Visual large-model low-code development platform and method

Abstract

The invention relates to the technical field of software development platforms, in particular to a visual large-model low-code development platform and method, and the visual large-model low-code development platform comprises a front-end interaction module, a rear-end processing module and a scheduling gateway. Meanwhile, the front-end interaction module comprises a SaaS environment running module, a large-model low-code editing module, a data acquisition module and a visual component, and can be used for project editing, scene configuration and window model configuration, so that technical support is provided for low-code rapid development and data docking; and meanwhile, the visual component can render and display scenes and window graphs, so that non-professional users can conveniently configure the service system. And the rear-end processing module comprises an authority configuration module, an SaaS control module and a data storage module, so that SaaS multi-tenant management is supported, the operation capability of a low-code development platform is improved, and the requirements of personalized solutions of various tenants can be met.

Description

Technical Field

[0001] The present invention relates to the technical field of software development platforms under large models, and specifically to a visual large model low-code development platform and method. Background Art

[0002] Low-code is a development method that delivers application programs with less code and at a faster speed. It automates the code that programmers usually do not want to develop, thereby simplifying the development process and providing a necessary opportunity for the digital transformation of enterprises.

[0003] The low-code platform operates through a graphical interface. Users can build application programs by dragging and dropping components and configuring modules without having to write complex code in depth. At the same time, the platform provides a rich set of preset components and modules, such as data sources, forms, processes, reports, etc. These components and modules can be used directly or customized and extended.

[0004] Chinese Patent Publication No. CN112835560A discloses a WEB multi-terminal low-code intelligent software development platform, which divides the functions of the development platform into zero-code development and low-code development. When switching functions, it is necessary to evaluate the number of required functions for software development and then select a development mode according to the evaluation results. This significantly prolongs the low-code software development process and speed. Summary of the Invention

[0005] The purpose of the present invention is to provide a visual large model low-code development platform and method for the above-mentioned problems.

[0006] The technical solution adopted by the present invention is as follows. A visual large model low-code development platform includes a front-end interaction module, a back-end processing module, and a scheduling gateway;

[0007] The front-end interaction module includes a SaaS environment operation module, a large model low-code editing module, a data acquisition module, and visual components;

[0008] The SaaS environment operation module is used to provide a SaaS operation mode and associate enterprise products, functions, or components with tenants;

[0009] The large model low-code editing module is used for project editing, scenario configuration, and window model configuration, and interacts with the scheduling gateway;

[0010] The data acquisition module is used to collect the data displayed on the front end of the current time-series platform and the data of external ports, and interacts with the scheduling gateway;

[0011] The visual components are used to generate visual scenarios and window graphics under the control of the back-end processing module;

[0012] The backend processing module includes a permission configuration module, a SaaS control module, and a data storage module;

[0013] The permission configuration module is used to configure the resource operation permissions and data access permissions of enterprises and tenants;

[0014] The data storage module is used to store the display data collected by the data collection module, the model data to be called by the SaaS control module, and the algorithm data to be called by the low-code editing module for large models;

[0015] The SaaS control module is used to process the data input by the low-code editing module for large models and the data collected by the data collection module, and call the content stored in the data storage module, and display it on the front end of the platform through visual components;

[0016] The scheduling gateway is used to connect the front-end interaction module and the backend processing module.

[0017] Optionally, a container component and an algorithm model are set in the low-code editing module for large models;

[0018] The container component contains multiple API groups, and the multiple API groups can be dragged;

[0019] The algorithm model includes large model code, code functions, and code connection processes.

[0020] Optionally, the API group contains data for processing and protection.

[0021] Optionally, project editing includes adding, deleting, modifying, and viewing the platform template through the container component;

[0022] The scenario configuration includes adding, deleting, modifying, and viewing the platform scenario resources through the container component;

[0023] The window model configuration includes adding, deleting, modifying, and viewing the platform window model resources through the container component;

[0024] When performing project editing, scenario configuration, and window model configuration, the algorithm model calls the API interfaces in the API group and connects to multiple processing backends.

[0025] Optionally, the low-code editing module for large models can provide one or more services among SpringCloud Stream service, RabbitMQ message queue service, Feign call service, Eureka registration service, and Config configuration service.

[0026] Optionally, the data collection module includes one or more of redis, MySQL, ElasticSearch, and Tencent COS.

[0027] Optionally, the visual component includes the material and UI of the platform template, and can form a pie chart and a line chart for display in one or more scenarios among the official account, APP, mini-program, and the platform itself.

[0028] Optionally, the scheduling gateway includes Nginx and SpringCloud Gateway.

[0029] This application also provides a visual large model low-code development method based on a visual large model low-code development platform. For the front-end interaction module, the method includes the following steps:

[0030] The SaaS environment operation module configures the environment parameters for the visual large model low-code development platform to run the SaaS system;

[0031] The large model low-code editing module receives service requests sent by enterprises or tenants, and sends permission verification to the scheduling gateway;

[0032] The large model low-code editing module provides project editing, scenario configuration, and window model configuration services for enterprises or tenants that pass the permission verification, and sends the request data generated by the services to the scheduling gateway;

[0033] The data collection module collects the data displayed on the front end of the platform and the data of the set external ports within the set time sequence, and sends the collected data to the scheduling gateway;

[0034] The visual component receives the visual display requirements or material and UI adjustment requirements sent by the scheduling gateway. For the visual display requirements, it generates the required graphics for display in one or more scenarios among the official account, APP, mini-program, and the platform itself; for the material and UI adjustment requirements, it generates the required material or UI for display in one or more scenarios among the official account, APP, mini-program, and the platform itself.

[0035] This application also provides a visual large model low-code development method based on a visual large model low-code development platform. For the back-end processing module, the method includes the following steps:

[0036] The permission configuration module receives the permission verification request or permission change request sent by the SaaS control module. For the permission verification request, it checks the stored permission information of enterprises and tenants and feeds back to the SaaS control module; for the permission change request, it changes the stored permission information of enterprises and tenants;

[0037] The data storage module receives the storage requirements or call requirements sent by the SaaS control module. For storage requirements, it receives and stores the storage data; for call requirements, it feeds back the model data to be called by the SaaS control module or the algorithm data to be called by the low-code editing module of the large model to the SaaS control module.

[0038] The SaaS control module receives the information passed by the front-end interaction module through the scheduling gateway. After processing, it feeds back to the scheduling gateway.

[0039] The beneficial effects of the present invention at least include one of the following;

[0040] 1. By setting up a visual large model low-code development platform including a front-end interaction module, a back-end processing module, and a scheduling gateway, and the front-end interaction module includes a SaaS environment operation module, a large model low-code editing module, a data acquisition module, and visual components, it can be used for project editing, scenario configuration, and window model configuration, providing technical support for low-code rapid development and data docking. At the same time, the visual components can render and display the scenarios and window graphics, facilitating non-professional users to configure the business system.

[0041] 2. The back-end processing module includes a permission configuration module, a SaaS control module, and a data storage module, thus supporting SaaS multi-tenant management, improving the operation ability of the low-code development platform, and also meeting the needs of various tenant personalized solution requirements. Description of the Drawings

[0042] Figure 1 It is a schematic diagram of the framework of a visual large model low-code development platform;

[0043] Figure 2 It is a low-code editing diagram;

[0044] Figure 3 It is a technical framework diagram;

[0045] Figure 4 It is a service framework diagram. Detailed Embodiments

[0046] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and shown here can be arranged and designed in various different configurations.

[0047] Accordingly, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0048] It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments may be combined with each other.

[0049] As Figure 1 shown, a visual large model low-code development platform includes a front-end interaction module, a back-end processing module, and a scheduling gateway;

[0050] The front-end interaction module includes a SaaS environment operation module, a large model low-code editing module, a data acquisition module, and visual components;

[0051] The SaaS environment operation module is used to provide a SaaS operation mode and associate enterprise products, functions, or components with tenants;

[0052] The large model low-code editing module is used for project editing, scenario configuration, and window model configuration, and interacts with the scheduling gateway;

[0053] The data acquisition module is used to collect the front-end display data and external port data of the current time series platform and interact with the scheduling gateway;

[0054] The visual components are used to generate visual scenarios and window graphics under the control of the back-end processing module;

[0055] The back-end processing module includes a permission configuration module, a SaaS control module, and a data storage module;

[0056] The permission configuration module is used to configure the resource operation permissions and data access permissions of enterprises and tenants;

[0057] The data storage module is used to store the display data collected by the data acquisition module, the model data to be called by the SaaS control module, and the algorithm data to be called by the large model low-code editing module;

[0058] The SaaS control module is used to process the data input by the large model low-code editing module and the data collected by the data acquisition module, call the content stored in the data storage module, and display it on the front-end of the platform through the visual components;

[0059] The scheduling gateway is used to connect the front-end interaction module and the back-end processing module.

[0060] The purpose of such a design is to provide a visual large model low-code development platform that includes a front-end interaction module, a back-end processing module, and a scheduling gateway. The front-end interaction module includes a SaaS environment operation module, a large model low-code editing module, a data collection module, and visual components. It can be used for project editing, scenario configuration, and window model configuration, providing technical support for low-code rapid development and data docking. At the same time, the visual components can render and display scenarios and window graphics, facilitating non-professional users to configure business systems. The back-end processing module includes a permission configuration module, a SaaS control module, and a data storage module, thus supporting SaaS multi-tenant management, enhancing the operation ability of the low-code development platform, and also meeting the needs of various tenant personalized solution requirements.

[0061] At the same time, in this embodiment, a specific composition of the large model low-code editing module is provided, in which a container component and an algorithm model are set in the large model low-code editing module;

[0062] The container component contains multiple API groups, and the multiple API groups can be dragged;

[0063] The algorithm model includes large model code, code functions, and code connection processes.

[0064] At the same time, the API group contains data for processing and protection.

[0065] The purpose of such a design is to enable non-coding professionals to directly achieve a business matching system through preset templates, scenarios, or window models without coding by setting draggable API groups. Its general design dragging process is to set module grouping and module design in the menu interface, and perform data model, function design, page design, and data association on the design page. On the advanced settings page, process design, extension buttons, data extraction display, data backwriting, custom pages, interface development, page scripts, and reminder settings are carried out. After non-coding professionals complete the above dragging or information entry work, the entire visual large model low-code platform can be put into use with functions online and authorized.

[0066] At the same time, in actual use, it can be retrained through the built-in large model code to adapt to low-code development in various scenarios. Such large models usually include existing network models such as convolutional neural networks, recurrent neural networks, long short-term memory networks, and bidirectional long short-term memory networks

[0067] At the same time, in this embodiment, as Figure 2 shown, the project editing functions implemented by the large model low-code editing module mainly include adding, deleting, modifying, and viewing platform templates;

[0068] The scenario configuration includes adding, deleting, modifying, and viewing platform scenario resources through container components;

[0069] The window model configuration includes adding, deleting, modifying, and viewing platform window model resources through container components.

[0070] Meanwhile, when performing project editing, scenario configuration, and window model configuration, the algorithm model calls the API interfaces in the API group and connects to multiple processing backends.

[0071] It should be noted that when using the large model low-code editing module, it includes code groups, filters, scripts, selectors, data write-back, data initialization, data models, extension buttons, data mapping, data extractors, data sources, and API interfaces. Among them, parts such as code groups, filters, scripts, selectors, data write-back, and data initialization are encapsulated in the algorithm model and presented in a visual form such as images for non-coding professionals to drag and configure.

[0072] The code group is mainly a set of fixed option values, also known as a data field selector;

[0073] The filter is mainly used to visually configure data conditions to automatically generate various SQLs;

[0074] The script is mainly a JavaScript script to achieve data interaction on the page;

[0075] The selector is mainly a dataset generated by the user operating system and available for other services to reference;

[0076] Data write-back mainly refers to writing or changing data in other tables after data operations are completed;

[0077] Data initialization mainly refers to other data that meets the conditions loaded when the page appears;

[0078] The data model mainly refers to the storage and fields of forms in the system;

[0079] The extension button is mainly used to configure its own business buttons on the existing basis to implement business logic;

[0080] Data mapping mainly refers to the operation of filling other data into the corresponding table when selecting a certain piece of data;

[0081] The data extractor is mainly used to extract data that meets the conditions from the associated table;

[0082] The data source is mainly different data models extracted into new data sources;

[0083] The interface is mainly the acquisition channel for data provided by third parties.

[0084] As Figure 3 and Figure 4 shown, the large model low-code editing module can provide one or more services among SpringCloud Stream service, RabbitMQ message queue service, Feign call service, Eureka registration service, and Config configuration service.

[0085] Meanwhile, the data collection module includes one or more of redis, MySQL, ElasticSearch, and Tencent COS.

[0086] Furthermore, the visual components include the materials and UI of the platform template, and can form pie charts and line charts for display in one or more scenarios among official accounts, APPs, mini-programs, and the platform itself.

[0087] Meanwhile, the scheduling gateway includes Nginx and SpringCloud Gateway.

[0088] The purpose of such a design is to form a four-layer structure on the service framework through the large model low-code editing module, data collection module, visual components, and scheduling gateway, namely the presentation layer, gateway layer, service layer, and data layer, so that the approval process, reports, data analysis, log monitoring, and multi-terminal synchronous interconnection of the entire visual large model low-code development platform are realized, and the data blood relationship is automatically formed to generate the data assets of enterprises or tenants.

[0089] In this embodiment, a visual large model low-code development method is also provided based on a visual large model low-code development platform, with the front-end interaction module end as the entry, including the following steps:

[0090] The SaaS environment operation module configures the environment parameters for the visual large model low-code development platform to run the SaaS system;

[0091] The large model low-code editing module receives service requests sent by enterprises or tenants and sends permission verification to the scheduling gateway;

[0092] The large model low-code editing module provides project editing, scenario configuration, and window model configuration services for enterprises or tenants that pass the permission verification, and sends the request data generated by the services to the scheduling gateway;

[0093] The data collection module collects the data displayed on the front end of the platform and the data of the set external ports within the set time sequence, and sends the collected data to the scheduling gateway;

[0094] The visual component receives visual display requirements or material and UI adjustment requirements sent by the scheduling gateway, generates the required graphics for visual display requirements, and displays them in one or more scenarios including the official account, APP, mini program, and the platform itself; and generates the required materials or UI for material and UI adjustment requirements, and displays them in one or more scenarios including the official account, APP, mini program, and the platform itself.

[0095] In this embodiment, a visual large model low-code development method is also provided based on a visual large model low-code development platform, and for the back-end processing module end, the following steps are included:

[0096] The permission configuration module receives the permission verification request or permission change request sent by the SaaS control module, verifies the stored permission information of the enterprise and tenants for the permission verification request, and feeds back to the SaaS control module; for the permission change request, changes the stored permission information of the enterprise and tenants;

[0097] The data storage module receives storage requirements or call requirements sent by the SaaS control module. For storage requirements, the storage data is received and stored; for call requirements, the model data to be called by the SaaS control module or the algorithm data to be called by the large model low-code editing module is fed back to the SaaS control module;

[0098] The SaaS control module receives the information transmitted by the front-end interaction module from the scheduling gateway, processes it, and then feeds it back to the scheduling gateway.

[0099] The purpose of this design is that when performing the above operations, the list page generated by the visual component can realize free switching of multiple styles, and the editing page generated by the visual component can realize free dragging and random combination of data model fields, system fields, process fields, sub-table one-to-many, sub-table one-to-one and tab grouping.

[0100] The combination of the permission configuration module and the SaaS control module constitutes three levels of security policies, namely data security policy, login security policy and physical security policy.

[0101] The data security policy includes whether custom data is stored in encrypted form, whether self-controlled fields display interference, authorization required for all buttons, detailed logs for all data operations, token access to documents, and configurable logical and physical deletion modes.

[0102] The login security policy includes setting strong passwords (using a mixture of Chinese, English and numbers), locking accounts if passwords are entered incorrectly, recording the IP addresses of login users, limiting or specifying specific login IP addresses, and all modules can be configured to allow for secondary logins.

[0103] Among them, the physical security policies include physical isolation between tenants, data stored physically, files stored in the cloud space, document watermarking, and anti-leeching.

[0104] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A visual large model low-code development platform, characterized by: Includes front-end interaction module, back-end processing module and scheduling gateway; The front-end interaction module includes a SaaS environment operation module, a large model low-code editing module, a data collection module and a visual component; The SaaS environment operation module is used to provide a SaaS operation mode and associate enterprise products, functions or components with tenants; The large model low-code editing module is used for project editing, scene configuration and window model configuration, and interacts with the scheduling gateway; The data acquisition module is used to collect the current timing platform front-end display data and external port data, and interact with the scheduling gateway; The visual component is used to generate visual scenes and window graphics under the control of the back-end processing module; The backend processing module includes a permission configuration module, a SaaS control module and a data storage module; The permission configuration module is used to configure the resource operation permissions and data access permissions of enterprises and tenants; The data storage module is used to store the display data collected by the data acquisition module, the model data to be called by the SaaS control module, and the algorithm data to be called by the large model low-code editing module; The SaaS control module is used to process the data input by the large model low-code editing module and the data collected by the data collection module, and call the content stored in the data storage module to display it on the front end of the platform through the visual component; The scheduling gateway is used to connect the front-end interaction module and the back-end processing module.

2. A visual large model low-code development platform according to claim 1, characterized in that: The large model low-code editing module is provided with a container component and an algorithm model; The container component includes multiple API groups, and the multiple API groups can be dragged; The algorithm model includes large model code, code functions and code connection process.

3. A visual large model low-code development platform according to claim 2, characterized in that: The API set includes data for processing and protection.

4. A visual large model low-code development platform according to claim 3, characterized in that: The project editing includes adding, deleting, modifying and viewing the platform template through the container component; The scene configuration includes adding, deleting, modifying and viewing the platform scene resources through the container component; The window model configuration includes adding, deleting, modifying and viewing the platform window model resources through the container component; When performing project editing, scene configuration, and window model configuration, the algorithm model calls the API interface in the API group and connects with multiple processing backends.

5. A visual large model low-code development platform according to claim 2, characterized in that: The large model low-code editing module can provide one or more services including SpringCloud Stream service, RabbitMQ message queue service, Feign call service, Eureka registration service and Config configuration service.

6. A visual large model low-code development platform according to claim 1, characterized in that: The data collection module includes one or more of redis, MySQL, ElasticSearch and Tencent COS.

7. A visual large model low-code development platform according to claim 1, characterized in that: The visual component includes the material and UI of the platform template, and can form a pie chart and a line chart for display in one or more scenarios of the official account, APP, mini program and the platform itself.

8. A visual large model low-code development platform according to claim 1, characterized in that: The scheduling gateway includes Nginx and SpringCloud Gateway.

9. A visual large model low-code development method, implemented based on a visual large model low-code development platform according to any one of claims 1 to 8, characterized in that: For the front-end interactive module, the following steps are included: The SaaS environment operation module configures the environment parameters of the visual large model low-code development platform to run the SaaS system; The large model low-code editing module receives a service request sent by an enterprise or tenant, and sends a permission verification to the scheduling gateway; The large model low-code editing module provides project editing, scene configuration and window model configuration services to enterprises or tenants that have passed the authority verification, and sends the request data generated by the service to the scheduling gateway; The data acquisition module collects the platform front-end display data and the set external port data within the set time sequence, and sends the collected data to the scheduling gateway; The visual component receives visual display requirements or material and UI adjustment requirements sent by the scheduling gateway, generates the required graphics for visual display requirements, and displays them in one or more scenarios including the official account, APP, mini program, and the platform itself; and generates the required materials or UI for material and UI adjustment requirements, and displays them in one or more scenarios including the official account, APP, mini program, and the platform itself.

10. A visual large model low-code development method, implemented based on a visual large model low-code development platform according to any one of claims 1 to 8, characterized in that: For the backend processing module, the following steps are included: The permission configuration module receives the permission verification request or permission change request sent by the SaaS control module, verifies the stored permission information of the enterprise and tenants for the permission verification request, and feeds back to the SaaS control module; for the permission change request, changes the stored permission information of the enterprise and tenants; The data storage module receives storage requirements or call requirements sent by the SaaS control module. For storage requirements, the storage data is received and stored; for call requirements, the model data to be called by the SaaS control module or the algorithm data to be called by the large model low-code editing module is fed back to the SaaS control module; The SaaS control module receives the information transmitted by the front-end interaction module from the scheduling gateway, processes it, and then feeds it back to the scheduling gateway.

Images