A method and system for monitoring the operating state of an injection molding device

Abstract

The present invention relates to the technical field of electronic digital data processing, and particularly relates to a method and system for monitoring the operating state of an injection molding device, including: collecting the multi-dimensional operating parameter time series of the injection molding device at different times, aggregating the parameters of the same dimension into a data set, and constructing an isolation tree based on each data set; after sorting the tree nodes of the isolation tree according to the data average value, assigning a tree node serial number to each item of data; regarding the combination of parameters of each dimension at the same time as a data point, calculating their degree of being in the same group by comparing the distribution relationship of any two data points on the isolation tree, and classifying the data points with a degree of being in the same group higher than a set threshold into the same group; calculating the degree of abnormality of the data points; if the degree of abnormality exceeds the set threshold, determining it as an abnormal data point and simultaneously sending out an alarm. The present invention solves the problem of low accuracy and reliability of state detection.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic digital data processing. More specifically, the present invention relates to a method and system for monitoring the operating state of an injection molding device. Background Art

[0002] An injection molding device is a key device for the production of plastic products, and the stability of its operating state directly affects product quality, production efficiency, and the service life of the device. However, during the injection molding process, the operating state of the device is usually affected by various complex factors, including the temperature, vibration, pressure, and speed of the device. These parameters have a high degree of non-linear correlation and coupling, and any abnormality may lead to product quality problems or equipment failures.

[0003] Currently, most methods for monitoring the operating state of injection molding devices mainly rely on manual inspections or simple monitoring of single operating parameters. For example, the temperature, pressure, or vibration is independently judged by setting thresholds. However, these traditional methods have obvious deficiencies. Manual inspections are not only inefficient and lack real-time performance but are also easily affected by insufficient human experience or subjective misjudgments. Single-parameter monitoring fails to fully consider the interaction between multiple parameters and its impact on the overall operating state. Especially in a complex production environment, it is easy to cause missed alarms or false alarms in anomaly detection. In addition, with the rapid development of industrial automation and intelligent technologies, the complexity of injection molding processes and equipment is increasing day by day. Existing monitoring technologies are unable to cope with diverse working conditions and high-frequency abnormal states, and cannot meet the higher requirements of modern production for equipment stability, anomaly response speed, and fault location accuracy.

[0004] The patent application document with the publication number CN118348925A discloses a device operating state monitoring system and a monitoring method. The patent application document obtains the external operating state and internal operating parameters of the device in real time through a video acquisition unit and a data acquisition unit. The processor combines the two to judge the overall operating state of the device and issues an early warning in case of anomalies, thereby achieving accurate positioning and rapid maintenance of device anomalies, improving production efficiency, and reducing risks and losses.

[0005] However, the above technical solution only simply combines and judges the operating state of the device by the processor, without fully considering the correlation between multi-dimensional operating parameters in complex scenarios and the interference of anomalies on the detection results, and it is difficult to effectively cope with diverse complex environments, resulting in problems of low accuracy and reliability in the state detection of injection molding equipment. Summary of the Invention

[0006] To solve the problem of low accuracy and reliability in the state detection of injection molding equipment proposed in the above background art, the present invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a method for monitoring the operating state of an injection molding device, including: collecting the time series of operating parameters of the injection molding device at different times, where the time series of operating parameters includes time series of parameters in multiple dimensions, and taking the parameters of the same dimension at different times as a data set; constructing isolation trees based on each data set respectively, and sorting the tree nodes based on the average value of the data in each tree node of the isolation tree to obtain the serial numbers of the tree nodes corresponding to each data in the data set; taking the parameters of each dimension at the same time as a data point, and calculating the degree of co-grouping of any two data points, where the degree of co-grouping of the th data point and the th data point , where, is the total number of parameter items in each dimension of the data point, is the serial number of the tree node corresponding to the th item of data in the th data point, is the serial number of the tree node corresponding to the th item of data in the th data point, is the total number of tree nodes corresponding to the th item of data; marking the data points with a degree of co-grouping greater than the set co-grouping threshold as co-group data points; calculating the degree of abnormality of the th data point , where, is the initial degree of abnormality of the th data point, is the number of co-group data points that are co-grouped with the th data point, is the average value of the total number of tree nodes corresponding to each item of data in the th data point, is the total number of data points; in response to the data point with a degree of abnormality greater than the set abnormality threshold being an abnormal data point.

[0008] The above technical solution can efficiently detect the abnormal state of injection molding equipment by constructing isolation trees and analyzing multi-dimensional parameters. By calculating the degree of group similarity and abnormality of data points, the system can accurately identify minor deviations or abnormal patterns during equipment operation, thereby improving the accuracy of equipment status monitoring. Since this method takes into account the co-variation of multi-dimensional parameters, it can comprehensively capture potential problems during equipment operation and avoid misjudgments that may be caused by single-dimensional parameters. In addition, combined with dynamically set abnormality thresholds and intelligent data analysis, the system has stronger adaptability and flexibility under different working conditions, further improving the reliability of detection. Overall, this technical solution effectively reduces the omission of equipment failures, improves the fault prediction ability of injection molding equipment, and provides more accurate and timely decision-making support for equipment maintenance and production scheduling.

[0009] Further, the construction of the isolation tree is specifically as follows: randomly select a feature at each tree node of the isolation tree, and randomly select a splitting value between the maximum and minimum values of the feature, and split the data set according to the splitting value; until the tree node reaches a predetermined maximum depth or the number of data in the tree node reaches a preset quantity threshold.

[0010] The above technical solution effectively constructs an efficient anomaly detection model by randomly selecting features and performing data splitting at each node of the isolation tree. Since the splitting of each node is based on random selection of features, this randomness helps to avoid overfitting and improves the adaptability of the model to different data distributions. By controlling the maximum depth of the tree and the threshold of the node data volume, the system can fully explore potential patterns in the data while ensuring computational efficiency. This construction method ensures the stability and efficiency of the isolation tree when dealing with large-scale and high-dimensional data, thereby improving the accuracy and sensitivity of anomaly detection.

[0011] Further, the time series of operating parameters includes the temperature time series, vibration time series, pressure time series, and speed time series of the injection molding equipment.

[0012] Further, the temperature time series of the injection molding equipment is collected by a temperature sensor, the vibration time series of the injection molding equipment is collected by a vibration sensor, the pressure time series of the injection molding equipment is collected by a pressure sensor, and the speed time series of the injection molding equipment is collected by a speed sensor.

[0013] Further, it also includes data cleaning of the time series of operating parameters.

[0014] The above technical solution can effectively remove noise, fill in missing values, and correct outliers through data cleaning, thereby avoiding inaccurate or incorrect data from interfering with subsequent analysis and prediction.

[0015] Further, the initial degree of abnormality of the data points is obtained by using the Isolation Forest algorithm.

[0016] Through the above technical solution, by using the Isolation Forest algorithm to obtain the initial degree of abnormality of the data points, the system can efficiently and accurately identify potential abnormalities in the operation of the equipment. The Isolation Forest algorithm evaluates abnormalities based on the isolation of data points in the feature space, and can effectively distinguish normal and abnormal data, especially performing well when dealing with high-dimensional and large-scale data.

[0017] Further, it also includes that when abnormal data points are detected, the injection molding equipment automatically issues an alarm.

[0018] Through the above technical solution, by automatically issuing an alarm when abnormal data points are detected, the response speed and fault prevention ability of the injection molding equipment are significantly improved. The system can real-time capture the abnormal state during the operation of the equipment and issue an alarm in a timely manner when the problem first appears, avoiding the further deterioration of equipment failures or affecting the production progress. This automated alarm mechanism reduces the dependence on manual monitoring, improves the monitoring efficiency and accuracy of the system, and also reduces losses caused by human negligence or delayed response. In addition, through timely alarms, operators can quickly take emergency measures, such as adjusting equipment parameters or performing maintenance, to ensure that the equipment operates in a safe state, thereby reducing downtime, improving production efficiency, and enhancing the reliability and sustainability of the equipment.

[0019] In a second aspect, the present invention provides an operating state monitoring system for an injection molding equipment, including a memory and a processor. Computer program instructions are stored in the memory, and when the computer program instructions are executed by the processor, the above-mentioned operating state monitoring method for an injection molding equipment is implemented.

[0020] The beneficial effects of the present invention are as follows:

[0021] By adopting multi-dimensional parameter time series analysis, isolation tree construction, anomaly degree calculation, and an automatic alarm mechanism, the present invention effectively improves the accuracy and reliability of the operation status monitoring of injection molding equipment. By performing real-time monitoring and data cleaning on parameters such as equipment temperature, vibration, pressure, and speed, the system can comprehensively capture the operation status of the equipment and promptly identify potential abnormal behaviors. The application of the isolation tree algorithm enhances the sensitivity and adaptability of anomaly detection. Especially when facing high-dimensional and large-scale data, it can efficiently distinguish normal and abnormal data points. The automatic alarm mechanism for abnormal data points further improves the response speed of the system, ensuring that an alarm can be issued and emergency measures can be taken in a timely manner when problems occur with the equipment, thereby reducing the occurrence of equipment failures, extending the service life of the equipment, and optimizing production efficiency. Overall, the present invention provides an intelligent and efficient operation monitoring and fault warning solution for injection molding equipment, greatly improving the operation safety and production stability of the equipment. Brief Description of the Drawings

[0022] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present invention will become readily understandable. In the drawings, several embodiments of the present invention are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0023] Figure 1 is a flowchart schematically showing a method for monitoring the operation status of an injection molding equipment according to an embodiment of the present invention;

[0024] Figure 2 is a block diagram schematically showing the structure of a system for monitoring the operation status of an injection molding equipment according to an embodiment of the present invention. Detailed Embodiments

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] Next, the detailed embodiments of the present invention will be described in detail with reference to the drawings.

[0027] An embodiment of a method for monitoring the operation status of an injection molding equipment.

[0028] As Figure 1 shown, a flowchart of a method for monitoring the operation status of an injection molding equipment according to an embodiment of the present invention includes the following steps:

[0029] S1: Collect the operating parameter timing of the injection molding equipment at different times, where the operating parameter timing includes parameter timing in multiple dimensions.

[0030] In one embodiment, the operating parameter sequence includes the temperature sequence, vibration sequence, pressure sequence and speed sequence of the injection molding equipment. These parameters are important indicators for evaluating the operating status of the injection molding equipment, and the change of each parameter may directly affect the working efficiency and product quality of the equipment. Therefore, collecting and analyzing these operating parameters is crucial for accurately monitoring the equipment status.

[0031] Specifically: Use temperature sensors to collect the temperature sequence of injection molding equipment. Temperature is one of the important factors affecting the injection molding process. Changes in equipment temperature may be directly related to equipment overheating, poor cooling and other faults; Use vibration sensors to collect the vibration sequence of injection molding equipment. Vibration is a common state feedback during equipment operation. Abnormal vibration fluctuations may be a precursor to equipment failure; Use pressure sensors to collect the pressure sequence of injection molding equipment. Pressure fluctuations may directly affect the molding accuracy and consistency of the product; Use speed sensors to collect the speed sequence of injection molding equipment. Changes in equipment operating speed are directly related to production efficiency and product consistency. By real-time monitoring of speed sequence, it is possible to ensure that the equipment runs smoothly at a predetermined speed. Once speed fluctuations or abnormalities occur, the system can issue an early warning in a timely manner to help operators adjust the equipment status.

[0032] After collecting the time series data of operating parameters such as temperature, vibration, pressure and speed, these data need to be cleaned. Data cleaning is a key link in the entire monitoring process. The purpose is to eliminate noise data, missing values and abnormal data in the collection process to ensure the accuracy and reliability of the analysis results. Through data cleaning, the influence of external environmental interference, sensor failure or error and other factors on data collection can be effectively eliminated.

[0033] During the data cleaning process, the collected raw data will first be filled with missing values and processed for outliers to ensure that the data is complete and has no obvious errors. In addition, the system will correct or remove abnormal fluctuations caused by sensor drift or external factors to prevent these interference factors from affecting subsequent fault diagnosis and prediction results. After data cleaning, the obtained data will have higher quality, providing reliable data support for subsequent equipment status monitoring, fault warning and maintenance decisions.

[0034] Through the above data collection and cleaning process, the system can obtain accurate and reliable time series data of equipment operating parameters. After cleaning, these data can provide high-quality input for anomaly detection, fault prediction and equipment health management, greatly improving the accuracy and real-time performance of equipment status monitoring.

[0035] S2: Take the parameters of each dimension at the same moment as a data point, and calculate the degree of belonging to the same group between any two data points.

[0036] In one embodiment, take the parameters of the same dimension at different moments as an independent data set; construct isolation trees based on each data set respectively, and sort the tree nodes based on the average value of the data in each tree node of the isolation tree to obtain the serial numbers of the tree nodes corresponding to the data in the data set.

[0037] Among them, constructing the isolation tree specifically means: randomly select a feature in each tree node of the isolation tree, and randomly select a splitting value between the maximum value and the minimum value of the feature, and split the data set according to the splitting value; until the tree node reaches the predetermined maximum depth or the number of data in the tree node reaches the preset quantity threshold.

[0038] Take the parameters of each dimension at the same moment as a data point, and calculate the degree of belonging to the same group of any two data points. Among them, the degree of belonging to the same group of the th data point and the th data point, where in the formula, is the total number of parameter items of each dimension in the data point, is the serial number of the tree node corresponding to the th item of data in the th data point, is the serial number of the tree node corresponding to the th item of data in the th data point, is the total number of tree nodes corresponding to the th item of data;

[0039] And mark the data points with the degree of belonging to the same group greater than the set threshold of belonging to the same group as data points belonging to the same group.

[0040] By calculating the degree of belonging to the same group of the parameters of each dimension at the same moment as data points, this technical solution can quantify the similarity between different data points, thereby effectively identifying similar or consistent operation modes. Specifically, by calculating the degree of belonging to the same group based on the difference in the serial numbers of the isolation tree nodes, the proximity of different data points in the multi-dimensional parameter space can be accurately evaluated. This method ensures a comprehensive measurement of the similarity of data points by comprehensively considering the tree node distribution and structure of each dimension of data, improving the model's ability to identify abnormal data points. In addition, this method of measuring belonging to the same group can effectively filter out noise, improving the accuracy of data analysis, and providing a more reliable basis for subsequent anomaly detection, fault prediction, and status monitoring.

[0041] S3: Calculate the degree of abnormality of each data point based on the belonging-to-the-same-group situation of each data point.

[0042] In one embodiment, the initial anomaly degree of the data points is obtained by using the Isolation Forest algorithm; the anomaly degree of the th data point , where is the initial anomaly degree of the th data point, is the number of data points in the same group as the th data point, is the average value of the total number of tree nodes corresponding to each item of data in the th data point, is the total number of data points.

[0043] By obtaining the initial anomaly degree of the data points by using the Isolation Forest algorithm and performing comprehensive calculations in combination with the number of data points in the same group, the average value of the tree node distribution, and the total number of data points, this technical solution can effectively quantify the anomaly degree of each data point. This method can accurately evaluate the anomaly degree of data points in the multi-dimensional parameter space and further optimize the anomaly detection results through the relationship with the data points in the same group. Through this comprehensive calculation, the solution enhances the sensitivity to minor anomalies, improves the accuracy and reliability of detection, and thus provides more accurate support for equipment fault warning, abnormal state identification, etc.

[0044] S4: Responding to data points with an anomaly degree greater than the set anomaly threshold as abnormal data points and automatically issuing an alarm.

[0045] The value of the above set anomaly threshold can be 0.65. Of course, it can also be determined according to the actual situation.

[0046] In one embodiment, when the anomaly degree of a certain data point exceeds the set anomaly threshold, the system automatically determines that the data point is an abnormal data point, promptly triggers an alarm, and notifies the operator or maintenance personnel to perform corresponding processing. The alarm can be issued not only locally on the device but also through the remote monitoring system to transmit information to the management platform or mobile terminal so that relevant personnel can take action in a timely manner. The triggering of the alarm can be linked with the control system of the device to automatically start predetermined emergency response measures, such as pausing production, adjusting parameters, or switching to standby equipment, etc., to ensure the continuity of production and the safety of the equipment.

[0047] The solution of the present invention significantly improves the monitoring accuracy and real-time performance of the operating status of injection molding equipment by combining the monitoring of multi-dimensional operation parameter time series, the construction of isolation tree models, and the automatic identification and alarm mechanism for abnormal data points. First, intelligent analysis of the multi-dimensional data of the equipment using isolation tree and isolation forest algorithms can efficiently identify minor abnormalities during equipment operation, avoiding potential faults that may be overlooked in traditional methods. Second, the automatic alarm mechanism ensures timely response to abnormal situations, reducing human delays and misjudgments, and improving the ability to prevent faults. In addition, through data cleaning and the fusion of multi-sensor data, the system further enhances the quality and reliability of the monitoring data, reducing the risk of false alarms caused by noise and outliers. Overall, this solution improves the accuracy of equipment fault prediction, reduces downtime, optimizes production efficiency, and enhances the intelligent level of equipment maintenance, with good application prospects.

[0048] An embodiment of an operating status monitoring system for injection molding equipment:

[0049] As Figure 2 shown, a structural block diagram of an operating status monitoring system for injection molding equipment according to an embodiment of the present invention includes a processor and a memory.

[0050] The present invention also provides an operating status monitoring system for injection molding equipment. As Figure 2 shown, the system includes a processor and a memory, and the memory stores computer program instructions. When the computer program instructions are executed by the processor, it implements an operating status monitoring method for injection molding equipment according to the above of the present invention.

[0051] The operating status monitoring system for injection molding equipment further includes other components well-known to those skilled in the art, such as a communication interface. Its settings and functions are known in the art, so they will not be elaborated here.

[0052] In the present invention, the foregoing memory may be any tangible medium that contains or stores a program, which can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any suitable magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory (RRAM), a dynamic random access memory (DRAM), a static random access memory (SRAM), an enhanced dynamic random access memory (EDRAM), a high-bandwidth memory (HBM), a hybrid memory cube (HMC), etc., or any other medium that can be used to store the required information and can be accessed by an application program, a module, or both. Any such computer storage medium may be part of the device or accessible or connectable to the device. Any application or module described in the present invention may be implemented using computer-readable / executable instructions that can be stored or otherwise held by such a computer-readable medium.

[0053] In the description of this specification, "a plurality of" and "several" mean at least two, such as two, three, or more, etc., unless otherwise specifically and clearly defined.

[0054] Although this specification has shown and described multiple embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will think of many changes, alterations, and alternative ways without departing from the spirit and concept of the present invention. It should be understood that various alternatives to the embodiments of the present invention described herein may be employed in the practice of the present invention.

Claims

1. A method for monitoring the operating state of an injection molding device, characterized in that, Including: Collect the operation parameter time series of the injection molding equipment at different times. The operation parameter time series includes parameter time series in multiple dimensions. The parameters of the same dimension at different times are used as a data set. Build isolation trees based on each data set respectively, and sort the tree nodes based on the average value of the data in each tree node of the isolation tree to obtain the serial numbers of the tree nodes corresponding to the data in the data set. Taking the parameter values of each dimension at the same moment as a data point, calculate the degree of belonging to the same group for any two data points. Among them, the degree of belonging to the same group between the th data point and the th data point is , where is the total number of parameter items of each dimension in the data point, is the serial number of the tree node corresponding to the th item of data in the th data point, is the serial number of the tree node corresponding to the th item of data in the th data point, and is the total number of tree nodes corresponding to the th item of data; Record the data points with a same-group degree greater than the set same-group threshold as same-group data points. Calculate the abnormality degree of the th data point , where is the initial abnormality degree of the th data point, is the number of data points in the same group as the th data point, is the average value of the total number of tree nodes corresponding to each item of data in the th data point, is the total number of data points; Respond that the data points with an abnormal degree greater than the set abnormal threshold are abnormal data points.

2. The operation status monitoring method for an injection molding device according to claim 1, characterized in that The building of the isolation tree is specifically as follows: Randomly select a feature at each tree node of the isolation tree, and randomly select a splitting value between the maximum value and the minimum value of the feature. Split the data set according to the splitting value. Until the tree node reaches the predetermined maximum depth or the number of data in the tree node reaches the preset quantity threshold.

3. The operating state monitoring method for an injection molding device according to claim 1, characterized in that, The operation parameter time series includes the temperature time series, vibration time series, pressure time series and speed time series of the injection molding equipment.

4. A method for monitoring the operating state of an injection molding device according to claim 3, characterized in that, Use a temperature sensor to collect the temperature time series of the injection molding equipment, use a vibration sensor to collect the vibration time series of the injection molding equipment, use a pressure sensor to collect the pressure time series of the injection molding equipment, and use a speed sensor to collect the speed time series of the injection molding equipment.

5. A method for monitoring the operating state of an injection molding device according to claim 1, characterized in that, It also includes cleaning the data of the operation parameter time series.

6. The operating state monitoring method for an injection molding device according to claim 1, characterized in that, Use the isolation forest algorithm to obtain the initial abnormal degree of the data points.

7. A method for monitoring the operating state of an injection molding device according to claim 1, characterized in that, It also includes that when abnormal data points are monitored, the injection molding equipment automatically issues an alarm.

8. An operating state monitoring system for an injection molding device, characterized in that, Including a memory and a processor. The memory stores computer program instructions. When the computer program instructions are executed by the processor, it realizes the operation state monitoring method for an injection molding equipment according to any one of claims 1 to 7.

Images