Coal mine dispatching voice information storage based on alliance chain and large language model

Abstract

The invention discloses coal mine dispatching voice information storage based on an alliance chain and a large language model, which comprises two information processing modules, and the two information processing modules are a voice information processing module and a data encryption and storage module based on an alliance block chain; the voice information processing module comprises a fine-tuned Wi sper model, a fine-tuned coal mine large language model and a corresponding abstract text, and the data encryption and storage module based on the alliance block chain further comprises a data review and traceability module. And the data review and traceability module comprises rule and standard formulation, data review, data exception processing, review result recording and data traceability. Compared with the prior art, the method has the advantages that the requirement of a storage space is remarkably reduced in a storage mode of uploading the abstract and the hash to the alliance chain, so that the storage cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of information storage, specifically to the storage of coal mine dispatching voice information based on the consortium blockchain and large language models. Background Art

[0002] In the underground coal mine environment, sensors and devices constitute an Internet of Things system, generating a large amount of information, and the storage and management of information become key issues. The central control room plays an important role in mining safety, production, and management. It highly integrates various information from different sensors in various parts of the coal mine underground. Traditionally, mining information is stored in the central control room for a certain period of time. For example, according to the "Coal Mine Safety Technology" regulations, dispatching voice information should be saved for 3 months. The dispatching of coal mine underground voice is a bridge for communication between the central control room and the underground working face. The dispatching process will exchange mining production, safety, and equipment information. Therefore, the information contained in the dispatching process is crucial for the risk prediction and accident traceability of the entire mine. The dispatching voice is also an important part of the mining Internet of Things system. However, this local information storage method has problems such as security risks, information opacity, large storage space occupation, and difficulty in tracing problems.

[0003] The patents in the prior art have the following disadvantages:

[0004] Regarding the data storage method and privacy protection method, after converting the voice information into text, it is directly stored in the blockchain. This approach reduces the occupation of storage space to a certain extent, but still has problems such as high storage cost and low efficiency. There may be a competitive relationship between different coal mine enterprises, and the original text usually contains sensitive information. Although directly storing in the blockchain can ensure the immutability and integrity of the data, there is a risk of privacy leakage.

[0005] Regarding the traceability of data, the original text contains a large amount of text content converted from voice, including a large number of long and complex details and background information. This makes it time-consuming and laborious for the regulatory authorities to screen and analyze when looking for information directly related to the accident, and it is difficult to trace, which is not conducive to the regulatory authorities to investigate the cause of the accident.

[0006] Regarding the transparency of data, since the text transcribed from voice information is usually unstructured and contains a large amount of redundant information, it is difficult to integrate and analyze the data, and it is difficult to form a comprehensive and accurate overview of the production situation. The regulatory authorities lack clear and unified standards and bases during the review process, resulting in subjectivity and uncertainty in the review results. At the same time, due to the low efficiency problem, the review results of the regulatory authorities may not be publicly disclosed in a timely and comprehensive manner, resulting in insufficient understanding of the review results by coal mine enterprises and the public. Summary of the Invention

[0007] The technical problem to be solved by the present invention is to overcome the above technical defects.

[0008] To solve the above problems, the technical solution of the present invention is: Coal mine dispatching voice information storage based on the alliance chain and large language model, including two information processing modules, which are the voice information processing module and the data encryption and storage module based on the alliance blockchain;

[0009] The voice information processing module includes a fine-tuned Wi sper model, a fine-tuned coal mine large language model, and corresponding abstract texts. An encoder is electrically connected to the fine-tuned Wi sper model, and a decoder is connected to the encoder for corresponding settings, for voice information processing;

[0010] The data encryption and storage module based on the alliance blockchain includes an application private data set and an alliance chain distributed ledger storage;

[0011] The alliance chain distributed ledger storage includes a certificate issuing agency, which authenticates and authorizes coal mine organizations and regulatory organizations. The coal mine organizations are connected to the alliance chain for storage. The alliance chain includes preprocessed data, a coal mine private data set, a consensus mechanism, and a public ledger. The alliance chain can provide data for the regulatory organization to read;

[0012] The data encryption and storage module based on the alliance blockchain further includes a data review and traceability module, which includes rule and standard formulation, data review, data exception handling, review result recording, and data traceability;

[0013] The rule and standard formulation needs to analyze the specific requirements and objectives of supervision, determine the data types for data review and the key indicators for review, and formulate specific standards for data review according to the requirements, including regulations on data format, content, scope, timestamp, etc. The RAFT consensus mechanism is used to verify the consistency and integrity of the data to ensure the accuracy and reliability of the data review;

[0014] The data review combines data format verification, time range verification, and a specific calculation formula to evaluate the gap score between the data and the standards formulated by the regulatory department;

[0015] The data exception handling is to call other services or systems to perform necessary exception handling operations, such as sending notice forms or emails, activating alarms, and generating corresponding review processing records. These records detail key information such as exception information, the coal mine where the data comes from, and the processing results, and are uploaded to the alliance chain;

[0016] The review result record covers data compliance, anomalies, and handling results. Each coal mine member on the consortium blockchain can access the regulatory audit information on the chain to learn about their audit situation and handling results;

[0017] Through inputting this traceability code, the data traceability can query traceability information such as the voice information and production records of the coal mine uploaded to the chain in the consortium blockchain system, and relevant information can be input and packaged into the consortium blockchain. As a peer entity of each coal mine organization, the regulatory department can quickly locate the source of the problem and take corresponding measures through traceability queries.

[0018] Furthermore, the distributed ledger storage of the consortium blockchain adopts hash operation to generate a unique hash value of the original text. The hash value, as the unique digital fingerprint of the original data, can verify whether the original data has been modified, thus effectively preventing data tampering.

[0019] Furthermore, the regulatory department uses the stored hash value to verify the integrity of the data, ensuring the authenticity and reliability of the data. This data storage and review mechanism not only speeds up the generation speed of coal mine scheduling information but also greatly enhances the security and credibility of the data.

[0020] The advantages of the present invention compared with the existing technology are as follows:

[0021] Through the storage method of uploading to the consortium blockchain by hash, the present invention significantly reduces the demand for storage space, thereby reducing storage costs. At the same time, using the hash value for data integrity verification effectively prevents data from being tampered with, ensuring the security and reliability of the data. In addition, the application of the privacy data isolation technology protects the business secrets of coal mine enterprises and enhances the privacy protection. The hash traceability function enables the traceability of data sources, provides convenience for the regulatory department, and improves the efficiency of problem traceability. Finally, the introduction of the regulatory review mechanism ensures the authenticity of the data and enables transparent supervision. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 is the structural diagram of the voice information processing module for storing coal mine scheduling voice information based on the consortium blockchain and large language model of the present invention.

[0023] Figure 2 is the data upload and storage diagram for storing coal mine scheduling voice information based on the consortium blockchain and large language model of the present invention.

[0024] Figure 3 is the flow chart of the review and traceability for storing coal mine scheduling voice information based on the consortium blockchain and large language model of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0025] The following further illustrates the detailed implementation manners of the present invention with reference to the drawings.

[0026] To make the content of the present invention easier to be clearly understood, the following will, in conjunction with the accompanying drawings in the embodiments of the present invention, clearly and completely describe the technical solutions in the embodiments of the present invention.

[0027] As Figures 1 to 3 shown, the storage of coal mine dispatching voice information based on the consortium blockchain and large language model includes two information processing modules, and the two information processing modules are a voice information processing module and a data encryption and storage module based on the consortium blockchain;

[0028] The voice information processing module includes a fine-tuned Wi sper model, a fine-tuned coal mine large language model, and corresponding abstract texts. An encoder is electrically connected to the fine-tuned Wi sper model, and a corresponding decoder is connected to the encoder for voice information processing;

[0029] The data encryption and storage module based on the consortium blockchain includes an application private data set and a consortium chain distributed ledger storage;

[0030] The consortium chain distributed ledger storage includes a certificate issuing agency, which authenticates and authorizes a coal mine organization and a supervision organization. The coal mine organization is connected to and stores data on the consortium chain. The consortium chain includes pre-processed data, a coal mine private data set, a consensus mechanism, and a public ledger. The consortium chain can provide data for the supervision organization to read;

[0031] The data encryption and storage module based on the consortium blockchain further includes a data review and traceability module, and the data review and traceability module includes rule and standard formulation, data review, data anomaly handling, review result recording, and data traceability;

[0032] The rule and standard formulation requires analyzing the specific requirements and objectives of supervision, determining the data types for data review and the key indicators for review, and formulating specific standards for data review according to the requirements, including regulations on data format, content, scope, timestamp, etc. The RAFT consensus mechanism is used to verify the consistency and integrity of the data to ensure the accuracy and reliability of the data review;

[0033] The data review combines data format verification, time range verification, and a specific calculation formula to evaluate the gap score between the data and the standards formulated by the supervision department;

[0034] The data anomaly handling is to call other services or systems to perform necessary anomaly handling operations, such as sending notice forms or emails, activating alarms, and generating corresponding review processing records. These records detail key information such as anomaly information, the coal mine where the data comes from, and the processing results, and are uploaded to the consortium chain;

[0035] The review result record covers data compliance, anomalies, and handling results. Each coal mine member on the alliance chain can access the regulatory review information on the chain to learn about their review situation and handling results;

[0036] Through the input of this traceability code, the data traceability can query traceability information such as the voice information and production records of the coal mine uploaded to the chain in the alliance chain system, and relevant information can be input and packaged into the alliance chain. As a peer entity of each coal mine organization, the regulatory department can quickly locate the source of the problem and take corresponding measures through traceability queries.

[0037] The distributed ledger storage of the alliance chain uses hash operations to generate a unique hash value for the original text. The hash value, as the unique digital fingerprint of the original data, can verify whether the original data has been modified, thus effectively preventing data tampering.

[0038] The regulatory department uses the stored hash value to verify the integrity of the data and ensure the authenticity and reliability of the data. This data storage and review mechanism not only speeds up the generation of coal mine scheduling information but also greatly enhances the security and credibility of the data.

[0039] In specific use, the storage and utilization of coal mine scheduling voice information based on the alliance chain and large language models utilize voice information processing, integrating Whisper ASR and fine-tuned large language models. An adjusted Whisper automatic speech recognition (ASR) model is used for speech-to-text conversion, and a fine-tuned coal mine large language model is combined to generate a text summary, thereby achieving a more accurate and refined scheduling information extraction effect.

[0040] Customized adjustment for the coal mine environment: For the specific coal mine environment, we have specifically optimized the Whisper ASR model and the large language model (LLM) to enable them to better understand and process professional terms and local dialects in the coal mine field, thereby improving the accuracy of recognition and summarization.

[0041] Data protection and storage mechanism based on the alliance blockchain: Application of private data sets: Using the private data set function of Hyperledger Fabric, independent data storage spaces are established for each coal mine enterprise to ensure the isolation and security of data among enterprises.

[0042] Hash verification storage: Record the hash value of the original text on the public ledger of the alliance blockchain. With the uniqueness of the hash value, verify the data integrity, effectively prevent data tampering, and endow the data with traceability.

[0043] Data review and traceability process: Introduce smart contracts to execute the data review process, and record the review results and process information on the consortium blockchain to achieve regulatory transparency. At the same time, through the traceability code, the source of the problem can be quickly and accurately traced.

[0044] System architecture: Modular construction: The system is divided into two major modules, namely voice information processing and data encryption and storage of the consortium blockchain, which is conducive to the development and subsequent maintenance of the system.

[0045] Consortium blockchain framework: Adopt the consortium blockchain technology framework, which not only realizes the secure isolation storage of data, but also ensures data integrity verification, traceability and transparency, and can meet the strict requirements of the coal mining industry for data management.

[0046] The above describes the present invention and its implementation manners. Such a description is not restrictive, and the actual structure is not limited thereto. Generally speaking, if those of ordinary skill in the art are inspired by it and design similar structural manners and embodiments without creative efforts without departing from the purpose of the present invention, they shall fall within the protection scope of the present invention.

Claims

1. Coal mine dispatching voice information storage based on consortium blockchain and large language models, characterized in that: It includes two information processing modules, which are a voice information processing module and a data encryption and storage module based on the alliance blockchain; The speech information processing module includes a fine-tuned Wisper model, a fine-tuned coal mine language model and a corresponding summary text, the fine-tuned Wisper model is electrically connected to an encoder, and the encoder is connected to a corresponding decoder for speech information processing; The data encryption and storage module based on the alliance blockchain includes application private data sets and alliance chain distributed ledger storage; The alliance chain distributed ledger storage includes a certificate issuing agency, whose certification authorization includes coal mine organizations and regulatory organizations. The coal mine organization is connected to the alliance chain for storage. The alliance chain includes pre-processed data, coal mine private data collection, consensus mechanism and public ledger. The alliance chain can provide regulatory organizations with data reading; The data encryption and storage module based on the alliance blockchain also includes a data review and traceability module, which includes rule and standard formulation, data review, data exception processing, review result recording and data tracing; The formulation of the rules and standards mentioned above requires analyzing the specific needs and objectives of supervision, determining the data types and key indicators of data review, and formulating specific standards for data review based on the needs, including regulations on data format, content, scope, timestamp, etc. The RAFT consensus mechanism is used to verify the consistency and integrity of data to ensure the accuracy and reliability of data review; The data review combines data format validation, time range validation, and a specific calculation formula to assess the gap score between the data and the standards set by the regulatory authorities; The data exception processing is to call other services or systems to perform necessary exception processing operations, such as sending notifications or emails, activating alarms, and generating corresponding review and processing records. These records record key information such as exception information, the coal mine where the data comes from, and processing results in detail, and are uploaded to the alliance chain; The review result records cover data compliance, abnormal situations and processing results. Each coal mine member of the alliance chain can access the regulatory review information on the chain to learn about its review status and processing results; By inputting this traceability code, the data traceability can query the voice information, production records and other traceability information uploaded to the chain by the coal mine in the alliance chain system, and can input relevant information and package it into the alliance chain. The regulatory department, as a peer entity of each coal mine organization, can quickly locate the source of the problem and take corresponding measures through traceability query.

2. The coal mine dispatching voice information storage based on the alliance chain and the large language model according to claim 1, characterized in that: The alliance chain distributed ledger storage adopts hash operation to generate a unique hash value of the original text. The hash value serves as the unique digital fingerprint of the original data, which can verify whether the original data has been modified, thereby effectively preventing data tampering.

3. The coal mine dispatching voice information storage based on the consortium blockchain and the large language model according to claim 1, characterized in that: The regulatory authorities use the stored hash values to verify the integrity of the data and ensure the authenticity and reliability of the data. This data storage and review mechanism not only speeds up the generation of coal mine dispatching information, but also greatly enhances the security and credibility of the data.

Images