Experiment intelligent assistant and method based on artificial intelligence

Abstract

The invention discloses an intelligent experiment assistant and method based on artificial intelligence, and relates to the technical field of artificial intelligence. The system adopts a five-layer architecture comprising a voice interaction layer, a gateway and access layer, a core service layer, a data service layer and an external system layer. The voice interaction layer is used for inputting and feeding back voice instructions; the core service layer integrates an intelligent service module, a business logic module and an industrial control module, and realizes voice recognition and understanding, experimental process management and industrial equipment control respectively; the data service layer supports multi-type data storage; and the external system layer realizes integration with an industrial system and enterprise authentication. According to the corresponding method, intelligent processing and task execution of voice instructions are realized through the system. According to the invention, traditional manual operation can be replaced by natural voice interaction, the operation convenience and safety are improved, intelligent decision-making and multi-device cooperation in a complex experiment scene are supported, and intelligent management of the whole experiment process is realized.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and more specifically, to an experimental intelligent assistant and method based on artificial intelligence. Background Technology

[0002] Traditional laboratory operations primarily rely on input devices such as keyboards and mice, interacting through graphical interfaces or command lines. This process is complex, unfriendly to non-professional users, prone to errors, and impacts experimental efficiency. Existing experimental management systems suffer from significant shortcomings in natural interaction and intelligent understanding, making them ill-suited for complex and ever-changing experimental scenarios.

[0003] Currently, while some laboratories have attempted to introduce voice control technology, most are limited to executing simple commands and lack a deep understanding of complex voice commands and the ability to engage in multi-turn interactions, thus failing to achieve intelligent experimental management throughout the entire process. Furthermore, existing systems still lag behind in multimodal information fusion, real-time decision-making, and efficient collaboration with industrial equipment.

[0004] Furthermore, existing experimental management systems are often deployed independently, making it difficult to achieve seamless integration with industrial automation equipment, enterprise certification systems, and distributed data platforms. This results in prominent information silos, hindering the construction of a unified, secure, and efficient intelligent experimental ecosystem and restricting the improvement of the overall automation and intelligence level of laboratories. Summary of the Invention

[0005] In view of the above-mentioned technical problems in related technologies, the present invention proposes an experimental intelligent assistant and method based on artificial intelligence, which can overcome the above-mentioned shortcomings of the prior art.

[0006] To achieve the above-mentioned technical objectives, the technical solution of the present invention is implemented as follows: An AI-based experimental intelligent assistant; This AI-based experimental intelligent assistant includes a voice interaction layer, a gateway and access layer, a core service layer, a data service layer, and an external system layer. The voice interaction layer is used to enable interaction between the user and the system; The gateway and access layer are used to achieve unified access and secure access to the system. The core service layer includes an intelligent service module, a business logic module, and an industrial control module; the intelligent service module is used to implement speech recognition and natural language understanding; the business logic module is used to implement experimental process control and information query; and the industrial control module is used to implement collaborative control and data acquisition with external industrial automation equipment. The data service layer is used to store and manage various types of data required for system operation; The external system layer is used to achieve integration and collaborative operation with external monitoring systems, industrial servers, laboratory equipment, and enterprise certification systems.

[0007] Furthermore, the voice interaction layer includes a voice input device and a voice output device, and the gateway and access layer includes at least one of an API gateway, a load balancer, a voice stream processing gateway, and a security authentication gateway.

[0008] Furthermore, the intelligent service module includes a voice processing service cluster, an AI model service cluster, and a decision and control service. The voice processing service cluster is used to recognize user voice commands. The business logic module includes an experiment management service, a user and session management service, and a query and retrieval service. The industrial control module includes a protocol adaptation service and an equipment control service.

[0009] Furthermore, the speech recognition algorithm integrated in the speech processing service cluster is the Whisper model.

[0010] Furthermore, the data service layer includes a relational database cluster, a time-series database cluster, and a vector database; the external system layer includes a WinCC monitoring system and an OPC UA server cluster.

[0011] Furthermore, the system also includes a security architecture for performing at least one of the following security measures: voice command authorization verification, industrial control operation auditing, and data transmission encryption.

[0012] According to another aspect of the present invention, an experimental intelligent assistant method based on artificial intelligence is provided; This AI-based experimental intelligent assistant method includes the following steps: Receive voice commands input by the user through the voice interaction layer; The voice commands are routed to the core service layer via the gateway and access layer. The intelligent service module of the core service layer recognizes and understands the voice commands and determines the type of operation to be performed. According to the operation type, the business logic module or industrial control module of the core service layer is invoked to perform the corresponding experimental control, information query or equipment control operation. The data service layer stores and manages the data generated during the execution process. It interacts with corresponding systems in the external system layer to complete device control, data acquisition, or identity authentication operations; The operation results are fed back to the user through the voice interaction layer.

[0013] Furthermore, the steps for recognizing and understanding voice commands include: recognizing the voice commands using the Whisper model and converting them into text commands; and performing natural language understanding on the text commands using an AI model service cluster to parse the user's intent.

[0014] Furthermore, the operations performed by calling the business logic module include at least one of experiment start, stop, status query, experiment step query, or knowledge base Q&A.

[0015] Furthermore, the operation performed by calling the industrial control module includes: converting control commands into industrial communication protocol commands through a protocol adaptation service, and sending them to external industrial automation equipment for execution through a device control service.

[0016] The beneficial effects of this invention are as follows: by constructing a five-layer discrete architecture and integrating speech recognition, natural language understanding and multimodal information fusion technologies, the experimental operation becomes more natural and convenient, and the system has the ability to understand complex instructions and make intelligent decisions; thereby realizing intelligent management of the entire experimental process, significantly improving experimental efficiency and operational safety, and enabling collaboration with industrial equipment and information systems to build an efficient and unified intelligent experimental platform. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a system architecture diagram of an artificial intelligence-based experimental intelligent assistant according to an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0020] like Figure 1 As shown in the embodiment of the present invention, an experimental intelligent assistant based on artificial intelligence includes a voice interaction layer, a gateway and access layer, a core service layer, a data service layer, and an external system layer. The voice interaction layer is used to enable interaction between the user and the system; The gateway and access layer are used to achieve unified access and secure access to the system. The core service layer includes an intelligent service module, a business logic module, and an industrial control module; the intelligent service module is used to implement speech recognition and natural language understanding; the business logic module is used to implement experimental process control and information query; and the industrial control module is used to implement collaborative control and data acquisition with external industrial automation equipment. The data service layer is used to store and manage various types of data required for system operation; The external system layer is used to achieve integration and collaborative operation with external monitoring systems, industrial servers, laboratory equipment, and enterprise certification systems.

[0021] According to an embodiment of the present invention, an artificial intelligence-based experimental intelligent assistant is provided. In a specific implementation, the voice interaction layer includes a voice input device and a voice output device, and the gateway and access layer includes at least one of an API gateway, a load balancer, a voice stream processing gateway, and a security authentication gateway.

[0022] According to an embodiment of the present invention, an artificial intelligence-based experimental intelligent assistant is provided. In a specific implementation, the intelligent service module includes a voice processing service cluster, an AI model service cluster, and a decision and control service. The voice processing service cluster is used to recognize user voice commands. The business logic module includes an experimental management service, a user and session management service, and a query and retrieval service. The industrial control module includes a protocol adaptation service and a device control service.

[0023] According to an embodiment of the present invention, an artificial intelligence-based experimental intelligent assistant is provided, in a specific implementation, wherein the speech recognition algorithm integrated in the speech processing service cluster is the Whisper model.

[0024] According to an embodiment of the present invention, an artificial intelligence-based experimental intelligent assistant is provided. In a specific implementation, the data service layer includes a relational database cluster, a time-series database cluster, and a vector database; the external system layer includes a WinCC monitoring system and an OPC UA server cluster.

[0025] According to an embodiment of the present invention, an artificial intelligence-based experimental intelligent assistant is provided. In a specific implementation, the system further includes a security architecture for performing at least one security measure among voice command permission verification, industrial control operation auditing, and data transmission encryption.

[0026] Secondly, according to an embodiment of the present invention, an experimental intelligent assistant method based on artificial intelligence includes the following steps: Receive voice commands input by the user through the voice interaction layer; The voice commands are routed to the core service layer via the gateway and access layer. The intelligent service module of the core service layer recognizes and understands the voice commands and determines the type of operation to be performed. According to the operation type, the business logic module or industrial control module of the core service layer is invoked to perform the corresponding experimental control, information query or equipment control operation. The data service layer stores and manages the data generated during the execution process. It interacts with corresponding systems in the external system layer to complete device control, data acquisition, or identity authentication operations; The operation results are fed back to the user through the voice interaction layer.

[0027] According to an embodiment of the present invention, an experimental intelligent assistant method based on artificial intelligence, in a specific implementation, the step of recognizing and understanding voice commands includes: recognizing the voice commands using a Whisper model and converting them into text commands; and performing natural language understanding on the text commands using an AI model service cluster to parse the user's intent.

[0028] According to an embodiment of the present invention, an artificial intelligence-based experimental intelligent assistant method is provided. In a specific implementation, the operation performed by calling the business logic module includes at least one of experiment start, stop, status query, experiment step query, or knowledge base question and answer.

[0029] According to an embodiment of the present invention, an artificial intelligence-based experimental intelligent assistant method, in a specific implementation, the operation of calling the industrial control module includes: converting control commands into industrial communication protocol commands through a protocol adaptation service, and sending them to external industrial automation equipment for execution through a device control service.

[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments. It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of protection of this invention.

[0031] Example 1 This embodiment provides an artificial intelligence-based experimental intelligent assistant, which adopts a five-layer split architecture, specifically including a voice interaction layer, a gateway and access layer, a core service layer, a data service layer, and an external system layer.

[0032] The voice interaction layer enables user interaction with the system and includes a voice input device, a voice output device, a web control interface, a mobile application, and a physical control panel. The voice input device employs an eight-microphone array with a spacing of 20 cm, a sampling frequency of 16 kHz, and a quantization bit depth of 16 bits. The voice output device consists of four full-range speakers with a maximum output power of 5 watts. The web control interface is implemented using HTML5 technology and supports cross-browser access; the mobile application is developed based on Android and iOS platforms and supports speech recognition and speech synthesis functions.

[0033] The gateway and access layer are used to achieve unified access and secure access to the system, and include an API gateway, load balancer, voice stream processing gateway, WebSocket connection manager, and security authentication gateway. The API gateway adopts a RESTful architecture and supports data interaction in JSON format. The load balancer uses a round-robin strategy to allocate service resources selectively. The voice stream processing gateway uses WebRTC to achieve real-time transmission of audio data. The WebSocket connection manager is used to establish and maintain real-time communication connections. The security authentication gateway supports the OAuth 2.0 protocol to ensure the security of user identities.

[0034] The core service layer comprises an intelligent service module, a business logic module, and an industrial control module. The intelligent service module implements speech recognition and natural language understanding, including a speech processing service cluster, an AI model service cluster, and a decision-making and control service. Addressing common issues in laboratory environments such as equipment noise, low-frequency vibration, and interference from multiple voices, this system introduces an environmental noise suppression module and a domain terminology enhancement recognition mechanism into the speech processing service cluster, building upon the Whisper model. Incremental training on audio samples from experimental scenarios improves the recognition accuracy of specific commands to 98.5%. The AI ​​model service cluster uses the Ollam large model with an inference speed of 10 times per second. The decision-making and control service includes an experimental process decision engine, an anomaly handling decision-maker, and an intelligent recommendation engine. The business logic module implements experimental process control and information retrieval, including an experimental management service, a user and session management service, and a query and retrieval service. The experimental management service, developed using Python-Django, enables visual management of the experimental process. The user and session management module uses a Redis cluster to store session states. The query and retrieval service utilizes Elasticsearch for efficient text retrieval. The industrial control module is used to achieve collaborative control and data acquisition with external industrial automation equipment. It includes protocol adaptation services and equipment control services. The protocol adaptation service supports Modbus TCP and OPC UA protocols. The equipment control service uses a PLC as middleware to achieve precise control of the experimental equipment.

[0035] The data service layer stores and manages various types of data required for system operation, including a relational database cluster, a time-series database cluster, a vector database, a cache cluster, and file storage services. The relational database cluster uses a MySQL master-slave replication architecture, supporting high-concurrency transaction processing. The time-series database cluster uses InfluxDB to store real-time experimental data. The vector database uses Milvus to store feature vectors of the experimental data. The cache cluster uses a Redis cluster to cache frequently accessed data. The file storage service implements efficient object storage through MinIO.

[0036] The external system layer is used to integrate and collaborate with external monitoring systems, industrial servers, laboratory equipment, and enterprise authentication systems. It includes the WinCC monitoring system, OPC UA server cluster, laboratory equipment network, and enterprise LDAP / AD authentication system. The WinCC monitoring system is implemented using Siemens S7-1200 series PLCs. The OPC UA server cluster adopts an open UA framework to achieve real-time data acquisition. The laboratory equipment network includes PLCs, sensors, and actuators, enabling full monitoring of the experimental process. The enterprise LDAP / AD authentication system uses Active Directory for centralized user management.

[0037] Example 2 This embodiment provides another AI-based experimental intelligent assistant, which also adopts a five-layer split architecture, including a voice interaction layer, a gateway and access layer, a core service layer, a data service layer, and an external system layer.

[0038] The voice interaction layer includes a voice input device, a voice output device, a web control interface, a mobile application, and a physical control panel. The voice input device uses a four-microphone array with a spacing of 15 cm, a sampling frequency of 48 kHz, and a quantization bit depth of 24 bits. The voice output device consists of two high-fidelity speakers with a maximum output power of 3 watts. The web control interface is implemented using the Vue.js framework and supports cross-browser access; the mobile application is developed based on Flutter and supports speech recognition and speech synthesis functions.

[0039] The gateway and access layer includes an API gateway, load balancer, voice stream processing gateway, WebSocket connection manager, and security authentication gateway. The API gateway uses the gRPC framework and supports data exchange in Protocol Buffers format. The load balancer employs a weighted random selection strategy to dynamically adjust service resource allocation. The voice stream processing gateway transmits audio data via the RTP protocol. The WebSocket connection manager is used to establish and maintain real-time communication connections. The security authentication gateway supports the JWT protocol to ensure user identity security.

[0040] The core service layer includes an intelligent service module, a business logic module, and an industrial control module. The intelligent service module includes a voice processing service cluster, an AI model service cluster, and a decision-making and control service. In this embodiment, the voice processing service cluster integrates the Kaldi speech recognition algorithm, achieving an accuracy rate of over 98%. The AI ​​model service cluster uses the Bert-large model, with an inference speed of 20 times / second. The decision-making and control service includes an experimental process decision engine, an anomaly handling decision engine, and an intelligent recommendation engine. Furthermore, the system has a built-in experimental process self-learning module that automatically identifies common operation sequences by analyzing historical experimental data and operation logs, generates experimental process templates, and supports context awareness and adaptive process jumps for voice commands, such as automatically loading corresponding parameters in the 'repeat the previous experiment' command. The business logic module includes an experiment management service, a user and session management service, and a query and retrieval service. The experiment management service is developed using Node.js to achieve visual management of the experimental process. The user and session management module uses MongoDB to store session states. The query and retrieval service uses Solr for efficient text retrieval. The industrial control module includes a protocol adaptation service and an equipment control service. Protocol adaptation services support the EtherCAT protocol. Device control services are available through the Siemens S7-200 series.

[0041] PLC implementation enables precise control of experimental equipment.

[0042] The data service layer includes a relational database cluster, a time-series database cluster, a vector database, a cache cluster, and file storage services. The relational database cluster uses a PostgreSQL master-slave replication architecture, supporting high-concurrency transaction processing. The time-series database cluster uses InfluxDB to store real-time experimental data. The vector database uses Milvus to store feature vectors of the experimental data. The cache cluster uses a Redis cluster to cache frequently accessed data. File storage services utilize Alibaba Cloud OSS for efficient object storage.

[0043] The external system layer includes the WinCC monitoring system, OPC UA server cluster, laboratory equipment network, and enterprise LDAP / AD authentication system. The WinCC monitoring system is implemented using Siemens SIMATIC S7-1500 series PLCs. The OPC UA server cluster adopts an open UA framework to achieve real-time data acquisition. The laboratory equipment network includes PLCs, sensors, and actuators, enabling end-to-end monitoring of the experimental process. The enterprise LDAP / AD authentication system uses Microsoft AD for centralized user management.

[0044] In summary, by utilizing the technical solutions described above in this invention, and by constructing a five-layer discrete architecture and integrating speech recognition, natural language understanding, and multimodal information fusion technologies, experimental operations become more natural and convenient, and the system possesses the ability to understand complex instructions and make intelligent decisions. This enables intelligent management of the entire experimental process, significantly improves experimental efficiency and operational safety, and facilitates collaboration with industrial equipment and information systems, thereby constructing a highly efficient and unified intelligent experimental platform.

[0045] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An artificial intelligence-based experimental intelligent assistant, characterized in that, It includes a voice interaction layer, a gateway and access layer, a core service layer, a data service layer, and an external system layer; The voice interaction layer is used to enable interaction between the user and the system; The gateway and access layer are used to achieve unified access and secure access to the system. The core service layer includes an intelligent service module, a business logic module, and an industrial control module; the intelligent service module is used to implement speech recognition and natural language understanding; the business logic module is used to implement experimental process control and information query; and the industrial control module is used to implement collaborative control and data acquisition with external industrial automation equipment. The data service layer is used to store and manage various types of data required for system operation; The external system layer is used to achieve integration and collaborative operation with external monitoring systems, industrial servers, laboratory equipment, and enterprise certification systems.

2. The artificial intelligence-based experimental intelligent assistant according to claim 1, characterized in that, The voice interaction layer includes a voice input device and a voice output device, and the gateway and access layer includes at least one of an API gateway, a load balancer, a voice stream processing gateway, and a security authentication gateway.

3. The artificial intelligence-based experimental intelligent assistant according to claim 1, characterized in that, The intelligent service module includes a voice processing service cluster, an AI model service cluster, and a decision and control service. The voice processing service cluster is used to recognize user voice commands. The business logic module includes an experiment management service, a user and session management service, and a query and retrieval service. The industrial control module includes a protocol adaptation service and an equipment control service.

4. The artificial intelligence-based experimental intelligent assistant according to claim 3, characterized in that, The speech recognition algorithm integrated in the speech processing service cluster is the Whisper model.

5. The artificial intelligence-based experimental intelligent assistant according to claim 1, characterized in that, The data service layer includes a relational database cluster, a time-series database cluster, and a vector database; the external system layer includes a WinCC monitoring system and an OPC UA server cluster.

6. The artificial intelligence-based experimental intelligent assistant according to claim 1, characterized in that, The system also includes a security architecture for performing at least one of the following security measures: voice command authorization verification, industrial control operation auditing, and data transmission encryption.

7. A method for an experimental intelligent assistant based on any one of claims 1 to 6, characterized in that, Includes the following steps: Receive voice commands input by the user through the voice interaction layer; The voice commands are routed to the core service layer via the gateway and access layer. The intelligent service module of the core service layer recognizes and understands the voice commands and determines the type of operation to be performed. According to the operation type, the business logic module or industrial control module of the core service layer is invoked to perform the corresponding experimental control, information query or equipment control operation. The data service layer stores and manages the data generated during the execution process. It interacts with corresponding systems in the external system layer to complete device control, data acquisition, or identity authentication operations; The operation results are fed back to the user through the voice interaction layer.

8. The method for an artificial intelligence-based experimental intelligent assistant according to claim 7, characterized in that, The steps for recognizing and understanding voice commands include: recognizing the voice commands using the Whisper model and converting them into text commands; and performing natural language understanding on the text commands using an AI model service cluster to parse the user's intent.

9. The method for an artificial intelligence-based experimental intelligent assistant according to claim 7, characterized in that, The operations performed by calling the business logic module include at least one of the following: starting the experiment, stopping the experiment, querying the status, querying the experiment steps, or answering questions in the knowledge base.

10. A method for an artificial intelligence-based experimental intelligent assistant according to claim 7, characterized in that, The operations performed by calling the industrial control module include: converting control commands into industrial communication protocol commands through a protocol adaptation service, and sending them to external industrial automation equipment for execution through a device control service.

Images