Injection molding production real-time monitoring and intelligent scheduling management system based on Internet of Things

Abstract

The invention relates to the technical field of injection molding production, in particular to an injection molding production real-time monitoring and intelligent scheduling management system based on the Internet of Things, which comprises a data acquisition module, a data fusion module, an anomaly detection module, a self-adaptive scheduling module and a feedback and learning module, the data acquisition module is used for acquiring running states, environmental parameters and logistics information of all equipment in the injection molding production process in real time; the data fusion module is used for carrying out preprocessing, standardization and time sequence alignment on data from different devices and different communication protocols; and the abnormal detection module is used for analyzing the fused data in real time by adopting an abnormal detection method based on a statistical threshold value and machine learning. In the invention, a full-process closed-loop system comprising a data acquisition module, a data fusion module, an abnormal detection module, a self-adaptive scheduling module, a feedback module, an online learning module and the like is constructed; the problems that in an existing system, data are heterogeneous, data quality is not high, anomaly detection lags behind, and scheduling decision is not intelligent are effectively solved.

Description

Technical Field

[0001] The present invention relates to the technical field of injection molding production, and specifically to a real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things. Background Art

[0002] A real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things is an advanced manufacturing system integrating sensor technology, data acquisition, wireless communication, and intelligent algorithms; By installing sensors on injection molding machines, molds, and other key equipment, this system can collect key parameters such as temperature, pressure, and speed during the production process in real time. Using Internet of Things technology, these data are wirelessly transmitted to the central control system and analyzed and processed through intelligent algorithms. The system can monitor the production status in real time, predict and identify potential production problems, automatically adjust production parameters to optimize production efficiency and product quality. In addition, the intelligent scheduling function can automatically plan production tasks and resource allocation according to order requirements, production capacity, and equipment status, reduce downtime, and improve production flexibility and response speed. Such a system helps manufacturing enterprises achieve digital and intelligent transformation and enhance competitiveness; Currently, Internet of Things technology has been widely applied in the injection molding production process to achieve real-time monitoring and data collection; However, existing Internet-of-Things-based injection molding production monitoring systems have problems such as heterogeneous data collection, low data quality, lagging anomaly detection, and non-intelligent scheduling decisions. These systems often struggle to achieve efficient and accurate information fusion in massive and multi-source data, resulting in untimely detection of abnormal events, thus affecting the real-time response of production scheduling and the rationality of resource allocation. The main technical problems arising therefrom are: how to construct a full-process, closed-loop adaptive monitoring and scheduling management system that can not only ensure high-quality data processing and accurate anomaly detection but also intelligently optimize scheduling strategies based on real-time status to achieve rapid response to anomalies and emergencies and improve the overall production efficiency of the system. Therefore, in view of the above problems, a real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things is proposed. Summary of the Invention

[0003] The purpose of the present invention is to provide a real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things to solve the problem of how to construct a full-process, closed-loop adaptive monitoring and scheduling management system that can not only ensure high-quality data processing and accurate anomaly detection but also intelligently optimize scheduling strategies based on real-time status to achieve rapid response to anomalies and emergencies and improve the overall production efficiency of the system.

[0004] To achieve the above purpose, the present invention provides the following technical solutions: The real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things includes a data acquisition module, a data fusion module, an anomaly detection module, an adaptive scheduling module, and a feedback and learning module; the data acquisition module is used to collect the operating status, environmental parameters, and logistics information of each device in the injection molding production process in real time; the data fusion module is used to preprocess, standardize, and align the time series of data from different devices and different communication protocols; the anomaly detection module uses an anomaly detection method based on statistical thresholds and machine learning to analyze the fused data in real time and issue an alarm when an atypical anomaly is detected; the adaptive scheduling module, after the anomaly detection module issues a warning, based on the device status, production tasks, and material supply situation, uses a rule library, a heuristic algorithm, and a reinforcement learning module to achieve dynamic rescheduling; the feedback and learning module is used to receive the feedback data after the scheduling execution and continuously optimize the scheduling strategy using an online learning algorithm.

[0005] As a further optimized content of the present invention, wherein: the adaptive scheduling module includes: A rule library module for storing conventional production scheduling rules and plans; A heuristic scheduling algorithm module that globally searches for a production scheduling scheme based on a genetic algorithm or an ant colony algorithm; A reinforcement learning module that iteratively updates the scheduling strategy using deep reinforcement learning based on the current environmental state and historical scheduling feedback.

[0006] As a further optimized content of the present invention, wherein: the data fusion module realizes the time-domain consistency of multi-source data by aligning and fusing the time series data from different sensors.

[0007] As a further optimized content of the present invention, wherein: the real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things includes a real-time monitoring and intelligent scheduling management method for injection molding production based on the Internet of Things, including the following steps: Step 1: Use Internet of Things devices to collect key parameter data in the injection molding production process in real time to form a data sequence , preprocess the collected data, and form a standardized time series data set ; Among them, represents the sampling time, and the data comes from injection molding machine temperature sensors, pressure sensors, and vibration sensors; Step 2: Use an anomaly scoring function to perform anomaly detection on the preprocessed data sequence. The formula of the anomaly scoring function is: ; In the formula, is a prediction model constructed based on historical data, is the robust scale factor calculated based on historical data. When it is determined as an abnormal event, where is the threshold, that is, the preset tolerance; Step 3: The abnormal detection triggers the adaptive scheduling algorithm to construct the current state vector of the production line , where represents the set of device online status and load rate, represents the task queue status, represents the material inventory information; When generating the candidate scheduling plan, the optimization objective function is adopted: ; In the formula, represents the expected delay of the th task, represents the risk factor of additional resource occupancy caused by the unexpected event related to this task, is the task weight coefficient, is the risk adjustment constant, is the total number of tasks; Step 4: Adopt the optimized reinforcement learning algorithm to update the scheduling strategy, and use the SARSA update formula: ; In the formula, is the current state vector , is the current scheduling action, is the scheduling feedback reward, is the learning rate, is the discount factor, is the new state after the scheduling execution, is the next scheduling action, is the risk penalty factor, is the risk compensation term, which is used to reflect the additional risk impact brought by abnormalities or unexpected events; Step 5: Verify the feasibility of the generated optimal scheduling plan through simulation or constraint solving technology. Its constraint conditions are: ; In the formula, represents the allocation amount of the th type of resource in the optimized scheduling plan, and are the minimum and maximum allowable allocation amounts of the th type of resource respectively, is the total number of resource types; After the verification is passed, the real-time production data of the scheduling execution result is fed back to the system as a new training sample to update the prediction model , Candidate Scheduling Generation and Reinforcement Learning Module , so as to achieve closed-loop adaptive scheduling and continuously improve the system's response ability to unconventional anomalies and emergencies.

[0008] As a further optimization content of the present invention, wherein: based on the real-time acquisition of key parameters in the injection molding production process, the system further performs noise elimination, outlier screening, and data completion processing on the original data to form a high-quality standardized time series data set.

[0009] As a further optimization content of the present invention, wherein: after the data preprocessing is completed, the system uses a prediction model constructed based on historical data and its corresponding robust scale factor to perform anomaly detection on the standardized data collected at each moment, and refine it into the anomaly analysis of single-device and multi-device associated data.

[0010] As a further optimization content of the present invention, wherein: when the anomaly detection module triggers a scheduling response, the production line state vector constructed by the system not only includes the online state and load rate information of the equipment, but also comprehensively considers the task queue state, material inventory situation, and equipment historical operation characteristics to achieve a multi-dimensional description of the overall state of the production line.

[0011] As a further optimization content of the present invention, wherein: when generating a candidate scheduling plan, the system adopts an optimization objective function based on risk adjustment according to the comprehensive consideration of task delay expectation, task weight, and additional risk factor, and dynamically allocates resources to balance the task delay and the additional risk brought by emergencies, so as to achieve the dual optimization of the scheduling plan in terms of efficiency and safety.

[0012] As a further optimization content of the present invention, wherein: while the system verifies the feasibility of the candidate scheduling plan through simulation or constraint solving technology, it also uses the reinforcement learning algorithm to continuously update the scheduling strategy. The reinforcement learning process not only comprehensively considers the scheduling feedback reward, but also introduces a risk compensation and penalty mechanism for realizing the closed-loop adaptive adjustment of the system operation state.

[0013] Compared with the prior art, the beneficial effects of the present invention are: 1. In the present invention, by constructing a full - process closed - loop system including modules such as data acquisition, data fusion, anomaly detection, adaptive scheduling, and feedback and online learning, the problems of data heterogeneity, low data quality, lagging anomaly detection, and non - intelligent scheduling decision - making in the existing system are effectively solved. This system improves data quality through high - precision pre - processing and time - series alignment of multi - source data; uses anomaly detection methods based on statistical thresholds and machine learning to accurately and timely identify anomaly events, thereby reducing the false - alarm rate; at the same time, by constructing a multi - dimensional state vector and introducing a rule base, heuristic algorithms, and reinforcement learning techniques, the dynamic optimization and adaptive adjustment of production scheduling are realized, effectively balancing task delay and risk factors. 2. In the present invention, the constructed system realizes the seamless connection of data acquisition, fusion, anomaly detection, and adaptive scheduling. Through the collaborative processing of multiple modules, precise monitoring of the entire injection - molding production process is achieved. The system can identify anomalies in real - time and dynamically optimize the scheduling plan, effectively reducing the risk of equipment failure and downtime, and improving resource utilization and the overall operation stability of the production line. 3. In the present invention, an intelligent scheduling strategy based on risk adjustment is adopted, and reinforcement learning is used to continuously optimize production scheduling to ensure the precise matching of equipment status, task queue, and material inventory. Through the closed - loop feedback and online learning mechanism, the balance between task delay and the risk of unexpected events is effectively achieved, significantly enhancing the safety and efficiency of injection - molding production. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is the system block diagram of the real - time monitoring and intelligent scheduling management system for injection - molding production based on the Internet of Things of the present invention; Figure 2 is the flowchart of the real - time monitoring and intelligent scheduling management method for injection - molding production based on the Internet of Things of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0015] Please refer to Figure 1-2, the present invention provides a technical solution: an Internet of Things-based real-time monitoring and intelligent scheduling management system for injection molding production, including a data acquisition module, a data fusion module, an anomaly detection module, an adaptive scheduling module, and a feedback and learning module; the data acquisition module is used to collect the operating status, environmental parameters, and logistics information of each device in the injection molding production process in real time; the data fusion module is used to preprocess, standardize, and align the time series of data from different devices and different communication protocols; the anomaly detection module uses an anomaly detection method based on statistical thresholds and machine learning to perform real-time analysis on the fused data and issue an alarm when an atypical anomaly is detected; the adaptive scheduling module, after the anomaly detection module issues a warning, based on the device status, production tasks, and material supply situation, uses a rule base, heuristic algorithms, and a reinforcement learning module to achieve dynamic rescheduling; the feedback and learning module is used to receive the feedback data after the scheduling execution and continuously optimize the scheduling strategy using an online learning algorithm. The system realizes the full-process closed-loop management from data acquisition to dynamic scheduling, ensuring that anomalies in the production process can be detected and responded to in a timely manner, thereby effectively reducing the risk of production interruption and improving resource utilization and overall production efficiency.

[0016] As a further technical solution of this scheme, the adaptive scheduling module includes: A rule base module for storing conventional production scheduling rules and plans; A heuristic scheduling algorithm module that globally searches for a production scheduling scheme based on a genetic algorithm or an ant colony algorithm; A reinforcement learning module that iteratively updates the scheduling strategy using deep reinforcement learning based on the current environmental state and historical scheduling feedback. Through the organic combination of multiple scheduling strategies, this module can quickly obtain an optimized scheduling scheme in a complex environment, realize the intelligence and self-optimization of scheduling decisions, and enhance the system's adaptability to the dynamic production environment; As a further technical solution of this scheme, the data fusion module realizes the time-domain consistency of multi-source data by aligning and fusing the time series data from different sensors. After realizing the time-domain consistency of multi-source data, the data quality and the accuracy of subsequent analysis can be greatly improved, providing a solid data support for anomaly detection and scheduling decisions, reducing the misjudgment rate, and enhancing the overall reliability of the system; As a further technical solution of this scheme, the Internet of Things-based real-time monitoring and intelligent scheduling management system for injection molding production includes an Internet of Things-based real-time monitoring and intelligent scheduling management method, including the following steps: Step 1: Use Internet of Things devices to collect key parameter data in the injection molding production process in real time to form a data sequence , preprocess the collected data to form a standardized time series data set ; Among them, Denote the sampling moment, and the data is sourced from the temperature sensor, pressure sensor, and vibration sensor of the injection molding machine; Step 2: Use an anomaly scoring function to perform anomaly detection on the preprocessed data sequence. The formula for the anomaly scoring function is: ; In the formula, is a prediction model constructed based on historical data, is a robust scale factor calculated based on historical data. When , it is determined as an abnormal event, where is the threshold, i.e., the preset tolerance; Step 3: When anomaly detection triggers the adaptive scheduling algorithm, construct the current state vector of the production line , where represents the set of equipment online status and load rate, represents the task queue status, represents the material inventory information; When generating a candidate scheduling plan, use the optimization objective function: ; In the formula, represents the expected delay of the th task, represents the risk factor of additional resource occupancy caused by unexpected events related to this task, is the task weight coefficient, is the risk adjustment constant, is the total number of tasks; Step 4: Use the optimized reinforcement learning algorithm to update the scheduling strategy, and use the SARSA update formula: ; In the formula, is the current state vector , is the current scheduling action, is the scheduling feedback reward, is the learning rate, is the discount factor, is the new state after scheduling execution, is the next scheduling action, is the risk penalty factor, is the risk compensation term, which is used to reflect the additional risk impact brought by anomalies or unexpected events; Step 5: Verify the feasibility of the generated optimal scheduling plan through simulation or constraint solving techniques. The constraint conditions are:

[0017] In the formula, represents the allocation quantity of the -th type of resource in the optimized scheduling plan, and are respectively the minimum and maximum allowable allocation quantities of the -th type of resource, is the total number of resource types; After passing the verification, the real-time production data of the scheduling execution result is fed back to the system and used as new training samples to update the prediction model and the candidate scheduling generation and reinforcement learning module , so as to realize closed-loop adaptive scheduling, continuously improve the system's response ability to unconventional anomalies and emergencies, clarify the data processing, anomaly detection, scheduling plan generation and scheduling strategy update processes of each step, and realize the full-process closed-loop control from data collection to decision feedback, which not only ensures real-time performance but also ensures the scientificity and efficiency of scheduling decisions, and significantly improves the production line's response and adjustment ability to abnormal situations; As a technical solution for further implementation of this plan, on the basis of real-time collecting key parameters in the injection molding production process, the system further performs noise elimination, outlier screening and data completion processing on the original data to form a high-quality standardized time series data set. By performing detailed preprocessing on the original data, the data quality is greatly improved, and the misjudgment risk caused by data anomalies is reduced, providing a solid and reliable data basis for subsequent anomaly detection and intelligent scheduling, thereby enhancing the stability and accuracy of the overall system; As a technical solution for further implementation of this plan, after the data preprocessing is completed, the system uses the prediction model constructed based on historical data and its corresponding robust scale factor to perform anomaly detection on the standardized data collected at each moment, and refine it into the anomaly analysis of single-device and multi-device associated data. Performing anomaly analysis on single-device and multi-device associated data simultaneously can not only capture local anomalies in a timely manner but also identify potential system-level problems, further improving the accuracy and sensitivity of anomaly detection and ensuring that problems in the production process can be effectively solved in the initial stage; As a technical solution for further implementation of this plan, when the anomaly detection module triggers a scheduling response, the production line state vector constructed by the system not only includes information on the online state and load rate of the equipment, but also comprehensively considers the task queue state, material inventory situation and the historical operation characteristics of the equipment to achieve a multi-dimensional description of the overall state of the production line. The constructed multi-dimensional state vector can comprehensively reflect the actual operation of the production line, enabling scheduling decisions to be based not on a single indicator but on multiple key factors, thus realizing more accurate and reasonable resource scheduling and risk control; As a further technical solution for the implementation of this solution, when generating a candidate scheduling plan, the system comprehensively considers the task delay expectation, task weight, and additional risk factor, and adopts an optimization objective function based on risk adjustment to dynamically allocate resources to balance the task delay and the additional risks brought by emergencies, so as to achieve the dual optimization of the scheduling plan in terms of efficiency and security. By introducing a risk adjustment mechanism, while taking into account the task completion efficiency, the additional resource occupation risks caused by emergencies are fully considered, making the scheduling result more robust and reliable, thus ensuring a high level of safety guarantee while the production process operates efficiently; As a further technical solution for the implementation of this solution, while the system verifies the feasibility of the candidate scheduling plan through simulation or constraint solving technology, it also uses the reinforcement learning algorithm to continuously update the scheduling strategy. The reinforcement learning process not only comprehensively considers the scheduling feedback reward, but also introduces a risk compensation and penalty mechanism to achieve the closed-loop adaptive adjustment of the system operating state. The method of combining simulation verification and online reinforcement learning can not only timely adjust the scheduling strategy to cope with the dynamically changing production environment, but also gradually optimize the decision-making model to achieve the self-evolution and continuous performance improvement of the system during long-term operation, further enhancing the robustness and adaptability of the system.

[0018] In this article, specific examples are used to elaborate on the principles and implementation methods of the present invention. The descriptions of the above examples are only used to help understand the method and its core idea of the present invention. The above are only the preferred implementation methods of the present invention. It should be noted that due to the limited nature of written expression and the objectively infinite specific structures, for those of ordinary skill in the art in this technical field, without departing from the principles of the present invention, several improvements, refinements or changes can be made, or the above technical features can be combined in an appropriate manner; these improvements, refinements, changes or combinations, or directly applying the concept and technical solution of the invention to other occasions without improvement, shall all be regarded as the protection scope of the present invention.

Claims

1. An injection molding production real-time monitoring and intelligent scheduling management system based on the Internet of Things, characterized in that It includes a data acquisition module, a data fusion module, an anomaly detection module, an adaptive scheduling module, and a feedback and learning module; The data acquisition module is used to collect the operating status, environmental parameters, and logistics information of each device in the injection molding production process in real time; The data fusion module is used to preprocess, standardize, and align the time series of data from different devices and different communication protocols; The anomaly detection module uses an anomaly detection method based on statistical thresholds and machine learning to perform real-time analysis on the fused data and issue an alarm when atypical anomalies are detected; After the anomaly detection module issues a warning, the adaptive scheduling module uses a rule base, a heuristic algorithm, and a reinforcement learning module to achieve dynamic rescheduling according to the device status, production tasks, and material supply situation; The feedback and learning module is used to receive the feedback data after the scheduling execution and continuously optimize the scheduling strategy using an online learning algorithm.

2. The real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things according to claim 1, wherein: The adaptive scheduling module includes: A rule base module for storing conventional production scheduling rules and plans; A heuristic scheduling algorithm module that globally searches for a production scheduling plan based on a genetic algorithm or an ant colony algorithm; A reinforcement learning module that iteratively updates the scheduling strategy using deep reinforcement learning based on the current environmental state and historical scheduling feedback.

3. The real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things according to claim 1, wherein: The data fusion module realizes the time-domain consistency of multi-source data by aligning and fusing the time series data from different sensors.

4. The real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things according to any one of claims 1 to 3, characterized in that: The real-time monitoring and intelligent scheduling management system for injection molding production based on the Internet of Things includes a real-time monitoring and intelligent scheduling management method for injection molding production based on the Internet of Things, which includes the following steps: Step 1: Use Internet of Things devices to collect key parameter data in the injection molding production process in real time to form a data sequence , preprocess the collected data to form a standardized time series data set ; Among them, represents the sampling moment, and the data is sourced from the temperature sensor, pressure sensor, and vibration sensor of the injection molding machine; Step 2: Use an anomaly scoring function to perform anomaly detection on the preprocessed data sequence. The formula of the anomaly scoring function is: ； In the formula, is a prediction model constructed based on historical data, is a robust scale factor calculated based on historical data. When , it is determined as an abnormal event, where is the threshold, that is, the tolerance set in advance; Step 3: The anomaly detection triggers the adaptive scheduling algorithm to construct the current state vector of the production line , where represents the set of equipment online status and load rates, represents the task queue status, represents the material inventory information; When generating a candidate scheduling plan, use an optimization objective function: ； In the formula, Indicates The expected delay of a task, The risk factor indicating the additional resource occupation caused by unexpected events related to the task. is the task weight coefficient, is the risk adjustment constant, is the total number of tasks; Step 4: Use the optimized reinforcement learning algorithm to update the scheduling strategy, and use the SARSA update formula: ； wherein, is the current state vector , is the current scheduling action, is the scheduling feedback reward, is the learning rate, is the discount factor, is the new state after scheduling execution, is the next scheduling action, is the risk penalty factor, is the risk compensation term, which is used to reflect the additional risk impact brought by anomalies or emergencies; Step 5: Verify the feasibility of the generated optimal scheduling plan through simulation or constraint solving technology. The constraint conditions are: ； In the formula, represents the allocation quantity of the -th type of resource in the optimized scheduling plan, and are respectively the minimum and maximum allowable allocation quantities of the -th type of resource, is the total number of resource types; after passing the verification, the real-time production data of the scheduling execution result is fed back to the system as new training samples for updating the prediction model , the candidate scheduling generation and reinforcement learning module , so as to realize closed-loop adaptive scheduling and continuously improve the system's response ability to unconventional anomalies and emergencies.

5. The real-time monitoring and intelligent scheduling management method for injection molding production based on the Internet of Things according to claim 4, characterized in that: Based on the real-time collection of key parameters in the injection molding production process, the system further performs noise elimination, outlier screening, and data completion processing on the original data to form a high-quality standardized time series data set.

6. The real-time monitoring and intelligent scheduling management method for injection molding production based on the Internet of Things according to claim 4, characterized in that: After the data preprocessing is completed, the system uses a prediction model constructed based on historical data and its corresponding robust scale factor to perform anomaly detection on the standardized data collected at each moment, and refine it into the anomaly analysis of single-device and multi-device related data.

7. The real-time monitoring and intelligent scheduling management method for injection molding production based on the Internet of Things according to claim 4, characterized in that: When the anomaly detection module triggers a scheduling response, the production line state vector constructed by the system comprehensively considers the task queue state, material inventory situation, and device historical operation characteristics in addition to the device online state and load rate information, realizing a multi-dimensional description of the overall state of the production line.

8. The real-time monitoring and intelligent scheduling management method for injection molding production based on the Internet of Things according to claim 4, characterized in that: When generating a candidate scheduling plan, the system comprehensively considers the task delay expectation, task weight, and additional risk factor, and uses an optimization objective function based on risk adjustment to dynamically allocate resources to balance the task delay and the additional risk brought by emergencies, so as to achieve the dual optimization of the scheduling plan in terms of efficiency and safety.

9. The real-time monitoring and intelligent scheduling management method for injection molding production based on the Internet of Things according to claim 4, characterized in that: While the system verifies the feasibility of candidate scheduling schemes through simulation or constraint solving techniques, it also continuously updates the scheduling strategy using a reinforcement learning algorithm. In the reinforcement learning process, not only is the scheduling feedback reward comprehensively considered, but also a risk compensation and penalty mechanism is introduced to achieve closed-loop adaptive adjustment of the system operating state.

Images