Data Acquisition Method, Device and Storage Medium for Industrial Internet of Things

Abstract

The present invention relates to the field of technical data acquisition, and specifically discloses a method, device and storage medium for data acquisition in the industrial Internet of Things, including the following steps: S1. Construct a dynamic sensor network to collect the operation data of industrial equipment in real time through multi-modal sensors; S2. Dynamically adjust the data acquisition frequency based on the environmental noise coefficient and the device health index; S3. Perform hierarchical compression and encryption on the collected data, and allocate the compression ratio and encryption intensity according to the importance of the data type; S4. Execute load balancing through edge computing nodes and allocate computing tasks to the optimal nodes; S5. Store the processed data in a distributed database according to spatio-temporal correlation and generate a multi-copy disaster recovery index. Through dynamic acquisition, intelligent processing, optimized storage and adaptive strategy adjustment, it realizes accurate and efficient data acquisition, flexibly guarantees data security and processing performance, improves storage query efficiency and disaster recovery ability, and maximizes the utilization of storage resources.

Description

Technical Field

[0001] The present invention relates to the field of technical data acquisition, and particularly to a data acquisition method, device and storage medium for industrial Internet of Things. Background Technique

[0002] With the rapid development of industrial Internet of Things, data acquisition, processing and storage in industrial production are facing unprecedented challenges. Traditional data acquisition methods have exposed many problems in complex industrial environments.

[0003] In the data acquisition link, previous sensor networks were often statically fixed and difficult to adapt to the dynamic changes of industrial equipment operating conditions. For example, in some large factories, different frequencies of data acquisition are required for equipment under different working conditions, but the traditional method cannot be adjusted in real time, resulting in either insufficient data acquisition affecting the monitoring accuracy of equipment or over-acquisition causing resource waste. Moreover, environmental noise interference seriously affects data quality, and traditional methods lack effective coping strategies, making it difficult to ensure the accuracy and reliability of the acquired data.

[0004] For data processing, on the one hand, different types of data have different importance in industrial production, but traditional data compression and encryption methods adopt a unified standard, which cannot specifically guarantee the security of important data and cannot reasonably allocate storage resources. On the other hand, the task allocation of edge computing nodes lacks an effective load balancing mechanism, and it is easy to have single-point overload, resulting in the degradation or even collapse of the performance of the entire system.

[0005] In terms of data storage, traditional distributed database storage methods lack full consideration of the spatio-temporal characteristics of industrial data, and the data query efficiency is low. When querying device data at a specific time and location, it is often necessary to traverse a large amount of irrelevant data, consuming a lot of time and resources. At the same time, the data disaster recovery ability is insufficient, and once the data is lost or damaged, it is difficult to recover quickly, bringing serious risks to industrial production.

[0006] In summary, the existing industrial Internet of Things data acquisition, processing and storage technologies have many defects and cannot meet the industrial production's requirements for efficient, reliable and secure data management. Summary of the Invention

[0007] The purpose of the present invention is to provide a data acquisition method, device and storage medium for industrial Internet of Things to solve the problems raised in the above background technique.

[0008] To achieve the above purpose, the present invention provides the following technical solution: A data acquisition method for industrial Internet of Things, including the following steps:

[0009] S1. Real-time acquisition: Construct a dynamic sensor network and collect the operation data of industrial equipment in real time through multi-modal sensors;

[0010] S2. Frequency adjustment: Based on the environmental noise coefficient and the device health index , dynamically adjust the data acquisition frequency , and the formula is:

[0011] ;

[0012] wherein, is the reference frequency, is the noise change rate, is the noise threshold, is the maximum value of the health index;

[0013] S3. Encryption and compression: Perform hierarchical compression and encryption on the collected data, and allocate the compression ratio and the encryption intensity according to the importance of the data type , satisfying:

[0014] ;

[0015] wherein, is the maximum value of the importance of the data type; is the maximum value of the encryption intensity;

[0016] S4. Load balancing: Perform load balancing through edge computing nodes, allocate computing tasks to the optimal nodes, and avoid single-point overload;

[0017] S5. Data storage: Store the processed data in a distributed database according to spatio-temporal correlation, and generate a multi-copy disaster recovery index.

[0018] Preferably, the calculation formula of the device health index is:

[0019] ;

[0020] wherein, is the weight of the th sensor parameter, is the real-time value of the th sensor, is the nominal value of the th sensor, is the maximum value of the th sensor parameter, is the minimum value of the th sensor parameter.

[0021] Preferably, during the hierarchical compression and encryption process, a dynamic switching strategy between the lightweight encryption algorithm LEA and the Advanced Encryption Standard AES is adopted. When the encryption strength exceeds a preset threshold , the AES encryption algorithm is enabled for encryption; otherwise, the lightweight encryption algorithm LEA is used. The conditions for the dynamic switching strategy are determined by the following formula:

[0022] ;

[0023] where represents the security factor, represents the current idle computing power of the edge node, represents the total computing power of the edge node.

[0024] Preferably, the load balancing dynamically selects edge nodes through task allocation weights , and the weight formula is:

[0025] ;

[0026] where is the optimization weight coefficient, is the latency of the th edge node, is the idle CPU computing power of the th edge node, is the task queue length of the th edge node, is the maximum length of the task queue of the th edge node.

[0027] Preferably, the generation of the multi-copy disaster tolerance index specifically includes the following steps: According to the data access frequency and importance , combined with the critical security factor preset by the system, calculate the number of copies , and the formula is:

[0028] .

[0029] The data acquisition device for industrial Internet of Things includes: a dynamic perception module, an edge computing gateway, and a distributed storage engine; the dynamic perception module is used to perform data acquisition and frequency adjustment; the edge computing gateway integrates a lightweight encryption chip and an FPGA accelerator to implement hierarchical compression encryption and load balancing; the distributed storage engine supports space-time partitioning and automatic scaling of replicas, and the storage structure satisfies:

[0030] , the data will be sliced according to time , device location identifier and data priority for organization.

[0031] Preferably, the edge computing gateway is built with a fault self - recovery mechanism. When a node fails, data is re - routed through the redundant path weight as follows:

[0032] ;

[0033] wherein, represents the bandwidth of the path, refers to the signal strength, is the number of hops, that is, the number of intermediate nodes that data needs to pass through from the source node to the target node, represents the delay.

[0034] Preferably, the distributed storage engine adopts column - based storage and time - window partitioning. The data query efficiency optimization formula is:

[0035] ;

[0036] wherein, represents the actual data query time, represents the benchmark query time, represents the cache hit rate, which refers to the proportion that can be directly obtained from the cache when querying data, represents the size of the data to be queried, that is, the amount of data involved in the query operation, represents the I / O speed, which refers to the data read - write speed of the storage device.

[0037] The data storage medium of the industrial Internet of Things stores a computer program executable by a processor. When executed by the processor, it can implement the steps of the data acquisition method of the industrial Internet of Things, and the computer program also has the function of dynamically adjusting the storage strategy based on data heat The calculation method of data heat is as follows:

[0038] ;

[0039] wherein, represents the number of accesses to the data within a specific time window ; represents the time elapsed since the data was last accessed; is the data half - life parameter, which is used to describe the rate at which data heat decays over time.

[0040] An industrial Internet of Things data acquisition method, device, and storage medium proposed by the present invention have the beneficial effects that:

[0041] 1. Dynamic and efficient data acquisition: By constructing a dynamic sensor network and dynamically adjusting the data acquisition frequency based on the environmental noise coefficient and device health index, it is possible to accurately collect data in a complex industrial environment, ensuring both the integrity and accuracy of the data while avoiding waste of resources, effectively improving the efficiency and quality of data acquisition, and providing strong support for the precise monitoring and fault prediction of industrial equipment.

[0042] 2. Intelligent and flexible data processing: The hierarchical compression and encryption strategy allocates the compression ratio and encryption intensity according to the importance of data types, which can not only ensure the security of important data but also reasonably utilize storage resources. At the same time, the dynamic switching between lightweight encryption algorithms and the Advanced Encryption Standard, combined with the load balancing mechanism of edge computing nodes, dynamically allocates tasks according to the node's latency, idle computing power, and task queue load, effectively avoiding single-point overload, greatly improving the overall performance and stability of the system, and ensuring the efficiency and reliability of data processing.

[0043] 3. Optimized data storage and management: By adopting columnar storage and time window partitioning, and storing data in combination with spatio-temporal correlation, the data query efficiency is significantly improved. The multi-copy disaster recovery index is generated according to the data access frequency and importance, enhancing the disaster recovery ability of the data and ensuring the security and reliability of the data. The distributed storage engine supports spatio-temporal partitioning and automatic scaling of replicas, further optimizing the data storage structure and facilitating data management and use.

[0044] 4. Adaptive storage strategy adjustment: Dynamically adjusting the storage strategy based on data popularity can reasonably allocate storage resources according to the actual usage of data. For frequently accessed hot data, a more efficient storage method is given to improve the data access speed; for less popular data, a more resource-saving storage method is adopted to reduce storage costs, achieving the maximum utilization of storage resources.

[0045] In summary, through dynamic acquisition, intelligent processing, optimized storage, and adaptive strategy adjustment, the present invention realizes accurate and efficient data acquisition, flexibly ensures data security and processing performance, improves storage query efficiency and disaster recovery ability, and maximizes the utilization of storage resources. Brief Description of the Drawings

[0046] Figure 1 It is a flowchart of the data acquisition method for the industrial Internet of Things of the present invention;

[0047] Figure 2 It is a schematic block diagram of the data acquisition device for the industrial Internet of Things of the present invention;

[0048] Figure 3 It is a schematic diagram of the dynamic adjustment of data popularity and storage strategy of the present invention. Detailed Embodiments

[0049] Next, in combination with the accompanying drawings in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0050] Please refer to Figures 1 - 3 , the present invention provides a technical solution: a data acquisition method for industrial Internet of Things, including the following steps:

[0051] S1. Real-time acquisition: Construct a dynamic sensor network to collect the operation data of industrial equipment in real time through multimodal sensors.

[0052] S2. Frequency adjustment: Based on the environmental noise coefficient and the device health index , dynamically adjust the data acquisition frequency , the formula is:

[0053] ; where is the reference frequency, which is the initial reference value of the data acquisition frequency; is the noise change rate, that is, the change degree of environmental noise within a certain time; is the noise threshold. When the noise change rate exceeds this threshold, it indicates that the interference degree of environmental noise on data acquisition has reached the level that requires adjustment of the acquisition frequency; is the maximum value of the health index, which is used to normalize the current device health index ; this formula comprehensively considers the influence of environmental noise and device health status on the acquisition frequency. When the environmental noise changes greatly ( increases) or the device health status is poor ( decreases), appropriately increase the acquisition frequency to more closely monitor the device operation status; otherwise, reduce the acquisition frequency to save resources.

[0054] S3. Encryption and compression: Perform hierarchical compression and encryption on the collected data, and allocate the compression rate and the encryption intensity according to the importance of data types , satisfying:

[0055] ; where is the maximum value of the importance of data types; is the maximum value of the encryption intensity;

[0056] For data with lower importance ( Smaller), a higher compression ratio is assigned ( The smaller, The larger), to save storage space, but the encryption strength is relatively low at the same time ( , The smaller, The smaller); while for data with high importance ( Larger), the compression ratio is reduced to ensure data integrity, and at the same time the encryption strength is increased to ensure data security. Among them, Is the maximum value of the importance of the data type, used to normalize the actual importance value For normalization processing; Is the maximum value of the encryption strength, used as an upper limit reference for determining the encryption strength of data with different importance levels.

[0057] S4. Load balancing: Execute load balancing through edge computing nodes, allocate computing tasks to the optimal nodes, and avoid single-point overload.

[0058] S5. Data storage: Store the processed data in a distributed database according to spatio-temporal correlation, and generate a multi-copy disaster recovery index.

[0059] The device health index The calculation formula is:

[0060] ; Among them, Is the weight of the th sensor parameter, Is the real-time value of the th sensor, Is the nominal value of the th sensor, Is the maximum value of the th sensor parameter, Is the minimum value of the th sensor parameter;

[0061] The device health index Is an index that comprehensively considers multiple sensor parameters, obtained by weighted summation of the evaluation results of each sensor parameter. The value range of this index is between 0 and 1. The closer the value is to 1, the better the health condition of the device; the closer the value is to 0, it indicates that the device has relatively serious faults or abnormalities and is in a poor health state;

[0062] Represents the number of sensors. In industrial equipment, multiple different types of sensors are usually installed to monitor the operating status of different aspects of the equipment, Is the total number of these sensors used to evaluate the health condition of the device; Is the Sensor parameter weight: Different sensors have different degrees of importance for judging the overall health status of the device. Therefore, a weight is assigned to each sensor. The value range of the weight is from 0 to 1, and the sum of the weights of all sensors satisfies ; For example, for a motor device, the weight of the temperature sensor is relatively high because too high motor temperature may cause serious failures, while the weights of some auxiliary sensors are low. is the real-time value of the -th sensor: That is, the data collected by the sensor at the current moment. For example, the motor temperature value collected by the temperature sensor in real time, the internal pressure value of the device collected by the pressure sensor in real time, etc. is the nominal value of the -th sensor, which is the value that the sensor parameter should reach when the device is in normal operation.

[0063] More specifically, , this part calculates the deviation degree between the real-time value and the nominal value of the -th sensor and normalizes it to the interval from 0 to 1. When , = 0, then = 0, indicating that the sensor parameter is in a normal state; when deviates from farther, the value is larger, the value is closer to 1, indicating that the sensor parameter deviates from the normal state more severely. is to convert the above deviation degree so that the closer the value is to 1, the more normal the -th sensor parameter is, and the closer the value is to 0, the more abnormal the sensor parameter is. considers the weight of the sensor and weights the normality of each sensor to highlight the influence of important sensors on the device health index. For example, when the parameter of a sensor with a larger weight is abnormal, the impact on the device health index will be greater.

[0064] By summing up the weighted normality of all sensors, the device health index is finally obtained. This index can intuitively reflect the overall health status of the device and can be used in the data acquisition frequency adjustment formula mentioned above to dynamically adjust the data acquisition frequency according to the device health status so as to more closely monitor the device status. For example, when the device health index is low, appropriately increase the data acquisition frequency to timely detect possible failures of the device; when the device health index is high, reduce the data acquisition frequency to save system resources.

[0065] During the hierarchical compression and encryption process, a dynamic switching strategy between the lightweight encryption algorithm LEA and the Advanced Encryption Standard AES is adopted. When the encryption intensity exceeds a preset threshold , the AES encryption algorithm is enabled for encryption; otherwise, the lightweight encryption algorithm LEA is used. The conditions for the dynamic switching strategy are determined by the following formula:

[0066] ; where represents the safety factor, which is a constant set according to actual needs, represents the current idle computing power of the edge node, represents the total computing power of the edge node.

[0067] The load balancing dynamically selects edge nodes through task assignment weights , and the weight formula is:

[0068] ; where is the optimization weight coefficient, is the th delay of the edge node, is the th idle CPU computing power of the edge node, is the th task queue length of the edge node, is the th maximum length of the task queue of the edge node;

[0069] The threshold is determined by the formula , where is the safety factor, which is a constant set according to actual needs. If the data security requirements are high, the value can be appropriately increased; if more attention is paid to resource utilization efficiency, the value can be decreased; represents the current idle computing power of the edge node, reflecting the amount of resources currently available for encryption calculation at this node; is the total computing power of the edge node. When the proportion of the idle computing power of the edge node is higher and the safety factor is set larger, the threshold is higher, and it is more inclined to use the AES encryption algorithm. Conversely, it is more inclined to use the LEA algorithm;

[0070] The weight formula comprehensively considers multiple factors affecting the processing ability of edge nodes; is the th delay of the edge node. The lower the delay the larger the value, multiplied by the optimization weight coefficient After that, the greater the proportion in the weight indicates that the node is more suitable for processing tasks; it reflects the consideration of the real-time nature of data processing. The larger it is, the higher the degree of emphasis the system places on node latency. represents the idle CPU computing power of the th edge node. The more idle computing power there is, multiply by the optimization weight coefficient and the greater the contribution to the weight means that the node has more resources to process new tasks. The size of reflects the degree of emphasis the system places on the utilization of node CPU resources. is the task queue length of the th edge node. is the maximum length of the task queue of this node. represents the load level of the current task queue. represents the remaining processing capacity, multiply by the optimization weight coefficient and then incorporate it into the calculation of the weight . The larger it is, the more the system pays attention to the load situation of the node task queue, avoiding allocating tasks to nodes with full task queues to prevent task backlogs.

[0071] The generation of the multi-copy disaster tolerance index specifically includes the following steps: According to the data access frequency and importance , combined with the critical safety factor preset by the system, calculate the number of copies , and the formula is:

[0072] ;

[0073] More specifically, Multiply the data access frequency, importance, and critical safety factor to obtain a value that comprehensively considers these three factors. The larger this value is, the higher the importance and access requirements of the data, and more copies are needed to ensure its security and availability; Take the logarithm to the base 2 of the above product. The role of taking the logarithm is to scale the product result so that the finally calculated number of copies is within a reasonable range. Taking the logarithm to the base 2 is because in computer storage and data processing, data is usually stored and transmitted in binary form, and this choice conforms to the characteristics of data storage and management; Use the ceiling function to process the logarithmic result to obtain the final number of copies . The ceiling function is used because the number of copies must be an integer and sufficient copies must be ensured to meet the disaster tolerance requirements. For example, if the calculated result is 2.3, then after taking the ceiling, it becomes 3, that is, 3 copies need to be created;

[0074] According to the characteristics of different data, an appropriate number of copies can be dynamically allocated to each data item. For data with high access frequency and high importance, more copies will be generated to ensure rapid data recovery in case of failures and guarantee business continuity; for data with low access frequency and low importance, fewer copies will be generated, thus reasonably utilizing storage resources while ensuring data security and reducing storage costs.

[0075] The data acquisition device for the industrial Internet of Things includes: a dynamic sensing module, an edge computing gateway, and a distributed storage engine; the dynamic sensing module is used to perform data acquisition and frequency adjustment; the edge computing gateway integrates a lightweight encryption chip and an FPGA accelerator to achieve hierarchical compression encryption and load balancing; the distributed storage engine supports spatio-temporal partitioning and automatic scaling of replicas, and the storage structure satisfies:

[0076] , the data will be organized according to time slices , device location identifiers and data priorities for organization.

[0077] The edge computing gateway has a built-in fault self-recovery mechanism. When a node fails, data is re-routed through the redundant path weight as follows: ; where represents the bandwidth of the path, refers to the signal strength, is the number of hops, that is, the number of intermediate nodes that the data needs to pass through from the source node to the target node, represents the delay;

[0078] The numerator represents the transmission capacity and stability of the path. The larger the bandwidth and the stronger the signal strength, the larger the value of the numerator, indicating that the performance of this path in data transmission is better; the denominator represents the transmission cost of the path. The more hops and the greater the delay, the larger the value of the denominator, indicating that this path will bring more overhead and delay in the data transmission process; therefore, the weight The larger the value, the better the comprehensive performance of this path in terms of transmission capacity, stability, and transmission cost, and it is a more suitable path for data transmission;

[0079] When a certain node in the edge computing gateway fails, the system will calculate the weights of each redundant path according to the above formula ; then, select the weight The maximum path is used as the new data transmission path to reroute the data to the target node; in this way, data transmission can be quickly and automatically restored, ensuring the normal operation of the industrial Internet of Things system and improving the reliability and stability of the system; at the same time, this mechanism can also dynamically select the optimal path according to the real-time state of the network, make full use of network resources, and improve the efficiency of data transmission.

[0080] The distributed storage engine adopts columnar storage and time window partitioning, and the data query efficiency optimization formula is:

[0081] ; where represents the actual data query time, represents the benchmark query time, represents the cache hit rate, which refers to the proportion of query data that can be directly obtained from the cache, represents the size of the data to be queried, that is, the amount of data involved in the query operation, represents the I / O speed, which refers to the data read and write speed of the storage device;

[0082] More specifically, this part reflects the impact of the cache on the query time. When the cache hit rate is 100%, , the time overhead of this part is 0, which means that all data can be quickly obtained from the cache; when the cache hit rate is 0%, , and at this time the query time includes the benchmark query time; reflects the time required to read data from the storage device. The larger the data volume and the slower the I / O speed, the longer this part of the time;

[0083] Through this formula, it can be analyzed that measures such as increasing the cache hit rate, reducing the amount of query data, and improving the I / O speed of the storage device can effectively reduce the actual data query time and improve the data query efficiency of the distributed storage engine.

[0084] The data storage medium of the industrial Internet of Things, in which there is a computer program executable by a processor stored. When executed by the processor, it can implement the steps of the data acquisition method of the industrial Internet of Things, and the computer program also has the function of dynamically adjusting the storage strategy based on the data heat The calculation method of the data heat is as follows:

[0085] ; where represents within a specific time window The number of accesses to the data within a preset time period (which is used to count the number of accesses to the data), this parameter reflects the degree of attention the data has received recently. The more accesses there are, the hotter the data is; Indicates the time elapsed since the data was last accessed. As time goes by, if the data has not been accessed for a long time, its heat will gradually decrease; Is the data half-life parameter, which is used to describe the rate at which the data heat decays over time, The smaller the value, the faster the data heat decays; The larger the value, the slower the data heat decays.

[0086] The computer program stored in the storage medium has two functions:

[0087] Implement the steps of the data acquisition method: This program can execute each step of the data acquisition method for the industrial Internet of Things, build a dynamic sensor network, collect real-time operation data of industrial equipment, dynamically adjust the acquisition frequency according to the environmental noise coefficient and the device health index, perform hierarchical compression and encryption on the acquired data, achieve load balancing through edge computing nodes, and store the processed data in a distributed database according to spatio-temporal correlation and generate a multi-copy disaster recovery index, etc. These steps together constitute a complete data acquisition, processing and storage process, ensuring that the industrial Internet of Things system can efficiently and accurately obtain and manage the operation data of the equipment;

[0088] Dynamically adjust the storage policy based on the data heat: This is an important feature of this program, aiming to optimize the allocation of storage resources according to the actual usage of the data.

[0089] More specifically, Calculate the average access frequency of the data within a specific time window, which reflects the current activity level of the data; Is an exponential decay function. As increases, the value of this function gradually decreases, so that the data heat decreases over time. By multiplying the two, an index that comprehensively reflects the data heat is obtained .

[0090] The computer program will dynamically adjust the storage policy of the data according to the calculated value. For data with higher heat, since it is frequently accessed, a more efficient storage method is adopted, such as storing it in a high-speed storage device (such as a solid-state drive) to improve the data access speed and reduce the query response time; for data with lower heat, a more space-saving storage method can be adopted, such as migrating it to a low-cost storage device such as a tape library to reduce the storage cost. In this way, while meeting the data access requirements, the optimal configuration of storage resources can be achieved, improving the overall performance and economy of the industrial Internet of Things system.

[0091] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. The data collection method of the industrial Internet of Things is characterized by: The following steps are involved: S1. Real-time collection: Build a dynamic sensor network to collect the operating data of industrial equipment in real time through multimodal sensors; S2. Frequency adjustment: based on environmental noise coefficient and device health index , dynamically adjust the data collection frequency , the formula is: ； in, is the reference frequency, is the noise change rate, is the noise threshold, is the maximum value of the health index; S3. Encryption and compression: Perform layered compression and encryption on the collected data, based on the importance of the data type. Allocate compression ratio and encryption strength ,satisfy: ； in, is the maximum value of the data type's importance; is the maximum value of encryption strength; S4. Load balancing: Load balancing is performed through edge computing nodes, and computing tasks are allocated to the optimal nodes to avoid single point overload; S5. Data storage: Store the processed data in a distributed database according to time and space correlation, and generate a multi-copy disaster recovery index.

2. The data collection method of the industrial Internet of Things according to claim 1 is characterized in that: The device health index The calculation formula is: ； in, For the sensor parameter weights, For the Real-time values ​​of sensors, For the The nominal value of the sensor, For the The maximum value of the sensor parameter, For the The minimum value of a sensor parameter.

3. The data collection method of the industrial Internet of Things according to claim 2 is characterized in that: In the layered compression encryption process, a dynamic switching strategy between the lightweight encryption algorithm LEA and the advanced encryption standard AES is adopted. Exceeding the preset threshold When the AES encryption algorithm is enabled, the lightweight encryption algorithm LEA is used for encryption. The condition for dynamic switching strategy is determined by the following formula: ； in, represents the safety factor, Indicates the current idle computing power of the edge node, Indicates the total computing power of edge nodes.

4. The data collection method of the industrial Internet of Things according to claim 3 is characterized in that: The load balancing is done by assigning weights to tasks. Dynamically select edge nodes, the weight formula is: ； in, To optimize the weight coefficient, For the The latency of the edge nodes, For the The idle CPU power of edge nodes, For the The length of the edge node task queue, For the The maximum length of the task queue of an edge node.

5. The data collection method of the industrial Internet of Things according to claim 4 is characterized in that: The generation of the multi-copy disaster recovery index specifically includes the following steps: and importance , combined with the system preset critical safety factor , calculate the number of copies , the formula is: 。 6. A data acquisition device for industrial Internet of Things, which is applied to the data acquisition method for industrial Internet of Things according to claim 5, characterized in that: include: Dynamic sensing module, used to perform data collection and frequency adjustment; Edge computing gateway, integrating lightweight encryption chip and FPGA accelerator to achieve layered compression encryption and load balancing; Distributed storage engine, supports time-space partitioning and automatic expansion and contraction of replicas, and the storage structure meets the following requirements: , the data will be divided into time slices , Equipment location identification and data priority Get organized.

7. The data acquisition device of the industrial Internet of Things according to claim 6, characterized in that: The edge computing gateway has a built-in fault self-recovery mechanism. When a node fails, the redundant path weight Rerouting data: ； in, represents the bandwidth of the path, Refers to the signal strength, is the number of hops, that is, the number of intermediate nodes that data needs to pass through from the source node to the target node. Represents delay.

8. The data acquisition device of the industrial Internet of Things according to claim 7, characterized in that: The distributed storage engine adopts column storage and time window partitioning, and the data query efficiency optimization formula is: ； in, Indicates the actual data query time. Indicates the benchmark query time, Indicates the cache hit rate, which refers to the proportion of data that can be directly obtained from the cache when querying data. Indicates the size of the data to be queried, that is, the amount of data involved in the query operation. Indicates I / O speed, which refers to the data reading and writing speed of the storage device.

9. A data storage medium for industrial Internet of Things, characterized in that: The data storage medium of the industrial Internet of Things stores a computer program that can be executed by a processor. When executed by the processor, the steps of the data collection method of the industrial Internet of Things described in any one of claims 1 to 5 can be implemented, and the computer program also has a function based on data heat. Dynamically adjust storage strategies to meet data heat The calculation method is as follows: ； in, Indicates that in the time window The number of accesses to data within the Indicates the time that has passed since the data was last accessed; It is a data half-life parameter, which is used to describe the speed at which data popularity decays over time.

Images