Intelligent production monitoring and scheduling system and method based on double fusion technology

Abstract

The invention discloses an intelligent production monitoring and scheduling system and method based on a double fusion technology, and belongs to the field of industrial automation. The system comprises an MEScada management platform, a device end bee executor and a database, and data interaction is achieved through standardized API communication. And the bee executor is deployed at the edge end of the equipment, comprises a polling dispatching center, an equipment controller and a data processing module, and processes acquired data by adopting a multi-level cache mechanism. The MEScada management platform is used for constructing a factory physical model, a process model and work order scheduling logic, and comprises an instruction designer, a view designer and a report designer. According to the method, equipment acquisition point locations are configured through a bee actuator, multi-level cache processing data is adopted, a factory model is constructed on an MEScada platform, equipment interaction logic is arranged, and production resources are dynamically scheduled based on real-time data. According to the invention, real-time monitoring and intelligent scheduling of the production process are realized, and the production efficiency and the resource utilization rate are improved.

Description

Technical Field

[0001] The present application relates to the field of industrial automation, and specifically to an intelligent production monitoring and scheduling system and method based on dual-industry fusion technology. Background Art

[0002] Intelligent manufacturing has become an important direction for the development of modern industry. In the manufacturing process, real-time monitoring and intelligent scheduling systems are of great significance for improving production efficiency, reducing costs and ensuring product quality.

[0003] However, existing production monitoring and scheduling systems suffer from the following major issues: First, IT and OT system integration is difficult, resulting in low data collection efficiency; second, the lack of an effective multi-level caching mechanism makes it difficult to cope with high-frequency data collection scenarios; third, inconsistent protocol standards lead to poor system interoperability; fourth, limited factory model building capabilities make it difficult to adapt to complex industrial scenarios; and fifth, a lack of flexible visual design tools results in a poor user experience. These issues have severely hindered the further development of intelligent manufacturing. Summary of the Invention

[0004] To solve the above problems, the present invention provides an intelligent production monitoring and scheduling system and method based on dual-informatization fusion technology.

[0005] The technical solution adopted by the present invention to solve its technical problems is to provide an intelligent production monitoring and scheduling system based on dual-informatization fusion technology, including three parts: the MEScada management platform, the device-side bee actuator, and a database. The MEScada management platform is used to build the factory physical model, process model, work order scheduling logic, and visual monitoring interface; the device-side bee actuator is independently deployed at the device edge to collect device data and perform data cleaning, protocol conversion, and multi-level cache processing; the database is used to store factory model data, process parameters, and real-time production data; the bee actuator communicates with the MEScada management platform through a standardized API and supports at least one of the MQTT, OPCUA, and HTTP protocols for data forwarding.

[0006] Furthermore, the bee executor includes: a polling scheduling center, which manages the concurrent collection tasks of multiple devices according to preset time slices and priorities; a device controller, which sends data requests to the PLC device and receives response data to the cache queue; and a data processing module, which performs rule verification and forwarding control on the cached data.

[0007] Preferably, the bee executor adopts a multi-level cache mechanism, including: a first-level memory ring cache area, which pre-allocates a fixed-size continuous memory space and writes real-time data streams through head and tail pointer loop overwriting; a second-level persistent cache area, which stores data in a non-volatile storage medium to ensure data reliability and peak shaving; a cache elimination module, which eliminates invalid data based on a time sliding window or least recently used strategy.

[0008] Furthermore, the MEScada management platform includes: an instruction designer, which arranges equipment control logic by visually dragging node modules, and the node modules include JavaScript script nodes, function blocks and business logic nodes; a view designer, which generates real-time production monitoring screens and work order progress visualization interfaces; and a report designer, which generates equipment OEE, quality pass rate and output statistical analysis reports.

[0009] Preferably, the scheduling logic configured by the instruction designer includes: responding to the production scheduling instruction, as a trigger condition, automatically issuing a scheduling instruction to the device side (AVG\collaborative robot\machine production equipment), and the instruction includes parameters such as the target device address and the device control value.

[0010] Furthermore, the factory physical model is a hierarchical structure, including four associated levels: factory, workshop, production line, and workstation, and each level is associated with corresponding production equipment, production process and material formula data.

[0011] The present invention also provides an intelligent production monitoring and scheduling method based on dual-industry fusion technology, comprising the following steps: configuring equipment collection points and alarm conditions through a bee actuator, and periodically polling PLC register data; using a multi-level cache mechanism to process the collected data, and forwarding it to the MEScada management platform through a protocol conversion module; building a factory physical model, process route, and work order task set on the MEScada management platform; arranging equipment interaction logic through an instruction designer, and issuing work order tasks and recipes or scheduling instructions to the edge end for execution;

[0012] Monitor the execution status of work orders based on real-time data and dynamically schedule equipment, AGVs, and robotic arms for collaborative operations.

[0013] Preferably, the multi-level cache mechanism specifically includes: writing high-frequency real-time data into a memory ring cache area, using a circular overwrite strategy to ensure timeliness; writing data that needs to be persisted into a disk buffer to prevent loss due to system power failure; and applying an elimination strategy based on the frequency of data changes to release storage space.

[0014] Furthermore, the dynamic scheduling includes: when it is detected that the material in the line-side warehouse is lower than a threshold, the AGV scheduling instruction is automatically triggered, and the robotic arm is linked to complete the loading operation.

[0015] Preferably, a data cleaning step is further included before data forwarding: the original data is converted by a calculation formula configured at the edge end, and the formula includes linear scaling, unit conversion or outlier filtering.

[0016] Due to the adoption of the above technical solution, the present application has at least one of the following beneficial effects compared with the prior art:

[0017] 1. Through the collaborative work of the MEScada management platform and the Bee actuator on the device side, seamless integration of OT and IT systems is achieved, solving problems such as difficult data integration, poor standardization and compatibility in traditional systems.

[0018] 2. The Bee Actuator supports multiple protocol access and can be flexibly configured according to the characteristics of different types of equipment to achieve unified access and transmission of digital signals of various types of equipment, wired and wireless, long and short distances.

[0019] 3. The multi-level caching mechanism effectively solves network delay and data concurrency problems, ensuring system stability and data integrity under high load conditions.

[0020] 4. The instruction designer quickly builds business processes through visual drag and drop, which greatly reduces the workload of developers and improves system configuration efficiency.

[0021] 5. Through data modeling such as factory modeling and process modeling, combined with predictive maintenance technology, the system can promptly detect and resolve quality problem factors, significantly improving product quality and production efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0023] in:

[0024] Figure 1 A schematic diagram of the framework of an intelligent production monitoring and scheduling system based on dual-informatization fusion technology provided in one embodiment of the present application;

[0025] Figure 2 A schematic diagram of the forwarding principle provided in one embodiment of the present application;

[0026] Figure 3 This is a schematic diagram of the collection and polling principle of an embodiment of the present application;

[0027] Figure 4 This is a schematic diagram of the cache design principle of an embodiment of the present application. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. It will be understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for ease of description, only some, rather than all, structures related to the present application are shown in the drawings. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0029] The terms "first," "second," and the like in this application are used to distinguish between different objects, not to describe a particular order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements, but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, product, or apparatus.

[0030] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0031] Example 1

[0032] like Figure 1 As shown, Figure 1 A schematic diagram of the framework of an intelligent production monitoring and scheduling system based on dual-informatization fusion technology provided in one embodiment of the present application, including a MEScada management platform, a device-side bee executor, and a database.

[0033] The MEScada management platform is used to build factory physical models, process models, work order scheduling logic, and visual monitoring interfaces. The MEScada management platform adopts a B / S architecture design, and all system functions can be accessed through a web browser without the need to install additional software on the client. The platform adopts a modular design, and its core functional modules include a factory modeling module, a process route configuration module, a work order management module, and a monitoring large screen module. The factory modeling module supports the construction of the factory physical structure by dragging and dropping, and defines equipment parameters and associations; the process route configuration module is used to set product production procedures, process parameters, and quality control points; the work order management module is responsible for production planning and scheduling, material requirement calculation, and production progress tracking; and the monitoring large screen module provides real-time production status visualization.

[0034] like Figure 2 As shown, Figure 2 A schematic diagram of the forwarding principle provided for an embodiment of the present application shows that the device-side bee executor is independently deployed at the edge of the device to collect device data and perform data cleaning, protocol conversion, and multi-level cache processing. The bee executor adopts a lightweight containerized design and can be deployed on an industrial-grade edge computing gateway or embedded device, supporting ARM and x86 architectures. The executor has built-in multiple industrial protocol parsing engines and can directly connect to mainstream PLC devices such as Siemens, Mitsubishi, and Omron to collect equipment operating status, process parameters, and alarm information. The bee executor has local data processing capabilities and can clean and convert the collected raw data, including unit conversion, linear calibration, outlier filtering and other operations to ensure data quality. At the same time, the executor implements a multi-level cache mechanism to effectively cope with network fluctuations and system peak load conditions.

[0035] The database is used to store factory model data, process parameters, and real-time production data. The system employs a hybrid database architecture, using a relational database to store structured factory models, process recipes, bills of materials, and other foundational data; a time-series database to store frequently collected equipment operation data; and a document-based database to store unstructured work order documents, operation records, and other information. The database cluster utilizes a master-slave replication architecture to ensure high data availability, and uses partitioning and sharding technologies to improve query performance even with large data volumes.

[0036] The Bee Actuator communicates with the MEScada management platform through a standardized API and supports at least one of the MQTT, OPCUA, and HTTP protocols for data forwarding. The system adopts the RESTful API design specification, which defines standard interfaces such as device registration, data reporting, and command issuance. At the data transmission level, the system supports multiple communication protocols: the MQTT protocol is suitable for scenarios with limited bandwidth and unstable network quality, adopts a publish / subscribe model, and supports QoS service quality assurance; the OPCUA protocol provides a unified industrial equipment communication standard, supports complex data models and security mechanisms; the HTTP protocol is suitable for simple integration scenarios and supports synchronous request / response mode. The system can flexibly select the most suitable communication protocol based on the network environment and business needs to ensure the reliability and real-time performance of data transmission.

[0037] The bee executor includes a polling scheduling center, a device controller and a data processing module.

[0038] The polling dispatch center manages the concurrent collection tasks of multiple devices according to preset time slices and priorities. The dispatch center uses a time wheel algorithm to achieve efficient task scheduling, and divides the collection tasks into three levels: high, medium, and low according to priority. The collection frequency of high-priority tasks (such as key equipment status and safety monitoring points) can reach 100ms / time, the collection frequency of medium-priority tasks (such as process parameters and production counts) is 1s / time, and the collection frequency of low-priority tasks (such as energy consumption data and environmental parameters) is 10s / time or longer. The dispatch center maintains a task queue and allocates computing resources according to task priority and time slice to ensure that the collection of critical data is prioritized under limited resource conditions. At the same time, the dispatch center implements an adaptive adjustment mechanism that can dynamically adjust the collection frequency according to system load and network conditions to avoid system overload.

[0039] The device controller sends data requests to the PLC device and receives response data into a cache queue. The controller supports multiple PLC communication drivers, including the Siemens S7 protocol, Mitsubishi MC protocol, and Omron FINS protocol, and can directly read and write PLC internal registers, data blocks, and I / O points. The controller uses an asynchronous I / O model, and a single controller instance can manage concurrent communications for up to 32 PLC devices, effectively improving system throughput. To cope with the complex network environment of industrial sites, the controller implements reliability mechanisms such as timeout retry and disconnection reconnection to ensure the continuity of data collection. The received device response data will be placed in a multi-level cache queue and stored according to data type and importance to provide data support for subsequent processing.

[0040] The data processing module performs rule verification and forwarding control on cached data. The processing module has a built-in rule engine that supports the configuration of multiple data verification rules, including value range checks, rate of change detection, data type verification, etc. For abnormal data that does not comply with the rules, the system will mark or filter it to ensure data quality. The processing module also supports data conversion functions, which can convert the original collected values ​​into engineering unit values ​​through linear transformation, table lookup conversion, etc. In terms of data forwarding, the processing module implements an intelligent distribution strategy, which can determine the forwarding time and forwarding target of data based on data importance, change range and system load, so as to avoid invalid data transmission occupying network bandwidth.

[0041] The Bee Executor adopts a multi-level cache mechanism, including a first-level memory ring cache, a second-level persistent cache, and a cache elimination module.

[0042] The first-level memory ring buffer pre-allocates a fixed-size continuous memory space, and uses head and tail pointers to cyclically overwrite and write real-time data streams. This cache adopts a lock-free ring buffer design, pre-allocates 16MB of continuous memory space, and can store approximately 1 million device data points. The buffer uses a head pointer and a tail pointer to identify the write position and the read position respectively. When the buffer is full, the new data will overwrite the earliest written data, forming a circular write mode. This design avoids frequent memory allocation and release operations, greatly improving the performance of high-frequency data processing. The ring buffer is mainly used to store data with high real-time requirements, such as equipment status, process parameters, etc., to ensure that this data can be processed and forwarded with minimal latency.

[0043] The second-level persistent cache stores data in non-volatile storage media to ensure data reliability and facilitate load balancing. This cache is implemented using an embedded database or structured file system, and data is written to a local SSD or industrial-grade flash memory. In the event of a network interruption or excessive load on the upper-layer system, newly collected data is temporarily stored in the persistent cache and uploaded in batches after the network is restored or the system load is reduced, thus achieving load balancing. Furthermore, the system regularly migrates important historical data from the memory cache to the persistent cache to prevent data loss caused by unexpected system restarts.

[0044] The cache elimination module eliminates invalid data based on a time sliding window or least recently used strategy. The module implements a variety of cache elimination algorithms and can select appropriate elimination strategies based on different data types. For real-time data with strong time series, a time-based sliding window strategy is adopted to only retain data from the most recent period (such as the last 24 hours); for configuration data and parameter data with irregular access patterns, an LRU (least recently used) strategy is adopted to prioritize the elimination of data that has not been accessed for a long time. The cache elimination module will regularly perform cleanup tasks to free up storage space and compress and archive data to improve storage efficiency. At the same time, the module also implements a data importance assessment mechanism to ensure that critical data is not eliminated prematurely.

[0045] The MEScada management platform includes instruction designer, view designer and report designer.

[0046] The instruction designer arranges the device control logic by visually dragging and dropping node modules, which include JavaScript script nodes, function blocks, and business logic nodes. The instruction designer uses a flowchart-style visual programming interface, and users can connect various functional nodes into a control process by dragging and dropping. The system provides a rich variety of node types: JavaScript script nodes support the writing of custom logic, and the built-in code editor provides syntax highlighting and automatic completion functions; function block nodes encapsulate commonly used mathematical operations, logical judgments, data conversion and other functions; business logic nodes provide special functions for specific production scenarios, such as work order initiation, equipment control, alarm processing, etc. The data flow and execution order are defined by connections between nodes, supporting complex processes such as conditional branching, loops, and parallel execution. The designed control logic can be saved as a template, supporting version management and permission control, and facilitating reuse and maintenance.

[0047] The view designer generates real-time production monitoring screens and work order progress visualization interfaces. The designer provides a drag-and-drop component library containing various display controls (such as dashboards, trend charts, status lights) and interactive controls (such as buttons, sliders, and input boxes). Users can freely combine these controls according to actual needs to build monitoring screens that conform to the actual layout of the factory. The designer supports multi-level screen design and can drill down from the global view of the factory to specific equipment details. The view component supports data binding and can directly associate real-time data points to automatically update the interface when the data changes. At the same time, the designer also provides dynamic effect configurations, such as color changes, size scaling, position movement, etc., to make the monitoring screen more intuitive. The designed monitoring interface supports responsive layout and can be adaptively displayed on display devices of different sizes.

[0048] The report designer generates statistical analysis reports on equipment OEE, quality pass rate and output. The designer provides a wealth of report templates, including statistical reports for different time dimensions such as daily, weekly, and monthly reports, as well as professional reports such as equipment efficiency analysis, quality traceability, and energy consumption analysis. Users can customize the report layout by adding elements such as tables, charts, and pivot tables by dragging and dropping. The report designer has a built-in data query builder that supports visually defining data filtering conditions, aggregation methods, and sorting rules without writing complex SQL statements. The system supports a variety of chart types, including bar charts, line charts, pie charts, radar charts, etc., which can intuitively display various production indicators. The designed reports can be exported to Excel, PDF and other formats, or they can be set to automatically generate at a scheduled time and sent to relevant personnel via email. The report data supports drill-down functions, and users can further view detailed records from the summary data to facilitate problem analysis and decision support.

[0049] The scheduling logic configured by the instruction designer responds to production scheduling instructions and, as a trigger, automatically issues scheduling instructions to devices (AVGs, collaborative robots, and production equipment). These instructions contain parameters such as the target device address and device control values. The system uses IoT sensors to monitor material inventory levels in real time at line warehouses. When a material's inventory level falls below a preset threshold (default: 20% of safety stock), the material replenishment process is triggered. The instruction designer automatically generates AGV scheduling instructions based on preconfigured scheduling rules. These instructions contain complete task information: the target warehouse address specifies the material pickup location, including the warehouse number, shelf area, and specific location coordinates; the outbound material details list the type, specification, and quantity of the material to be replenished, supporting batch delivery of multiple materials; and the delivery route calculates the optimal route based on real-time factory conditions, avoiding congested areas and temporary obstacles. The generated scheduling instructions are directly transmitted to the AGV control system through a system interface. Upon receiving the instructions, the AGV automatically executes material pickup, transport, and unloading operations, ensuring timely material replenishment and continuous production line operation.

[0050] The physical model of a factory is a hierarchical structure, consisting of four levels of association: factory, workshop, production line, and workstation, with each level associated with corresponding production equipment, production processes, and material formula data. The factory level is the top-level structure, containing attributes such as basic enterprise information, geographic location, and overall layout, and is associated with shared resources such as the energy management system and warehousing and logistics system at the factory level. The workshop level is the factory's main production unit, containing attributes such as workshop type, area, and person in charge, and is associated with workshop-level equipment resources, personnel scheduling, maintenance plans, and other information. The production line level represents a specific production line, containing attributes such as production line number, type, and production capacity, and is associated with production line-level process routes, quality control points, equipment groups, and other data. The workstation level is the most basic production unit, containing attributes such as workstation number, job type, and operator, and is associated with specific production equipment, fixtures, work instructions, and other resources.

[0051] Each level is associated with the corresponding production equipment, recording the equipment's basic information, technical parameters, maintenance records, and real-time status. At the same time, each level is also associated with the corresponding production process data, including process routes, process parameters, quality standards, and testing methods. Furthermore, material recipe data is also managed hierarchically, recording information such as the raw materials, auxiliary materials, and packaging materials required for each product, as well as their usage, substitution relationships, and inventory strategies. This hierarchical model structure enables the system to manage and monitor the production process at different granularities, enabling both a grasp of the overall production status and an in-depth understanding of the details of specific workstations.

[0052] Example 2

[0053] An intelligent production monitoring and scheduling method based on dual-informatization fusion technology includes the following steps:

[0054] like Figure 3 As shown, Figure 3 This is a schematic diagram of the collection and polling principle of an embodiment of the present application. The bee executor configures the equipment collection points and alarm conditions, and periodically polls the PLC register data. First, in the configuration interface of the bee executor, the technician configures the data points that need to be collected according to the characteristics of the production equipment and monitoring requirements. The configuration content includes the point address, data type, collection cycle and data description. At the same time, the alarm conditions are configured for key parameters, including upper and lower limit alarms, rate of change alarms and status alarms. After the configuration is completed, the bee executor sends a data read request to the PLC through the industrial Ethernet or fieldbus network according to the preset collection cycle to obtain real-time data of the target register. In order to improve the collection efficiency, the system adopts a batch reading strategy to merge multiple points with consecutive addresses into one request to reduce communication overhead. For key data points, the system implements a redundant collection mechanism to obtain data through multiple paths to ensure data reliability.

[0055] like Figure 4As shown, Figure 4 This is a schematic diagram of the cache design principle of an embodiment of the present application. A multi-level cache mechanism is used to process collected data and forward it to the MEScada management platform through a protocol conversion module. The collected raw data first enters a ring memory cache area, which adopts a lock-free design and supports high-concurrency write and read operations. For important data that needs to be persisted, the system writes it to a persistent cache area stored locally to prevent data loss due to system restart or power outage. The cached data undergoes preliminary processing by the data processing module, including unit conversion, linear calibration, outlier filtering, and other operations to ensure data quality. The processed data is converted to MQTT, OPCUA, or HTTP protocol format according to the configured communication method through the protocol conversion module and sent to the MEScada management platform. The protocol conversion module dynamically selects the most suitable transmission strategy based on network conditions and data importance: for control data with high real-time requirements, low-latency communication methods are preferred; for large quantities of historical data, batch compression transmission is used to improve transmission efficiency. When a network anomaly is detected, the system automatically activates the local caching mechanism and uploads the data in batches after the network is restored to ensure data integrity.

[0056] Build a factory physical model, process routing, and work order task sets on the MEScada management platform. Using the MEScada platform's modeling tools, administrators construct a four-level hierarchical factory physical model, encompassing the factory, workshop, production line, and workstation levels. At each level, configure the corresponding production equipment, personnel resources, and auxiliary facilities information. Next, use the process editor to define the product's production process routing, including information such as process sequence, process parameters, quality inspection points, and material requirements. The process routing supports conditional branching and parallel processes, adapting to complex production scenarios. Then, based on production plans and customer orders, create work order task sets to clearly define production targets, quantities, delivery dates, and priorities. Based on the process routing and resource availability, the system automatically calculates the earliest start time, latest completion time, and critical path for a work order, assisting in production scheduling decisions. Work order task sets support hierarchical management, allowing large orders to be split into multiple sub-work orders, which can be assigned to different production lines or workstations for parallel production, improving production efficiency.

[0057] The instruction designer orchestrates device interaction logic and issues work orders to the edge for execution. Using the instruction designer's visual interface, technicians drag and drop various functional nodes to orchestrate device control and interaction logic. This control logic includes functions such as device start / stop control, parameter adjustment, status monitoring, and exception handling. After system verification, the orchestrated control logic is associated with the work order task to form a complete execution plan. When the work order is scheduled for execution, the system issues the corresponding control instructions to the edge Bee executor. Upon receiving the instructions, the executor, based on pre-set execution conditions and timing requirements, sends control commands to the production equipment, driving it to operate according to process requirements. Simultaneously, the executor continuously monitors the equipment's operating status and process parameters to ensure that the production process meets expectations. If an anomaly is detected, the executor automatically implements emergency measures or alerts higher-level systems based on pre-set handling strategies to ensure production safety.

[0058] The system monitors the execution status of work orders based on real-time data and dynamically schedules AGVs and robotic arms for coordinated operations. Using real-time data collected by bee actuators, the system continuously monitors the execution status of work orders, including information such as equipment operating conditions, production progress, quality indicators, and material consumption. Based on this real-time data, the system assesses deviations from the current production plan and dynamically adjusts production resource allocation according to pre-set scheduling rules. When the system detects that material inventory in the lineside warehouse falls below a threshold, it automatically triggers an AGV dispatch instruction, assigning the appropriate AGV to the central warehouse to retrieve materials and deliver them to the workstations that need replenishment. Simultaneously, the system coordinates the robotic arms' work schedules to ensure that materials are immediately loaded upon arrival, minimizing production interruption. In scenarios with multiple AGVs and robotic arms, the system uses optimization algorithms to calculate the optimal task allocation plan, avoiding resource conflicts and path interference, thereby improving logistics efficiency. Furthermore, the system dynamically adjusts the task ordering of AGVs and robotic arms based on changing production priorities, ensuring that critical work orders are prioritized.

[0059] The multi-level caching mechanism specifically includes writing high-frequency real-time data into a circular memory cache area and adopting a circular overwrite strategy to ensure timeliness; writing data that needs to be persisted into a disk buffer to prevent loss during system power outages; and applying an elimination strategy based on the frequency of data changes to release storage space.

[0060] For high-frequency real-time data, such as equipment status signals and key process parameters, the system writes it to a circular memory buffer. This buffer pre-allocates contiguous memory blocks of a fixed size and manages the data write and read locations using head and tail pointers. When new data arrives, the system writes it to the current tail pointer position and then updates the tail pointer. When the buffer is full, the new data overwrites the oldest data, forming a circular overwrite pattern.

[0061] For data that needs to be persisted, such as production records, quality data, and energy consumption data, the system writes it to the disk buffer. The disk buffer adopts a log structure design, and new data is written to the end of the file in an appended manner to avoid performance degradation caused by random writes. The system adopts a batch write strategy, merging multiple data in a short period of time into a single write operation to reduce the number of I / O operations. To prevent data loss caused by system power outages, the buffer implements a transaction log mechanism to ensure data integrity and consistency. The capacity of the disk buffer can be configured according to the storage capacity of the device, and the default is 20% of the total storage space. When the network is interrupted or the upper-level system is unavailable, the newly collected data will be continuously written to the disk buffer and uploaded in batches after the conditions are restored, realizing reliable data transmission.

[0062] The system applies an elimination strategy based on the frequency of data changes to free up storage space. For the memory cache, the system uses a time-based elimination strategy, retaining only data from the most recent period (configurable, default is 4 hours); for the disk buffer, the system implements a multi-level elimination strategy: First, data that has been successfully uploaded to the upper-level system will be marked as eligible for elimination; second, the system will analyze the access pattern and change frequency of the data, prioritizing the elimination of data with low change frequency and long periods of inaccessibility; finally, when the storage space approaches the threshold (default is 80%), the system will trigger an emergency elimination mechanism to eliminate low-importance historical data based on the data importance score. Before elimination, the system will compress and archive the data, saving summary information of the historical data with a higher compression ratio, balancing storage efficiency and data availability.

[0063] Dynamic scheduling involves automatically triggering AGV dispatch instructions and linking robotic arms to complete material loading operations when the system detects that material levels in the lineside warehouse are below a threshold. The system uses IoT sensors to monitor material inventory levels in real time. Each material has a set safety stock threshold, calculated based on its consumption rate, replenishment cycle, and criticality. When the system detects that a material's inventory level falls below a preset threshold, it triggers the material replenishment process.

[0064] The system first queries the central warehouse for material inventory to confirm the availability and storage location of the required materials. Then, based on the current AGV's location, load status, and task queue, it selects the most appropriate AGV to perform the delivery task. The system generates complete AGV scheduling instructions, including the material collection location (warehouse area, shelf number, cargo location coordinates), material information (material code, quantity, batch), and delivery destination (workshop number, production line number, workstation number). At the same time, the system calculates the optimal delivery route, taking into account the factory layout, current channel congestion, and the operating trajectories of other AGVs to avoid path conflicts.

[0065] After receiving the dispatch instruction, the AGV automatically navigates to the designated material collection location, confirms the target material through the built-in recognition system, and then completes the material collection operation. After the material collection is completed, the AGV goes to the target workstation along the planned path. At the same time, the system sends a pre-processing instruction to the robotic arm at the target workstation, and the robotic arm prepares to receive the material in advance. When the AGV arrives near the target workstation, the system coordinates the precise docking of the AGV and the robotic arm. The robotic arm selects the appropriate grasping method based on the material type and specifications, removes the material from the AGV, and accurately places it at the feed port or designated location of the production equipment. The entire process realizes the fully automated distribution of materials from the warehouse to the production line without the need for human intervention, greatly improving logistics efficiency and production continuity.

[0066] Before data forwarding, a data cleaning step is also included: the raw data is converted using a calculation formula configured at the edge, which includes linear scaling, unit conversion, or outlier filtering.

[0067] Data cleaning is a key step in ensuring system data quality. After the bee actuator collects the raw data, it goes through the following processing steps:

[0068] First, the system performs linear scaling on the raw data. Many industrial sensors output standard signals (such as 4-20mA current signals or 0-10V voltage signals), which need to be converted into actual physical quantities. The edge end is pre-configured with a linear transformation formula: Y = aX + b, where X is the original acquisition value, Y is the converted physical quantity, and a and b are the proportional coefficient and offset. For example, the temperature sensor outputs a 4-20mA signal corresponding to the 0-100°C temperature range, then a = (100-0) / (20-4) = 6.25, b = 0-6.25×4 = -25, and the conversion formula is Y = 6.25X-25; the system supports configuring different conversion parameters for different ranges and accuracy requirements to ensure data accuracy.

[0069] Secondly, the system performs unit conversion operations. Equipment and systems on industrial sites may use different units of measurement, which need to be uniformly converted to standard units. The edge has a built-in library of commonly used unit conversion formulas that supports unit conversion of physical quantities such as length, mass, temperature, pressure, and flow. For example, converting imperial units to metric units (inches to millimeters, pounds to kilograms, etc.), or interchange different temperature units (Celsius, Fahrenheit, Kelvin). Users can select the appropriate conversion formula through the configuration interface, or customize complex conversion relationships to meet special process requirements.

[0070] Finally, the system applies an outlier filtering mechanism to identify and process anomalies in the data. The edge supports a variety of anomaly detection methods: the range check method sets reasonable upper and lower limits for the data, and values ​​outside the range are marked as anomalies; the rate of change check method calculates the rate of change of adjacent data points to identify mutation points; the statistical method uses moving average and standard deviation to detect anomalies that deviate from the normal distribution. For detected outliers, the system can execute different processing strategies based on the configuration: directly discard, replace with the last valid value, use interpolation to estimate a reasonable value, or retain but add an anomaly mark. These processes ensure that the data transmitted to the upper-level system is of high quality and reliability, providing a solid foundation for subsequent analysis and decision-making.

[0071] It should be noted that both the first embodiment and the second embodiment are a type of intelligent production monitoring and scheduling system based on dual-informatization fusion technology.

[0072] In the several embodiments provided in this application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules or units is merely a logical functional division. In actual implementation, other division methods may be used, such as combining or integrating multiple units or components into another system, or ignoring or not implementing certain features.

[0073] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of this embodiment.

[0074] In addition, each functional unit in each embodiment of the present application may be integrated into a processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The above-mentioned integrated units may be implemented in the form of hardware or software functional units.

[0075] The above description is only an implementation method of the present application and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the description and drawings of this application, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. An intelligent production monitoring and scheduling system based on dual-industry fusion technology, characterized in that: include: MEScada management platform, used to build factory physical models, process models, work order scheduling logic, and visual monitoring interfaces; Device-side bee actuators are independently deployed at the edge of the device to collect device data and perform data cleaning, protocol conversion, and multi-level cache processing; Database, used to store factory model data, process parameters and real-time production data; The bee actuator communicates with the MEScada management platform through a standardized API and supports at least one of the MQTT, OPCUA, and HTTP protocols for data forwarding.

2. The intelligent production monitoring and scheduling system according to claim 1, characterized in that: The bee actuator comprises: Polling scheduling center to manage concurrent collection tasks of multiple devices according to preset time slices and priorities; The device controller sends data requests to the PLC device and receives response data into the buffer queue; The data processing module performs rule verification and forwarding control on cached data.

3. The intelligent production monitoring and scheduling system according to claim 2, characterized in that: The bee executor adopts a multi-level cache mechanism, including: The first-level memory ring buffer area pre-allocates a fixed-size continuous memory space and writes the real-time data stream through the head and tail pointer loop overwriting; The second-level persistent cache stores data in non-volatile storage media to ensure data reliability and smooth out peak traffic. The cache elimination module eliminates invalid data based on a time sliding window or least recently used strategy.

4. The intelligent production monitoring and scheduling system according to claim 1, characterized in that: The MEScada management platform includes: The instruction designer arranges device control logic by visually dragging and dropping node modules, including JavaScript script nodes, function blocks, and business logic nodes; View designer, generating real-time production monitoring screens and work order progress visualization interfaces; Report designer, generates equipment OEE, quality pass rate and output statistical analysis reports.

5. The intelligent production monitoring and scheduling system according to claim 4, characterized in that: The scheduling logic configured by the instruction designer includes: In response to production scheduling instructions, as a trigger condition, scheduling instructions are automatically issued to the device end (AVG\cobot\machine production equipment). The instructions contain parameters such as target device address and device control value.

6. The intelligent production monitoring and scheduling system according to claim 1, characterized in that: The factory physical model is a hierarchical structure, including four associated levels: factory, workshop, production line, and workstation, and each level is associated with corresponding production equipment, production process and material formula data.

7. An intelligent production monitoring and scheduling method based on dual-industry fusion technology, characterized in that: The following steps are involved: Configure the device to collect position and alarm conditions through the bee actuator, and periodically poll the PLC register data; A multi-level cache mechanism is used to process collected data and forward it to the MEScada management platform through a protocol conversion module; Build factory physical models, process routes, and work order task sets on the MEScada management platform; Use the instruction designer to orchestrate device interaction logic and issue work order tasks and recipes or scheduling instructions to the edge for execution; Monitor the execution status of work orders based on real-time data and dynamically schedule equipment, AGVs, and robotic arms for collaborative operations.

8. The intelligent production monitoring and scheduling method according to claim 7, characterized in that: The multi-level cache mechanism specifically includes: Write high-frequency real-time data into the memory ring buffer, and use a circular overwriting strategy to ensure timeliness; Write data that needs to be persisted into the disk buffer to prevent loss in the event of a system power outage; Apply an elimination strategy based on the frequency of data changes to free up storage space.

9. The intelligent production monitoring and scheduling method according to claim 7, characterized in that: The dynamic scheduling includes: When it is detected that the material level in the line-side warehouse is below the threshold, the AGV scheduling instruction is automatically triggered, and the robotic arm is linked to complete the loading operation.

10. The intelligent production monitoring and scheduling method according to claim 7, characterized in that: Before data forwarding, the following steps are also included: The raw data is transformed using edge-configured calculation formulas, including linear scaling, unit conversion, or outlier filtering.

Images