AI algorithm engine system and method applied to Internet of Things platform

Abstract

The invention discloses an AI algorithm engine system and method applied to an Internet of Things platform, and relates to the field of AI algorithms. Comprising an algorithm center module and an algorithm application module, the algorithm center module comprises an algorithm template management unit, and the algorithm template management unit provides creation, editing, storage and retrieval functions of an algorithm template; a user can customize algorithm templates, including algorithm types, input parameters and output formats, according to requirements; and the sending template definition unit generates a non-executable algorithm template management page and a non-executable algorithm template definition page according to an algorithm template operation instruction input by a user, so that the user can check and modify the non-executable algorithm template management page and the non-executable algorithm template definition page. The algorithm production efficiency is improved, the preset algorithm template is constructed through the algorithm center module, then the algorithm template is quoted by the algorithm application module to carry out data binding, an executable algorithm is constructed, intervention of developers is not needed, and the algorithm production efficiency is greatly improved.

Description

Technical Field

[0001] The present invention relates to the field of AI algorithms, and in particular to an AI algorithm engine system and method applied to an Internet of Things platform. Background Art

[0002] The traditional IoT platform is the connection between things, realizing an integrated platform with capabilities such as device management, data security communication and message subscription. It supports connecting massive devices downward and collecting device data to the cloud; it provides cloud API upward, and the server can send instructions to the device by calling the cloud API to achieve remote control. However, in the context of big data, the key is to integrate AI (artificial intelligence) technology and IoT (Internet of Things) technology to realize an intelligent IoT platform. The AIization of IoT platforms is an inevitable trend, continuously mining data value, providing knowledge services, and using AI capabilities to realize the closed loop of perception, cognition, decision-making, and action of people and things. The existing IoT platforms mainly have the following key problems: they do not have algorithm capabilities, or they have algorithm capabilities but the algorithm system is independent of the IoT platform, and algorithm production requires customized development, which has high implementation costs and low deployment and implementation efficiency.

[0003] In order to solve the above problems, after searching, the Chinese patent publication number CN116088842A discloses an AI algorithm engine system and method applied to the Internet of Things platform, which includes: an algorithm center module and an algorithm application module; the algorithm center module outputs a non-executable algorithm template management page and an algorithm template definition page according to the algorithm template operation instructions input by the user; the algorithm application module outputs an executable algorithm management page and an algorithm definition page according to the algorithm operation instructions input by the user.

[0004] The above patent has the following shortcomings: it relies on customized development, which is not only time-consuming and labor-intensive, but also difficult to adapt to rapidly changing needs.

[0005] To this end, the present invention proposes an AI algorithm engine system and method applied to the Internet of Things platform Summary of the invention

[0006] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an AI algorithm engine system and method applied to the Internet of Things platform.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] An AI algorithm engine system applied to an Internet of Things platform, comprising: an algorithm center module and an algorithm application module.

[0009] The algorithm center module includes:

[0010] Algorithm template management unit, which provides the creation, editing, storage and retrieval functions of algorithm templates; users can customize algorithm templates according to their needs, including algorithm type, input parameters, and output format;

[0011] The template definition unit generates a non-executable algorithm template management page and an algorithm template definition page according to the algorithm template operation instruction input by the user for the user to view and modify;

[0012] Operator library, which provides operator configuration page and operator configuration information, including various types of operators such as mathematical operations, logical judgment, and data processing, for calling when building algorithm templates;

[0013] The algorithm application module includes:

[0014] The algorithm management unit outputs an executable algorithm management page and an algorithm definition page according to the algorithm operation instructions input by the user. The user can select an algorithm template on this page, perform data binding, and build an executable algorithm instance;

[0015] The data management unit provides a data calculation configuration page for configuring the data source and data preprocessing rules required by the algorithm;

[0016] Model management unit, including the model training page, supports users to upload data sets, perform model training, and save trained models;

[0017] The business management unit provides a business execution page and deploys the trained model to actual business scenarios to achieve intelligent decision-making and optimization.

[0018] Preferably: the algorithm template management unit is stored in JSON or XML format to facilitate parsing and expansion, and defines the basic structure of the algorithm, including an input parameter list, an output parameter list, and algorithm logic steps.

[0019] Preferably: the algorithm template management unit supports multiple data source access methods, including MySQL database or PostgreSQL database, Kafka message queue or RabbitMQ message queue, and file system.

[0020] Preferably: the operator library includes:

[0021] Mathematical operators, including one or more combinations of addition operators, subtraction operators, multiplication operators, division operators, exponential operators, and logarithmic operators;

[0022] Logical judgment operators, including one or more combinations of equal judgment operators, greater than judgment operators, less than judgment operators, and judgment operators, or judgment operators and non-judgment operators;

[0023] Data processing operators, including one or more combinations of filtering operators, sorting operators, aggregation operators, and transformation operators.

[0024] Preferably: the model management unit adopts any one of the learning frameworks of TensorFlow, PyTorch, and scikit-learn, and has a built-in model evaluation tool that supports cross-validation, confusion matrix, and ROC curve evaluation methods.

[0025] Preferably: the service management unit includes:

[0026] Business rule configuration: Users can configure business rules according to business needs, combine trained models with business rules, and implement complex business logic;

[0027] Business monitoring: Provides business monitoring functions to monitor business operation status in real time, including model inference speed, accuracy, and error rate indicators;

[0028] Exception handling: During the business execution process, if you encounter data anomalies, model reasoning failures, etc., the system can automatically handle the exceptions and record the exception information for subsequent analysis.

[0029] A method for implementing an AI algorithm engine system applied to an Internet of Things platform comprises the following steps:

[0030] S1: Algorithm template construction;

[0031] S2: Algorithm instance construction;

[0032] S3: Model training and deployment;

[0033] S4: Continuous optimization and monitoring.

[0034] Preferably: the step S1 comprises the following steps:

[0035] S11: The user logs in to the IoT platform and enters the algorithm center module;

[0036] S12: On the algorithm template management page, create a new algorithm template and define the algorithm type, input parameters, and output format;

[0037] S13: Use the operators in the operator library to build the algorithm logic and generate the algorithm template definition page.

[0038] Preferably: the step S2 comprises the following steps:

[0039] S21: The user enters the algorithm application module and selects a constructed algorithm template;

[0040] S22: On the algorithm management page, perform data binding, specify the data source and preprocessing rules;

[0041] S23: Build an executable algorithm instance and perform test verification.

[0042] Preferably: the step S3 comprises the following steps:

[0043] S31: The user uploads the data set to the model management page for model training;

[0044] S32: After training is completed, save the model and configure business execution parameters;

[0045] S33: Deploy the trained model to actual business scenarios to achieve intelligent decision-making and optimization.

[0046] The beneficial effects of the present invention are:

[0047] 1. The present invention improves algorithm production efficiency: a preset algorithm template is constructed through the algorithm center module, and then the algorithm application module references the algorithm template for data binding to construct an executable algorithm without the intervention of developers, which greatly improves the algorithm production efficiency.

[0048] 2. The present invention reduces the cost of customized development: it realizes the automated production capability of algorithms and reduces the human and material input for customized development.

[0049] 3. The present invention improves the level of business intelligence: the trained model is deployed to actual business scenarios to achieve intelligent decision-making and optimization, thereby improving the overall business intelligence level. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a flow chart of a method for implementing an AI algorithm engine system applied to an Internet of Things platform proposed by the present invention. DETAILED DESCRIPTION

[0051] The technical solution of the present invention is further described in detail below in conjunction with specific implementation methods.

[0052] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", and "set" should be understood in a broad sense, for example, they can be fixedly connected or set, or detachably connected or set, or integrally connected or set. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0053] Embodiment 1:

[0054] An AI algorithm engine system applied to an Internet of Things platform, comprising: an algorithm center module and an algorithm application module.

[0055] The algorithm center module includes:

[0056] Algorithm template management unit, which provides the creation, editing, storage and retrieval functions of algorithm templates; users can customize algorithm templates according to their needs, including algorithm type, input parameters, and output format;

[0057] The template definition unit generates a non-executable algorithm template management page and an algorithm template definition page according to the algorithm template operation instruction input by the user for the user to view and modify;

[0058] Operator library, which provides operator configuration page and operator configuration information, including various types of operators such as mathematical operations, logical judgment, and data processing, for calling when building algorithm templates;

[0059] The algorithm application module includes:

[0060] The algorithm management unit outputs an executable algorithm management page and an algorithm definition page according to the algorithm operation instructions input by the user. The user can select an algorithm template on this page, perform data binding, and build an executable algorithm instance;

[0061] The data management unit provides a data calculation configuration page for configuring the data source and data preprocessing rules required by the algorithm;

[0062] Model management unit, including the model training page, supports users to upload data sets, perform model training, and save trained models;

[0063] The business management unit provides a business execution page and deploys the trained model to actual business scenarios to achieve intelligent decision-making and optimization.

[0064] The algorithm template management unit is stored in JSON format for easy parsing and expansion, and defines the basic structure of the algorithm, including an input parameter list, an output parameter list, and algorithm logic steps.

[0065] The algorithm template management unit supports multiple data source access methods, including MySQL database, Kafka message queue, and file system.

[0066] The operator library includes:

[0067] Mathematical operators, including addition operators, subtraction operators, multiplication operators, division operators, exponential operators, and logarithmic operators;

[0068] Logical judgment operators, including equal judgment operators, greater than judgment operators, less than judgment operators, and judgment operators, or judgment operators, and non-judgment operators;

[0069] Data processing operators include filtering processing operators, sorting processing operators, aggregation processing operators, and transformation processing operators.

[0070] Embodiment 2:

[0071] An AI algorithm engine system applied to an Internet of Things platform, comprising: an algorithm center module and an algorithm application module.

[0072] The algorithm center module includes:

[0073] Algorithm template management unit, which provides the creation, editing, storage and retrieval functions of algorithm templates; users can customize algorithm templates according to their needs, including algorithm type, input parameters, and output format;

[0074] The template definition unit generates a non-executable algorithm template management page and an algorithm template definition page according to the algorithm template operation instruction input by the user for the user to view and modify;

[0075] Operator library, which provides operator configuration page and operator configuration information, including various types of operators such as mathematical operations, logical judgment, and data processing, for calling when building algorithm templates;

[0076] The algorithm application module includes:

[0077] The algorithm management unit outputs an executable algorithm management page and an algorithm definition page according to the algorithm operation instructions input by the user. The user can select an algorithm template on this page, perform data binding, and build an executable algorithm instance;

[0078] The data management unit provides a data calculation configuration page for configuring the data source and data preprocessing rules required by the algorithm;

[0079] Model management unit, including the model training page, supports users to upload data sets, perform model training, and save trained models;

[0080] The business management unit provides a business execution page and deploys the trained model to actual business scenarios to achieve intelligent decision-making and optimization.

[0081] The algorithm template management unit is stored in XML format for easy parsing and expansion, and defines the basic structure of the algorithm, including an input parameter list, an output parameter list, and algorithm logic steps.

[0082] The algorithm template management unit supports multiple data source access methods, including PostgreSQL database, RabbitMQ message queue, and file system.

[0083] The operator library includes:

[0084] Mathematical operators, including addition operators, subtraction operators, multiplication operators, division operators, exponential operators, and logarithmic operators;

[0085] Logical judgment operators, including equal judgment operators, greater than judgment operators, less than judgment operators, and judgment operators, or judgment operators, and non-judgment operators;

[0086] Data processing operators include filtering processing operators, sorting processing operators, aggregation processing operators, and transformation processing operators.

[0087] Embodiment 3:

[0088] An AI algorithm engine system applied to an Internet of Things platform, comprising: an algorithm center module and an algorithm application module.

[0089] The algorithm center module includes:

[0090] Algorithm template management unit, which provides the creation, editing, storage and retrieval functions of algorithm templates; users can customize algorithm templates according to their needs, including algorithm type, input parameters, and output format;

[0091] The template definition unit generates a non-executable algorithm template management page and an algorithm template definition page according to the algorithm template operation instruction input by the user for the user to view and modify;

[0092] Operator library, which provides operator configuration page and operator configuration information, including various types of operators such as mathematical operations, logical judgment, and data processing, for calling when building algorithm templates;

[0093] The algorithm application module includes:

[0094] The algorithm management unit outputs an executable algorithm management page and an algorithm definition page according to the algorithm operation instructions input by the user. The user can select an algorithm template on this page, perform data binding, and build an executable algorithm instance;

[0095] The data management unit provides a data calculation configuration page for configuring the data source and data preprocessing rules required by the algorithm;

[0096] Model management unit, including the model training page, supports users to upload data sets, perform model training, and save trained models;

[0097] The business management unit provides a business execution page and deploys the trained model to actual business scenarios to achieve intelligent decision-making and optimization.

[0098] The algorithm template management unit is stored in JSON or XML format for easy parsing and expansion, and defines the basic structure of the algorithm, including an input parameter list, an output parameter list, and algorithm logic steps.

[0099] The algorithm template management unit supports multiple data source access methods, including MySQL database or PostgreSQL database, Kafka message queue or RabbitMQ message queue, and file system.

[0100] The operator library includes:

[0101] Mathematical operators, including addition operators, subtraction operators, multiplication operators, division operators, exponential operators, and logarithmic operators;

[0102] Logical judgment operators, including equal judgment operators, greater than judgment operators, less than judgment operators, and judgment operators, or judgment operators, and non-judgment operators;

[0103] Data processing operators include filtering processing operators, sorting processing operators, aggregation processing operators, and transformation processing operators.

[0104] The model management unit adopts the TensorFlow learning framework, and has a built-in model evaluation tool that supports cross-validation, confusion matrix, and ROC curve evaluation methods.

[0105] The service management unit includes:

[0106] Business rule configuration: Users can configure business rules according to business needs, combine trained models with business rules, and implement complex business logic;

[0107] Business monitoring: Provides business monitoring functions to monitor business operation status in real time, including model inference speed, accuracy, and error rate indicators;

[0108] Exception handling: During the business execution process, if you encounter data anomalies, model reasoning failures, etc., the system can automatically handle the exceptions and record the exception information for subsequent analysis.

[0109] Embodiment 4:

[0110] An AI algorithm engine system applied to an Internet of Things platform, comprising: an algorithm center module and an algorithm application module.

[0111] The algorithm center module includes:

[0112] Algorithm template management unit, which provides the creation, editing, storage and retrieval functions of algorithm templates; users can customize algorithm templates according to their needs, including algorithm type, input parameters, and output format;

[0113] The template definition unit generates a non-executable algorithm template management page and an algorithm template definition page according to the algorithm template operation instruction input by the user for the user to view and modify;

[0114] Operator library, which provides operator configuration page and operator configuration information, including various types of operators such as mathematical operations, logical judgment, and data processing, for calling when building algorithm templates;

[0115] The algorithm application module includes:

[0116] The algorithm management unit outputs an executable algorithm management page and an algorithm definition page according to the algorithm operation instructions input by the user. The user can select an algorithm template on this page, perform data binding, and build an executable algorithm instance;

[0117] The data management unit provides a data calculation configuration page for configuring the data source and data preprocessing rules required by the algorithm;

[0118] Model management unit, including the model training page, supports users to upload data sets, perform model training, and save trained models;

[0119] The business management unit provides a business execution page and deploys the trained model to actual business scenarios to achieve intelligent decision-making and optimization.

[0120] The algorithm template management unit is stored in JSON or XML format for easy parsing and expansion, and defines the basic structure of the algorithm, including an input parameter list, an output parameter list, and algorithm logic steps.

[0121] The algorithm template management unit supports multiple data source access methods, including MySQL database or PostgreSQL database, Kafka message queue or RabbitMQ message queue, and file system.

[0122] The operator library includes:

[0123] Mathematical operators, including addition operators, subtraction operators, multiplication operators, division operators, exponential operators, and logarithmic operators;

[0124] Logical judgment operators, including equal judgment operators, greater than judgment operators, less than judgment operators, and judgment operators, or judgment operators, and non-judgment operators;

[0125] Data processing operators include filtering processing operators, sorting processing operators, aggregation processing operators, and transformation processing operators.

[0126] The model management unit adopts the PyTorch learning framework, and has a built-in model evaluation tool that supports cross-validation, confusion matrix, and ROC curve evaluation methods.

[0127] The service management unit includes:

[0128] Business rule configuration: Users can configure business rules according to business needs, combine trained models with business rules, and implement complex business logic;

[0129] Business monitoring: Provides business monitoring functions to monitor business operation status in real time, including model inference speed, accuracy, and error rate indicators;

[0130] Exception handling: During the business execution process, if data anomalies, model reasoning failures, etc. are encountered, the system can automatically handle the exceptions and record the exception information for subsequent analysis

[0131] Embodiment 5:

[0132] An AI algorithm engine system applied to an Internet of Things platform, comprising: an algorithm center module and an algorithm application module.

[0133] The algorithm center module includes:

[0134] Algorithm template management unit, which provides the creation, editing, storage and retrieval functions of algorithm templates; users can customize algorithm templates according to their needs, including algorithm type, input parameters, and output format;

[0135] The template definition unit generates a non-executable algorithm template management page and an algorithm template definition page according to the algorithm template operation instruction input by the user for the user to view and modify;

[0136] Operator library, which provides operator configuration page and operator configuration information, including various types of operators such as mathematical operations, logical judgment, and data processing, for calling when building algorithm templates;

[0137] The algorithm application module includes:

[0138] The algorithm management unit outputs an executable algorithm management page and an algorithm definition page according to the algorithm operation instructions input by the user. The user can select an algorithm template on this page, perform data binding, and build an executable algorithm instance;

[0139] The data management unit provides a data calculation configuration page for configuring the data source and data preprocessing rules required by the algorithm;

[0140] Model management unit, including the model training page, supports users to upload data sets, perform model training, and save trained models;

[0141] The business management unit provides a business execution page and deploys the trained model to actual business scenarios to achieve intelligent decision-making and optimization.

[0142] The algorithm template management unit is stored in JSON or XML format for easy parsing and expansion, and defines the basic structure of the algorithm, including an input parameter list, an output parameter list, and algorithm logic steps.

[0143] The algorithm template management unit supports multiple data source access methods, including MySQL database or PostgreSQL database, Kafka message queue or RabbitMQ message queue, and file system.

[0144] The operator library includes:

[0145] Mathematical operators, including addition operators, subtraction operators, multiplication operators, division operators, exponential operators, and logarithmic operators;

[0146] Logical judgment operators, including equal judgment operators, greater than judgment operators, less than judgment operators, and judgment operators, or judgment operators, and non-judgment operators;

[0147] Data processing operators include filtering processing operators, sorting processing operators, aggregation processing operators, and transformation processing operators.

[0148] The model management unit adopts the scikit-learn learning framework, and has a built-in model evaluation tool that supports cross-validation, confusion matrix, and ROC curve evaluation methods.

[0149] The service management unit includes:

[0150] Business rule configuration: Users can configure business rules according to business needs, combine trained models with business rules, and implement complex business logic;

[0151] Business monitoring: Provides business monitoring functions to monitor business operation status in real time, including model inference speed, accuracy, and error rate indicators;

[0152] Exception handling: During the business execution process, if you encounter data anomalies, model reasoning failures, etc., the system can automatically handle the exceptions and record the exception information for subsequent analysis.

[0153] Embodiment 6:

[0154] A method for implementing an AI algorithm engine system applied to an Internet of Things platform comprises the following steps:

[0155] S1: Algorithm template construction;

[0156] S2: Algorithm instance construction;

[0157] S3: Model training and deployment;

[0158] S4: Continuous optimization and monitoring.

[0159] The step S1 comprises the following steps:

[0160] S11: The user logs in to the IoT platform and enters the algorithm center module;

[0161] S12: On the algorithm template management page, create a new algorithm template and define the algorithm type, input parameters, and output format;

[0162] S13: Use the operators in the operator library to build the algorithm logic and generate the algorithm template definition page.

[0163] The step S2 comprises the following steps:

[0164] S21: The user enters the algorithm application module and selects a constructed algorithm template;

[0165] S22: On the algorithm management page, perform data binding, specify the data source and preprocessing rules;

[0166] S23: Build an executable algorithm instance and perform test verification.

[0167] The step S3 comprises the following steps:

[0168] S31: The user uploads the data set to the model management page for model training;

[0169] S32: After training is completed, save the model and configure business execution parameters;

[0170] S33: Deploy the trained model to actual business scenarios to achieve intelligent decision-making.

[0171] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An AI algorithm engine system applied to an Internet of Things platform, characterized in that: include: Algorithm center module and algorithm application module, The algorithm center module includes: Algorithm template management unit, which provides the creation, editing, storage and retrieval functions of algorithm templates; users can customize algorithm templates according to their needs, including algorithm type, input parameters, and output format; The template definition unit generates a non-executable algorithm template management page and an algorithm template definition page according to the algorithm template operation instruction input by the user for the user to view and modify; Operator library, which provides operator configuration page and operator configuration information, including various types of operators such as mathematical operations, logical judgment, and data processing, for calling when building algorithm templates; The algorithm application module includes: The algorithm management unit outputs an executable algorithm management page and an algorithm definition page according to the algorithm operation instructions input by the user. The user can select an algorithm template on this page, perform data binding, and build an executable algorithm instance; The data management unit provides a data calculation configuration page for configuring the data source and data preprocessing rules required by the algorithm; Model management unit, including the model training page, supports users to upload data sets, perform model training, and save trained models; The business management unit provides a business execution page and deploys the trained model to actual business scenarios to achieve intelligent decision-making and optimization.

2. According to claim 1, the AI ​​algorithm engine system applied to the Internet of Things platform is characterized in that: The algorithm template management unit is stored in JSON or XML format for easy parsing and expansion, and defines the basic structure of the algorithm, including an input parameter list, an output parameter list, and algorithm logic steps.

3. The AI ​​algorithm engine system applied to the Internet of Things platform according to claim 1, characterized in that: The algorithm template management unit supports multiple data source access methods, including MySQL database or PostgreSQL database, Kafka message queue or RabbitMQ message queue, and file system.

4. The AI ​​algorithm engine system applied to the Internet of Things platform according to claim 1, characterized in that: The operator library includes: Mathematical operators, including one or more combinations of addition operators, subtraction operators, multiplication operators, division operators, exponential operators, and logarithmic operators; Logical judgment operators, including one or more combinations of equal judgment operators, greater than judgment operators, less than judgment operators, and judgment operators, or judgment operators and non-judgment operators; Data processing operators, including one or more combinations of filtering operators, sorting operators, aggregation operators, and transformation operators.

5. The AI ​​algorithm engine system applied to the Internet of Things platform according to claim 1, characterized in that: The model management unit adopts any one of the learning frameworks of TensorFlow, PyTorch, and scikit-learn, and has a built-in model evaluation tool that supports cross-validation, confusion matrix, and ROC curve evaluation methods.

6. The AI ​​algorithm engine system applied to the Internet of Things platform according to claim 1, characterized in that: The service management unit includes: Business rule configuration: Users can configure business rules according to business needs, combine trained models with business rules, and implement complex business logic; Business monitoring: Provides business monitoring functions to monitor business operation status in real time, including model inference speed, accuracy, and error rate indicators; Exception handling: During the business execution process, if you encounter data anomalies, model reasoning failures, etc., the system can automatically handle the exceptions and record the exception information for subsequent analysis.

7. A method for implementing an AI algorithm engine system applied to an Internet of Things platform, which is a method for implementing an AI algorithm engine system applied to an Internet of Things platform as described in any one of claims 1 to 6, characterized in that: The following steps are involved: S1: Algorithm template construction; S2: Algorithm instance construction; S3: Model training and deployment; S4: Continuous optimization and monitoring.

8. The method for implementing an AI algorithm engine system applied to an Internet of Things platform according to claim 7, characterized in that: The step S1 comprises the following steps: S11: The user logs in to the IoT platform and enters the algorithm center module; S12: On the algorithm template management page, create a new algorithm template and define the algorithm type, input parameters, and output format; S13: Use the operators in the operator library to build the algorithm logic and generate the algorithm template definition page.

9. The method for implementing an AI algorithm engine system applied to an Internet of Things platform according to claim 7, characterized in that: The step S2 comprises the following steps: S21: The user enters the algorithm application module and selects a constructed algorithm template; S22: On the algorithm management page, perform data binding, specify the data source and preprocessing rules; S23: Build an executable algorithm instance and perform test verification.

10. The method for implementing an AI algorithm engine system applied to an Internet of Things platform according to claim 7, characterized in that: The step S3 comprises the following steps: S31: The user uploads the data set to the model management page for model training; S32: After training is completed, save the model and configure business execution parameters; S33: Deploy the trained model to actual business scenarios to achieve intelligent decision-making and optimization.

Images