Fire-fighting inspection and acceptance system and acceptance method thereof

Abstract

The invention provides a fire-fighting inspection and acceptance system and an acceptance method thereof, and belongs to the technical field of fire-fighting safety informatization, and the system specifically comprises four core units: an interaction unit, a data unit, an analysis unit and an execution unit. The interaction unit comprises a voiceprint recognition module, a voice instruction receiving module and a touch / gesture interaction module; the interaction unit is used for receiving an instruction of an operator and completing direct or indirect interaction operation between the operator and the system; the data unit comprises a multi-source data storage module, a fire-fighting knowledge graph area and a dynamic template library; the analysis unit comprises an NLP semantic analysis module, a compliance analysis engine and a report generation engine; the execution unit comprises a data display module, an Internet of Things control module and a mobile terminal cooperation module; according to the fire-fighting inspection and acceptance system and the acceptance method thereof, the problems of high manual dependency, standard execution deviation and insufficient traceability of the conventional acceptance mode at present can be effectively solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of fire safety informatization, and particularly relates to a fire inspection and acceptance system and an acceptance method thereof. Background Art

[0002] Fire inspection (fire inspection or fire acceptance) is an important link to ensure that buildings, places or facilities meet fire safety standards, aiming to prevent fires and ensure the safety of people's lives and property. In this process, the management party needs to submit a fire acceptance application to the local fire department and provide specific information of the area to be inspected, such as the panoramic view of the building, design drawings, fire emergency plans, etc.

[0003] Managers also need to conduct functional tests and linkage tests on the fire protection facilities in the inspection site, simulate fire alarms, and verify whether the fire protection facilities are linked according to the preset logic to ensure that the fire protection equipment can meet the basic usage requirements in case of danger.

[0004] Finally, based on the inspection results, the manager issues an inspection report to conduct a comprehensive evaluation of the inspected site.

[0005] Currently, the following core problems exist in traditional fire acceptance: High manual dependence: The inspection relies on paper lists, and report compilation is time-consuming and prone to missing key items.

[0006] Technical fragmentation: The data collection, analysis, and report generation links are isolated, lacking full-link collaboration.

[0007] Deviation in standard execution: Inspectors have inconsistent understandings of specifications (such as GB50116-2022), resulting in strong subjectivity in acceptance results.

[0008] Insufficient traceability: Operation records lack biometric binding, making it difficult to trace responsibilities. Summary of the Invention

[0009] In view of this, the present invention proposes a fire inspection and acceptance system and an acceptance method thereof, which can effectively solve the problems of high manual dependence, deviation in standard execution, and insufficient traceability existing in the current conventional acceptance methods in the background art.

[0010] The present invention is implemented as follows: The present invention proposes a fire inspection and acceptance system, which specifically includes four core units: an interaction unit, a data unit, an analysis unit, and an execution unit; The interaction unit includes a voiceprint recognition module, a voice command receiving module, and a touch / gesture interaction module; the interaction unit is used to receive operator instructions and complete direct or indirect interaction operations between the operator and the system; The data unit includes a multi-source data storage module, a fire protection knowledge graph area, and a dynamic template library; the data unit is used to store the data required for inspection, and at the same time, in cooperation with the interaction unit, retrieve the corresponding data required by the operator; The analysis unit includes an NLP semantic parsing module, a compliance analysis engine, and a report generation engine; the analysis unit is used to analyze and process the voice commands received by the system, convert them into electrical signals that the system can recognize, and the system controls the execution unit to make a response according to this signal; The execution unit includes a data display module, an Internet of Things control module, and a mobile terminal collaboration module; the execution unit is used to control the corresponding module to complete the response action according to the electrical signal sent by the analysis unit after processing.

[0011] Furthermore, the interaction unit is also equipped with a voiceprint recognition engine; using MFCC feature extraction combined with a deep residual network algorithm to process the received sound wave signals, while supporting live detection, ensuring that the misrecognition rate of the system is ≤0.01% and the rejection rate is ≤1.5%, and at the same time, it can endow the system with the ability to resist recording attacks; Among them, the voice command receiving module includes offline keyword recognition, supporting fire protection term correction; The touch / gesture includes a capacitive touch projection interface, supporting gesture zooming and circle selection marking.

[0012] Furthermore, the data unit is involved in the storage and retrieval of data. The content stored is a fire protection knowledge graph. The data structure of the fire protection knowledge graph includes fire protection equipment, building components, and the installation position compliance thresholds between the fire protection equipment and the building components; the data unit also has an automatic update function, directly synchronizing and storing the latest national standards to the local database and replacing the old version of the data; The data unit also includes a dynamic template library. The template types of the dynamic template library include completion acceptance, annual inspection, and temporary spot check; at the same time, the template library can also automatically match inspection items with sensor data fields through NLP parsing.

[0013] Furthermore, the specific content of the analysis unit includes: The analysis unit is equipped with a compliance analysis engine. The compliance analysis engine specifically performs dynamic rule loading based on the Drools rule engine, while ensuring that the analysis unit supports complex logic verification; The analysis unit has a real-time alarm function. When major hidden dangers are detected, it triggers an audible and visual alarm and completes the freezing operation process; Meanwhile, the analysis unit is also equipped with a report generation engine to output the analysis results, associate with the 3D model coordinates, and mark the locations of potential hazard points; the output formats are PDF, Excel, and HTML visualization dashboards; the generated analysis results are synchronously accompanied by compliance scores.

[0014] Beneficial effects of adopting the above further solution: By embedding the Drools rule engine into the fire inspection system, the following can be achieved: Automated compliance inspection, replacing manual item-by-item checking of regulations, and improving the inspection efficiency (e.g., completing the automated preliminary inspection of a single building within 10 minutes); Dynamic adaptability, quickly responding to new regulations (e.g., the fire protection standard for electric vehicle charging facilities in 2024); Interactive experience, reducing the professional threshold through real-time feedback and intelligent suggestions (e.g., assisting non-professional personnel in operation) and enabling more intelligent interaction; Recording the results of each rule execution to meet the requirements of fire audits.

[0015] Furthermore, the specific content of the execution unit includes: Data display module: The data display module performs rapid positioning based on physical marker points, with an overall positioning error within ±3 cm; the system uses a short-throw laser projector with a standard brightness of 4000 lumens and a resolution of 1920×1080. To ensure normal use in environments with high light intensities such as outdoors, the projector also has the ability to resist ambient light interference; The interaction layer of the data display module has the function of superimposing real-time data, and virtual operation buttons are also added to the system for convenient interaction; Internet of Things control module: Utilize the network for linkage self-check and automatically trigger tests during acceptance; according to the trigger results, record the fire protection facilities and update the record results to the report in real time.

[0016] Furthermore, the present invention also provides an acceptance method for a fire inspection and acceptance system, and the specific steps of the acceptance method are as follows: Step S1: Voiceprint feature storage. The administrator starts the voiceprint storage program of the system, inputs multiple groups of audio data of multiple operators into the system. The system receives the audio data, extracts the corresponding voiceprint features, and stores the voiceprint features correspondingly; the voiceprint corresponds to the operator's identity one by one, and the operator's permissions are synchronously corresponding; Step S2: System startup. The operator sends a startup instruction to the system. The voiceprint recognition module in the system interaction unit recognizes and matches the operator and determines the operator's permissions; meanwhile, the system's infrared detection is started to detect whether the operator's location is within 5 meters; if so, the system starts; if not, the system does not respond; Step S3: Complete the inspection process. The operator issues an instruction to the system. The voice instruction receiving module in the interaction unit receives the sound signal, converts it into an electrical signal, and transmits it to the NLP semantic parsing module in the analysis unit. The NLP semantic parsing module parses the semantics of the sound signal and controls the execution unit to make corresponding responses according to the semantics. This process will be repeated multiple times during the inspection process until the inspection process is completed; Step S4: Generate a report after the inspection is completed, and synchronously generate a rectification work order; the inspection report will be automatically stored, and the operator can export and print the inspection report; the rectification work order will be automatically assigned to the rectification person in charge and automatically archived.

[0017] Furthermore, step S3 also includes data annotation and data collection during the inspection process, specifically including automatically retrieving sensor data, projecting and displaying and highlighting abnormal items.

[0018] Furthermore, the inspection report in step S4 includes an identity QR code containing the personal information of the inspector.

[0019] Furthermore, in step S4, the maintenance personnel repair the fire safety problems that occur in the inspection site according to the rectification work order, and overwrite the archive of the system rectification work order after the repair is completed.

[0020] The beneficial effects of the present invention are as follows: Full-process automation: The time-consuming from inspection to report generation is reduced by 70%.

[0021] High-precision compliance: The dynamic rule engine ensures 100% compliance with the latest fire protection standards.

[0022] Strong traceability: The operation log is bound with the voiceprint biometric feature, meeting the legal audit requirements.

[0023] Low-cost deployment: The hardware cost is reduced by 50% compared with similar solutions, adapting to small and medium-sized enterprises. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present invention, and therefore should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0025] Figure 1 It is a schematic diagram of the composition of the fire inspection system; Figure 2 It is a schematic diagram of the process for adding management personnel; Figure 3 It is a logic flow chart of the fire inspection; Figure 4 For fire inspection, the associated diagrams of each engine are provided. Specific implementation manners

[0026] In order to make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0027] As Figure 1 shown, a fire inspection and acceptance system provided by the present invention specifically includes four core units: an interaction unit, a data unit, an analysis unit, and an execution unit; The interaction unit includes a voiceprint recognition module, a voice command receiving module, and a touch / gesture interaction module; the interaction unit uses the voiceprint recognition module to determine the permissions of the interacting personnel; the interaction unit uses the voice command receiving module to receive the commands issued by the interacting personnel; the interaction unit relies on the touch / gesture interaction module to complete the operation interaction between the interacting personnel and the system; The data unit includes a multi-source data storage module, a fire knowledge graph area, and a dynamic template library; the data unit uses the multi-source data storage module to store the data required for inspection, the data unit uses the fire knowledge graph area to synchronize with the fire information promulgated by the state in real time and cover the storage, and the data unit uses the dynamic template library to store and call the fire inspection templates; The analysis unit includes an NLP semantic parsing module, a compliance analysis engine, and a report generation engine; the analysis unit uses the NLP semantic parsing module to analyze and process the command signals such as voice information received by the interaction unit; the analysis unit uses the compliance analysis engine to perform compliance checks on the fire facility data and assist in generating rectification work orders; the analysis unit uses the report generation engine to generate inspection reports; The execution unit includes a data display module, an Internet of Things control module, and a mobile terminal collaboration module; the execution unit uses the data display module to display the data that needs to be compared during the inspection to the inspection personnel or system users, such as drawings, specifications, etc.; the execution unit uses the Internet of Things control module to communicate with the network, combines the network data with the inspection equipment, and assists the acceptance personnel to complete the inspection process; the execution unit uses the mobile terminal collaboration module to send the entire inspection process to the mobile terminal, and at the same time establishes a remote connection, so that the management personnel do not need to go to the inspection site to understand the problems that occur in the entire inspection process and can communicate with the acceptance personnel who are conducting the inspection at any time.

[0028] Specifically, the interaction unit further includes the following contents: The interaction unit is equipped with a voiceprint recognition engine; it uses the MFCC feature extraction + deep residual network (ResNet-34) algorithm to process the received sound wave signals and supports liveness detection (lip movement synchronization verification) at the same time; the system is also equipped with infrared sensing to detect whether there is an operator within a range of 5m; by using the method of multi-faceted synchronous detection, it ensures that the false recognition rate (FAR) of the system ≤ 0.01% and the rejection rate (FRR) ≤ 1.5%, and at the same time can endow the system with the ability to resist recording attacks; The existence of the interaction unit enables the system to have the ability of multi-modal interaction, and the staff can adjust and control the system through voice commands or touch / gesture operations; Voice command receiving module: offline keyword recognition (such as "drawing viewing", "report generation", "fault marking"), supports correction of fire fighting terms; Touch / gesture interaction module: capacitive touch projection interface, supports gesture zooming, circle selection and annotation (accuracy ±2mm).

[0029] The data unit includes the following: The data unit is mainly involved in the storage and retrieval of data. The content of the stored data mainly includes the design drawings of the building; the data structure of the fire fighting knowledge graph area mainly includes fire fighting equipment (fire extinguishers, smoke detectors, etc.), building components (evacuation passages, fire prevention zones), and the compliance threshold of the installation positions between the above parts (such as the pressure range of fire extinguishers is 0.8-1.2MPa); the data unit also has an automatic update function, directly synchronizing the latest national standards (such as the latest revised version of GB50016-2014 "Code for Fire Protection Design of Buildings") to the local database and replacing the old version of the data; The data unit also includes a dynamic template library. The template types in the template library mainly include completion acceptance, annual inspection, and temporary spot check; at the same time, the template library can automatically match inspection items with sensor data fields through NLP parsing (such as "evacuation passage width" - laser rangefinder reading).

[0030] The specific content of the analysis unit includes: The analysis unit is equipped with a compliance analysis engine. The engine specifically performs dynamic rule loading based on the Drools rule engine, and at the same time ensures that the analysis unit supports complex logic verification (such as "if the smoke detector alarms, then the emergency lighting must be activated"); converts fire fighting regulations and industry standards into executable Drools rules; The analysis unit has real-time alarms. When major hidden dangers are detected (such as the water pressure of the fire hydrant < 0.5MPa), it triggers an audible and visual alarm and freezes the operation process; The simultaneous analysis unit is also equipped with a report generation engine to output the analysis results, associate with the 3D model coordinates, and mark the locations of potential hazard points; the output formats are mainly PDF (digitally signed), Excel (structured data), and HTML visualization dashboards; the generated analysis results are synchronously accompanied by compliance scores (such as "qualified rate 92.3%").

[0031] The specific content of the execution unit includes: (1) Data display module: The data display module performs rapid positioning based on physical marker points (QR codes / infrared reflective stickers), and the overall positioning error is within ±3 cm; the system uses a short-focus laser projector with a standard brightness of 4000 lumens and a resolution of 1920×1080. To ensure normal use in environments with high light intensity such as outdoors, the projector also has the ability to resist ambient light interference; The interaction layer of the data display module has the function of superimposing real-time data (pressure, temperature). At the same time, for the convenience of interaction, virtual operation buttons (such as "generate a work order with one click") are also added to the system.

[0032] (2) Internet of Things controller: The execution unit also includes an Internet of Things controller, which uses the network for linkage self-check and automatically triggers tests during acceptance (such as sprinkler system water testing, emergency lighting startup); according to the trigger results, records of fire protection facilities are made (such as "sprinkler response time: 2.1 s, meeting the requirement of ≤3 s"), and the recorded results are updated to the report in real time.

[0033] As Figures 2-4 shown, the specific acceptance method involved in the system is as follows: Step S1: Voiceprint feature storage. The administrator starts the voiceprint storage program of the system, inputs multiple groups of audio data of multiple operators into the system. The system receives the audio data, extracts the corresponding voiceprint features, and stores the voiceprint features correspondingly; the voiceprint corresponds to the operator's identity one by one, and the operator's permissions are synchronized correspondingly; The management personnel input their own voice texts into the system through a voice recording device and make corresponding annotations. The system extracts the voiceprint features in the voice texts through the MFCC (Mel Frequency Cepstral Coefficient) algorithm to determine the number of management personnel and the corresponding voiceprint features for subsequent use; Step S2: System startup. The operator sends a startup instruction to the system. The voiceprint recognition module in the system interaction unit recognizes and matches the operator and determines the operator's permissions; at the same time, the system's infrared detection is started to detect whether the operator's location is within 5 meters; if so, the system starts; if not, the system does not respond; Start the acceptance process. The administrator interacts with the system, uses the voiceprint stored in Step 1 to activate the system, and simultaneously issues a voice command to the system. Administrator's voice command: "Start the monthly inspection of the fire protection facilities in Building B". The system verifies the voiceprint and loads the corresponding inspection template (GB50016-2022).

[0034] Step S3: Complete the inspection process. The operator issues a command to the system. The voice command receiving module in the interaction unit receives the sound signal, converts it into an electrical signal, and transmits it to the NLP semantic parsing module in the analysis unit. The NLP semantic parsing module parses the semantics of the sound signal and controls the execution unit to make corresponding responses according to the semantics. This process will be repeated multiple times during the inspection process until the inspection is completed. The system automatically retrieves sensor data (such as the pressure value of the fire extinguisher), projects and displays it, and highlights the abnormal items. The inspector circles the hidden danger area with a gesture and supplements the description verbally: "The door closer of Fire Door No. 3 is damaged".

[0035] Step S4: Generate a report after the inspection is completed and simultaneously generate a rectification work order. The inspection report will be automatically stored, and the operator can export and print the inspection report. The rectification work order will be automatically assigned to the rectification person in charge and automatically archived.

[0036] The system compares the threshold in the knowledge graph, marks the unqualified items, and generates rectification suggestions (such as "Replace the door closer, refer to the standard GA93-2023"). Output a PDF report, which also includes a voiceprint verification QR code (scanning the code can trace the identity of the operator). Hidden danger heat map: The TOP5 hidden danger points are marked in the 3D model. Rectification work order: The rectification person in charge notifies the maintenance personnel to rectify the facilities according to the specific description in the rectification work order, and uploads the status of the rectified facilities. The rectification person in charge checks the completion of the rectification and makes a mark in the system to complete the process loop. Rectification tracking and loop closure: The maintenance personnel scan the code to update the status, and the system automatically triggers a re-inspection task. After the acceptance is passed, the report is archived and encrypted for storage (AES-256), and the retention period is ≥10 years.

[0037] The above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A fire inspection and acceptance system, characterized in that, The system specifically includes four core units: an interaction unit, a data unit, an analysis unit, and an execution unit; The interaction unit includes a voiceprint recognition module, a voice command receiving module, and a touch / gesture interaction module; the interaction unit is used to receive the operator's commands and complete the direct or indirect interaction operations between the operator and the system; The data unit includes a multi-source data storage module, a fire protection knowledge graph area, and a dynamic template library; the data unit is used to store the data required for inspection, and at the same time cooperate with the interaction unit to retrieve the corresponding data required by the operator; The analysis unit includes an NLP semantic parsing module, a compliance analysis engine, and a report generation engine; the analysis unit is used to analyze and process the voice commands received by the system, convert them into electrical signals that the system can recognize, and the system controls the execution unit to make a response according to this signal; The execution unit includes a data display module, an Internet of Things control module, and a mobile terminal cooperation module; the execution unit is used to control the corresponding module to complete the response action according to the electrical signal processed and sent by the analysis unit.

2. The fire inspection and acceptance system according to claim 1, wherein, The interaction unit is also equipped with a voiceprint recognition engine; using MFCC feature extraction combined with a deep residual network algorithm to process the received sound wave signal, and at the same time supporting live detection, ensuring that the misrecognition rate of the system ≤ 0.01% and the rejection rate ≤ 1.5%, and at the same time being able to endow the system with the ability to resist recording attacks; Among them, the voice command receiving module includes offline keyword recognition and supports correction of fire protection terms; The touch / gesture interaction module includes a capacitive touch projection interface and supports gesture zooming and circle selection marking.

3. A fire inspection and acceptance system according to claim 1, characterized in that, The data unit is involved in the storage and retrieval of data. The content stored is a fire protection knowledge graph. The data structure of the fire protection knowledge graph includes fire protection equipment, building components, and the compliance threshold of the installation positions between the fire protection equipment and the building components; the data unit also has an automatic update function, directly synchronizing and storing the latest national standards to the local database and replacing the old version of the data; The data unit also includes a dynamic template library. The template types of the dynamic template library include completion acceptance, annual inspection, and temporary spot check; at the same time, the template library can also automatically match inspection items with sensor data fields through NLP parsing.

4. A fire inspection and acceptance system according to claim 3, characterized in that, The specific content of the analysis unit includes: The analysis unit is equipped with the compliance analysis engine. The compliance analysis engine specifically realizes dynamic rule loading based on the Drools rule engine, and at the same time ensures that the analysis unit supports complex logic verification; The analysis unit has a real-time alarm function. When major hidden dangers are detected, it triggers an audible and visual alarm and completes the freezing operation process; At the same time, the analysis unit is also equipped with the report generation engine. The report generation engine is used to output the analysis results, associate the 3D model coordinates, and mark the specific positions of the hidden danger points; the output format is PDF, Excel, HTML, and a visualization dashboard; the generated analysis results are synchronously accompanied by a compliance score.

5. A fire inspection and acceptance system according to claim 3, wherein The specific content of the execution unit includes: Data display module: The data display module performs rapid positioning based on physical marker points, and the overall positioning error is within ±3 cm. The system uses a short-throw laser projector with a standard brightness of 4000 lumens and a resolution of 1920×1080. To ensure normal use in environments with high light intensity such as outdoors, the projector also has the ability to resist ambient light interference. The interaction layer of the data display module has the function of superimposing real-time data. At the same time, for the convenience of interaction, virtual operation buttons are also added to the system. Internet of Things control module: Use the network for linkage self-check and automatically trigger the test during acceptance. According to the trigger result, record the fire-fighting facilities and update the record result to the report in real time.

6. A method for inspection and acceptance of a fire inspection and acceptance system, characterized in that, The acceptance method is applicable to a fire inspection and acceptance system described in any one of claims 1-5, and the specific steps of the acceptance method are as follows: Step S1: Voiceprint feature storage. The administrator starts the voiceprint storage program of the system, inputs multiple groups of audio data of multiple operators into the system. The system receives the audio data, extracts the corresponding voiceprint features, and stores the voiceprint features correspondingly. The voiceprint corresponds to the operator's identity one by one, and synchronously corresponds to the operator's permissions. Step S2: System startup. The operator sends a startup instruction to the system. The voiceprint recognition module in the system interaction unit recognizes and matches the operator and determines the operator's permissions. At the same time, the system starts infrared detection to detect whether the operator's location is within 5 meters. If so, the system starts. If not, the system does not respond. Step S3: Complete the inspection process. The operator sends an instruction to the system. The voice instruction receiving module in the interaction unit receives the sound signal, converts it into an electrical signal and transmits it to the NLP semantic parsing module of the analysis unit. The NLP semantic parsing module parses the semantics of the sound signal and controls the execution unit to make corresponding responses according to the semantics. This process will be repeated multiple times during the inspection process until the inspection process is completed. Step S4: Generate a report after the inspection is completed and synchronously generate a rectification work order. The inspection report will be automatically stored, and the operator can export and print the inspection report. The rectification work order will be automatically assigned to the rectification person in charge and automatically archived.

7. The inspection and acceptance method of a fire inspection and acceptance system according to claim 6, characterized in that, The said step S3 also includes data annotation and data collection during the inspection process, specifically including automatically retrieving sensor data, projecting and highlighting abnormal items.

8. The inspection and acceptance method of a fire inspection and acceptance system according to claim 6, characterized in that, The inspection report in the said step S4 includes an identity QR code containing the personal information of the inspector.

9. The inspection and acceptance method of a fire inspection and acceptance system according to claim 6, characterized in that, In the said step S4, the rectification person in charge notifies the specific maintenance personnel to repair the fire safety problems that occur in the inspection site according to the rectification work order. After the repair is completed, the archive of the system rectification work order is overwritten.

Images