Smart Factory Equipment Data Acquisition System Based on Extended Access

Abstract

The application provides a kind of smart factory equipment data acquisition system based on extended access, belongs to data acquisition technical field.The system includes adapter, data acquisition module, data processing module, communication transmission network.Adapter connects factory equipment to adapt different interface and protocol;Data acquisition module connects adapter to collect operation data, and connects management system to obtain production shift data;Data processing module selects transmission and time sequence synchronization processing to operation data;Communication transmission network contains transmission link and relay node, relay node includes power supply, network access, frequency prediction and path optimization unit, can realize power supply monitoring, transmission mode selection, load level determination, high-risk node shunt and standby path generation, suitable for efficient collection and transmission of smart factory equipment data.The application realizes wide coverage, high reliability and efficient transmission storage of smart factory equipment data acquisition through modular design and dynamic optimization mechanism.

Description

Technical Field

[0001] This invention relates to the field of data acquisition technology, and more specifically to a smart factory equipment data acquisition system based on extended access. Background Technology

[0002] In the wave of digital transformation, smart factories have become an inevitable trend in the development of modern manufacturing. With the deep integration of information technology and manufacturing, factories have increasingly stringent requirements for the collection, transmission, and management of equipment data. Efficient and accurate data collection and transmission are the key foundation for optimizing production processes, real-time monitoring of equipment status, fault early warning, and intelligent decision-making, playing a decisive role in improving factory production efficiency, reducing costs, and enhancing competitiveness.

[0003] Existing smart factory data acquisition systems still have significant shortcomings. On the one hand, in the data transmission stage, traditional communication networks suffer from poor transmission stability when faced with the complex electromagnetic environment, large factory areas, and diverse equipment distribution, easily leading to signal interruptions and delays. Moreover, the massive amount of factory data results in a large amount of redundant data being transmitted and stored, wasting resources and reducing data processing efficiency. Summary of the Invention

[0004] This invention provides a smart factory equipment data acquisition system based on extended access, which solves the problems of low data transmission efficiency and poor transmission path reliability in the prior art.

[0005] To achieve the above objectives, embodiments of the present invention provide a smart factory equipment data acquisition system based on extended access, comprising: an adapter for connecting factory equipment to adapt to different interface types and communication protocols; a data acquisition module for connecting to the adapter to collect operational data of the factory equipment and connecting to the factory's management system to obtain production shift data; a data processing module for selectively transmitting and storing the collected operational data and performing time-series synchronization processing; and a communication transmission network including multiple transmission links and relay nodes. The transmission links include various transmission methods and are connected to the data acquisition module to provide a channel for data transmission. The relay nodes are arranged in the transmission links for forwarding data within the transmission links. The relay nodes include a power supply... The system comprises a power supply unit, a network access unit, a frequency prediction unit, and a path optimization unit. The power supply unit supplies power to the relay nodes and monitors their own power status. The network access unit selects the transmission mode based on the network performance parameters of each transmission mode. The frequency prediction unit predicts the usage frequency of each relay node within a preset time period by constructing a prediction model and determines the load level of each relay node. It also adjusts the operating status of the corresponding relay node according to a preset adjustment strategy. The path optimization unit performs data diversion for high-risk nodes using a preset data diversion strategy. It also generates corresponding backup transmission paths for relay nodes with power status below a first preset threshold, based on the power status and usage frequency of each relay node, to replace the initial transmission path.

[0006] Optionally, the selection of the transmission method includes: adjusting the basic weights of each dimension of the network performance parameters based on the production shift data; calculating the deviation value of each dimension of the network performance parameters from a preset benchmark value; converting the deviation value of each dimension of the data into a corresponding score through a preset conversion rule; obtaining a comprehensive score by weighted summation based on the score corresponding to each dimension of the data and the adjusted basic weights; and selecting the transmission method with the highest comprehensive score for data transmission.

[0007] Optionally, the construction of the prediction model includes: dividing the operation data into time periods based on the operation data processed by the data processing module and the production shift data; statistically analyzing the feature quantities of the operation data in each time period and comparing the feature quantities of the same time period in each production shift to obtain a comparison result; selecting the difference quantities that are related to the equipment operation status and exceed a preset range as features based on the comparison result; and constructing a prediction model using a simple regression algorithm based on the selected features.

[0008] Optionally, each load level is configured with a corresponding usage frequency range, and the frequency prediction unit determines the load level for each relay node based on the predicted usage frequency and the usage frequency range.

[0009] Optionally, the adjustment strategy includes: calculating the target communication power of the relay node based on its load level using a preset power adjustment function; and adjusting the communication power of the relay node based on the calculated target communication power.

[0010] Optionally, the relay node has a built-in sensor for collecting its own operating parameters. The traffic splitting strategy includes: for the relay node with the highest load level, calculating the corresponding load factor based on the relay node's operating parameters; when the load factor exceeds a second preset threshold, determining the relay node as a high-risk node; for the high-risk node, generating a traffic splitting path using a preset path generation algorithm; and based on the generated traffic splitting path, splitting the data currently transmitted by the high-risk node to the adjacent relay node with the lowest load level.

[0011] Optionally, the generation of the backup transmission path includes: for relay nodes whose power status is lower than a first preset threshold, constructing a path generation model based on the power status and usage frequency of each relay node, and generating multiple candidate paths through the path generation algorithm; changing the weights of power status and usage frequency according to the production shift data; calculating the path score of each candidate path based on the changed weights and a preset scoring function; and selecting the candidate path with the highest path score as the backup transmission path.

[0012] Optionally, the data acquisition module is further used to obtain historical operating data of factory equipment and the time difference between factory equipment through the factory management system. The selection of transmission and storage includes: Step S1: Based on the historical operating data of factory equipment, the correlation strength between different factory equipment operating parameters is identified through machine learning algorithms; Step S2: According to the correlation strength, the factory equipment is divided into multiple groups of related equipment; Step S3: For each group of related equipment, the operating data of each equipment is redundantly filtered; Step S4: Based on the filtered operating data, a mapping relationship model of the operating data in each group of related equipment is established, and the master equipment and slave equipment in each group of related equipment are determined. Each mapping relationship model contains multiple conversion formulas and feature indexes and mapping association identifiers of each equipment. The conversion formulas are used for mapping the operating data between the master equipment and the slave equipment; Step S5: For each master equipment and slave equipment, the difference data between their operating data is determined through a preset analysis algorithm; Step S6: The complete operating data of the master equipment and the difference data of the slave equipment are selected and transmitted through the communication transmission network; Step S7: The operating parameters of the factory equipment are stored through a preset storage strategy.

[0013] Optionally, the time synchronization includes: establishing a time sequence association model for the factory equipment based on the time difference between the equipment and a preset time sequence association rule base; calculating the time point of the state change of the corresponding associated equipment based on the time sequence association model when the state of the factory equipment changes; and generating a time synchronization signal based on the time point of the state change to adjust the acquisition frequency of the data acquisition module for the corresponding associated equipment.

[0014] Optionally, the storage strategy described in step S7 includes: for the operating data of the master device, only the feature index and mapping association identifier of the operating data are stored; for the complete operating data of the slave device and the difference data of the master device, the complete data is stored.

[0015] This invention provides a smart factory equipment data acquisition system based on extended access. Through multi-dimensional modular design and collaborative optimization, it significantly improves the compatibility, reliability, and efficiency of the smart factory equipment data acquisition system. Adapters enable interface and protocol adaptation for various types of equipment, solving the problem of heterogeneous equipment access. The data acquisition module, combined with production shift data, provides more comprehensive time-dimensional information for subsequent analysis. The selective transmission and time-series synchronization mechanism of the data processing module reduces redundant data transmission and ensures time consistency of data from associated devices. The communication transmission network, through dynamic adjustment of relay nodes, effectively copes with network fluctuations and node failures, achieving intelligent scheduling and dynamic redundancy of data transmission, thus improving the overall adaptability and stability of the system in complex industrial environments. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0017] Figure 1 This is a framework diagram of a smart factory equipment data acquisition system based on extended access provided in an embodiment of the present invention;

[0018] Figure 2 This is a data splitting flowchart provided in an embodiment of the present invention;

[0019] Figure 3 This is a flowchart of the alternative path generation process provided in an embodiment of the present invention;

[0020] Figure 4 This is a flowchart of the selection of transmission and storage provided in an embodiment of the present invention;

[0021] Figure 5 This is a timing synchronization flowchart provided in an embodiment of the present invention. Detailed Implementation

[0022] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0023] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0024] As mentioned above, with the expansion of smart factories and the surge in the number of connected devices, existing data acquisition systems suffer from uncontrollable energy consumption and insufficient load balancing of relay nodes. Traditional systems often employ fixed transmission methods and energy consumption patterns, making it difficult to dynamically adjust according to node load. This leads to some nodes becoming overloaded and paralyzed while others remain idle. Simultaneously, the stability and integrity of data transmission are also affected, hindering the improvement of factory production efficiency. Therefore, it is essential to develop a highly efficient and stable smart factory equipment data acquisition system.

[0025] To address this problem, this invention provides a smart factory equipment data acquisition system based on extended access. By acquiring equipment operation data and production shift data, it dynamically selects transmission methods, predicts relay node load and adjusts energy consumption, and uses path optimization to achieve load balancing and backup path generation. This effectively solves the problems of uncontrollable energy consumption and insufficient load balancing, improves the stability and integrity of data transmission, and ensures the efficient and orderly operation of smart factory production.

[0026] The following is combined with Figures 1-5 This invention is described in detail.

[0027] Figure 1 This is a framework diagram of a smart factory equipment data acquisition system based on extended access, provided in an embodiment of the present invention. Figure 1 As shown, this embodiment of the invention provides a smart factory equipment data acquisition system based on extended access, including: an adapter for connecting factory equipment to adapt to different interface types and communication protocols; a data acquisition module for connecting to the adapter to collect operating data of the factory equipment and connecting to the factory's management system to obtain production shift data; a data processing module for selectively transmitting and storing the collected operating data and performing time-series synchronization processing; and a communication transmission network including multiple transmission links and relay nodes. The transmission links connect to the data acquisition module and include multiple transmission methods to provide channels for data transmission. The relay nodes are arranged in the transmission links for forwarding data within the transmission links. The relay nodes include a power supply unit, The system comprises a network access unit, a frequency prediction unit, and a path optimization unit. The power supply unit supplies power to the relay nodes and monitors their own power status. The network access unit selects a transmission mode based on the network performance parameters of each transmission mode. The frequency prediction unit predicts the usage frequency of each relay node within a preset time period by constructing a prediction model and determines the load level of each relay node. It also adjusts the operating status of the corresponding relay node according to a preset adjustment strategy. The path optimization unit performs data diversion for high-risk nodes using a preset data diversion strategy. It also generates a corresponding backup transmission path for relay nodes whose power status is below a first preset threshold, based on the power status and usage frequency of each relay node, to replace the initial transmission path.

[0028] The adapter refers to the hardware interface that connects to factory equipment and collects operational data. It supports industrial communication protocols such as Modbus, Profinet, and OPCUA, and can convert raw data output from different factory equipment into a unified JSON format. The data acquisition module is the hardware device that connects the adapter to the factory management system, possessing multi-channel data acquisition capabilities. The data processing module consists of a processor and related software, capable of filtering operational data, determining the data to be transmitted, and performing time calibration on the operational data from different devices to ensure data synchronization over time, providing an accurate time-series basis for subsequent analysis.

[0029] In a communication transmission network, the transmission link refers to the physical channel used for data transmission. The communication transmission network contains multiple transmission links, and each transmission link includes various transmission methods, such as Bluetooth Mesh transmission and LoRaWAN transmission. These multiple transmission methods can be switched according to the actual scenario.

[0030] Among them, relay nodes refer to network devices deployed in transmission links to enhance signals and forward data. Their power supply units are battery-powered and can monitor the power status in real time, i.e., the remaining power percentage.

[0031] The smart factory equipment data acquisition system based on extended access provided in this invention achieves compatibility with multiple interfaces and protocols through adapters, allowing data from different types of factory equipment to be smoothly accessed. The data acquisition module, combined with production shift data, provides time-dimensional information that aligns with actual production for data processing and transmission. The data processing module selects transmission and performs time-series synchronization processing, reducing redundant data and ensuring data time consistency. Multiple transmission modes in the communication transmission network can be flexibly switched, relay nodes work collaboratively, the power supply unit ensures power stability and monitors power consumption, the network access unit selects the optimal transmission mode, the frequency prediction unit adjusts node operating status to achieve load balancing, and the path optimization unit performs data diversion and generates backup paths to ensure reliable transmission. Overall, the system can efficiently and stably complete the acquisition and transmission of smart factory equipment data, adapting to complex factory environments.

[0032] Preferably, the selection of the transmission method includes: adjusting the basic weights of each dimension of the network performance parameters based on the production shift data; calculating the deviation value of each dimension of the network performance parameters from a preset benchmark value; converting the deviation value of each dimension of the data into a corresponding score through a preset conversion rule; obtaining a comprehensive score by weighted summation based on the score corresponding to each dimension of the data and the adjusted basic weights; and selecting the transmission method with the highest comprehensive score for data transmission.

[0033] Network performance parameters can include signal strength and transmission delay, packet loss rate, bandwidth utilization, and node energy consumption. Production shift data refers to the time periods of different shifts in the factory production process, which can be obtained through the production scheduling interface of the factory management system. This data reflects the differences in actual network transmission performance requirements at different production stages, with production periods divided into peak and off-peak periods. During peak periods, due to the dense production tasks, the requirements for real-time and stable data transmission are higher; therefore, the weights of transmission delay and packet loss rate, which are related to transmission timeliness, are increased. During off-peak periods, with fewer production tasks, the focus is on reducing system energy consumption to save costs; therefore, the weight of node energy consumption is increased, while the weights of transmission delay and packet loss rate are decreased, and the weights of signal strength and bandwidth utilization remain unchanged. The preset conversion rules can be implemented using non-linear functions, the core of which is to convert the deviation values ​​of each dimension parameter into scores through an exponential mapping relationship. This function employs differentiated mapping logic for different types of network performance parameters: For signal strength, when the actual value is better than the benchmark, the score increases exponentially with the increase of the absolute value of the deviation; when the actual value is worse than the benchmark, the score decreases exponentially with the increase of the deviation. For parameters such as transmission delay, packet loss rate, bandwidth utilization, and node energy consumption, where lower values ​​indicate better performance, the same function form as signal strength is used but with an inverse mapping: the score increases exponentially when the actual value is better than the benchmark, and decreases exponentially when the actual value is worse than the benchmark.

[0034] The preferred embodiment of this invention proposes a transmission mode selection mechanism that, by combining production shift data with the basic weights of various dimensions of network performance parameters, allows the weight settings to better suit the production needs of different shifts. Based on the network performance parameters and benchmark values, a deviation value is calculated and converted into a score. This score is then combined with the adjusted weights and summed to obtain a comprehensive score. The transmission mode with the highest comprehensive score is selected based on this score. This ensures that the selected transmission mode has optimal network performance under the current production shift, effectively improving the efficiency and stability of data transmission, avoiding problems such as delays and packet loss caused by improper transmission mode selection, and better adapting to the production data transmission needs of the factory at different times.

[0035] For example, assuming there are two transmission methods: Bluetooth Mesh transmission and LoRaWAN transmission, at 10:00 AM, the factory's basic weights are set as follows: signal strength 20%, transmission latency 25%, packet loss rate 20%, bandwidth utilization 15%, and node power consumption 20%. The first preset threshold is 10 minutes, and the peak period is from 8:00 AM to 8:00 PM. According to the adjustment rules, the weight of transmission latency is increased to 37.5%, the weight of packet loss rate is increased to 26%, the weight of node power consumption is reduced to 10%, and the weights of signal strength and bandwidth utilization remain unchanged. After normalization, the weights of each dimension are: signal strength 18.5%, transmission latency 34.8%, packet loss rate 24.1%, bandwidth utilization 13.9%, and node power consumption 8.7%. At this time, the network performance parameters of LoRaWAN obtained through the performance monitoring subunit are as follows: signal strength 65dBm (baseline value 70dBm), transmission latency 40ms (baseline value 50ms), packet loss rate 0.8% (baseline value 1%), bandwidth utilization 25% (baseline value 30%), and node power consumption 18mA (baseline value 20mA); the parameters of Bluetooth Mesh are as follows: signal strength 72dBm, transmission latency 60ms, packet loss rate 1.2%, bandwidth utilization 35%, and node power consumption 15mA. According to the preset conversion rules, the scores for each dimension of LoRaWAN are as follows: signal strength 90 points, transmission latency 85 points, packet loss rate 88 points, bandwidth utilization 80 points, and node power consumption 85 points. The weighted sum of these scores is approximately (90×18.5%) + (85×34.8%) + (88×24.1%) + (80×13.9%) + (85×8.7%) ≈ 85.6 points. Similarly, the overall score for Bluetooth Mesh is approximately 66.2 points. Since the overall score of LoRaWAN is higher than that of Bluetooth Mesh, the network access unit selects LoRaWAN for transmission.

[0036] Preferably, the construction of the prediction model includes: dividing the operation data into time periods based on the operation data processed by the data processing module and the production shift data; statistically analyzing the feature quantities of the operation data in each time period and comparing the feature quantities of the same time period in each production shift to obtain a comparison result; selecting the difference quantities that are related to the equipment operation status and exceed a preset range as features based on the comparison result; and constructing a prediction model using a simple regression algorithm based on the selected features.

[0037] The time period is divided into task-intensive and task-sparse periods. During the task-intensive period, equipment operates frequently, and the data transmission volume of relay nodes is large; during the task-sparse period, equipment operates less frequently, and the data transmission volume of relay nodes is small. Specific features include the total number of bytes transmitted per unit time, the average interval between data frames, the CPU utilization rate of relay nodes, and memory utilization rate. The comparison results refer to the numerical differences of various features across different shifts within the same time period. For example, in the task-intensive period, shift A transmits 100MB of data per unit time, while shift B transmits 80MB, a difference of 20MB. The simple regression algorithm for constructing the predictive model refers to establishing a mathematical relationship between feature variables and the frequency of relay node usage using linear or multinomial regression methods. Specifically, this can be achieved by fitting parameters using the least squares method, transforming multidimensional data features into quantifiable predictive indicators.

[0038] In a preferred embodiment of the present invention, the operational data is first divided into task-intensive and sparse segments according to the production rhythm, based on production shift data. Features such as data transmission volume and node resource occupancy rate are statistically analyzed for each time period. By comparing the differences in features during the same time period across different shifts, features exceeding a preset range and related to equipment operating status are selected. Then, a model is constructed using a univariate or multivariate linear regression algorithm to predict the future usage frequency of relay nodes. This process, by combining actual production time segmentation and feature selection, ensures the relevance and effectiveness of the model input. The application of simple regression algorithms reduces computational complexity while maintaining prediction accuracy, enabling relay nodes to anticipate load levels and adjust their operating status accordingly, thereby achieving load balancing, avoiding data transmission congestion caused by node overload, and improving the stability and efficiency of the communication transmission network.

[0039] Preferably, each load level is configured with a corresponding usage frequency range, and the frequency prediction unit determines the corresponding load level for each relay node based on the predicted usage frequency and the usage frequency range.

[0040] More preferably, the adjustment strategy includes: calculating the target communication power of the relay node based on the load level of the relay node using a preset power adjustment function; and adjusting the communication power of the relay node based on the calculated target communication power.

[0041] Specifically, the load level is divided into three levels: light load, medium load, and heavy load. Light load corresponds to a frequency range of [0, 150] times / minute, medium load to [150, 350] times / minute, and heavy load to [350, +∞] times / minute. The frequency prediction unit compares the predicted usage frequency within a preset time period with the corresponding range. For example, if the predicted value is 145 times / minute, falling within the [0, 150] times / minute range, the relay node is determined to be light load. The power adjustment function can be expressed as:

[0042] (1)

[0043] in, Indicates the target communication power. Indicates basic communication power. Indicates the current load level (1~3). Indicates base load level 1. Indicates the load adjustment coefficient. Indicates the power sensitivity coefficient. Indicates battery status.

[0044] The preferred embodiment of this invention clarifies the process of determining the load level and the strategy for adjusting the power of relay nodes. On the one hand, by dividing the usage frequency of relay nodes into different intervals, the frequency prediction unit can accurately determine the load level based on the predicted usage frequency. On the other hand, the target communication power is calculated through a preset power adjustment function (1), and dynamic power adjustment is achieved by combining the load adjustment coefficient and the power sensitivity coefficient, and a step-by-step fine-tuning is used to ensure a smooth transition. The combination of the two means that, on the one hand, the accurate determination of the load level is achieved through the clear division of frequency intervals, providing a reliable basis for power adjustment; on the other hand, the dynamic power adjustment based on the load level effectively balances the energy consumption of relay nodes while ensuring the quality of data transmission, avoiding signal instability caused by power fluctuations, and significantly improving the stability, energy efficiency, and adaptability of relay node operation in the smart factory equipment data acquisition system.

[0045] For example, the frequency prediction unit predicts that the predicted usage frequency of a certain relay node is 380 times / minute. Based on the division of the frequency range, the relay node is determined to be overloaded. Then, using Equation (1), combined with the overload load level 3 and 80% of the remaining power, the target communication power is calculated to be 60% higher than the base value. The target value is gradually achieved through step-by-step fine-tuning of ±1dBm each time. This satisfies the signal strength requirements of a large amount of data transmission during peak periods, avoids signal fluctuations caused by a sudden increase in power, and ensures the stability and efficiency of data transmission because the power is sufficient and the power is not excessively limited.

[0046] like Figure 2As shown, preferably, the relay node has a built-in sensor for collecting its own operating parameters. The traffic splitting strategy includes: for the relay node with the highest load level, calculating the corresponding load factor based on the relay node's operating parameters; when the load factor exceeds a second preset threshold, determining the relay node as a high-risk node; for the high-risk node, generating a splitting path using a preset path generation algorithm; and according to the generated splitting path, diverting the data currently transmitted by the high-risk node to the adjacent relay node with the lowest load level.

[0047] The highest load level relay node refers to a heavily loaded relay node, while the lowest load level adjacent relay node refers to a relay node directly connected to the current node with a light load level. Operating parameters include: relay node CPU utilization, relay node queue length for data transmission, and relay node's maximum data cache size. The path generation algorithm can use Dijkstra's algorithm, which has the advantage of quickly finding the shortest path from the starting point to all other nodes in a weighted graph. It prioritizes nodes with low transmission latency and low packet loss rate for path construction, and simultaneously determines the traffic splitting ratio based on the load factor difference; for example, for every 10% increase in the difference, the splitting ratio increases by 20%, smoothly guiding excess data traffic to other relay nodes.

[0048] The load factor can be calculated using the following formula:

[0049] (2)

[0050] in, This indicates the CPU utilization of the relay node. This indicates the length of the queue for transmitting data at the relay node. This indicates the maximum data cache limit for relay nodes. This represents the weighting coefficient.

[0051] The preferred embodiment of this invention constructs a two-level decision-making mechanism for dynamic load balancing: when a heavily loaded relay node appears, its load factor is calculated according to formula (2). This formula innovatively integrates two-dimensional indicators of computing resource occupancy and data backlog, breaking through the limitations of the traditional single CPU criterion. Then, the load factor is compared with a second preset threshold to accurately determine high-risk nodes. Once a high-risk node is found, the system immediately triggers the calculation of the optimal load balancing path based on the path generation algorithm, migrating the data to the lightly loaded node in milliseconds. Through the above technical solution, the data transmission interruption problem caused by high-load relay nodes is effectively solved. Through dynamic load balancing and path optimization, the load pressure of key nodes is reduced, while avoiding adjacent nodes from entering an overload state due to sudden traffic, thereby improving the overall fault tolerance and data transmission efficiency of the system.

[0052] For example, relay node A is identified as a heavily loaded node, with its CPU utilization at 70%, a queue length of 800 data entries to be transmitted, a node data cache limit of 1000 entries, and a weighting coefficient of 0.5. Using formula (1), the load factor is calculated as 0.7 × 0.5 + (800 / 1000) × (1 - 0.5) = 0.35 + 0.4 = 0.75. Assuming the second preset threshold is 0.8, node A does not currently reach a high-risk level. Subsequently, node A's load increases, and node A is monitored again. =0.85, =900, the calculated load factor is 0.875, exceeding the threshold, and it is identified as a high-risk overload node. At this time, the path optimization unit uses the path generation algorithm to select relay node B, whose transmission delay is 15% lower than the original path, as a backup path. Since the load factor difference is 0.075, 15% of the traffic is diverted to node B according to the rules, reducing the load factor of node A to 0.74, ensuring that data transmission is not blocked, and demonstrating the efficient adjustment role of this mechanism in high-load scenarios.

[0053] like Figure 3 As shown, preferably, the generation of the backup transmission path includes: for relay nodes whose power status is lower than a first preset threshold, constructing a path generation model based on the power status and usage frequency of each relay node, with the node's power status as the basis and usage frequency as the constraint, and generating multiple candidate paths through the path generation algorithm; changing the weights of power status and usage frequency according to the production shift data; calculating the path score of each candidate path based on the changed weights and combined with a preset scoring function; and selecting the candidate path with the highest path score as the backup transmission path.

[0054] The relay node's power status represents the percentage of its remaining power, used to determine if the node is in a low-power state. Backup transmission paths are generated based on a real-time network topology graph, using Dijkstra's algorithm to calculate the shortest path from the source node to the target node, while avoiding marked high-risk nodes.

[0055] The scoring function can be expressed as:

[0056] (3)

[0057] in, This represents the average percentage of battery status across all relay nodes along the path. This indicates the maximum predicted usage frequency of relay nodes along the path. This represents a dynamic weighting coefficient, determined by the production shift, during peak hours (8:00 to 20:00). It is 0.8 during off-peak periods. It is 0.3.

[0058] The preferred embodiment of this invention constructs a complete backup path generation mechanism, which fundamentally avoids the risk of data interruption caused by relay node failure by generating backup paths during operation: when the power status of a relay node is lower than a first preset threshold, a path generation model is first constructed based on the power status and usage frequency of each relay node, with the node power status as the basis and the usage frequency as the constraint condition. Then, multiple candidate paths are generated through a path generation algorithm. Subsequently, the weights of power status and usage frequency are dynamically adjusted in combination with production shift data (e.g., usage frequency is emphasized during peak periods and power status is emphasized during off-peak periods), and the candidate paths are evaluated according to the scoring function (3). Finally, the path with the highest score is selected as the backup transmission path. This technical solution addresses the specific scenario of low power in relay nodes. By incorporating power status and usage frequency into the path generation logic, it ensures that the backup path avoids low-power relay nodes and adapts to the needs of different production periods through dynamic weights, so that the generated backup path achieves optimal reliability and adaptability. This effectively reduces the risk of data transmission interruption caused by relay node power failure and improves the fault tolerance and operational stability of the smart factory equipment data acquisition system.

[0059] For example, a communication transmission network contains four relay nodes. At 10 PM in a smart factory, the battery level of one relay node is detected to be 18%, lower than a preset value (e.g., 20%), classifying it as a low-battery node. At this point, the path optimization unit activates a backup path generation mechanism: First, a path generation model is constructed based on the node's battery level and usage frequency (constraint: usage frequency ≤ 60 times / minute). Then, a path generation algorithm generates three candidate paths from the remaining four nodes: Path 1 includes node A (80% battery level, 50 times / minute usage) and node B (75% battery level, 45 times / minute usage); Path 2 includes node D (60% battery level, 55 times / minute usage) and node E (70% battery level, 58 times / minute usage); Path 3 includes nodes A and E. The current time period is the night shift. The comprehensive score of the three candidate paths is calculated according to formula (3) as 0.3. The comprehensive score of path 1 is 0.9325, the comprehensive score of path 2 is 0.783, and the comprehensive score of path 3 is 0.813. The system finally selects path 1 with the highest comprehensive score as the backup transmission path and uses this path to replace the initial path to continue data transmission. This path meets the usage frequency constraint and has better comprehensive performance in terms of power status and usage frequency, thus ensuring the stability of data transmission in low-power scenarios at night.

[0060] like Figure 4As shown, preferably, the data acquisition module is further used to obtain historical operating data of factory equipment and the time difference between factory equipment through the factory management system. The selection of transmission and storage includes: Step S1: Based on the historical operating data of factory equipment, the correlation strength between different factory equipment operating parameters is identified through machine learning algorithms; Step S2: According to the correlation strength, the factory equipment is divided into multiple groups of related equipment; Step S3: For each group of related equipment, the operating data of each equipment is redundantly filtered; Step S4: Based on the filtered operating data, a mapping relationship model of the operating data in each group of related equipment is established, and the master equipment and slave equipment in each group of related equipment are determined. Each mapping relationship model contains multiple conversion formulas and the feature index and mapping association identifier of each equipment. The conversion formulas are used for mapping the operating data between the master equipment and the slave equipment; Step S5: For each master equipment and slave equipment, the difference data between their operating data is determined through a preset analysis algorithm; Step S6: Select the complete operating data of the master equipment and the difference data of the slave equipment, and transmit them through the communication transmission network; Step S7: Store the operating parameters of the factory equipment through a preset storage strategy.

[0061] The time difference in equipment flow between factories refers to the time interval taken for a product to flow from one piece of equipment to the next related equipment. For example, a 5-minute time difference for a part to flow from processing equipment A to assembly equipment B reflects the production coordination rhythm between equipment. Correlation strength measures the degree of mutual influence between the operating parameters of two or more pieces of equipment. It can be represented by a value between 0 and 1; the closer the value is to 1, the stronger the correlation between the parameters. For example, a temperature increase in equipment A and an increase in energy consumption in equipment B may have a high correlation strength. Machine learning algorithms can use association rule algorithms (such as the Apriori algorithm) or correlation analysis algorithms (such as the Pearson correlation coefficient) to calculate and output the correlation strength between different equipment operating parameters by training on historical operating data. Redundancy filtering removes duplicate, invalid, or highly repetitive information from equipment operating data, or information that is highly repetitive with parameters of other equipment in the same group. For example, ambient temperature data collected simultaneously by equipment A and equipment B may be retained only once to reduce the amount of data. Conversion formulas are mathematical expressions or algorithmic rules used to describe the quantitative relationship between the operating parameters of master and slave equipment, and they support bidirectional calculation. For example, the conversion formula "T=0.8P+25" between the heating power (P) of the main equipment and the temperature (T) of the slave equipment can be transformed into "P=(T-25) / 0.8". This allows the slave equipment temperature to be calculated from the main equipment power, and vice versa, achieving a two-way mapping. Feature indexes are key information tags extracted from operational data that uniquely identify the relationship between the data and specific parameters of the corresponding equipment, covering two-way associations. For example, the vibration frequency data of the slave equipment can be extracted as "high-frequency vibration - main equipment speed 1500 r / min", and the speed data of the main equipment can be extracted as "1500 r / min - slave equipment high-frequency vibration". This index allows for quick location of relevant parameters and conversion formulas for two-way mapping. Mapping association identifiers are unique codes or symbols assigned to each pair of master and slave equipment two-way mapping relationships, clearly distinguishing two-way mapping entries between different equipment. This identifier allows for precise retrieval of the corresponding two-way conversion formulas and feature indexes. Analysis algorithms can employ feature selection algorithms (such as the ReliefF algorithm) to filter out discrepancies in the operational data of the master and slave equipment. Discrepancy data refers to equipment operation data that cannot be fully derived from the bidirectional conversion formula of the mapping relationship model.

[0062] Specifically, the process for selecting transmission and storage is as follows: First, based on the historical operating data of the factory equipment, machine learning algorithms are used to identify the correlation strength of different equipment operating parameters, and the equipment is divided into multiple groups of related equipment. After redundancy filtering of the operating data of each group of related equipment, a mapping relationship model including conversion formulas, feature indexes, and mapping association identifiers is established, and the master and slave equipment within the group are determined. Then, a preset analysis algorithm is used to extract the difference data between the master and slave equipment that cannot be derived from the mapping model. Subsequently, the complete operating data of the master equipment and the difference data of the slave equipment are selected for transmission through the communication transmission network. Finally, the remaining data is stored according to the storage strategy.

[0063] The preferred embodiments of this invention propose a selection of transmission and storage methods. By using machine learning algorithms to identify the correlation of equipment operating parameters and group related equipment, precise equipment grouping is achieved. Redundancy filtering of equipment data within a group reduces unnecessary data transmission and lowers the load on the communication network. A mapping relationship model is established to clarify association rules and transformation logic, providing a reliable basis for determining master and slave devices. Selecting the transmission of complete master device data and differential data from slave devices ensures data validity and integrity while further simplifying the amount of transmitted data and improving data transmission efficiency. Simultaneously, the extraction of differential data ensures that core status information of slave devices is not lost. Overall, this makes the smart factory equipment data acquisition system more efficient and accurate in the data transmission process, better adapting to the complex equipment data management needs of factories.

[0064] like Figure 5 As shown, preferably, the time synchronization includes: establishing a time sequence association model for the factory equipment based on the time difference between the equipment and a preset time sequence association rule library; calculating the time point of the state change of the corresponding associated equipment based on the time sequence association model when the state of the factory equipment changes; and generating a time synchronization signal based on the time point of the state change to adjust the acquisition frequency of the data acquisition module for the corresponding associated equipment.

[0065] The temporal association rule base is a database storing time-related rules formed between devices due to material flow and process connections. For example, rules like "After device A completes processing, it flows to device B after 2 minutes, and device B immediately starts running" provide the foundation for establishing a temporal association model. The runtime temporal association model, built upon the flow time difference and the temporal association rule base, describes the temporal logical relationship of state changes between devices. It shows when related devices will undergo corresponding state changes after a certain device's state changes. Device state changes include changes in state such as from running to stopping, from low-speed operation to high-speed operation, and malfunctions. The state change time point is calculated by the temporal association model, referring to the specific time when a related device's state changes due to a change in the state of another device, accurate to the second. For example, after device A's state changes, device B will undergo a state change 30 seconds later.

[0066] In a preferred embodiment of the present invention, a runtime sequence association model is established based on the time difference between equipment flow and a time sequence association rule base. This model can accurately capture the temporal logical relationship between changes in equipment state and changes in the state of associated equipment. When the equipment state changes, the model calculates the time point of the state change of associated equipment and generates a time synchronization signal to adjust the data acquisition frequency. This ensures that key data can be accurately collected before and after changes in the state of associated equipment, avoiding data omissions or redundancy caused by time asynchrony. Furthermore, by dynamically adjusting the acquisition frequency, the model improves the targeting and efficiency of data acquisition and reduces unnecessary resource consumption. This enhances the data synchronization accuracy and overall operational efficiency of the smart factory equipment data acquisition system in equipment collaborative operation scenarios.

[0067] For example, on an electronic component production line in a smart factory, the time difference between the placement equipment and the inspection equipment is 15 seconds. The timing association rule base records that "after the placement equipment completes the component placement, the inspection equipment will start optical inspection 15 seconds later." Based on this, the timing association model can clearly reflect the linkage between the two states. When the placement equipment switches from "running" to "waiting for material," the timing association model calculates that the inspection equipment will enter the "inspection" state 15 seconds later. It then generates a time synchronization signal, increasing the data acquisition frequency of the data acquisition module from once per 30 seconds to once per second. This accurately captures key parameters such as the light source intensity and lens focal length when the inspection equipment starts up. This avoids the loss of initial inspection data due to time asynchrony and prevents high-frequency acquisition during non-inspection periods from consuming system resources. It effectively improves the data acquisition quality and resource utilization efficiency when equipment collaborates in the smart factory production line.

[0068] Preferably, the storage strategy described in step S7 includes: for the operating data of the master device, only the feature index and mapping association identifier of the operating data are stored; for the complete operating data of the slave device and the difference data of the master device, the complete data is stored.

[0069] Specifically, this storage strategy is based on the mapping relationship between master and slave device data: all operational data of the master device only stores feature indexes (key tags for quickly associating mapping relationships) and mapping association identifiers (unique codes to distinguish different mapping entries), without storing the full data; the complete operational data of the slave device (all unfiltered raw parameters and auxiliary information) and the differential data of the master device (parameters that cannot be derived from the slave device data through the mapping relationship model) must be stored in their entirety to ensure that core information is not lost. Furthermore, this storage strategy corresponds inversely to the transmission logic: during transmission, the complete data of the master device is the core, supplemented by key feature data of the slave device; while during storage, only the feature indexes and mapping association identifiers of the master device data are retained, and the complete operational data of the slave device and the differential data of the master device are stored. This reverse design makes the core data transmitted and the baseline data stored complementary. The transmitted data focuses on real-time analysis efficiency, while the stored data focuses on the integrity of the mapping baseline, providing bidirectional support for data verification and subsequent retrieval.

[0070] The preferred embodiment of this invention provides a storage strategy that significantly reduces the amount of stored data and lowers storage resource consumption by storing only the feature index and identifier of the master device data. It stores complete slave device data as a mapping benchmark and master device difference data as irreplaceable core information. This ensures that data can be deduced and reconstructed through the mapping relationship model while guaranteeing the comprehensiveness of device status analysis, achieving a balance between storage efficiency and data integrity. The reverse design of transmitted and stored data allows for rapid verification of the accuracy of the data mapping relationship by cross-referencing the transmitted complete master device data with the stored complete slave device data and master device difference data. Simultaneously, storing only the master device index while fully preserving the slave device data reduces storage redundancy of duplicate master device data and uses the complete slave device data as a stable benchmark, facilitating subsequent reconstruction of the master device's historical operating status based on the stored data, thus improving data verification efficiency and optimizing storage resource allocation.

[0071] The smart factory equipment data acquisition system based on extended access provided by this invention achieves unified access for heterogeneous devices by adapting to multiple types of interfaces and protocols through adapters. The data acquisition module collects operational data in conjunction with production shift data, providing time-dimensional production background information for subsequent analysis. The data processing module optimizes data transmission based on the correlation strength of devices by selective transmission and adjusts the acquisition frequency through time synchronization, improving the targeting and timeliness of data acquisition. The communication transmission network dynamically adjusts the load level and communication power through the frequency prediction unit of relay nodes, and combines the path optimization unit to divert high-risk nodes and generate backup paths, ensuring the stability of data transmission and optimizing network resource allocation. The data storage module adopts a differentiated storage strategy to reduce data redundancy and lower storage costs. The collaborative work of each module significantly improves the system in terms of equipment compatibility, data acquisition efficiency, transmission stability, and storage optimization, enhancing the comprehensiveness, accuracy, and reliability of smart factory equipment data acquisition and effectively meeting the needs of intelligent manufacturing in complex industrial environments.

[0072] The above description is merely a preferred embodiment of the technical solution of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A smart factory equipment data acquisition system based on extended access, characterized in that, The application relates to a data transmission system for a factory, comprising: an adapter connected to factory equipment to adapt different interface types and communication protocols; a data acquisition module connected to the adapter to acquire operation data of the factory equipment and connected to a management system of the factory to acquire production shift data of the factory; the data acquisition module is further used to acquire historical operation data of the factory equipment and a flow transfer time difference between the factory equipment through the management system of the factory; a data processing module used to select transmission and storage and time sequence synchronization processing of the acquired operation data; a communication transmission network comprising a plurality of transmission links and relay nodes, wherein the transmission links comprise a plurality of transmission modes, and the transmission links are connected to the data acquisition module to provide a channel for data transmission; the relay nodes are arranged in the transmission links and used to forward data in the transmission links; wherein the relay nodes comprise a power supply unit, a network access unit, a frequency prediction unit and a path optimization unit; the power supply unit is used to supply power to the relay nodes and monitor the power state of the relay nodes; the network access unit is used to select a transmission mode according to network performance parameters of each transmission mode; the frequency prediction unit is used to predict the usage frequency of each relay node in a future preset time and determine the load level of each relay node by constructing a prediction model, and is further used to adjust the operation state of the corresponding relay node according to a preset adjustment strategy; the path optimization unit is used to perform data shunting on high-risk nodes through a preset data shunting strategy, and is further used to generate a corresponding backup transmission path for a relay node with a power state lower than a first preset threshold to replace an initial transmission path according to the power state and the usage frequency of each relay node; wherein the selection transmission and storage of the operation data by the data processing module comprises the following steps: step S1: recognizing the correlation intensity between different factory equipment operation parameters based on historical operation data of the factory equipment through a machine learning algorithm; step S2: dividing the factory equipment into a plurality of groups of associated equipment according to the correlation intensity; step S3: performing redundancy filtering on the operation data of each equipment for each group of associated equipment; step S4: establishing a mapping relationship model of the operation data in each group of associated equipment based on the filtered operation data, and determining a master device and a slave device in each group of associated equipment, wherein each mapping relationship model internally comprises a plurality of conversion formulas and a characteristic index and a mapping association identifier of each device, and the conversion formulas are used for mapping operation data between the master device and the slave device; step S5: determining difference data between the operation data of each master device and slave device through a preset analysis algorithm; step S6: selecting complete operation data of the master device and difference data of the slave device to transmit through the communication transmission network; and step S7: storing operation parameters of the factory equipment through a preset storage strategy.

2. The smart factory equipment data collection system of claim 1, wherein, The selection of the transmission mode comprises: adjusting the basic weight of each dimension data in the network performance parameters according to the production shift data; calculating the deviation value of each dimension data based on each dimension data in the network performance parameters and a preset reference value; and The bias values of the dimensional data are converted into corresponding scores through preset conversion rules; Based on the corresponding scores of the dimensional data and the adjusted basic weights, a comprehensive score is obtained through weighted summation; The transmission mode with the highest comprehensive score is selected for data transmission.

3. The smart factory equipment data collection system of claim 1, wherein, The construction of the prediction model comprises: Based on the running data processed by the data processing module, the running data is divided into time periods in combination with the production shift data; The characteristic quantities of the running data in each time period are counted, and the characteristic quantities of the same time period in each production shift are compared to obtain a comparison result; According to the comparison result, a difference quantity related to the equipment running state and exceeding a preset range is selected as a feature; According to the selected feature, a prediction model is constructed through a simple regression algorithm.

4. The smart factory equipment data collection system of claim 1, wherein, Each of the load levels is configured with a corresponding usage frequency interval, and the frequency prediction unit determines the load level of each relay node according to the predicted usage frequency and the usage frequency interval.

5. The smart factory equipment data collection system of claim 1, wherein, The adjustment strategy comprises: According to the load level of the relay node, the target communication power of the relay node is calculated through a preset power adjustment function; According to the calculated target communication power, the communication power of the relay node is adjusted.

6. The smart factory equipment data collection system of claim 1, wherein, The relay node is built-in with a sensor for collecting its own running parameters, and the shunting strategy comprises: For the relay node with the highest load level, the corresponding load factor is calculated according to the running parameters of the relay node; When the load factor exceeds a second preset threshold, the relay node is determined as a high-risk node; For the high-risk node, a shunting path is generated using a preset path generation algorithm; According to the generated shunting path, the data currently transmitted by the high-risk node is shunted to the adjacent relay node with the lowest load level.

7. The smart factory equipment data collection system of claim 6, wherein, The generation of the backup transmission path comprises: For the relay node with the power state lower than a first preset threshold, a path generation model is constructed based on the power state and usage frequency of each relay node, with the power state of the node as the basis and the usage frequency as the constraint condition, and a plurality of candidate paths are generated through the path generation algorithm; According to the production shift data, the weights of the power state and the usage frequency are changed; Based on the changed weights, the path scores of the candidate paths are calculated in combination with a preset scoring function; The candidate path with the highest path score is selected as the backup transmission path.

8. The smart factory equipment data collection system of claim 1, wherein, The time sequence synchronization comprises: According to the flow time difference between the factory devices, a running time sequence association model of the factory devices is established through a preset time sequence association rule base; When the state of a factory device changes, the state change time point of the corresponding associated device of the device is calculated based on the time sequence association model; According to the state change time point, a time synchronization signal is generated to adjust the collection frequency of the data collection module for the corresponding associated device of the device.

9. The smart factory equipment data collection system of claim 1, wherein, The storage strategy in step S7 comprises: For the running data of the master device, only the characteristic index and the mapping association identifier of the running data are stored; for the complete running data of the slave device and the difference data of the master device, complete storage is performed.

Images