Grain and oil production line tracking detection method and system
By constructing local and real-time cumulative chains and combining data from IoT devices, the system quantifies the production capacity and rhythm disorder of different zones, dynamically monitors the spread of congestion in grain and oil production lines, solves the problem of inaccurate monitoring caused by differences in production capacity between zones, and improves the intelligence and early warning capabilities of the production lines.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- JIANGXI IND & TRADE VOCATIONAL & TECH COLLEGE (JIANGXI PROVINCIAL GRAIN CADRE SCHOOL JIANGXI PROVINCIAL GRAIN WORKERS SECONDARY VOCATIONAL SCHOOL)
- Filing Date
- 2025-12-25
- Publication Date
- 2026-05-12
AI Technical Summary
Existing technologies cannot effectively quantify the differences in production capacity between different zones, resulting in inaccurate production line monitoring and congestion detection. This may lead to delayed early warnings or excessive intervention, affecting overall production capacity and product quality.
By constructing local and real-time cumulative chains, the congestion response rate and diffusion impact of each partition can be identified. By combining historical and real-time production data collected by IoT devices, the production load and rhythm disorder of each partition can be quantified, and potential abnormal partitions can be dynamically monitored and warned.
It has improved the level of dynamic monitoring and intelligence of grain and oil production lines, reduced the risk of local congestion spreading into production line-level blockage, and achieved accurate early warning and scheduling optimization for downstream areas.
Smart Images
Figure CN122022255A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of grain and oil production data analysis technology, and in particular to a method and system for tracking and detecting grain and oil production lines. Background Technology
[0002] In grain and oil production, production lines typically consist of multiple zones arranged sequentially, each corresponding to different processes such as cleaning, processing, cooling, and packaging. The continuity and stability of the production line are crucial to production efficiency and product quality. Due to differences in processes, equipment automation levels, manual operation efficiency, and buffer capacity limitations among zones, the same upstream input may produce different congestion responses in different zones. For example, for the same amount of work-in-process inventory, a highly automated zone may be able to quickly clear the backlog, while a zone with low automation or a higher proportion of manual operation may quickly reach its processing limit, leading to production line delays, downstream process disruptions, and even affecting overall capacity and product quality.
[0003] Some production line monitoring and congestion detection methods treat each zone as the same processing unit, lacking effective quantification of the differences in production capacity between zones. Furthermore, when faced with uncertainties such as transportation delays and buffer zone congestion, they cannot accurately determine whether congestion in one zone will affect downstream zones, potentially leading to delayed warnings or excessive intervention. Therefore, accurately identifying the actual processing capacity of a zone under congestion conditions and its impact on downstream zones, while considering the differences in production characteristics across zones, has become a key issue for improving real-time production line monitoring and optimizing production scheduling. Summary of the Invention
[0004] To address the aforementioned technical issues, this invention proposes a method for tracking and detecting congestion response levels in grain and oil production lines based on historical data analysis. This method identifies the potential congestion status of different zones and their potential impact on downstream areas, thereby providing a reliable basis for real-time congestion early warning and scheduling optimization of production lines.
[0005] To achieve the above objectives, the present invention provides the following technical solution: A method for tracking and detecting grain and oil production lines includes: Historical grain and oil production data from multiple zones in the grain and oil production line are collected using IoT devices. Based on the historical grain and oil production data, product receiving data and product output data for each zone are extracted. A local cumulative chain for each zone is constructed based on the product receiving data and product output data. By combining local cumulative chains to identify production acceleration events in historical grain and oil production data, the congestion response rate and congestion response duration of each production acceleration event are extracted, and congestion resistance data for each partition are generated based on the congestion response rate and congestion response duration. Real-time congestion detection is performed on multiple zones in the grain and oil production line. Real-time grain and oil production data of each zone are collected, and a receiving interval sequence and an output interval sequence are constructed for each zone. Based on the receiving interval sequence and the output interval sequence, the production deviation status of each zone is determined. A real-time cumulative chain is constructed for each zone based on real-time grain and oil production data. Candidate abnormal zones are determined based on production deviation status and real-time cumulative chains. Congestion diffusion impact detection is performed on each candidate abnormal zone based on congestion resistance data to obtain the congestion diffusion detection results of the grain and oil production line.
[0006] Preferably, the congestion response rate and congestion response duration for each production acceleration event are extracted, and congestion resistance data for each partition is generated based on the congestion response rate and congestion response duration, including: The local accumulation chain includes the cumulative work-in-process quantity of a partition within multiple sliding windows. Based on historical grain and oil production data, the partition's basic production rate within multiple sliding windows is determined. The work-in-process accumulation rate of each sliding window is calculated based on the local accumulation chain. Combining the partition's basic production rate and work-in-process accumulation rate, multiple abnormal windows for each partition are determined, and multiple production acceleration events for each partition are constructed. The peak production rate of each production acceleration event is extracted as the congestion response rate of the production acceleration event. The duration of the production acceleration event at the peak production rate is determined to obtain the congestion response duration of the production acceleration event. The congestion response rates and congestion response durations of multiple production acceleration events in a partition are aggregated to generate congestion resistance data for each partition, including short-term accelerated production rate and short-term accelerated production duration.
[0007] Preferably, determining the production deviation state of each partition based on the received interval sequence and the output interval sequence includes: Each element in the receiving interval sequence is the difference in the receiving rate of the product's received data within two adjacent sliding windows, and each element in the output interval sequence is the difference in the output rate of the product's output data within two adjacent sliding windows. The receiving interval variation parameter of the partition is calculated based on the receiving interval sequence, and the output interval variation parameter of the partition is calculated based on the output interval sequence. The production rhythm difference of the partition is detected based on the receiving interval sequence and the output interval sequence, and the production rhythm disorder parameter of the partition is calculated. The production congestion status parameter of the partition is determined based on the receiving interval variation parameter, the output interval variation parameter and the production rhythm disorder parameter, and the production deviation status information of each partition is output.
[0008] Preferably, the process of determining candidate anomalous partitions based on production deviation status and real-time cumulative chains, and detecting the congestion diffusion impact on each candidate anomalous partition based on congestion resistance data, includes: The real-time accumulation rate of each partition is calculated based on the real-time accumulation chain. The real-time accumulation rate of the partition and the production congestion status parameters are combined to mark multiple partitions as abnormal, resulting in multiple candidate abnormal partitions. The diffusion impact partition of each candidate abnormal partition is determined. Based on the congestion resistance data of the diffusion impact partition, the congestion diffusion impact monitoring window and diffusion impact output of the candidate abnormal partition are determined. Based on the congestion diffusion impact monitoring window, the sliding impact monitoring of the candidate abnormal partition is carried out. The diffusion impact early warning data of the candidate abnormal partition is generated by combining the diffusion impact output.
[0009] Preferably, the sliding impact monitoring of candidate abnormal zones is performed based on the congestion diffusion impact monitoring window, and the diffusion impact early warning data of candidate abnormal zones is generated by combining the diffusion impact output, including: Based on the congestion diffusion impact monitoring window, the real-time grain and oil production data of candidate abnormal zones are processed by sliding window. The local output of real-time grain and oil production data in multiple congestion diffusion impact monitoring window areas is calculated. If the difference between the diffusion impact output of the candidate abnormal zone and the local output is less than the diffusion impact warning threshold, the diffusion impact warning data of the candidate abnormal zone is generated.
[0010] Preferably, the cross-correlation coefficient between the receiving interval sequence and the output interval sequence of the partition is calculated as a parameter for the production rhythm disorder of the partition.
[0011] A grain and oil production line tracking and detection system includes: The grain and oil production data preprocessing module is used to collect historical grain and oil production data from multiple zones in the grain and oil production line based on IoT devices, extract product receiving data and product output data from each zone based on the historical grain and oil production data, and construct a local cumulative chain for each zone based on the product receiving data and product output data. The congestion resistance analysis module is used to identify production acceleration events in historical grain and oil production data by combining local cumulative chains, extract the congestion response rate and congestion response duration of each production acceleration event, and generate congestion resistance data for each partition based on the congestion response rate and congestion response duration. The production status deviation identification module is used to perform real-time congestion detection on multiple zones in the grain and oil production line, collect real-time grain and oil production data of each zone, construct the receiving interval sequence and output interval sequence of each zone, and determine the production deviation status of each zone based on the receiving interval sequence and output interval sequence. The congestion diffusion detection module is used to construct a real-time cumulative chain for each partition based on real-time grain and oil production data, determine candidate abnormal partitions based on production deviation status and real-time cumulative chain, and perform congestion diffusion impact detection on each candidate abnormal partition based on congestion resistance data to obtain the congestion diffusion detection results of the grain and oil production line.
[0012] Compared with the prior art, the beneficial effects of the embodiments of the present invention are at least one of the following: This invention quantifies the production load of different zones by constructing a local cumulative chain, extracts congestion resistance data of each zone by combining historical production acceleration events, and characterizes the tolerance of different zones to sudden congestion. It quantifies the production rhythm disorder state of zones in real time based on the receive / output interval sequence, capturing early deviation states of input-output disconnection. It integrates the real-time cumulative chain with production deviation to locate candidate abnormal zones, and dynamically calculates the diffusion impact monitoring window and early warning threshold based on the congestion resistance of downstream zones. By performing quantitative modeling of zone capacity differences, it identifies potential anomalies during the production rhythm disorder stage and adaptively adjusts the early warning mechanism based on downstream resistance, reducing the risk of local congestion spreading to production line-level blockage and improving the intelligence level of dynamic monitoring of grain and oil production lines. Attached Figure Description
[0013] Figure 1 This is a flowchart of a grain and oil production line tracking and detection method provided in one embodiment of the present invention.
[0014] Figure 2 This is a structural diagram of a grain and oil production line tracking and detection system provided in one embodiment of the present invention. Detailed Implementation
[0015] To enable those skilled in the art to better understand the technical solutions of this invention, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention.
[0016] Please see Figure 1 One embodiment of the present invention provides a method for tracking and detecting grain and oil production lines, comprising the following steps: Step S1: Collect historical grain and oil production data from multiple zones in the grain and oil production line using IoT devices. Extract product receiving data and product output data from each zone based on the historical grain and oil production data. Construct a local cumulative chain for each zone based on the product receiving data and product output data.
[0017] Specifically, historical grain and oil production data is collected through multiple IoT devices deployed in various sections of the production line. These IoT devices include, but are not limited to, hopper counters and conveyor belt flow sensors for measuring the quantity or weight of semi-finished products, barcode scanners and RFID devices for marking the flow of each batch of products or materials, and temperature, humidity, or pressure sensors for assisting in determining the production status of each process. The collected data can be uploaded to an industrial cloud platform or internet platform via an industrial network for centralized storage and management, providing a data foundation for further analysis between different sections.
[0018] Since different zones output different types of semi-finished products, their quantities also correspond differently to the output of final products. Therefore, the raw data can be converted into a comparable, unified metric. For example, based on historical data, the proportion of final products corresponding to the semi-finished products output by each zone can be statistically analyzed, and the production volume of different zones can be normalized according to this proportion. For instance, if 100 semi-finished products in zone A can form one unit of final product, and 400 semi-finished products in zone B can form one unit of final product, then the output of zones A and B can be converted into a unified number of final product units, facilitating comparisons between data from different zones.
[0019] After normalizing the data, we obtain product receiving data including the time and quantity of work-in-process arriving in this partition from upstream, and product output data including the time and quantity of work-in-process completed in the partition and flowing downstream. We calculate the difference between the work-in-process receiving and output quantities for each partition in different time periods to obtain the cumulative amount of work-in-process for the partition in multiple time periods. That is, the cumulative amount of work-in-process produced by the partition is normalized to the same dimension. This represents the total cumulative amount of products in the partition when the actual output is lower than the input during the production process, so as to construct a local cumulative chain that represents the production load status of the partition in different time periods.
[0020] Step S2: Combine local cumulative chain identification of production acceleration events in historical grain and oil production data, extract the congestion response rate and congestion response duration of each production acceleration event, and generate congestion resistance data for each partition based on the congestion response rate and congestion response duration.
[0021] Specifically, by using local cumulative chains to identify periods in historical data where the reception volume of each partition is significantly higher than the average processing capacity of the partition in a short period of time, these periods are marked as production acceleration events. The peak production rate of the partition in different production acceleration events and the duration of maintaining the peak production rate are analyzed. Finally, the data of multiple production acceleration events are aggregated to obtain congestion resistance data that characterizes the partition's tolerance or resistance to congestion.
[0022] In this process, the identification of production acceleration events can be achieved by determining the basic production rate of each zone across multiple sliding windows based on historical grain and oil production data. The basic production rate can be calculated by averaging the production rate under normal production conditions over the most recent day or seven days (i.e., the number of product units produced per unit time). Furthermore, the cumulative work-in-process (WIP) rate corresponding to multiple cumulative WIP quantities is calculated based on local cumulative chains. For example, the difference in cumulative WIP quantities between two sliding windows represents the cumulative WIP rate within a single window. When the cumulative WIP rate of a zone within a certain sliding window is greater than 0, and the actual production rate within that window is greater than a certain multiple of the basic production rate, it indicates that the zone is in a production acceleration state within that sliding window. This sliding window is marked as an abnormal window. After identifying multiple abnormal windows for each zone, consecutive abnormal windows are recorded as a single production acceleration event, indicating that the zone experiences input pressure exceeding normal production capacity and a short-term production acceleration phenomenon during that time period.
[0023] After constructing multiple production acceleration events, the peak production rate corresponding to each acceleration event is determined and denoted as the congestion response rate of the acceleration event. Furthermore, the duration for which the production rate maintained by a partition at a certain proportion of the peak production rate is extracted and denoted as the congestion response duration of the acceleration event, thus obtaining the congestion response rate and congestion response duration for multiple samples. Based on this, the data from these samples are aggregated. For example, based on preset quantiles, multiple representative short-term accelerated production rates and short-term accelerated production durations are determined to obtain congestion resistance data for each partition. This indicates that during normal production, when faced with product congestion (i.e., when the input is higher than the normal production speed), the partition can temporarily increase its production rate and maintain it for a certain period, thus partially alleviating product backlog through short-term acceleration. The larger the short-term accelerated production rate and short-term accelerated production duration, the faster the partition can recover from backlog and the less likely it is to affect downstream partitions.
[0024] Step S3: Perform real-time congestion detection on multiple zones in the grain and oil production line, collect real-time grain and oil production data for each zone, construct the receiving interval sequence and output interval sequence for each zone, and determine the production deviation status of each zone based on the receiving interval sequence and output interval sequence.
[0025] Specifically, IoT devices are used to monitor the congestion status of different zones in the grain and oil production line in real time. For the collected real-time grain and oil production data, including product receiving data and product output data during the real-time production process, the difference in the semi-finished product receiving rate between two adjacent sliding windows and the difference in the semi-finished product output rate between two adjacent sliding windows are extracted. These are combined according to the time sequence relationship to obtain the receiving interval sequence and the output interval sequence for each zone. Based on the interval sequence representing the differences between the upstream and downstream of the zone, abnormal changes in the production rhythm of the zone are identified, and the production deviation status of different zones is determined.
[0026] In this process, the receiving interval variation parameter of the partition is calculated based on the receiving interval sequence, and the output interval variation parameter of the partition is calculated based on the output interval sequence. The interval variation parameter can be the ratio of the variance to the mean of the interval sequence. In the early stages of congestion, clustering may occur within the partition. For example, upstream arrivals become more erratic due to the influx of semi-finished product clusters, while downstream departures become more even due to increased input. This indicates a more stable output due to sufficient raw material supply, representing an early signal of congestion. Simultaneously, production rhythm differences are detected within the partition based on the receiving and output interval sequences. For example, the cross-correlation between the two sequences is calculated to obtain the partition's production rhythm disorder parameter, indicating whether the upstream input rhythm can match the downstream rhythm after time shift. Finally, the production congestion state parameter of the partition is determined based on the receiving interval variation parameter, output interval variation parameter, and production rhythm disorder parameter, where the production congestion state parameter = (receiving interval variation parameter / output interval variation parameter) * (1 - production rhythm disorder parameter). As congestion intensifies, upstream input becomes more chaotic than downstream output, and the rhythm difference between the two ends gradually increases, resulting in an increasing trend in the production congestion state parameter. The production congestion state parameter can characterize the production deviation state of a partition, that is, the quantitative information on the degree of asynchronous rhythm between input and output.
[0027] Step S4: Construct a real-time cumulative chain for each partition based on real-time grain and oil production data. Determine candidate abnormal partitions based on production deviation status and real-time cumulative chains. Perform congestion diffusion impact detection on each candidate abnormal partition based on congestion resistance data to obtain the congestion diffusion detection results of the grain and oil production line.
[0028] Specifically, the construction method of the real-time accumulation chain is the same as that of the local accumulation chain described above. For the identification of candidate abnormal partitions, the real-time accumulation rate representing the partition can be calculated based on the multiple accumulated work-in-process quantities in the real-time accumulation chain. The production congestion state parameter and the real-time accumulation rate are then combined to identify whether different partitions have potential congestion. During the production process, there may be some noise information, such as brief product output stacking, causing the real-time accumulation rate to be greater than 0. Therefore, one of the marking conditions for candidate abnormal partitions can be a real-time accumulation rate greater than 0 or a baseline noise level for the accumulation rate determined based on historical data. Further, the production congestion state parameter is limited to being greater than a pre-set threshold, which can be determined by the quantile of the production congestion state parameter in historical normal operation data, for example, 90%. This process identifies and marks multiple candidate abnormal partitions with abnormal growth trends.
[0029] For multiple candidate anomalous partitions identified through labeling, and combining the congestion resistance data of each partition, the potential diffusion impact of congestion in each candidate anomalous partition on downstream partitions after transmission is further analyzed. This process first determines the diffusion impact partition for each candidate anomalous partition, i.e., the next partition adjacent to the candidate anomalous partition. Based on the congestion resistance data of the diffusion impact partition, the congestion diffusion impact monitoring window and diffusion impact output of the candidate anomalous partition are determined. Sliding impact monitoring is then performed on the candidate anomalous partition based on the congestion diffusion impact monitoring window.
[0030] Specifically, the congestion diffusion impact monitoring window is the time range corresponding to the short-term accelerated production duration in the congestion resistance data, that is, the longest duration for which the downstream zone can maintain the peak production rate in the short-term accelerated mode. The diffusion impact output represents the maximum output that the zone can handle while maintaining the short-term accelerated state, which measures the upper limit of the output of the zone to resist the sudden increase in upstream input within a short time window.
[0031] For the sliding impact monitoring process, specifically, the real-time grain and oil production data of candidate abnormal zones are processed using a sliding window based on the congestion diffusion impact monitoring window. The local output of real-time grain and oil production data in multiple congestion diffusion impact monitoring window areas is statistically analyzed. If the difference between the diffusion impact output and the local output of the candidate abnormal zone is less than the diffusion impact warning threshold, meaning the output output of the candidate abnormal zone within the window is close to the maximum output of the diffusion impact zone under short-term acceleration, and the input volume of the upstream zone is close to the short-term limit output of the downstream zone within a time window that the downstream zone can only support, this indicates that after a certain period, the downstream zone is likely to face a short-term input volume exceeding its maximum capacity. In this case, equipment or workstation congestion may occur, leading to increased waiting time and preventing materials from reaching the production point in a timely manner, thus affecting the normal production balance of the zone. For example, a decrease in output may occur. In this scenario, diffusion impact warning data of candidate abnormal zones is generated for early warning, rather than waiting until the downstream actually starts to accumulate before discovering the problem. This facilitates timely adjustments by grain and oil production line managers, minimizing the impact of local accumulation on the normal production of grain and oil products, achieving dynamic monitoring of the overall production safety of the production line, and improving the accuracy and reliability of congestion detection.
[0032] Please see Figure 2 Based on the same inventive concept, one embodiment of this invention also provides a grain and oil production line tracking and detection system, comprising: The grain and oil production data preprocessing module is used to collect historical grain and oil production data from multiple zones in the grain and oil production line based on IoT devices, extract product receiving data and product output data from each zone based on the historical grain and oil production data, and construct a local cumulative chain for each zone based on the product receiving data and product output data. The congestion resistance analysis module is used to identify production acceleration events in historical grain and oil production data by combining local cumulative chains, extract the congestion response rate and congestion response duration of each production acceleration event, and generate congestion resistance data for each partition based on the congestion response rate and congestion response duration. The production status deviation identification module is used to perform real-time congestion detection on multiple zones in the grain and oil production line, collect real-time grain and oil production data of each zone, construct the receiving interval sequence and output interval sequence of each zone, and determine the production deviation status of each zone based on the receiving interval sequence and output interval sequence. The congestion diffusion detection module is used to construct a real-time cumulative chain for each partition based on real-time grain and oil production data, determine candidate abnormal partitions based on production deviation status and real-time cumulative chain, and perform congestion diffusion impact detection on each candidate abnormal partition based on congestion resistance data to obtain the congestion diffusion detection results of the grain and oil production line.
[0033] The above are merely specific embodiments of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art. Parts not described in detail in this specification are prior art known to those skilled in the art.
Claims
1. A method for tracking and detecting grain and oil production lines, characterized in that, include: Historical grain and oil production data from multiple zones in the grain and oil production line are collected using IoT devices. Based on the historical grain and oil production data, product receiving data and product output data for each zone are extracted. A local cumulative chain for each zone is constructed based on the product receiving data and product output data. By combining local cumulative chains to identify production acceleration events in historical grain and oil production data, the congestion response rate and congestion response duration of each production acceleration event are extracted, and congestion resistance data for each partition are generated based on the congestion response rate and congestion response duration. Real-time congestion detection is performed on multiple zones in the grain and oil production line. Real-time grain and oil production data of each zone are collected, and a receiving interval sequence and an output interval sequence are constructed for each zone. Based on the receiving interval sequence and the output interval sequence, the production deviation status of each zone is determined. A real-time cumulative chain is constructed for each zone based on real-time grain and oil production data. Candidate abnormal zones are determined based on production deviation status and real-time cumulative chains. Congestion diffusion impact detection is performed on each candidate abnormal zone based on congestion resistance data to obtain the congestion diffusion detection results of the grain and oil production line.
2. The method for tracking and detecting grain and oil production lines according to claim 1, characterized in that, Extract the congestion response rate and congestion response duration for each production acceleration event, and generate congestion resistance data for each partition based on the congestion response rate and congestion response duration, including: The local accumulation chain includes the cumulative work-in-process quantity of a partition within multiple sliding windows. Based on historical grain and oil production data, the partition's basic production rate within multiple sliding windows is determined. The work-in-process accumulation rate of each sliding window is calculated based on the local accumulation chain. Combining the partition's basic production rate and work-in-process accumulation rate, multiple abnormal windows for each partition are determined, and multiple production acceleration events for each partition are constructed. The peak production rate of each production acceleration event is extracted as the congestion response rate of the production acceleration event. The duration of the production acceleration event at the peak production rate is determined to obtain the congestion response duration of the production acceleration event. The congestion response rates and congestion response durations of multiple production acceleration events in a partition are aggregated to generate congestion resistance data for each partition, including short-term accelerated production rate and short-term accelerated production duration.
3. The method for tracking and detecting grain and oil production lines according to claim 2, characterized in that, Determining the production deviation status of each partition based on the received interval sequence and the output interval sequence includes: Each element in the receiving interval sequence is the difference in the receiving rate of the product's received data within two adjacent sliding windows, and each element in the output interval sequence is the difference in the output rate of the product's output data within two adjacent sliding windows. The receiving interval variation parameter of the partition is calculated based on the receiving interval sequence, and the output interval variation parameter of the partition is calculated based on the output interval sequence. The production rhythm difference of the partition is detected based on the receiving interval sequence and the output interval sequence, and the production rhythm disorder parameter of the partition is calculated. The production congestion status parameter of the partition is determined based on the receiving interval variation parameter, the output interval variation parameter and the production rhythm disorder parameter, and the production deviation status information of each partition is output.
4. The method for tracking and detecting grain and oil production lines according to claim 3, characterized in that, Candidate anomalous partitions are identified based on production deviation status and real-time cumulative chains. Congestion diffusion impact detection is then performed on each candidate anomalous partition based on congestion resistance data, including: The real-time accumulation rate of each partition is calculated based on the real-time accumulation chain. The real-time accumulation rate of the partition and the production congestion status parameters are combined to mark multiple partitions as abnormal, resulting in multiple candidate abnormal partitions. The diffusion impact partition of each candidate abnormal partition is determined. Based on the congestion resistance data of the diffusion impact partition, the congestion diffusion impact monitoring window and diffusion impact output of the candidate abnormal partition are determined. Based on the congestion diffusion impact monitoring window, the sliding impact monitoring of the candidate abnormal partition is carried out. The diffusion impact early warning data of the candidate abnormal partition is generated by combining the diffusion impact output.
5. The method for tracking and detecting grain and oil production lines according to claim 4, characterized in that, Based on the congestion diffusion impact monitoring window, sliding impact monitoring is performed on candidate abnormal zones. Combined with the diffusion impact output, diffusion impact early warning data for candidate abnormal zones is generated, including: Based on the congestion diffusion impact monitoring window, the real-time grain and oil production data of candidate abnormal zones are processed by sliding window. The local output of real-time grain and oil production data in multiple congestion diffusion impact monitoring window areas is calculated. If the difference between the diffusion impact output of the candidate abnormal zone and the local output is less than the diffusion impact warning threshold, the diffusion impact warning data of the candidate abnormal zone is generated.
6. The method for tracking and detecting grain and oil production lines according to claim 3, characterized in that, The cross-correlation coefficient between the receiving interval sequence and the output interval sequence of the partition is calculated as a parameter for the production rhythm disorder of the partition.
7. A grain and oil production line tracking and detection system, characterized in that, The system is used to implement the grain and oil production line tracking and detection method according to any one of claims 1-6, comprising: The grain and oil production data preprocessing module is used to collect historical grain and oil production data from multiple zones in the grain and oil production line based on IoT devices, extract product receiving data and product output data from each zone based on the historical grain and oil production data, and construct a local cumulative chain for each zone based on the product receiving data and product output data. The congestion resistance analysis module is used to identify production acceleration events in historical grain and oil production data by combining local cumulative chains, extract the congestion response rate and congestion response duration of each production acceleration event, and generate congestion resistance data for each partition based on the congestion response rate and congestion response duration. The production status deviation identification module is used to perform real-time congestion detection on multiple zones in the grain and oil production line, collect real-time grain and oil production data of each zone, construct the receiving interval sequence and output interval sequence of each zone, and determine the production deviation status of each zone based on the receiving interval sequence and output interval sequence. The congestion diffusion detection module is used to construct a real-time cumulative chain for each partition based on real-time grain and oil production data, determine candidate abnormal partitions based on production deviation status and real-time cumulative chain, and perform congestion diffusion impact detection on each candidate abnormal partition based on congestion resistance data to obtain the congestion diffusion detection results of the grain and oil production line.