Polar expedition cross-border data sharing and verification method and system

Abstract

This invention relates to the field of data processing technology and discloses a method and system for cross-border data sharing and verification in polar scientific expeditions. The method first acquires raw data from polar scientific expeditions, then uses bidirectional data mapping rules to perform format conversion, index calibration, and initial screening of invalid data to obtain standardized data. Next, based on the sensitivity level of the standardized data, encryption is performed using hierarchical encryption rules, and cross-border transmission of the encrypted data is achieved via a polar satellite communication link. After decrypting the received encrypted data, parallel verification is performed using dual-path verification rules to obtain corresponding data verification results. Finally, combining the verification results with hierarchical access rules, corresponding access permissions are configured for the verified data, and cross-border sharing is completed. This invention enables refined and secure sharing of cross-border data from polar scientific expeditions, effectively optimizing the overall processing and application effects of cross-border data in polar scientific expeditions.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method and system for cross-border data sharing and verification in polar scientific expeditions. Background Technology

[0002] With the deepening of international cooperation in polar scientific research, the number of joint polar scientific expeditions conducted by multiple countries has continued to increase. Relying on various scientific research platforms, a large amount of measured data has been accumulated in areas such as sea ice monitoring, ocean temperature, salinity and depth profile detection, and polar ecological surveys. This data is the core data foundation supporting the commercial operation of polar shipping routes, research on polar climate evolution, and ecological environmental protection.

[0003] However, existing cross-border data sharing and verification models for polar scientific expeditions suffer from numerous structural and technical shortcomings. The core issues lie in standard barriers, security risks, insufficient accuracy, and a lack of collaboration. The polar scientific data observation standards of the collaborating parties originate from their respective research systems, exhibiting significant differences in areas such as CTD (Conductivity, Temperature, Depth) profiler data verification thresholds, sea ice thickness observation accuracy, and sediment analysis index definitions. This prevents direct data exchange, requiring manual, one-to-one data conversion, with a single expedition's data exchange cycle lasting 7-10 days and easily introducing human error. Current cross-border data transmission largely relies on general network channels, lacking dedicated encryption mechanisms for sensitive polar scientific data, posing risks of data leakage and tampering. Data verification is a single-entity, one-way verification process, resulting in high verification errors and severely impacting the reliability of research conclusions. Furthermore, general data processing tools are ill-suited to the complex polar conditions and the differences in standards between collaborating parties, leading to delayed data sharing, long verification cycles, and a lack of long-term optimized solutions for cross-border collaboration. This hinders the formation of a standardized cooperation paradigm and restricts the deepening of international cooperation in polar scientific expeditions. Summary of the Invention

[0004] To solve, or at least partially solve, the above-mentioned technical problems, the present invention provides a method for cross-border data sharing and verification in polar scientific expeditions.

[0005] In a first aspect, the present invention provides a method for cross-border data sharing and verification in polar scientific expeditions, comprising the following steps: S1. Obtain raw data from polar scientific expeditions, call pre-set bidirectional data mapping rules, perform format conversion, index calibration, and initial screening of invalid data on the raw data from polar scientific expeditions to obtain standardized data; S2. Based on the sensitivity level of the standardized data, a pre-set hierarchical encryption rule is invoked to encrypt the standardized data, and the encrypted data is transmitted across borders via a polar satellite communication link. S3. Decrypt the encrypted data received after cross-border transmission, call the pre-set dual-path verification rules to perform parallel verification on the decrypted data, and obtain the corresponding data verification result. S4. Based on the data verification results and the pre-set hierarchical access rules, complete the cross-border sharing of data that has passed the verification by matching access permissions.

[0006] Optionally, the bidirectional data mapping rules include the corresponding conversion relationship of index definitions, format parameters, and verification thresholds for core data types in polar scientific research, as well as the correction rules for systematic deviations of the sensor system, and a rule extension interface is reserved.

[0007] Optionally, in step S1, the initial screening of invalid data specifically includes the following steps: Based on the physical threshold of the polar environment, invalid data points in the original polar scientific expedition data that do not meet the physical threshold of the polar environment are removed; An unsupervised anomaly detection algorithm is used to identify hidden anomalies in the original polar scientific research data. Multiple anomaly detection algorithms are combined for cross-validation. The verified hidden anomalies are marked as pending review, and the causes of the anomalies and corresponding metadata are recorded simultaneously.

[0008] Optionally, in step S2, the encryption process for the standardized data specifically includes the following steps: The standardized data at the ordinary level is encrypted using a symmetric encryption algorithm; For the standardized data of the sensitive level, a hybrid mode combining asymmetric encryption algorithm and symmetric encryption algorithm is used for encryption; For the standardized data at the core level, dual-key verification is added on top of the hybrid encryption method.

[0009] Optionally, in step S2, the cross-border transmission tunnel adopts a layered protocol, the control channel adopts the TCP protocol, the data channel adopts the UDP protocol, and the transmission process completes two-way authentication based on digital certificates; the cross-border transmission of encrypted data through the polar satellite communication link specifically includes the following steps: The signal quality of the dual-path polar satellite communication link is monitored in real time, and the primary and backup links are switched according to the signal quality, or the encrypted data packets are split and transmitted in parallel through the dual-path links. The data packet size is adjusted in real time according to the signal quality. Forward error correction coding is added to the data packet before transmission. During transmission, a check digest is generated according to a fixed data length. The check digest is used by the receiving end for real-time comparison. Retransmit lost or tampered data packets.

[0010] Optionally, in step S3, the step of invoking a pre-set dual-path verification rule to perform parallel verification of the decrypted data specifically includes the following steps: The decrypted data is verified simultaneously using two verification algorithms. One verification algorithm verifies the consistency of data trends, while the other verification algorithm verifies extreme values ​​and abnormal patterns in the data. For decrypted data whose verification results are within a pre-set error threshold, a qualified data verification result is generated; for decrypted data whose verification results exceed the error threshold, the cause of the difference is located and a review process is initiated. and, The hidden anomaly data marked as pending review undergoes encryption, cross-border transmission, and decryption processing synchronously with the standardized data; after step S3, the method further includes: The hidden abnormal data marked as pending verification and the decrypted data whose parallel verification exceeds the error threshold are manually verified. Data that is determined to be valid after verification is added to the data verification result, and data that is determined to be invalid after verification is removed.

[0011] Optionally, a pre-trained random forest classification model is embedded in the dual-path verification rules. The input features of the random forest classification model include data difference features, data statistical features, spatiotemporal and environmental background features, and verification status features. The output results of the random forest classification model are divided into four categories: qualified data, abnormal data, special phenomena, and standard divergence.

[0012] Optionally, step S4 specifically includes the following steps: Access to qualified Level 1 public data will be granted to all partner institutions. Access to qualified Level 2 restricted data will be granted after joint authorization from the project leaders of both parties. Access to qualified Level 3 core data is granted only to designated core personnel and requires dual key verification.

[0013] Optionally, the method further includes the following steps: S5. Collect the correction results of manual review, update the corresponding parameters of the bidirectional data mapping rule and the dual-path verification rule in real time, and use the data samples completed by manual review as newly labeled data to incrementally train the random forest classification model. S6. Summarize the correction results and data application feedback information collected in the current period according to a fixed period, and perform a full iterative update on the bidirectional data mapping rules, the hierarchical encryption rules, the dual-path verification rules, and the hierarchical access rules. Retrain the random forest classification model using all the newly labeled data accumulated in the current period, adjust the feature dimensions of the random forest classification model, and optimize the data processing and verification logic.

[0014] Secondly, the present invention also provides a cross-border data sharing and verification system for polar scientific expeditions, comprising: The acquisition module is used to acquire raw data from polar scientific expeditions, call pre-set bidirectional data mapping rules, perform format conversion, index calibration, and initial screening of invalid data on the raw data from polar scientific expeditions, and obtain standardized data. The transmission module is used to encrypt the standardized data according to the sensitivity level of the standardized data by calling a pre-set hierarchical encryption rule, and to complete the cross-border transmission of the encrypted data through the polar satellite communication link. The verification module is used to decrypt the encrypted data received after cross-border transmission, and to perform parallel verification of the decrypted data by calling the pre-set dual-path verification rules to obtain the corresponding data verification result. The matching module is used to perform cross-border sharing of qualified data by matching access permissions based on the data verification results and pre-set hierarchical access rules.

[0015] The method provided by this invention has the following beneficial effects: The technical solution of this invention, by invoking bidirectional data mapping rules to perform format conversion, index calibration, and initial screening of invalid data for raw polar scientific expedition data, can eliminate barriers to data exchange between different sources, reduce the workload of manual data conversion, avoid errors introduced by manual operation, and improve the efficiency and consistency of data standardization processing. By matching corresponding encryption methods according to the sensitivity level of standardized data, it can adapt to the protection needs of different data. Combined with polar satellite communication links to complete cross-border transmission, it can adapt to the special communication environment of the polar regions, reduce the risk of leakage, tampering, and loss during data transmission, and ensure the security and stability of cross-border data transmission. By performing parallel verification of decrypted data through dual-path verification rules, it can integrate the technical advantages of different verification dimensions, improve the ability to identify abnormal data, reduce verification errors, ensure the reliability of cross-border shared scientific expedition data, and provide accurate data support for joint scientific research. By completing cross-border data sharing based on data verification results and hierarchical access rules, it can achieve refined management of data sharing, ensure the controllability of the data sharing process, and adapt to the different data sharing needs of cross-border scientific expedition cooperation.

[0016] Building upon this foundation, by incorporating correction rules for systematic sensor biases into the bidirectional data mapping rules, the accuracy of data calibration can be further improved, reducing the impact of inherent equipment biases on data consistency. Through the design of cross-validation of multiple anomaly detection algorithms, the accuracy of identifying hidden anomalies can be enhanced, reducing missed and false detections. The application of layered protocol design and forward error correction coding further adapts to the weak signal and high interference communication environment of polar regions, improving data transmission stability. Embedding a random forest classification model enhances the intelligence of data verification, accurately pinpointing the causes of data discrepancies and reducing the workload of manual review. The design of rules and model updates driven by manual review results and application feedback allows the data processing and verification logic to continuously adapt to actual scientific research needs, improving the long-term adaptability of the solution. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of a method for cross-border data sharing and verification in polar scientific expeditions, provided by an embodiment of the present invention. Figure 2 This is a schematic diagram of encrypted transmission logic provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. 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.

[0019] See Figure 1 This invention provides a method for cross-border data sharing and verification in polar scientific expeditions, comprising the following steps: S1. Obtain raw data from polar scientific expeditions, call pre-set two-way data mapping rules, perform format conversion, index calibration, and initial screening of invalid data on the raw data from polar scientific expeditions to obtain standardized data; In some implementations, the bidirectional data mapping rules include the corresponding conversion relationships of index definitions, format parameters, and verification thresholds for core data types in polar scientific research, as well as correction rules for systematic deviations of sensors, and a reserved rule extension interface.

[0020] In some implementations, the initial screening of invalid data in S1 specifically includes the following steps: Based on the physical threshold of the polar environment, invalid data points in the original data of polar scientific expeditions that do not meet the physical threshold of the polar environment are removed; An unsupervised anomaly detection algorithm is used to identify hidden anomalies in the raw data of polar scientific expeditions. Multiple anomaly detection algorithms are combined for cross-validation. The verified hidden anomalies are marked as pending review, and the causes of the anomalies and corresponding metadata are recorded simultaneously.

[0021] Specifically, during polar scientific expeditions, raw data collected by different partners often suffers from inconsistencies in indicator definitions, format specifications, and verification standards. Manual data conversion is time-consuming and prone to human error. Furthermore, data collected in the complex polar environment is susceptible to invalid and outlier values, affecting the usability and accuracy of subsequent data. To address these issues, pre-defined two-way data mapping rules can be established. These rules include the corresponding conversion relationships between indicator definitions, format parameters, and verification thresholds for core data types used in polar scientific expeditions, as well as correction rules for systematic biases in sensors. A rule extension interface is also provided to allow for rule supplementation based on newly added data types and data acquisition equipment types.

[0022] After acquiring raw data from polar scientific expeditions, bidirectional data mapping rules are invoked to perform format conversion, index calibration, and initial screening of invalid data, ultimately yielding standardized data. During format conversion, the system adapts to various data formats output by different scientific research equipment, completing a unified conversion from unstructured to structured data, ensuring a consistent processing foundation for data from different sources. During index calibration, bidirectional data mapping rules are used to perform unified conversion of data units. Simultaneously, for sensor data with known systematic biases, corresponding correction rules are invoked to complete batch calibration, eliminating data errors caused by inherent equipment biases and aligning with the verification threshold standards of different partners. The calibration process uses a linear correction formula for calculation, as follows: ; in, The raw data output by the sensor. For calibrated and standardized data, coefficients , It is obtained by fitting historical calibration data, which can be adapted to the deviation characteristics of different models of acquisition equipment to complete batch calibration.

[0023] Based on previous sensor comparison tests conducted by the joint laboratory, it was found that the AML-X200 CTD sensor exhibits a known systematic positive bias in temperature observations in Arctic summer low-temperature water (-2°C ~ 5°C). Through synchronous comparison tests with a benchmark sensor calibrated in a higher-precision laboratory in the same constant-temperature water bath, it was confirmed that this bias mainly manifests as a fixed zero-point drift and a slight gain error.

[0024] To correct this deviation, the technical team used 50 sets of historical synchronous calibration data obtained from the aforementioned comparative test to perform linear fitting using the least squares method, thereby determining a dedicated calibration coefficient for this sensor model. The resulting calibration formula is: Ycorrected = 0.98 * Xraw + 0.1. In this formula, Xraw represents the original temperature observation value (°C) of the AML-X200 CTD; coefficient a = 0.98 is used to compensate for the slight gain error of approximately 2% overestimation of the sensor's overall observation value; coefficient b = 0.1 is used to compensate for its fixed +0.1°C positive zero-point drift. This formula and its parameters, as a device-specific calibration rule, are written into the "bidirectional data mapping library" using "Device Model: AML-X200, Parameter: Temperature" as the index key.

[0025] In subsequent joint scientific expeditions, the system automatically invoked this rule to perform real-time batch calibration of a batch of temperature, salinity, and depth profile data collected by the AML-X200 sensor. Taking the expedition sites as an example, a comparison of the calibration process and results for some data is shown in Table 1: Table 1

[0026] Quantitative analysis of the calibration effect showed that the average absolute error between the original data and the true reference value before calibration was 0.23°C. After applying the calibration formula, the average absolute error between the calibrated data and the reference value significantly decreased to 0.02°C, a reduction of approximately 91%. This result effectively eliminated the systematic bias of this sensor model, ensuring that the data accuracy met the stringent requirement of joint research that temperature data errors are typically less than ±0.05°C.

[0027] This case study fully demonstrates the linear correction formula. From fitting specific coefficients to historical calibration data to integrating them into the database as automatically executed rules, the entire process culminates in high-precision batch calibration within the actual data stream. This not only validates the effectiveness of the calibration method, reducing specific sensor errors by approximately 91%, but also demonstrates the high degree of automation in the process—rules can be repeatedly invoked without manual intervention once they are entered into the database. Furthermore, by transforming expert experience and laboratory calibration results into reusable digital rules, it substantially solves the practical problem of systematic sensor bias mentioned in the background technology, enhancing the practical value of the solution.

[0028] The initial screening of invalid data is carried out in two steps. First, based on the physical threshold of the polar environment, invalid data points in the original polar scientific expedition data that do not meet the physical threshold of the polar environment are removed, completing the initial data purification. Then, an unsupervised anomaly detection algorithm is used to identify hidden anomalies in the original polar scientific expedition data. Multiple anomaly detection algorithms are combined for cross-validation to improve the accuracy of anomaly identification. The verified hidden anomalies are marked as pending review, and the reasons for the anomalies and corresponding metadata are recorded simultaneously. They are not directly removed, but are reviewed and judged in subsequent steps.

[0029] Through the above processing flow, the raw data of polar scientific expeditions from different sources can be processed automatically and in a standardized manner. The data processing efficiency is greatly improved compared with manual processing, the accuracy of data standardization processing is high, the time spent on data docking for a single voyage is greatly shortened, invalid data is effectively filtered out, hidden abnormal data is accurately identified, and errors caused by manual processing are avoided, providing a unified and high-quality data foundation for subsequent data transmission, verification and sharing.

[0030] S2. Based on the sensitivity level of the standardized data, the pre-set hierarchical encryption rules are invoked to encrypt the standardized data, and the encrypted data is transmitted across borders via the polar satellite communication link. In some implementations, S2 involves encrypting the standardized data, specifically including the following steps: For standardized data at the ordinary level, a symmetric encryption algorithm is used for encryption; For standardized data of a sensitive level, a hybrid mode combining asymmetric and symmetric encryption algorithms is used for encryption. For standardized data at the core level, dual-key verification is added on top of the hybrid encryption method.

[0031] In some implementations, in S2, the cross-border transmission tunnel uses a layered protocol, the control channel uses the TCP protocol, and the data channel uses the UDP protocol. The transmission process is based on two-way authentication using digital certificates. The cross-border transmission of encrypted data is completed via a polar satellite communication link, specifically including the following steps: Real-time monitoring of the signal quality of the dual polar satellite communication links; switching between primary and backup links based on signal quality; or splitting encrypted data packets for parallel transmission through the dual links. The data packet size is adjusted in real time according to the signal quality. Forward error correction coding is added to the data packet before transmission. During transmission, a check digest is generated according to a fixed data length. The check digest is used by the receiving end for real-time comparison. Retransmit lost or tampered data packets.

[0032] Specifically, after completing standardized data processing, it is necessary to address the problems of weak satellite communication signals, strong interference, and poor link stability in the polar environment. At the same time, it is necessary to prevent data leakage, tampering, and loss during cross-border transmission. General transmission methods cannot be adapted to the complex polar communication environment, nor can they meet the differentiated protection requirements of data with different sensitivity levels. This can easily lead to problems such as high transmission delays, high packet loss rates, and insufficient data security protection.

[0033] like Figure 2 As shown, in this invention, for standardized data with different sensitivity levels, pre-set hierarchical encryption rules are invoked to complete the encryption process. The hierarchical encryption rules are divided into three levels of protection strategies according to the data sensitivity level, with different encryption processing methods corresponding to different levels. For standardized data of ordinary sensitivity level, a symmetric encryption algorithm is used for encryption. This type of data is mostly publicly available polar routine meteorological observation data and publicly available statistical data on sea ice extent, which does not contain sensitive information. The encryption processing focuses on ensuring data integrity. The symmetric encryption algorithm can be AES-128, or it can be replaced with the national cryptographic algorithm SM4 according to the compliance requirements of both parties, while maintaining a consistent security protection level. For standardized data of a sensitive level, a hybrid encryption mode combining asymmetric and symmetric encryption algorithms is used. This type of data mainly includes temperature, salinity, and depth profile data from polar shipping route surveys, measured sea ice thickness data, and routine marine ecological monitoring data, involving core scientific research data not publicly disclosed by both parties. In the hybrid encryption mode, the session key is transmitted through an asymmetric encryption algorithm, while the data itself is encrypted using a symmetric encryption algorithm, balancing encryption efficiency and security. The asymmetric encryption algorithm can be RSA or replaced with the Chinese national standard SM2 algorithm, and the symmetric encryption algorithm can be AES-256 or replaced with the Chinese national standard SM4 algorithm. For standardized data of a core level, dual key verification is added to the hybrid encryption mode. This type of data mainly includes location and monitoring data of sensitive polar ecological protection areas and measured data of key points in core shipping routes, involving extremely high security protection requirements. Dual key verification requires the authorized keys of both parties to be verified simultaneously to complete data decryption and reading, avoiding data security issues caused by the leakage of one party's key.

[0034] After encryption, a cross-border transmission tunnel is established. This tunnel employs a layered protocol design. The control channel uses TCP to ensure the reliability of handshake signals, key negotiation, link control, and metadata data transmission, preventing the loss of critical control information. The data channel uses UDP, with a reliable transmission mechanism implemented at the application layer. Sequence number marking, reception confirmation, and selective retransmission adapt to the intermittent, weak link communication environment of the polar regions, reducing end-to-end latency. During tunnel construction, the entire transmission process utilizes digital certificates for two-way authentication. Both parties pre-issue compliant digital certificates, and the initiating and receiving ends perform two-way certificate verification before tunnel construction. Only verified ends can complete tunnel construction and data transmission, preventing data leakage risks due to unauthorized terminal access. The two ends of the transmission tunnel connect to the multi-mode satellite communication terminal on the polar research vessel and a ground gateway station, respectively. The ground gateway station connects to the dedicated research networks of both parties. The two ends are connected via a dedicated cross-border communication link, mitigating security risks associated with the public internet.

[0035] Encrypted data is transmitted across borders via polar satellite communication links. During transmission, the signal quality of the dual polar satellite communication links is monitored in real time, with monitoring indicators including signal-to-noise ratio (SNR), bit error rate (BER), round-trip time (RTT), and available bandwidth. When the SNR of the primary link consistently falls below a set threshold, or the BER continuously exceeds a set upper limit for a set duration, the system automatically switches the data stream to the backup link. The switching process maintains uninterrupted session connectivity, and upper-layer data processing remains unaffected. When the signal quality of both links meets the transmission requirements, the encrypted data packets can be split and transmitted in parallel through the dual links. At the receiving end, the data packets are reassembled, aggregating link bandwidth and improving the transmission efficiency of large data volumes.

[0036] During transmission, the size of data packets is adjusted in real time based on the monitored link signal quality. When the link signal quality is good, large packets of 8KB to 16KB are used for transmission to reduce protocol overhead. When the link signal quality is poor, small packets of 1KB to 2KB are used to reduce the retransmission cost caused by single packet errors. Before sending data packets, forward error correction coding is added to all data packets, allowing the receiver to directly correct a small number of transmission errors without initiating retransmission requests, further adapting to the high error rate communication environment. During transmission, a checksum is generated at a fixed data length. A set of SHA-256 checksums is generated every 10KB of data transmitted. The checksums are transmitted synchronously to the receiver along with the data packets. The receiver compares the received data packets in real time to confirm data integrity. For lost or tampered data packets detected during the comparison, the receiver only initiates a retransmission request for the abnormal data packets. Upon receiving the request, the sender completes the retransmission of the corresponding data packets without retransmitting all transmitted data, further reducing transmission latency.

[0037] Through the above encryption and transmission processing, it can be adapted to the complex satellite communication environment in the polar regions. The end-to-end transmission latency can be stably controlled at a low level. At the same time, it can achieve differentiated security protection for data with different sensitivity levels, greatly reducing the risk of data leakage, tampering, and loss during cross-border transmission, and meeting the timeliness and security requirements of cross-border data transmission in polar scientific expeditions.

[0038] S3. Decrypt the encrypted data received after cross-border transmission, call the pre-set dual-path verification rules to perform parallel verification on the decrypted data, and obtain the corresponding data verification result. In some implementations, in step S3, a pre-defined dual-path verification rule is invoked to perform parallel verification of the decrypted data, specifically including the following steps: Two verification algorithms are invoked simultaneously to verify the decrypted data. One verification algorithm verifies the consistency of data trends, while the other verification algorithm verifies extreme values ​​and abnormal patterns in the data. For decrypted data whose verification results are within the pre-set error threshold, generate a qualified data verification result; for decrypted data whose verification results exceed the error threshold, locate the cause of the difference and push the review process. and, Hidden anomaly data marked as pending review undergoes encryption, cross-border transmission, and decryption processing simultaneously with standardized data; after S3, the method also includes: Hidden abnormal data marked as pending verification and decrypted data whose parallel verification exceeds the error threshold are manually verified. Data that is deemed valid after verification is added to the data verification results, and data that is deemed invalid after verification is removed.

[0039] In some implementations, a pre-trained random forest classification model is embedded in the dual-path verification rules. The input features of the random forest classification model include data difference features, data statistical features, spatiotemporal and environmental background features, and verification status features. The output results of the random forest classification model are divided into four categories: qualified data, abnormal data, special phenomena, and standard divergence.

[0040] Specifically, after encrypted data is transmitted across borders and delivered to the receiving end, existing conventional verification methods mostly adopt a one-way verification logic of a single entity. This cannot integrate the verification experience and technical advantages accumulated by different partners in the field of polar scientific research. It is difficult to accurately distinguish whether the data differences are due to equipment errors, collection anomalies, or real special polar environmental phenomena. Verification errors often exceed reasonable ranges. At the same time, there is a lack of standardized and unified processing procedures for hidden abnormal data marked as pending verification in the early stage. This can easily lead to the accidental deletion of valid data and the omission of abnormal data, affecting the reliability of polar scientific research data and the accuracy of subsequent research conclusions.

[0041] In this invention, the receiving end first decrypts the encrypted data received after cross-border transmission, and simultaneously completes the format alignment of the hidden abnormal data marked as pending verification, which was encrypted, transmitted, and decrypted synchronously with the standardized data. Then, it calls a pre-set dual-path verification rule to perform parallel verification of the decrypted data. During the parallel verification process, two verification algorithms are simultaneously invoked to perform full data verification. One verification algorithm verifies the consistency of data trends, focusing on comparing the continuous change patterns of the same observation sequence and the change trends of historical data from the same station during the same period, verifying the rationality of the overall data changes. For example, for temperature, salinity, and depth profile data collected in a certain sea area, this algorithm will verify whether the temperature and salinity change trends corresponding to the vertical depth conform to the basic distribution patterns of polar sea water. The other verification algorithm verifies extreme values ​​and abnormal patterns in the data, focusing on identifying extreme values ​​that exceed the normal observation range and abnormal data distribution patterns that do not conform to the characteristics of the polar environment. For example, for sea ice concentration observation data, this algorithm will focus on identifying extreme anomalies in local sea areas and eliminating invalid data caused by sensor malfunctions and environmental interference.

[0042] After the two-way verification algorithm completes the full data verification, it generates corresponding data difference comparison results. For decrypted data whose verification results are within the pre-set error threshold, a qualified data verification result is generated and directly enters the subsequent sharing stage. For decrypted data whose verification results exceed the error threshold, the system automatically locates the corresponding data segment and related influencing factors that caused the difference and simultaneously pushes the review process. During the difference location and review judgment process, a pre-trained random forest classification model embedded in the dual-way verification rules completes the intelligent auxiliary judgment, avoiding the subjective bias and inefficiency of manual review.

[0043] The random forest classification model is pre-trained based on archived data accumulated from historical joint polar scientific expeditions and comparative test data under controlled conditions. All training samples have been jointly reviewed and labeled by research experts from both collaborating parties to ensure the accuracy of sample labeling. The model's input features include four core dimensions: the first is data difference features, including the absolute and relative differences between the two verification results, and the distribution of these differences in historical data from the same period and region; the second is data statistical features, including the mean, variance, and correlation of the verification data sequence, and the degree of deviation from the historical average; the third is spatiotemporal and environmental background features, including the time and geographical location of data collection, and the corresponding environmental observation parameters; the fourth is verification status features, including threshold alarm information triggered during the verification process and the preliminary determination of the difference type. The model's output results are divided into four fixed categories: data qualified, data abnormal, special phenomenon, and standard divergence, with each category corresponding to preset processing guidelines. Among these, "data qualified" corresponds to scenarios where the differences are within a reasonable error range and the data is acceptable; "data abnormal" corresponds to scenarios where one party's data may have technical anomalies; "special phenomena" corresponds to scenarios where the differences may reveal real special ocean phenomena; and "standard disagreements" corresponds to scenarios where the differences stem from inherent disagreements in the standards or methods of both parties, requiring the initiation of arbitration procedures. For example, for a set of sea ice thickness observation data exceeding the error threshold, after the model completes feature extraction and analysis, if the output result is a special phenomenon, it will mark that the sea ice changes in the corresponding sea area may have special causes, prompting researchers to pay close attention, rather than directly determining it as invalid data; if the output result is data abnormal, it will mark the possible causes of the data abnormality and forward it to the technical personnel of the corresponding data collection party for confirmation. During model training, node splitting is performed by randomly selecting features to improve generalization ability and avoid overfitting problems, enabling the model to maintain a stable judgment accuracy in different polar scientific expedition scenarios.

[0044] For each set of decrypted data exceeding the error threshold, the system synchronously generates a corresponding intelligent diagnostic report. The report comprehensively covers the entire automatic verification process, the judgment conclusions of the random forest classification model, root cause analysis suggestions for data discrepancies, and subsequent standardized processing guidelines. The automatic verification process fully records the verification process of both verification algorithms, the threshold information triggering alarms, the specific values ​​of data discrepancies, and the location information of the corresponding data segments. This eliminates the need for manual re-verification of the entire dataset, allowing direct identification of the core discrepancies. The model judgment conclusions clearly indicate the output classification results of the random forest classification model, along with the core judgment criteria for the corresponding classification, including key features triggering the classification and comparison results with similar historical data, providing standardized reference for manual review. The root cause analysis suggestions combine data acquisition equipment information, spatiotemporal environmental background, and historical records of similar anomalies to provide possible causes of the discrepancies. For example, for anomalies in a set of marine temperature, salinity, and depth profile data, different analysis suggestions are provided, including sensor equipment failure, interference from the acquisition environment, special polar hydrological phenomena, and data transmission errors, narrowing the scope of manual review. The follow-up processing guidance section provides standardized processing recommendations based on the classification results output by the model, including reference methods for data correction, corresponding review and approval processes, and precautions for data use. The intelligent diagnostic report is simultaneously pushed to the review process along with the corresponding data, significantly reducing the workload of manual review, improving review efficiency and consistency of judgment results, and avoiding review biases caused by differences in human experience.

[0045] All hidden anomaly data marked as pending review, as well as decrypted data whose parallel verification exceeded the error threshold, will enter a unified manual review process. This review will be jointly conducted by the relevant technical and research personnel from both collaborating parties. Data deemed valid after review will be added to the corresponding data verification results and simultaneously included in the available data range for subsequent sharing processes. Data deemed invalid will be uniformly removed, and the reasons for invalidity and corresponding metadata will be recorded to provide a basis for optimizing subsequent data processing rules.

[0046] Through the above verification process, the polar scientific research verification experience and technical advantages of different partners can be fully integrated, which can significantly reduce the average error of data verification, achieve a high accuracy rate of automatic verification, and greatly improve the efficiency and accuracy of data verification. At the same time, it can accurately distinguish between data anomalies and real polar special phenomena, avoid the accidental deletion of valid scientific research data, and provide reliable quality support for the cross-border sharing of polar scientific research data.

[0047] S4. Based on the data verification results and the pre-set hierarchical access rules, complete the cross-border sharing of data that has passed the verification and has the required access permissions.

[0048] In some implementations, S4 specifically includes the following steps: Access to qualified Level 1 public data will be granted to all partner institutions. Access to qualified Level 2 restricted data will be granted after joint authorization from the project leaders of both parties. Access to qualified Level 3 core data is granted only to designated core personnel and requires dual key verification.

[0049] Specifically, in the process of cross-border cooperation in polar scientific research, different types of scientific research data correspond to different open scopes and security control requirements. The general data sharing model cannot achieve fine-grained permission division, which easily leads to unauthorized access and data use beyond the scope. At the same time, it lacks complete operation traceability capabilities, making it impossible to trace and audit the entire data access process, and it is also difficult to effectively collect relevant feedback from researchers during the data use process, which cannot match the actual control needs of cross-border polar scientific research cooperation.

[0050] In this invention, after data verification is completed, cross-border sharing of verified data is achieved according to pre-set hierarchical access rules, matching access permissions. The hierarchical access rules set up a three-level access permission division standard based on the data's sensitivity level and the scope of scientific research cooperation. Each level corresponds to a clear scope of access, authorization process, and access control requirements, ensuring that different types of scientific research data are shared and used within a compliant scope.

[0051] Level 1 data is publicly available data. Verified Level 1 publicly available data is accessible to all collaborating institutions. This type of data mainly includes routine meteorological observation data from polar seas, publicly available monthly statistics on sea ice extent, and basic data on the public navigation environment of polar waterways. It does not involve unpublished scientific research results or sensitive location information and has no special usage restrictions. All institutions and research teams participating in polar scientific expeditions can directly query, download, and use the corresponding data compliantly through the sharing platform after completing basic identity registration, without additional authorization and approval processes, ensuring efficient sharing of basic scientific research data.

[0052] The second level is restricted data. Access to restricted data that passes verification is granted only after joint authorization from the project leaders of both collaborating parties. This type of data mainly includes measured temperature, salinity, and depth profile data collected during joint scientific expeditions, fixed-point observation data of sea ice thickness, and routine marine ecological monitoring data. This data represents core research data not publicly disclosed by either collaborating party and is only accessible to participating teams within the corresponding joint scientific expedition projects. When submitting a data access request, researchers must clearly indicate the intended use, scope, and duration of the data. The application materials must be simultaneously sent to the corresponding project leaders of both collaborating parties. Access to the corresponding data will only be granted after both parties have completed authorization confirmation. During the authorization process, download permissions, editing permissions, and access validity periods can be set separately for the data to prevent unauthorized use.

[0053] The third level is core data, which, once verified, is accessible only to designated core personnel and requires dual-key verification. This type of data primarily includes location and monitoring data for sensitive polar ecological protection areas, high-precision measured data of key risk points in polar shipping routes, and unpublished core polar scientific research experimental data. It has the highest security control level and is only accessible to core research personnel and project leaders pre-designated by both collaborating parties. After submitting an access request, relevant personnel must complete basic identity verification and pass dual-key verification. Access to the corresponding data is only unlocked when both parties' independent authorization keys are simultaneously verified. Furthermore, this type of data only supports online viewing and limited data analysis; full data download access is not available to minimize the risk of core data leakage.

[0054] Throughout the entire data sharing process, the system automatically generates complete access logs. These logs include the accessor's identity information, access time, accessed data scope, specific actions performed, and corresponding authorization information. All access logs are stored in an immutable manner, ensuring full traceability and auditability. In cases of unauthorized data access or use beyond permitted scope, the retained access logs can accurately pinpoint the corresponding operational steps and personnel, guaranteeing compliance throughout the data sharing process. Researchers can simultaneously submit feedback related to data application during access and use, including information on data accuracy, availability, and suitability. This feedback is synchronously linked to corresponding data identifiers and collection information, providing a reference for optimizing subsequent data processing and verification rules.

[0055] This hierarchical and shared management model enables refined access control of cross-border data in polar scientific expeditions. It not only ensures the security of data with different levels of sensitivity but also efficiently meets the data sharing needs of joint scientific expeditions. The authorization and review efficiency of data sharing is significantly improved compared to the manual offline connection model. It can adapt to the actual use needs of cross-border polar scientific expedition cooperation and provide stable data support for the smooth progress of joint scientific expeditions.

[0056] In some implementations, the method further includes the following steps: S5. Collect the correction results of manual review, update the corresponding parameters of bidirectional data mapping rules and dual-path verification rules in real time, and use the data samples completed by manual review as newly labeled data to incrementally train the random forest classification model. S6. Summarize the correction results of manual review and data application feedback collected in the current period according to a fixed period, and perform full iterative updates on the bidirectional data mapping rules, hierarchical encryption rules, dual-path verification rules and hierarchical access rules. Retrain the random forest classification model using all newly added labeled data accumulated in the current period, adjust the feature dimensions of the random forest classification model, and optimize the data processing and verification logic.

[0057] Specifically, in the process of cross-border cooperation in polar scientific expeditions, fixed data processing rules, verification models, and access control standards cannot adapt to the ever-increasing collection equipment, data types, observation scenarios, and cooperation needs. The experience accumulated during manual review cannot be effectively transformed into standardized processing logic. Problems such as data deviation and anomaly identification deviation will repeatedly occur. At the same time, the application feedback put forward by researchers during data use cannot be effectively implemented. As a result, the relevant rules for data processing, verification, and sharing gradually become disconnected from the actual scientific expedition needs, affecting the stability and adaptability of the solution in the long term.

[0058] In this invention, throughout the entire data verification and sharing process, the system automatically collects the correction results from manual verification, including correction parameters for data deviations, judgment results for abnormal data, annotation information for special phenomena, and unified rules for standard discrepancies. After collection, the system updates the corresponding parameters of the bidirectional data mapping rules and dual-path verification rules in real time based on the correction content. For example, during manual verification, it was discovered that the data collected by a newly deployed temperature, salinity, and depth profilometer had a fixed systematic deviation. After joint confirmation by both parties, data correction was completed. For the deviation in salinity observation under low-temperature conditions, a salinity compensation correction formula was used to correct it, as follows: ; in, This is the raw salinity data. To collect water temperature data synchronously, The compensation function, built based on water temperature, can eliminate salinity measurement bias caused by the low polar temperatures after correction. The system automatically adds the corresponding correction rules for this type of equipment to the bidirectional data mapping rules, allowing subsequent data collected by similar equipment to be automatically corrected without manual reprocessing. For example, if manual review determines that a set of data exceeding the error threshold is a genuine special hydrological phenomenon caused by a polar front, the system will record the characteristics of this set of data along with the determination result, using it as a valid labeled sample. After collecting manually reviewed data samples, the system will use them as new labeled data for incremental training of the random forest classification model. During incremental training, only the decision tree node parameters of the model are adjusted based on the newly added labeled data, without changing the basic framework of the pre-trained model. This allows for rapid integration of manual review experience, improving the model's accuracy in identifying similar abnormal data and special phenomena, without affecting the model's original stable judgment ability, thus avoiding performance fluctuations. For instance, for abnormal sea ice concentration data during the summer melting season confirmed by multiple cruises, incremental training can effectively improve the model's accuracy in identifying similar melting phenomena, significantly reducing the workload of manual review of similar data.

[0059] During the quarterly joint review of historical joint scientific expedition data, technical experts from both sides jointly identified a technical problem: when a new type of profiling buoy operated in the Arctic surface low-temperature waters (below -1.5°C), the salinity values ​​measured by its built-in conductivity sensor exhibited a systematic nonlinear negative bias. By comparing this data with data measured by a high-precision CTD (Conductivity To Diode) simultaneously deployed by our side and calibrated specifically for low-temperature conditions in our laboratory, it was found that this bias increased as the water temperature decreased, reaching approximately -0.02 PSU at -1.5°C and increasing to approximately -0.05 PSU at -1.8°C. Joint technical diagnosis confirmed that this bias stemmed from the inherent physical characteristic drift of the sensor's sensitive element within a specific low-temperature range.

[0060] To address this issue, the technical team extracted over 100 sets of synchronous comparative data obtained by the buoy and our CTD at multiple low-temperature sites, forming a historical calibration dataset. Using water temperature (T, in °C) as the independent variable and salinity deviation (ΔS, i.e., reference salinity minus original salinity) as the dependent variable, curve fitting was performed to obtain a targeted compensation function: f(T) = 0.03 * (T + 1.5) 2 - 0.02. This function is applicable to water temperatures between -1.8°C and -1.5°C. Based on this, a complete calibration rule is generated: when the device model is "Profiler-RU2023", the observed parameter is "salinity", and the water temperature meets the above conditions, the system will automatically perform the calibration calculation. This environmentally-aware correction rule was immediately added to the "bidirectional data mapping library" with the index key (device model: Profiler-RU2023, parameter: salinity, condition: low temperature).

[0061] During subsequent scientific expeditions, the system automatically invoked this new rule to perform real-time compensation on data collected by the same type of buoy in low-temperature environments. Table 2 shows the verification results of some measured data from the low-temperature stations during the expeditions: Table 2

[0062] The application effect was quantitatively analyzed: Before calibration, the mean absolute error between the original salinity data and the reference value was 0.045 PSU; after calibration using the compensation function, the mean absolute error was significantly reduced to 0.008 PSU. This means that the salinity observation error was reduced by approximately 82%, and the data accuracy was greatly improved. The results fully meet the technical standard that oceanographic research typically requires salinity data errors to be less than ±0.02 PSU.

[0063] The system will periodically summarize all manually reviewed correction results and data application feedback collected during the current period. This period can be set to quarterly based on the joint scientific expedition schedule, or a special summary iteration can be conducted after the completion of a large-scale joint scientific expedition. After the summary is completed, the system will collaborate with technical and research personnel from both collaborating parties to conduct a full-scale iterative update of the bidirectional data mapping rules, hierarchical encryption rules, dual-path verification rules, and hierarchical access rules. This includes supplementing conversion rules for newly added data types, adjusting the hierarchical standards for encryption protection, optimizing the error thresholds for data verification, and improving the access control requirements for data sharing, ensuring that all rules match the current needs of the scientific expedition collaboration. For example, after summarizing application feedback for a quarter, if it is found that the collaborating parties have added data types related to polar sediment surveys, the system will supplement the indicator definitions, format conversions, and bidirectional data mapping rules corresponding to the verification thresholds for this type of data during the full-scale iteration process. Similarly, for upgrades and adjustments to the polar satellite communication link, the system will optimize the algorithm adaptation requirements in the hierarchical encryption rules and improve the encryption protection logic during transmission.

[0064] During the full iteration process, all newly labeled data accumulated in the current period will be used, combined with the original basic training sample set, to retrain the random forest classification model. Simultaneously, based on the data application feedback and verification results of the current period, the feature dimensions of the random forest classification model will be adjusted. For example, environmental features specific to polar seasons and status features of new data collection equipment will be added, redundant features that do not significantly improve the model's judgment will be removed, and the model's feature weight allocation will be optimized. This comprehensively improves the model's judgment accuracy and generalization ability, and optimizes the overall data processing and verification logic. Through a combination of real-time updates and periodic iterations, continuous optimization of data processing rules and verification models can be achieved. The scope of data standardization processing can cover more than 95% of commonly used data types in polar scientific expeditions, and the false judgment rate of data verification can be reduced to an extremely low level. This adapts to the ever-changing observation needs and application scenarios in polar scientific expedition cooperation, providing continuous and stable technical support for long-term cross-border polar scientific expedition cooperation.

[0065] The preceding text has fully described the complete implementation process of the cross-border data sharing and verification method for polar scientific expeditions. To adapt to different scales of polar scientific expedition cooperation, compliance requirements, hardware conditions and application scenarios, the following provides multiple alternative implementation paths. All alternative paths do not change the core logic of the method and can be flexibly selected according to actual needs.

[0066] For cross-border cooperation scenarios with explicit encryption algorithm compliance requirements, the algorithms in the encryption processing stage can be replaced accordingly. The AES-128 algorithm used for ordinary-level data can be replaced with the national standard SM4 algorithm. The RSA and AES hybrid encryption mode used for sensitive and core-level data can be replaced with an ECC and SM4 hybrid encryption mode, or a national standard SM2 and SM4 hybrid encryption mode. The data security protection level remains unchanged after the replacement, and it can adapt to the encryption compliance requirements of the corresponding region. For scenarios with limited computing resources on polar research vessels, the classification model embedded in the dual-path verification rules can be replaced. The random forest classification model can be replaced with a lightweight support vector machine model, or a simplified neural network model. Only the corresponding labeled data is needed to complete model training and parameter optimization to achieve the same intelligent classification and judgment effect, while significantly reducing the computing power consumption of model operation and adapting to the hardware conditions of the vessel terminal.

[0067] For polar scientific research projects with few participants and small-scale collaborations, the tiered access rules can be simplified and adjusted. The three-tiered sharing permissions can be reduced to two levels: Level 1 is public data, accessible to all participating parties; Level 2 is restricted data, accessible only after joint authorization from the project leaders of both parties. This simplified access control model is suitable for the actual needs of small projects without affecting core data security control. For scenarios with unstable satellite coverage in high-latitude polar regions, the transmission links can be adjusted. Dual-path polar satellite communication links can be replaced with multi-mode redundant satellite links, compatible with more types of satellite communication systems, improving the continuity of link coverage and transmission stability.

[0068] All of the above implementation paths have been tested and verified in actual polar joint scientific expeditions, and can stably adapt to the complex polar operating environment and various requirements of cross-border cooperation. To intuitively demonstrate the improvement effect of the solutions compared with the original technologies, the core performance indicators measured in the experiments are summarized in the comparison results shown in Table 3: Regarding the methods and basis for obtaining the comparative data in Table 3, the technical team derived these findings through constructing a simulation testing environment, designing control experiments, and conducting empirical analysis using historical data. The specific implementation process is as follows: During the quantitative evaluation phase, to simulate real-world application scenarios, the technical team developed a simulated data generation script based on historical joint scientific expedition metadata. This script generated a test dataset that conformed to the original data formats of both parties and included typical anomaly patterns (such as sensor drift and transmission errors). Simultaneously, the core processing modules of this solution, including a bidirectional mapping library, an encrypted transmission logic simulator, and a dual-path verification module integrating a random forest model, were deployed in an independent sandbox environment to ensure consistency between the test logic and the design scheme. The test platform was equipped with detailed performance monitoring scripts to accurately record the processing time and computational resource consumption of each stage, and automatically compared the output results with preset standard answers.

[0069] The test used a controlled experimental design, divided into two groups: Control group: Used to simulate the existing technical solutions described in the background section. In this group of tests, the automated processing chain of this invention was not enabled. Instead, key time-consuming steps of manual interface conversion were simulated using scripts, and a single threshold verification rule was adopted. The baseline of 7-10 days for manual interface conversion was derived from a survey of the actual time taken by several senior data processing engineers to process similar amounts of data, with the median of 8.5 days taken as a representative comparison benchmark. The verification error of approximately 20% was an empirical value derived by retrospectively analyzing historical archived data and statistically analyzing the proportion of data points confirmed as misjudged or missed under a single standard verification.

[0070] Experimental group: Under the same hardware resource configuration and test data input, the complete automated processing flow of this invention was run in its entirety.

[0071] Based on the above experiments, the core performance indicators were collected and analyzed: Docking efficiency (1.5 days): Through batch processing tests on data from more than 10 simulated voyages, the average time required from data input to completion of the entire standardization process (S1) was calculated to be approximately 1.5 days. This time covers all automated steps, including rule mapping, automatic calibration, and initial screening.

[0072] Verification error (3.8%): Known categories and quantities of anomalous and special phenomenon samples are injected into the test data. After running the dual-path verification (including a random forest model) of this invention, the system outputs the judgment result for each sample. The overall error rate is obtained by calculating the proportion of samples misjudged (normal data is judged as anomalous) and missed (true anomalies are not identified) to the total number of test samples. After multiple rounds of testing and averaging, this error rate stabilizes at around 3.8%, which is significantly lower than the error level of the control group.

[0073] To ensure the simulation results have practical reference value, the initial bidirectional data mapping rule entries and the training set of the random forest model used in the test were all derived from archived, quality-controlled data from real joint scientific expeditions. The core processing logic and code modules were also consistent with the planned deployment version. Therefore, the simulation results can effectively deduce and demonstrate the expected performance improvement of the invention in practical application scenarios.

[0074] Table 3

[0075] Compared to existing conventional processing methods, significant performance improvements have been achieved in several aspects. In terms of data standardization, it enables automated adaptation and conversion of raw polar scientific research data from different sources, significantly shortens the data docking cycle for a single voyage, greatly improves processing efficiency compared to manual processing, and achieves an ideal level of accuracy in data standardization, significantly reducing the workload and human error associated with manual docking. Regarding cross-border data transmission, it is adaptable to the weak signal and high error rate satellite communication environment in the polar regions, with end-to-end transmission latency consistently controlled at a low level, ensuring the integrity of data transmission and significantly reducing the risk of data leakage, tampering, and loss.

[0076] In terms of data verification, the system fully integrates the polar scientific research verification experience of different partners, significantly reducing the average error of data verification. The accuracy of automatic verification is high, accurately distinguishing between data anomalies and real polar environmental phenomena, avoiding the accidental deletion of valid scientific data, and providing reliable data support for subsequent scientific research analysis. Regarding long-term adaptability, through a combination of real-time updates and periodic iterations, it can continuously adapt to newly added data types, acquisition equipment, and collaborative needs. The adaptability of data standardization processing can cover more than 95% of commonly used data types in polar scientific research, and the misjudgment rate of data verification can be reduced to an extremely low level, providing continuous and stable technical support for long-term cross-border polar scientific research cooperation.

[0077] This invention also provides a cross-border data sharing and verification system for polar scientific expeditions, comprising: The acquisition module is used to acquire raw data from polar scientific expeditions, call pre-set bidirectional data mapping rules, perform format conversion, index calibration, and initial screening of invalid data on the raw data from polar scientific expeditions, and obtain standardized data. The transmission module is used to encrypt standardized data by calling pre-set hierarchical encryption rules according to the sensitivity level of the standardized data, and complete the cross-border transmission of encrypted data through the polar satellite communication link. The verification module is used to decrypt the encrypted data received after cross-border transmission. It calls the pre-set dual-path verification rules to perform parallel verification on the decrypted data and obtain the corresponding data verification results. The matching module is used to match access permissions for cross-border sharing of data that has passed verification based on the data verification results and pre-set hierarchical access rules.

[0078] The system embodiments provided in this invention correspond to the method embodiments described above and have at least some of the same technical features. Therefore, they can also achieve at least some of the same technical effects, which will not be repeated here.

[0079] The above description is merely a preferred embodiment of the present invention and the technical principles employed. The present invention is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the scope of protection of the present invention. Therefore, although the present invention has been described in detail through the above embodiments, the present invention is not limited to the above embodiments, and may include many other equivalent embodiments without departing from the concept of the present invention.

Claims

1. A method for cross-border data sharing and verification in polar scientific expeditions, characterized in that, Includes the following steps: S1. Obtain raw data from polar scientific expeditions, call pre-set bidirectional data mapping rules, perform format conversion, index calibration, and initial screening of invalid data on the raw data from polar scientific expeditions to obtain standardized data; S2. Based on the sensitivity level of the standardized data, a pre-set hierarchical encryption rule is invoked to encrypt the standardized data, and the encrypted data is transmitted across borders via a polar satellite communication link. S3. Decrypt the encrypted data received after cross-border transmission, call the pre-set dual-path verification rules to perform parallel verification on the decrypted data, and obtain the corresponding data verification result. S4. Based on the data verification results and the pre-set hierarchical access rules, complete the cross-border sharing of data that has passed the verification by matching access permissions.

2. The method according to claim 1, characterized in that, The bidirectional data mapping rules include the corresponding conversion relationships of indicator definitions, format parameters, and verification thresholds for core data types in polar scientific research, as well as correction rules for systematic deviations of sensors, and a reserved rule extension interface.

3. The method according to claim 1, characterized in that, In step S1, the initial screening of invalid data specifically includes the following steps: Based on the physical threshold of the polar environment, invalid data points in the original polar scientific expedition data that do not meet the physical threshold of the polar environment are removed; An unsupervised anomaly detection algorithm is used to identify hidden anomalies in the original polar scientific research data. Multiple anomaly detection algorithms are combined for cross-validation. The verified hidden anomalies are marked as pending review, and the causes of the anomalies and corresponding metadata are recorded simultaneously.

4. The method according to claim 1, characterized in that, In step S2, the encryption process for the standardized data specifically includes the following steps: The standardized data at the ordinary level is encrypted using a symmetric encryption algorithm; For the standardized data of the sensitive level, a hybrid mode combining asymmetric encryption algorithm and symmetric encryption algorithm is used for encryption; For the standardized data at the core level, dual-key verification is added on top of the hybrid encryption method.

5. The method according to claim 1, characterized in that, In step S2, the cross-border transmission tunnel adopts a layered protocol, the control channel adopts the TCP protocol, and the data channel adopts the UDP protocol. The transmission process is based on digital certificates to complete two-way identity authentication. The cross-border transmission of encrypted data through the polar satellite communication link specifically includes the following steps: The signal quality of the dual-path polar satellite communication link is monitored in real time, and the primary and backup links are switched according to the signal quality, or the encrypted data packets are split and transmitted in parallel through the dual-path links. The data packet size is adjusted in real time according to the signal quality. Forward error correction coding is added to the data packet before transmission. During transmission, a check digest is generated according to a fixed data length. The check digest is used by the receiving end for real-time comparison. Retransmit lost or tampered data packets.

6. The method according to claim 3, characterized in that, In step S3, the step of calling a pre-set dual-path verification rule to perform parallel verification of the decrypted data specifically includes the following steps: The decrypted data is verified simultaneously using two verification algorithms. One verification algorithm verifies the consistency of data trends, while the other verification algorithm verifies extreme values ​​and abnormal patterns in the data. For decrypted data whose verification results are within a pre-set error threshold, a qualified data verification result is generated; for decrypted data whose verification results exceed the error threshold, the cause of the difference is located and a review process is initiated. and, The hidden anomaly data marked as pending review undergoes encryption, cross-border transmission, and decryption processing synchronously with the standardized data; after step S3, the method further includes: The hidden abnormal data marked as pending verification and the decrypted data whose parallel verification exceeds the error threshold are manually verified. Data that is determined to be valid after verification is added to the data verification result, and data that is determined to be invalid after verification is removed.

7. The method according to claim 6, characterized in that, The dual-path verification rule embeds a pre-trained random forest classification model. The input features of the random forest classification model include data difference features, data statistical features, spatiotemporal and environmental background features, and verification status features. The output results of the random forest classification model are divided into four categories: qualified data, abnormal data, special phenomena, and standard divergence.

8. The method according to claim 1, characterized in that, S4 specifically includes the following steps: Access to qualified Level 1 public data will be granted to all partner institutions. Access to qualified Level 2 restricted data will be granted after joint authorization from the project leaders of both parties. Access to qualified Level 3 core data is granted only to designated core personnel and requires dual key verification.

9. The method according to claim 7, characterized in that, The method further includes the following steps: S5. Collect the correction results of manual review, update the corresponding parameters of the bidirectional data mapping rule and the dual-path verification rule in real time, and use the data samples completed by manual review as newly labeled data to incrementally train the random forest classification model. S6. Summarize the correction results and data application feedback information collected in the current period according to a fixed period, and perform a full iterative update on the bidirectional data mapping rules, the hierarchical encryption rules, the dual-path verification rules, and the hierarchical access rules. Retrain the random forest classification model using all the newly labeled data accumulated in the current period, adjust the feature dimensions of the random forest classification model, and optimize the data processing and verification logic.

10. A cross-border data sharing and verification system for polar scientific expeditions, used to implement the method described in any one of claims 1-9, characterized in that, include: The acquisition module is used to acquire raw data from polar scientific expeditions, call pre-set bidirectional data mapping rules, perform format conversion, index calibration, and initial screening of invalid data on the raw data from polar scientific expeditions, and obtain standardized data. The transmission module is used to encrypt the standardized data according to the sensitivity level of the standardized data by calling a pre-set hierarchical encryption rule, and to complete the cross-border transmission of the encrypted data through the polar satellite communication link. The verification module is used to decrypt the encrypted data received after cross-border transmission, and to perform parallel verification of the decrypted data by calling the pre-set dual-path verification rules to obtain the corresponding data verification result. The matching module is used to perform cross-border sharing of qualified data by matching access permissions based on the data verification results and pre-set hierarchical access rules.

Images