axis, represents a constant, represents the derivative with respect to time ;

[0067] Calculate the total energy distribution of all intrinsic mode functions. The calculation expression is: , In the formula, represents the th intrinsic mode function at the time point of the instantaneous amplitude, represents the total energy distribution of all intrinsic mode functions, represents the total number of intrinsic mode functions. Calculate the difference between the maximum instantaneous frequency and the minimum instantaneous frequency, and calculate the ratio of the total energy distribution of all intrinsic mode functions to the difference between the maximum instantaneous frequency and the minimum instantaneous frequency to obtain the power stability characteristic value.

[0068] It should be noted that: the power factor analysis module described in the present invention performs sliding window statistical analysis on the power factor change rate data of each axis, constructs a power factor dynamic sequence reflecting the dynamic characteristics of energy conversion, combines empirical mode decomposition and Hilbert transform technology, extracts the multi-scale instantaneous frequency and amplitude characteristics of the signal, and finally calculates and obtains the power stability characteristic value through the ratio of the total energy distribution to the frequency fluctuation range, so as to accurately identify abnormal states during the energy conversion process. The innovation of this technical solution lies in: for the first time, the Hilbert-Huang transform method (HHT) is introduced into the energy stability analysis of PCB drilling machines, effectively solving the problem that traditional Fourier transform cannot handle non-linear and non-stationary signals; by extracting local statistical features through a sliding window and combining them with Hilbert-Huang transform analysis, the sensitivity and analysis ability of the system to energy fluctuations under complex working conditions are enhanced; the proposed power stability characteristic value comprehensively considers two dimensions of energy intensity and frequency change, improving the robustness and accuracy of abnormal identification. This technology significantly improves the refinement level of energy conversion process monitoring, provides efficient and reliable technical support for equipment energy consumption management and fault warning, and has good engineering promotion value.

[0069] In the abnormal identification module, the current abnormality eigenvalue and the power stability eigenvalue are constructed into a composite energy consumption abnormal feature vector and input into the machine learning model to determine whether there is an abnormal energy consumption state at present, specifically including:

[0070] Obtain the current abnormality eigenvalue and the power stability eigenvalue of each axis during the operation of the equipment, construct the current abnormality eigenvalue and the power stability eigenvalue into a composite energy consumption abnormal feature vector as the input of the machine learning model, minimize the error between the predicted abnormal energy consumption score and the actual abnormal energy consumption score as the training objective of the machine learning model, and output the abnormal energy consumption score according to the trained machine learning model. The machine learning model is a random forest model.

[0071] The training process of the machine learning model is as follows:

[0072] Extract the current anomaly degree eigenvalues and power stability eigenvalues of each axis from the historical data of the equipment operation. Construct a composite energy consumption anomaly feature vector for each set of collected eigenvalues as the input features of the model. Combine the corresponding actual anomaly energy consumption score labels as the target output to form a complete training sample set. Use the random forest algorithm to model the above training samples. The random forest has good anti-overfitting ability and the ability to capture non-linear relationships by constructing multiple decision trees and integrating their prediction results. It is suitable for complex energy consumption anomaly recognition tasks. During the training process, continuously adjust the model parameters (such as the number of trees, maximum depth), and evaluate the model performance through cross-validation to minimize the error (such as mean square error) between the predicted anomaly energy consumption score and the actual score. Finally, the trained random forest model can be used to input the newly collected feature vector in real-time and output the corresponding anomaly energy consumption score, so as to realize the intelligent evaluation of the energy consumption status of the 12-axis PCB drilling machine.

[0073] The judgment of whether there is an abnormal energy consumption status currently specifically includes:

[0074] Compare the abnormal energy consumption scores of each axis during the current equipment operation with the preset threshold, and judge whether the abnormal energy consumption score is greater than or equal to the preset threshold. If so, there is an abnormal energy consumption status currently; if not, there is no abnormal energy consumption status currently.

[0075] In the dynamic scheduling module, according to the anomaly recognition result, intelligently adjust the working modes of the 12 motion axes to achieve dynamic optimization of energy consumption and adaptive scheduling of resources. Specifically, it includes:

[0076] For the motion axes determined to be in an abnormal energy consumption status, the system automatically triggers energy-saving optimization strategies, including reducing the acceleration, deceleration, and operating speed of the axis, and switching to the low-power standby mode;

[0077] For the motion axes determined to be in a normal energy consumption status, keep the original working parameters unchanged and continue to execute the current production task;

[0078] After the adjusted working mode runs for a period of time, the system collects relevant parameters again and repeats the anomaly recognition process to form a closed-loop feedback mechanism, ensuring that each adjustment can effectively improve the energy consumption status and prevent unnecessary performance degradation caused by a single misjudgment.

[0079] Working principle of the present invention: By deploying current sensors and power monitoring modules on each motion axis driver, differential current data of input and output and data of power factor change rate are collected in real time; by normalizing the differential current data and constructing a differential current fluctuation sequence, combining with the Bayesian change point detection algorithm to identify the sudden change points of the energy state, and calculating the divergence between the empirical distributions before and after the sudden change, and then using the ratio of the divergence mean to the standard deviation as the current anomaly characteristic value to evaluate whether the motor operation state is abnormal; at the same time, for the power factor change rate data, the sliding window statistical analysis method is adopted, local statistical features are extracted and a power factor dynamic sequence is constructed, and further the instantaneous frequency and amplitude information of each order of intrinsic mode functions are obtained through empirical mode decomposition and Hilbert transform, and the power stability characteristic value is calculated based on the ratio of the total energy distribution to the frequency fluctuation range to realize the identification of abnormal energy conversion process; then the current anomaly characteristic value and the power stability characteristic value are fused into a composite energy consumption anomaly characteristic vector, input into a machine learning model trained based on the random forest algorithm, the abnormal energy consumption score of each axis is output, and compared with a preset threshold to judge whether there is an abnormal energy consumption state currently; finally, according to the identification result, the working modes of 12 motion axes are intelligently adjusted, including reducing the operation parameters of the abnormal axis or switching to the low power consumption mode, and continuously optimizing the scheduling strategy through a closed-loop feedback mechanism to realize the dynamic optimization of energy consumption and the adaptive scheduling of resources. This system combines Bayesian change point detection, Hilbert-Huang transform (HHT) signal analysis and machine learning for the first time and applies it to the energy consumption management field of multi-axis drilling equipment, significantly improving the abnormal identification accuracy and response speed, and having good engineering practicability and promotion value.

[0080] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by software simulation of collecting a large amount of data to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0081] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire or wirelessly (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that includes one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0082] It should be understood that the term "and / or" in this document is merely a description of the association relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. In addition, the character " / " in this document generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship, which can be specifically understood by referring to the context.

[0083] It should be understood that in various embodiments of the present application, the order numbers of the above processes do not indicate the order of execution. The order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0084] The above has described in detail an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered as limiting the scope of implementation of the present invention. Any equivalent changes and improvements made within the scope of the application of the present invention should still fall within the scope covered by the patent of the present invention.

Claims

1. An energy consumption management system for a 12-axis PCB drilling machine based on dynamic resource scheduling, characterized in that Including: A data acquisition module, which is used to collect real-time input and output differential current data and power factor change rate data of each axis through current sensors and power monitoring modules deployed on the drive of each moving axis of the drilling machine; A differential current analysis module, which is used to normalize the collected differential current data, construct a differential current fluctuation sequence, calculate the current anomaly eigenvalue, and evaluate whether the motor operating state is abnormal; A power factor analysis module, which is used to perform sliding window statistical analysis on the power factor change rate data, construct a power factor dynamic sequence, calculate the power stability eigenvalue, and identify abnormal conditions in the energy conversion process; An anomaly identification module, which constructs the current anomaly eigenvalue and the power stability eigenvalue into a composite energy consumption anomaly feature vector and inputs it into a machine learning model to determine whether there is an abnormal energy consumption state currently; A dynamic scheduling module, which intelligently adjusts the working modes of 12 moving axes according to the anomaly identification result to achieve dynamic optimization of energy consumption and adaptive scheduling of resources.

2. The energy consumption management system for a 12-axis PCB drilling machine based on dynamic resource scheduling according to claim 1, wherein The evaluation of whether the motor operating state is abnormal specifically includes: Collecting real-time input and output differential current data, normalizing the collected differential current data, constructing a differential current fluctuation sequence, analyzing the differential current fluctuation sequence, calculating the current anomaly eigenvalue according to the analysis result, and judging whether the current anomaly eigenvalue is greater than or equal to a preset threshold. If so, the motor operating state is abnormal; if not, the motor operating state is normal.

3. The energy consumption management system of the 12-axis PCB drilling machine based on resource dynamic scheduling according to claim 2, wherein, The process of obtaining the current anomaly eigenvalue is: Collecting real-time input and output differential current data through current sensors deployed on the drive of each moving axis, normalizing the collected differential current data, arranging the normalized differential current values in chronological order to obtain a normalized differential current sequence under a unified scale; Among them, the normalization process is to calculate the ratio of the difference between the differential current value and the average value of the differential current data to the standard deviation of the differential current data to obtain the normalized differential current value; Based on the normalized difference current sequence, the Bayesian change point detection algorithm is used to identify the mutation points of the energy state therein , and calculate the divergence between the empirical distributions of the two segments of data before and after the mutation; Calculate the ratio of the mean value of the divergence to the standard deviation value among all the change point positions to obtain the current anomaly characteristic value.

4. The energy consumption management system for a 12-axis PCB drilling machine based on resource dynamic scheduling according to claim 1, wherein The identification of abnormal conditions in the energy conversion process specifically includes: Collecting real-time power factor change rate data of each axis, performing sliding window statistical analysis on the power factor change rate data, constructing a power factor dynamic sequence, analyzing the power factor dynamic sequence, calculating the power stability eigenvalue according to the analysis result, and judging whether the power stability eigenvalue is greater than or equal to a preset threshold. If so, there is an abnormality in the energy conversion process; if not, the energy conversion process is normal.

5. The energy consumption management system for a 12-axis PCB drilling machine based on dynamic resource scheduling according to claim 1, wherein, The process of obtaining the power stability eigenvalue is: During the operation of the device, the power factor of each axis is sampled in real time, and the change rate within a unit time is calculated to obtain a power factor change rate sequence. The sliding window technique is used to segment the power factor change rate sequence, with the window size set to and the step size set to . Statistical analysis is performed on the data within each window to extract statistical features such as local mean, variance, and extreme values, and a power factor dynamic sequence is constructed. The statistical features within all windows are arranged in chronological order to form a new time series, denoted as the power factor dynamic sequence. Empirical mode decomposition is performed on the power factor dynamic sequence to decompose it into several intrinsic mode functions, and Hilbert transform is applied to each intrinsic mode function to obtain its corresponding analytic signal; Calculating its instantaneous amplitude and instantaneous frequency according to the Hilbert transform result; Calculating the total energy distribution of all intrinsic mode functions according to the instantaneous amplitude of all intrinsic mode functions, calculating the difference between the maximum instantaneous frequency and the minimum instantaneous frequency, and calculating the ratio of the total energy distribution of all intrinsic mode functions to the difference between the maximum instantaneous frequency and the minimum instantaneous frequency to obtain the power stability eigenvalue.

6. The energy consumption management system of a 12-axis PCB drilling machine based on dynamic resource scheduling according to claim 1, wherein, Constructing the current anomaly eigenvalue and the power stability eigenvalue into a composite energy consumption anomaly feature vector and inputting it into a machine learning model specifically includes: Obtaining the current anomaly eigenvalue and the power stability eigenvalue of each axis during the operation of the device, constructing the current anomaly eigenvalue and the power stability eigenvalue into a composite energy consumption anomaly feature vector as the input of the machine learning model, minimizing the error between the predicted abnormal energy consumption score and the actual abnormal energy consumption score as the training objective of the machine learning model, and outputting the abnormal energy consumption score according to the trained machine learning model. The machine learning model is a random forest model.

7. The energy consumption management system for a 12-axis PCB drilling machine based on dynamic resource scheduling according to claim 1, characterized in that The training process of the machine learning model is as follows: Extract the current anomaly eigenvalue and the power stability eigenvalue of each axis from the historical data of the device operation, construct a composite energy consumption anomaly feature vector for each set of collected eigenvalues as the input feature of the model; combine the corresponding actual abnormal energy consumption score label as the target output to form a complete training sample set, and use the random forest algorithm to model the above training samples. The random forest constructs multiple decision trees and integrates their prediction results. During the training process, continuously adjust the model parameters including: the number of trees and the maximum depth, and evaluate the model performance through cross-validation to minimize the error between the predicted abnormal energy consumption score and the actual score.

8. The energy consumption management system of a 12-axis PCB drilling machine based on dynamic resource scheduling according to claim 1, characterized in that Judging whether there is an abnormal energy consumption state currently specifically includes: Compare the abnormal energy consumption score of each axis during the current device operation with a preset threshold, and judge whether the abnormal energy consumption score is greater than or equal to the preset threshold. If so, there is an abnormal energy consumption state currently; if not, there is no abnormal energy consumption state currently.

9. The energy consumption management system for a 12-axis PCB drilling machine based on resource dynamic scheduling according to claim 1, characterized in that, Intelligently adjusting the working modes of 12 motion axes according to the anomaly recognition result specifically includes: For the motion axes determined to be in an abnormal energy consumption state, the system automatically triggers an energy-saving optimization strategy, including reducing the acceleration, deceleration, and operating speed of the axis, and switching to a low-power standby mode; For the motion axes determined to be in a normal energy consumption state, maintain the original working parameters unchanged and continue to execute the current production task; After the adjusted working mode runs for a period of time, the system collects relevant parameters again and repeats the anomaly recognition process to form a closed-loop feedback mechanism to prevent performance degradation caused by a single misjudgment.

Images