Search

CN-121998245-A - Intelligent laboratory full-flow control method based on multi-mode sensing and strict control logic

CN121998245ACN 121998245 ACN121998245 ACN 121998245ACN-121998245-A

Abstract

The invention relates to a full-flow control method of an intelligent laboratory based on multi-mode sensing and strict control logic, which is used for realizing instrument reservation, reagent full-life-cycle management, room intelligent control, environment and behavior real-time monitoring and deep fusion and data linkage of abnormal causality reasoning by constructing a unified business logic center and multi-mode sensing and strict control logic closed loop, effectively eliminating the defects of system fracture, information island and management passivity in the prior art, actively identifying intent deviation and causality abnormality in the full flow of the experiment by the system, triggering hierarchical intervention measures, effectively improving resource collaborative scheduling capability, real-time of safety admission control and overall operation and maintenance efficiency of the laboratory, remarkably reducing management risk and manual intervention cost, and meeting the requirements of modern laboratories on intelligent and closed-loop safety control.

Inventors

  • HE XIAOFEI

Assignees

  • 江苏洛达米尔智能科技有限公司

Dates

Publication Date
20260508
Application Date
20260122

Claims (10)

  1. 1. A full-flow control method for intelligent laboratory based on multi-mode sensing and strict control logic is characterized by comprising the following steps of, Responding to a user to initiate a reservation application for laboratory resources, generating an experimental intention abstract token comprising an experimental expected range and constraint rules after approval passes, and issuing the experimental intention abstract token to an edge computing node and a corresponding control terminal; responding to the approach of the user to the target resource, executing multi-mode identity verification and position presence verification, and updating a resource state machine and unlocking the resource only when all verification results are passed; in the process of experiment, constructing and dynamically updating a causal relation model in real time based on the acquired multi-source heterogeneous sensing data, and checking the intention similarity and causal consistency of the current operation state by combining an experiment intention abstract token; When an intent deviation or causal abnormality is detected, triggering hierarchical active intervention according to the severity of the abnormality until the experiment is finished and data asset compliance archiving and overall process log generation are completed.
  2. 2. The intelligent laboratory full-flow control method based on multi-modal awareness and stringent control logic of claim 1, wherein generating the experimental intent summary token comprising experimental intent ranges and constraint rules comprises, Extracting experimental objective description, a reagent leading list and corresponding rules in a preset standard experimental template library from the reserved application; Packaging the extracted content into an experimental intent summary token, the token comprising a range of permitted-to-use devices, reagent combination constraint rules, environmental safety constraints, and expected operational sequence features; And issuing the experimental intention abstract token to an edge computing node and a corresponding instrument or reagent cabinet control terminal to serve as a reference for subsequent real-time verification.
  3. 3. The intelligent laboratory full-flow control method based on multi-modal awareness and severe control logic of claim 2, wherein the performing multi-modal identity verification and location presence verification only when all verification results are passed, updating the resource state machine and unlocking the resource comprises, When the user is perceived to be close to a target resource, biological feature recognition is executed, an infrared image and an RGB image are collected at the same time, and texture depth and reflectivity are calculated through fused image data so as to finish living body detection; After the living body detection is passed, a dynamic interaction instruction is randomly generated and issued, and the matching degree of the facial key point displacement of the user and the instruction is collected and verified; aiming at the high-risk reagent cabinet, double-person double-lock logic is further executed, and biological characteristic information of two different persons with authority is acquired in a preset time window, so that a state machine enters a double-person presence state; and comparing the real-time positioning coordinates of the intelligent tablet with a preset electronic fence area, and transmitting a verification passing signal to a resource state machine to execute unlocking operation only when the coordinates are positioned in an effective operation area.
  4. 4. The intelligent laboratory full-flow control method based on multi-modal sensing and tightly-controlled logic of claim 3, wherein the real-time construction and dynamic updating of causal relationship model based on the collected multi-source heterogeneous sensing data, and the verification of intent similarity and causal consistency of current operation state by combining with experimental intent abstract token comprises, Collecting environment sensor data through an MQTT protocol, collecting video data through a video stream analysis gateway, collecting wearable equipment coordinate data through a positioning gateway, and performing time stamp alignment and space-time fusion on all data streams to generate a real-time digital twin snapshot; Analyzing the video stream frame by utilizing the target detection model, identifying personnel and protective equipment types, calculating the spatial relationship among the detection frames to judge the wearing compliance, and taking the compliance judgment result as the updated input of the causal graph nodes; Constructing and dynamically updating a small causal graph in an edge computing node based on the digital twin snapshot in real time, wherein the causal graph comprises an operation event node, an environment variable node, a reagent state node and an instrument reading node, and dynamically updating a graph structure according to a preset causal dependency relationship; And calculating the similarity between the current operation sequence and the experimental intention abstract token and the deviation degree of actual reading and expected causal relation in the causal graph, and carrying out self-adaptive optimization on the verification threshold and the causal graph structural weight by using historical intervention event data through a lightweight online learning model.
  5. 5. The intelligent laboratory full-flow control method based on multimodal perception and severity logic of claim 4, wherein when an intent deviation or causal anomaly is detected, triggering a hierarchical active intervention based on anomaly severity comprises, When the intention similarity is lower than an expected threshold value or the causal deviation is in a light range, preferentially generating a voice warning signal and a lamplight guiding signal, and collecting multi-source feedback data such as user response time, operation correction track, intention recovery state and the like; The feedback data is statistically analyzed based on a preset sliding time window, the intention correction rate and the abnormal recurrence frequency are calculated, and an adaptability index representing the effectiveness and the interference degree of the current warning strategy is generated; When the abnormality is continuous or the causal deviation is aggravated, generating a device power supply suspension instruction or a reagent cabinet locking instruction based on the updated causal graph and the adjusted intervention strategy, and recording an intervention event to a process log; When the causal graph shows a high risk causal chain, a forced rollback instruction for reversible operation is generated, including resetting the valve state, restoring the initial flow setting or powering off, and returning the rollback execution result to the resource state machine to update the resource state.
  6. 6. The intelligent laboratory full-flow control method based on multi-modal sensing and stringent control logic of claim 1 or 4, wherein for the reservation application containing the reagent use list, further comprising performing list constraint verification and causal risk deduction at reagent tie-up, in particular comprising, Reading the information of the actually taken reagent through RFID, and comparing the information with a pre-approved acceptance list for consistency of the product; based on the comparison result, checking whether the conditions of temperature, humidity, concentration and the like acquired by the environmental sensor in the current cabinet meet the storage requirement of the corresponding reagent in the list; If the environmental conditions meet the requirements, further detecting whether the conditions that the mutually exclusive chemicals in the list are taken out in the same time period exist or not, and generating a conflict signal; Constructing or updating a reagent-related small causal graph by utilizing the conflict signal, deducing a potential risk path, immediately generating and executing an intervention instruction for prohibiting further ex-warehouse, locking a cabinet body or giving an audible and visual alarm according to the risk level of the deduced path; and recording the intervention execution result to the whole process log so as to maintain the consistency of the whole flow data and the state.
  7. 7. The intelligent laboratory full-flow management and control method based on multi-modal awareness and stringent control logic of claim 1, wherein the completion of data asset compliance archiving and full-process log generation comprises, When a user uploads experimental data, checking the time validity and login state of a data access token bound with the reservation in real time; after the verification is passed, writing the experimental data into a storage space which is strongly bound with the reserved record, and recording a writing event into a process log; after the experiment is finished, the resource state machine is automatically restored to an idle state, all tokens and locks generated in the reservation process are released, and a complete process log comprising a multi-mode sensing event record, a causal anomaly detection record, an intervention operation record and a rollback operation record is generated so as to support subsequent audit and tracing.
  8. 8. The intelligent laboratory full-flow control method based on the multi-mode sensing and strict control logic is characterized by comprising an intention abstract generation module, a multi-mode admission verification module, a causal real-time monitoring module and a hierarchical intervention execution module; The intent abstract generation module is used for responding to the request of a user for reservation of laboratory resources, generating an experiment intent abstract token comprising an experiment expected range and constraint rules after approval is passed, and issuing the experiment intent abstract token to an edge computing node and a corresponding control terminal; The multi-mode access verification module is used for responding to the approach of a user to a target resource, executing multi-mode identity verification and position presence verification, and updating a resource state machine and unlocking the resource only when all verification results are passed; The causal real-time monitoring module is used for constructing and dynamically updating a causal relation model in real time based on the acquired multi-source heterogeneous sensing data in the experimental process, and carrying out intention similarity and causal consistency verification on the current operation state by combining an experimental intention abstract token; And the hierarchical intervention execution module is used for triggering hierarchical active intervention according to the severity degree of the abnormality when the intent deviation or causal abnormality is detected, and completing the data asset compliance archiving and the whole process log generation until the experiment is finished.
  9. 9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, and the processor implements the steps of the multi-modal awareness and severe control logic based intelligent laboratory full-flow management method of any one of claims 1-7 when the program is executed on the processor.
  10. 10. A storage medium storing a computer program, wherein the computer program when executed by a processor implements the steps of the intelligent laboratory full-flow management and control method based on multi-modal awareness and severe control logic as set forth in any one of claims 1-7.

Description

Intelligent laboratory full-flow control method based on multi-mode sensing and strict control logic Technical Field The invention belongs to the technical field of intelligent laboratory management, and particularly relates to a full-flow control method of an intelligent laboratory based on multi-mode sensing and strict control logic. Background With the continuous expansion of the scale of higher education and scientific research and the continuous deepening of the innovation of experimental teaching, the laboratory is used as a core place of scientific research and teaching, and the management requirement of the laboratory is accelerated from the traditional paper recording and manual inspection to the digitization and intellectualization directions. In recent years, the promotion of smart campus construction has prompted some laboratories to introduce informationized management systems, such as LIMS (laboratory information management system), entrance guard attendance systems, environmental monitoring platforms, etc., which implement the functions of instrument reservation, data recording and basic security monitoring to some extent. However, the existing laboratory management technology still has the obvious defects that firstly, functional modules are highly fragmented, subsystems (such as instrument reservation, reagent management and control, room scheduling and environment monitoring) are often deployed independently, data island phenomenon is serious, unified business logic middle platform and real-time intercommunication mechanism are lacked, resource scheduling conflict frequently occurs and information updating is delayed, secondly, management means mainly comprise passive recording, deep fusion and intelligent analysis capability on multi-source heterogeneous data are lacked, active early warning and intervention cannot be carried out when high-risk reagent operation, personnel behavior abnormality or environment risk occurs, thirdly, system expansibility and compatibility are poor, hardware interfaces and protocol standards are not unified, multi-manufacturer equipment is difficult to access, later function upgrading and cross-system integration cost is high, and complexity of laboratory personnel, financial, object and multi-dimensional cross management is difficult to adapt. Together, these drawbacks lead to the overall inefficiency and difficulty in real-time closed-loop control of laboratory safety hazards, which have not met the requirements of modern laboratories for high safety, efficient synergy and low risk operation and maintenance. Disclosure of Invention The invention aims to provide an intelligent laboratory full-flow management and control method based on multi-mode sensing and strict control logic, which aims to solve the defects that resource scheduling conflict is frequent, safety access control is lagged and overall operation and maintenance efficiency is low due to the fact that all functional subsystems are independent of each other, data cannot be mutually communicated and shared in real time and management means depend on manual intervention in the existing laboratory management technology. To achieve one of the above objects, one embodiment of the present invention provides a method for intelligent laboratory full-flow control based on multi-modal sensing and stringent control logic, the method comprising, Responding to a user to initiate a reservation application for laboratory resources, generating an experimental intention abstract token comprising an experimental expected range and constraint rules after approval passes, and issuing the experimental intention abstract token to an edge computing node and a corresponding control terminal; responding to the approach of the user to the target resource, executing multi-mode identity verification and position presence verification, and updating a resource state machine and unlocking the resource only when all verification results are passed; in the process of experiment, constructing and dynamically updating a causal relation model in real time based on the acquired multi-source heterogeneous sensing data, and checking the intention similarity and causal consistency of the current operation state by combining an experiment intention abstract token; When an intent deviation or causal abnormality is detected, triggering hierarchical active intervention according to the severity of the abnormality until the experiment is finished and data asset compliance archiving and overall process log generation are completed. As a further refinement of an embodiment of the present invention, the method further comprises, the generating an experimental intent summary token comprising experimental expected ranges and constraint rules comprises, Extracting experimental objective description, a reagent leading list and corresponding rules in a preset standard experimental template library from the reserved application; Packaging the extract