Systems and methods for storage configuration based on industrial internet of things data center

1 . A system for storage configuration based on an industrial internet of things (IoT) data center, comprising: an industrial IoT management platform, an industrial IoT sensing network platform, and an industrial IoT perception control platform, wherein the industrial IoT management platform is communicatively connected to the industrial IoT perception control platform via the industrial IoT sensing network platform;

the industrial IoT management platform is configured to:

determine a data upload parameter based on data attribute information, historical retrieval records, and historical processing records of data to be processed, including:

determining upload analysis data via a data allocation model, based on data popularity, the data attribute information, the historical retrieval records, the historical processing records, processing capacity of the industrial IoT management platform, and a data acquisition frequency, wherein the data allocation model is a neural network model; and

determining the data upload parameter based on the upload analysis data, including:

determining data acquisition quality based on device operating status, device health status, and data acquisition bandwidth quality; and

determining the data upload parameter based on the data acquisition quality and the upload analysis data;

generate database configuration parameters based on the data upload parameter and a data acquisition frequency, wherein the database configuration parameters further include cache update parameters for a cache database, and the cache update parameters include an update period and an updated volume, including:

determining an estimated retrieval time based on the historical retrieval records and the historical processing records;

determining a space release rate based on the estimated retrieval time, including:

determining a data retrieval threshold based on the data attribute information, a type of a current database, and the processing capacity;

designating data to be processed whose count of retrieved times reaching the data retrieval threshold during a next preset period as data to be processed that needs to be cleaned, and determining the space release rate based on the data retrieval threshold and the estimated retrieval time;

determining the database configuration parameters based on the data upload parameter, the space release rate, and the data acquisition frequency;

determining the cache update parameters via an update parameter determination model based on data attribute information of cached data in the cache database, a type of an enterprise user, the historical retrieval records, the historical processing records, the device health status, and the device operating status, wherein the update parameter determination model is a recurrent neural network model, wherein:

the update period in the cache update parameters is positively correlated with the data acquisition quality and the data acquisition bandwidth quality, and negatively correlated with a data storage rate; and

updating configuration parameters of a second storage component according to the cache update parameters;

in response to determining that future remaining space is less than a preset space threshold, reallocate a size of storage space of a database, so that the future remaining space exceeds the preset space threshold;

query an acquisition frequency table based on the database configuration parameters of a previous period to update the data acquisition frequency for a current period;

generate, based on the data upload parameter, a data upload instruction, and transmit the data upload instruction to the industrial IoT sensing network platform;

generate operating parameters of a second storage component based on cache database configuration parameters;

control a second storage controller to allocate storage space to a cache database based on the operating parameters of the second storage component; and

generate, based on sub-database configuration parameters, a sub-database configuration instruction, and issue the sub-database configuration instruction to the industrial IoT sensing network platform;

obtain, from a first storage controller, total storage space allocated or released by a sub-database to a task process within a preset time period, and designate a sub-database whose total storage space allocated or released exceeds a preset threshold as a target sub-database; and

modify configuration files and structured query language (SQL) of the target sub-database to adjust a maximum count of connections allowed by the target sub-database.

2 . The system according to claim 1 , wherein the industrial IoT management platform is further configured to:

during each preset period, update, based on the database configuration parameters of the previous period, the data acquisition frequency for the current period to obtain an updated frequency; and

control a data acquisition device via the industrial IoT perception control platform to perform data acquisition according to the updated frequency, wherein the data acquisition device includes at least one of a production device or an auxiliary device deployed on a production line.

3 . The system according to claim 1 , wherein a training phase of the data allocation model includes an initial training phase and a reinforcement training phase;

during the initial training phase, a training sample dataset is obtained based on a generic dataset on a cloud platform; and

during the reinforcement training phase, the training sample dataset is obtained based on a historical dataset of a target enterprise.

4 . The system according to claim 3 , wherein the training sample dataset includes a plurality of sample subsets, training data within a same sample subset share a same label, and training data across different sample subsets are assigned distinct learning rates.

5 . The system according to claim 1 , wherein the industrial IoT management platform is further configured to:

determine a timeliness requirement degree of the data to be processed based on a length of waiting processing time and a retrieved index;

in response to the timeliness requirement degree being greater than a first threshold, upload the data to be processed to a processing unit of the industrial IoT management platform;

in response to the timeliness requirement degree being greater than a second threshold and less than the first threshold, upload the data to be processed to the cache database of the industrial IoT management platform; and

in response to the timeliness requirement degree being less than the second threshold, upload the data to be processed to the sub-database of the industrial IoT sensing network platform;

wherein the length of waiting processing time is a length of time from a time that the data to be processed is uploaded to the sub-database to a time that the industrial IoT management platform begins to process the data to be processed, the retrieved index is used to characterize a frequency at which the data to be processed is retrieved, the timeliness requirement degree characterizes a timeliness of the data to be processed, the timeliness requirement degree is positively correlated with the retrieved index and negatively correlated with the length of waiting processing time, and the first threshold is greater than the second threshold.

6 . The system according to claim 1 , wherein the preset period refers to a period for updating the database configuration parameters, and the preset period is negatively correlated to the acquisition frequency of the data to be processed.

7 . The system according to claim 1 , wherein the upload analysis data is represented as a vector including a first value, a second value, and a third value, the first value, the second value, and the third value representing a recommendation degree for uploading the data to be processed to the sub-database, the cache database, and a processing unit, respectively.

8 . The system according to claim 1 , wherein the industrial IoT management platform is further configured to:

construct an acquisition quality vector based on the device operating status, the device health status, and the data acquisition bandwidth quality;

retrieve, from a quality vector database, a reference vector that has a highest similarity to the acquisition quality vector; and

designate a reference acquisition quality corresponding to the reference vector as the data acquisition quality.

9 . A method for storage configuration based on an industrial internet of things (IoT) data center, wherein the method is performed by an industrial IoT management platform of a system for storage configuration based on an industrial IoT data center, the system includes the industrial IoT management platform, an industrial IoT sensing network platform, and an industrial IoT perception control platform, wherein the industrial IoT management platform is communicatively connected to the industrial IoT perception control platform via the industrial IoT sensing network platform;

the method comprising:

determining a data upload parameter based on data attribute information, historical retrieval records, and historical processing records of data to be processed, including:

determining upload analysis data via a data allocation model, based on data popularity, the data attribute information, the historical retrieval records, the historical processing records, processing capacity of the industrial IoT management platform, and a data acquisition frequency, wherein the data allocation model is a neural network model; and

determining the data upload parameter based on the upload analysis data, including:

determining data acquisition quality based on device operating status, device health status, and data acquisition bandwidth quality; and

determining the data upload parameter based on the data acquisition quality and the upload analysis data;

generating database configuration parameters based on the data upload parameter and a data acquisition frequency, wherein the database configuration parameters further include cache update parameters for a cache database, and the cache update parameters include an update period and an updated volume, including:

determining an estimated retrieval time based on the historical retrieval records and the historical processing records;

determining a space release rate based on the estimated retrieval time, including:

determining a data retrieval threshold based on the data attribute information, a type of a current database, and the processing capacity;

designating data to be processed whose count of retrieved times reaching the data retrieval threshold during a next preset period as data to be processed that needs to be cleaned, and determining the space release rate based on the data retrieval threshold and the estimated retrieval time;

determining the database configuration parameters based on the data upload parameter, the space release rate, and the data acquisition frequency;

determining the cache update parameters via an update parameter determination model based on data attribute information of cached data in the cache database, a type of an enterprise user, the historical retrieval records, the historical processing records, the device health status, and the device operating status, wherein the update parameter determination model is a recurrent neural network model, wherein:

the update period in the cache update parameters is positively correlated with the data acquisition quality and the data acquisition bandwidth quality, and negatively correlated with a data storage rate;

updating configuration parameters of a second storage component according to the cache update parameters; and

in response to determining that future remaining space is less than a preset space threshold, reallocating a size of storage space of a database, so that the future remaining space exceeds the preset space threshold;

querying an acquisition frequency table based on the database configuration parameters of a previous period to update the data acquisition frequency for a current period;

generating, based on the data upload parameter, a data upload instruction, and transmitting the data upload instruction to the industrial IoT sensing network platform;

generating operating parameters of a second storage component based on cache database configuration parameters;

controlling a second storage controller to allocate storage space to a cache database based on the operating parameters of the second storage component; and

generating, based on sub-database configuration parameters, a sub-database configuration instruction, and issuing the sub-database configuration instruction to the industrial IoT sensing network platform;

obtaining, from a first storage controller, total storage space allocated or released by a sub-database to a task process within a preset time period, and designating a sub-database whose total storage space allocated or released exceeds a preset threshold as a target sub-database; and

modifying configuration files and structured query language (SQL) of the target sub-database to adjust a maximum count of connections allowed by the target sub-database.

10 . The method according to claim 9 , wherein the method further comprises:

during each preset period, updating, based on the database configuration parameters of the previous period, the data acquisition frequency for the current period to obtain an updated frequency; and

controlling a data acquisition device via the industrial IoT perception control platform to perform data acquisition according to the updated frequency, wherein the data acquisition device includes at least one of a production device or auxiliary device deployed on a production line.

11 . The method according to claim 9 , wherein a training phase of the data allocation model includes an initial training phase and a reinforcement training phase;

during the initial training phase, a training sample dataset is obtained based on a generic dataset on a cloud platform; and

during the reinforcement training phase, the training sample dataset is obtained based on a historical dataset of a target enterprise.

12 . The method according to claim 11 , wherein the training sample dataset includes a plurality of sample subsets, training data within a same sample subset share a same label, and training data across different sample subsets are assigned distinct learning rates.

13 . The method according to claim 9 , wherein the method further comprises:

determining a timeliness requirement degree of the data to be processed based on a length of waiting processing time and a retrieved index;

in response to the timeliness requirement degree being greater than a first threshold, uploading the data to be processed to a processing unit of the industrial IoT management platform;

in response to the timeliness requirement degree being greater than a second threshold and less than the first threshold, uploading the data to be processed to the cache database of the industrial IoT management platform; and

in response to the timeliness requirement degree being less than the second threshold, uploading the data to be processed to the sub-database of the industrial IoT sensing network platform;

wherein the length of waiting processing time is a length of time from a time that the data to be processed is uploaded to the sub-database to a time that the industrial IoT management platform begins to process the data to be processed, the retrieved index is used to characterize a frequency at which the data to be processed is retrieved, the timeliness requirement degree characterizes a timeliness of the data to be processed, the timeliness requirement degree is positively correlated with the retrieved index and negatively correlated with the length of waiting processing time, and the first threshold is greater than the second threshold.

14 . The method according to claim 9 , wherein:

the preset period refers to a period for updating the database configuration parameters, and the preset period is negatively correlated to an acquisition frequency of the data to be processed.

15 . The method according to claim 9 , wherein the upload analysis data is represented as a vector including a first value, a second value, and a third value, and the first value, the second value, and the third value representing a recommendation degree for uploading the data to be processed to the sub-database, the cache database, and a processing unit, respectively.

16 . The method according to claim 9 , wherein determining the data acquisition quality includes:

constructing an acquisition quality vector based on the device operating status, the device health status, and the data acquisition bandwidth quality;

retrieving, from a quality vector database, a reference vector that has a highest similarity to the acquisition quality vector; and

designating a reference acquisition quality corresponding to the reference vector as the data acquisition quality.