Radioactivity detection information report generation method, device and equipment

Abstract

The invention relates to the technical field of seafood radioactivity detection, in particular to a radioactivity detection information report generation method, device and equipment. Performing abnormal value detection on historical radioactivity detection data based on a quartile distance method; constructing a risk assessment model based on an analytic hierarchy process, an abnormal value judgment lower limit threshold and an abnormal value judgment upper limit threshold; performing logical analysis on the real-time radioactivity detection data according to the risk assessment model and a preset logical reasoning algorithm; when the logic analysis result meets the logic relation condition, constructing a radioactivity detection report template according to the logic analysis result; performing data filling on the radioactivity detection report template according to the real-time radioactivity detection data to obtain a radioactivity detection information report; according to the scheme, through abnormal identification, model adaptability improvement, comprehensive risk quantification and automatic report generation, marine product safety and consumer health are guaranteed, and supervision decision is assisted.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of seafood radioactivity detection, and specifically relates to a radioactivity detection information report generation method, device and equipment. BACKGROUND

[0002] In the field of seafood radioactivity detection, the existing technology has significant defects in the report generation link. Generally, a fixed template is used to generate a report, and the report content and structure are not adjusted flexibly according to the logical relationship of the detection data. This fixed template lacks pertinence and cannot be dynamically optimized according to the diversity and complexity of the detection data in different scenarios. For example, when dealing with different types of seafood or samples with different pollution levels, the report fails to highlight key risk indicators or provide targeted recommendations. This fixed report mode not only reduces detection efficiency, but also may lead to misjudgment due to the inability to accurately reflect the detection results. Therefore, there is an urgent need for a method that can automatically evaluate risks and generate flexible reports to improve detection efficiency and accuracy and ensure the scientificity and reliability of food safety supervision. SUMMARY

[0003] To solve the above-mentioned shortcomings of the prior art, the present application proposes a radioactivity detection information report generation method.

[0004] To solve the above-mentioned technical problems, the technical solutions adopted by the present application are as follows: A radioactivity detection information report generation method, comprising: obtaining historical radioactivity detection data, and performing abnormal value detection on the historical radioactivity detection data based on a quartile range method to obtain an abnormal value determination lower threshold value and an abnormal value determination upper threshold value; constructing a risk assessment model based on an analytic hierarchy process method, the abnormal value determination lower threshold value and the abnormal value determination upper threshold value; obtaining real-time radioactivity detection data; performing logical analysis on the real-time radioactivity detection data according to a risk assessment model and a preset logical reasoning algorithm to obtain a logical analysis result; determining whether the logical analysis result meets a preset logical relationship condition; when the logical analysis result meets the logical relationship condition, constructing a radioactivity detection report template according to the logical analysis result; and filling data in the radioactivity detection report template according to the real-time radioactivity detection data to obtain a radioactivity detection information report.

[0005] Furthermore, the outlier detection of historical radioactivity detection data based on the interquartile range method to obtain a lower threshold and an upper threshold for outlier determination includes: sorting the historical radioactivity detection data according to a preset sorting algorithm and a preset baseline quantile to obtain sorted detection data; performing outlier detection on the sorted detection data based on the interquartile range method to obtain a first quartile and a second quartile; calculating the difference between the second quartile and the first quartile to obtain the interquartile range; calculating the lower threshold for outlier determination based on a preset outlier determination coefficient, the first quartile, and the interquartile range; and calculating the upper threshold for outlier determination based on the outlier determination coefficient, the second quartile, and the interquartile range.

[0006] Furthermore, the risk assessment model constructed based on the analytic hierarchy process (AHP), the lower threshold for outlier determination, and the upper threshold for outlier determination includes: constructing a primary assessment model based on the AHP and a preset fuzzy comprehensive risk assessment index; extracting all features from historical radioactive detection data to obtain sample feature vectors; generating kernel function parameters based on the lower threshold for outlier determination and the upper threshold for outlier determination; and training the primary assessment model based on the sample feature vectors and the kernel function parameters to obtain the risk assessment model.

[0007] Furthermore, the step of performing logical analysis on real-time radioactivity detection data based on a risk assessment model and a preset logical reasoning algorithm to obtain logical analysis results includes: preprocessing the real-time radioactivity detection data to obtain cross-modal aligned feature vectors; performing risk assessment on the cross-modal aligned feature vectors based on the risk assessment model to obtain a risk assessment score; and performing logical analysis on the cross-modal fused feature vectors based on the logical reasoning algorithm and the risk assessment score to obtain logical analysis results.

[0008] Furthermore, the preprocessing of real-time radioactivity detection data to obtain cross-modal aligned feature vectors includes: extracting features from the real-time radioactivity detection data based on the covariance method to obtain energy spectrum features; obtaining a knowledge information table according to a preset intelligent chart selection algorithm; extracting features from the real-time radioactivity detection data based on a preset cosine similarity algorithm and the knowledge information table to obtain knowledge information features; and aligning the knowledge information features and energy spectrum features based on a preset attention mechanism to obtain cross-modal aligned feature vectors.

[0009] Furthermore, the alignment processing of knowledge information features and energy spectrum features based on a preset attention mechanism to obtain a cross-modal aligned feature vector includes: filtering the knowledge information features according to a preset confidence level and a preset wavelet transform denoising algorithm to obtain high-confidence knowledge information features; performing weight analysis on the energy spectrum features and high-confidence knowledge information features based on the attention mechanism to obtain adaptive weight features; and aligning the high-confidence knowledge information features and energy spectrum features according to the adaptive weight features to obtain a cross-modal aligned feature vector.

[0010] Furthermore, the step of constructing a radioactive detection report template based on the logical analysis results includes: extracting features from cross-modal aligned feature vectors based on a preset text feature extraction algorithm and a preset occurrence frequency threshold to obtain text features and data distribution features; generating design data items and configuring a learning framework based on the text features, data distribution features, and preset report display requirements; and constructing a radioactive detection report template based on the logical analysis results, design data items, and configuring the learning framework.

[0011] Furthermore, after the step of filling the radioactivity detection report template with data based on real-time radioactivity detection data to obtain a radioactivity detection information report, the method further includes: performing a hash operation on the radioactivity detection information report based on a preset hash function to obtain a hash value; obtaining the acquisition time and acquisition device number of the real-time radioactivity detection data; packaging the radioactivity detection information report, hash value, acquisition time, and acquisition device number into blocks according to a preset consensus algorithm to obtain blockchain evidence; and performing sharded storage of the blockchain evidence based on a preset consortium blockchain to obtain sharded on-chain evidence.

[0012] Furthermore, a radioactive detection information report generation device includes: an outlier detection module, used to acquire historical radioactive detection data and perform outlier detection on the historical radioactive detection data based on the interquartile range method to obtain a lower threshold and an upper threshold for outlier judgment; a model building module, used to build a risk assessment model based on the analytic hierarchy process (AHP), the lower threshold and the upper threshold for outlier judgment; a data acquisition module, used to acquire real-time radioactive detection data; a logic analysis module, used to perform logical analysis on the real-time radioactive detection data according to the risk assessment model and a preset logical reasoning algorithm to obtain a logical analysis result; a condition judgment module, used to judge whether the logical analysis result meets preset logical relationship conditions; a template generation module, used to build a radioactive detection report template based on the logical analysis result when the logical analysis result meets the logical relationship conditions; and a detection information report generation module, used to fill the radioactive detection report template with data based on the real-time radioactive detection data to obtain a radioactive detection information report.

[0013] Furthermore, a radioactive detection information report generation device includes: a memory and at least one processor, wherein the memory stores instructions; at least one processor invokes the instructions in the memory to cause the radioactive detection information report generation device to perform the various steps of a radioactive detection information report generation method as described in any one of the above descriptions.

[0014] The beneficial effects of the method for generating a radioactive detection information report according to the present invention are as follows: The interquartile range method is used to analyze historical data, effectively identifying outliers caused by factors such as equipment failure and sample contamination. The resulting outlier threshold serves as a real-time monitoring benchmark and is deeply integrated into the risk assessment model, significantly improving the model's sensitivity and adaptability to data fluctuations. The risk assessment model, constructed using the analytic hierarchy process (AHP), considers multiple factors such as radionuclide concentration, seafood type, and testing environment. Scientific weighting enables comprehensive quantitative risk assessment, making the results more objective and accurate. Logical reasoning algorithms are used for in-depth analysis of real-time data. If logical relationship conditions are met, the system automatically generates targeted and adaptive testing report templates, providing a scientific and efficient decision support tool for food safety supervision and effectively safeguarding seafood quality and consumer health. Attached Figure Description

[0015] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a first flowchart of a method for generating a radioactive detection information report according to an embodiment of the present invention; Figure 2 This is a second flowchart of a method for generating a radioactive detection information report provided in an embodiment of the present invention; Figure 3 This is a third flowchart of a method for generating a radioactive detection information report provided in an embodiment of the present invention; Figure 4 This is a fourth flowchart of a method for generating a radioactive detection information report provided in an embodiment of the present invention; Figure 5 The fifth flowchart of a method for generating a radioactive detection information report provided in an embodiment of the present invention; Figure 6 The sixth flowchart of a method for generating a radioactive detection information report provided in an embodiment of the present invention; Figure 7 The seventh flowchart of a method for generating a radioactive detection information report provided in an embodiment of the present invention; Figure 8The eighth flowchart of a method for generating a radioactive detection information report provided in an embodiment of the present invention; Figure 9 This is a schematic diagram of the structure of a radioactive detection information report generation device provided in an embodiment of the present invention; Figure 10 This is a schematic diagram of a radioactive detection information report generation device provided in an embodiment of the present invention. Detailed Implementation

[0016] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0018] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 An embodiment of a method for generating a radioactive detection information report according to the present invention includes: 101. Obtain historical radioactivity detection data and perform outlier detection on the historical radioactivity detection data based on the interquartile range method to obtain the lower threshold and upper threshold for outlier judgment. In this embodiment, this method can effectively identify outliers in historical data, providing a reliable basis for identifying subsequent real-time outliers. Outliers in historical radioactivity detection data often originate from equipment failure, sample contamination, operational errors, and other issues. The interquartile range method is based on the quantile characteristics of data, is not affected by extreme values, and can locate outliers. The obtained upper and lower thresholds for outlier judgment can be used as a benchmark for real-time data monitoring. The construction of subsequent risk assessment models depends on the training and optimization of the upper and lower thresholds for outlier judgment and historical data. 102. A risk assessment model was constructed based on the analytic hierarchy process (AHP), the lower threshold for outlier detection, and the upper threshold for outlier detection. In this embodiment, the analytic hierarchy process (AHP) decomposes multiple influencing factors of risk assessment into target layer, criterion layer, and indicator layer. A judgment matrix is ​​constructed through the target layer, criterion layer, and indicator layer to determine the weight of each factor. The outlier threshold serves as an important basis for measuring the risk of data fluctuation and is incorporated into the model. By comprehensively considering factors such as different radionuclide concentrations, seafood types, and detection environment, a scientific assessment of radioactive risk can be achieved. 103. Obtain real-time radioactivity detection data; 104. Based on the risk assessment model and the preset logical reasoning algorithm, perform logical analysis on the real-time radioactivity detection data to obtain the logical analysis results; In this embodiment, the logical reasoning algorithm performs in-depth analysis of real-time data based on the output of the risk assessment model and in combination with preset rules (such as the risk level corresponding to the excessive concentration of different nuclides, the risk sensitivity of different types of seafood, etc.), to obtain logical analysis results and clarify the risk status and key information reflected by the current detection data. 105. Determine whether the logical analysis result meets the preset logical relationship conditions; 106. When the logical analysis results meet the logical relationship conditions, a radioactivity detection report template is constructed based on the logical analysis results; 107. Fill the radioactivity detection report template with data based on real-time radioactivity detection data to obtain a radioactivity detection information report; In this embodiment, the logical relationship conditions include: risk level association conditions (for example, when the logical analysis results show that the risk level corresponding to the concentration of a certain type of radionuclide reaches "high risk" (such as exceeding the upper limit threshold for outlier judgment and exceeding the standard limit ratio range), the logical relationship conditions are met, and the construction of a high-risk report template can be triggered; if it is "medium risk", then the corresponding medium-risk report template generation rules apply), nuclide concentration combination conditions, detection environment and data linkage conditions, and historical data comparison conditions; the content structure and display focus of the detection report template are generated according to the logical analysis results, and finally, the real-time radioactive detection data is filled into the report template to generate a complete radioactive detection information report; In this embodiment, the interquartile range method is used to analyze historical data, effectively identifying outliers caused by factors such as equipment failure and sample contamination. Simultaneously, the outlier threshold generated based on this method serves as a real-time monitoring benchmark and is deeply integrated into the risk assessment model, significantly improving the model's sensitivity and adaptability to data fluctuation risks. The risk assessment model, constructed using the analytic hierarchy process (AHP), considers multiple factors such as radionuclide concentration, seafood type, and testing environment. Scientific weighting achieves a comprehensive quantitative assessment of risk, making the assessment results more objective and accurate. Logical reasoning algorithms are used to perform in-depth analysis of real-time data. If logical relationship conditions are met, the system automatically generates targeted and adaptive testing report templates, providing a scientific and efficient decision support tool for food safety supervision and effectively safeguarding the quality and safety of seafood and consumer health.

[0019] Please see Figure 2 A second embodiment of a method for generating a radioactive detection information report according to an embodiment of the present invention includes: 201. Sort the historical radioactivity detection data according to the preset sorting algorithm and preset benchmark quantiles to obtain sorted detection data; In this embodiment, the selection of sorting algorithm needs to comprehensively consider the data scale and computational efficiency. Quick sort has an average time complexity of O(nlogn) and is suitable for large-scale data processing. Merge sort has the characteristic of stable sorting, which can ensure that the relative order of elements with the same value remains unchanged. The setting of the benchmark quantile can help determine the key range of data sorting. For example, in massive historical data, data in key time periods or under specific detection conditions can be selected as the benchmark for sorting, thereby obtaining the sorted detection data. This step lays the foundation for subsequent analysis based on the data order characteristics, making the data present an ordered distribution state, which is convenient for quantile calculation and outlier identification. 202. Outlier detection is performed on the sorted detection data based on the interquartile range method to obtain the first quartile and the second quartile; 203. Calculate the interquartile range by subtracting the first quartile from the second quartile. In this embodiment, the formula for calculating the interquartile range is as follows: ,in, Interquartile range, It is the second quantile. It is the first quantile; 204. The lower threshold for outlier determination is calculated based on the preset outlier determination coefficient, the first quartile, and the interquartile range. In this embodiment, the formula for calculating the lower limit threshold for outlier detection is as follows: , The formula for calculating the upper limit threshold for outlier detection is as follows: ,in, This is the outlier determination coefficient; 205. The upper limit threshold for outlier determination is calculated based on the outlier determination coefficient, the second quartile, and the interquartile range. In this embodiment, a preset sorting algorithm is used to flexibly select the sorting method according to the data scale and computational needs. The sorting of key data is focused on the benchmark quantile, laying a solid orderly foundation for subsequent analysis. The first and second quartiles and the interquartile range are calculated based on the interquartile range method to build a stable outlier detection framework. The algorithm is not affected by extreme values ​​and can locate abnormal data. This solution provides real and effective data support for radioactive risk assessment, trend analysis, etc., enhances the credibility of radioactive detection information reports, and provides strong data evidence and decision-making basis for scenarios such as food safety supervision and seafood trade.

[0020] Please see Figure 3 A third embodiment of a method for generating a radioactive detection information report according to the present invention includes: 301. A preliminary assessment model is constructed based on the analytic hierarchy process and pre-defined fuzzy comprehensive risk assessment indicators; In this embodiment, the multiple risk influencing factors of the fuzzy comprehensive risk assessment index are decomposed into target layer, criterion layer and indicator layer by the hierarchical analysis method. A judgment matrix is ​​constructed by the target layer, criterion layer and indicator layer. The weight of each risk influencing factor can be determined by the judgment matrix, thereby constructing a primary assessment model and initially establishing a risk assessment framework. 302. Perform full feature extraction on historical radioactivity detection data to obtain sample feature vectors; 303. Generate kernel function parameters based on the lower and upper thresholds for outlier detection; 304. Train the primary assessment model based on the sample feature vector and kernel function parameters to obtain the risk assessment model; In this embodiment, full feature extraction is performed on historical radioactivity detection data to comprehensively mine the information contained in the data. In addition to conventional radionuclide concentration data, it also includes multi-dimensional features such as the physicochemical properties of the sample (e.g., water content and fat content, which may affect the detection and distribution of radioactive substances) and detection process parameters (e.g., detection equipment model and detection time interval). Through feature engineering technology, this information is transformed into sample feature vectors, with each vector representing a feature set of a historical detection sample, providing rich data input for model training. Kernel function parameters are configured to enable the model to adapt to the data distribution characteristics of different datasets. In this embodiment, by combining hierarchical analysis and fuzzy comprehensive risk assessment, risk factors are scientifically decomposed and weights are determined, effectively handling uncertainties in the assessment and establishing a rigorous primary assessment framework. Full feature extraction is performed on historical data, covering multi-dimensional information such as nuclide concentration, sample physicochemical properties, and detection parameters, providing comprehensive data support for model training. Kernel function parameters are generated based on outlier thresholds, enabling the model to adapt to different data distributions. The resulting risk assessment model not only improves the accuracy and reliability of risk assessment but also enhances the model's adaptability to complex scenarios, providing scientific conclusions for radioactive detection reports and strong decision-making support for food safety supervision and seafood trade.

[0021] Please see Figure 4 A fourth embodiment of a method for generating a radioactive detection information report according to the present invention includes: 401. Preprocess the real-time radioactivity detection data to obtain cross-modal aligned feature vectors; In this embodiment, radioactivity detection data from different modalities are mapped to the same feature space to construct cross-modal aligned feature vectors, enabling the data from each modality to complement and co-express each other, providing a unified data format for subsequent analysis; 402. Perform risk assessment on the cross-modal alignment feature vectors based on the risk assessment model to obtain a risk assessment score; In this embodiment, the model comprehensively considers multiple dimensions of information such as radionuclide concentration, sample physicochemical properties, and detection environment based on the input feature vector. Through internal weight calculation and nonlinear mapping, it outputs a risk assessment score, which quantifies the degree of radioactivity risk of the real-time detected sample and provides a quantitative basis for subsequent logical analysis. 403. Perform logical analysis on the cross-modal fusion feature vector based on logical reasoning algorithms and risk assessment scores to obtain the logical analysis results; In this embodiment, the logical reasoning algorithm pre-sets a variety of logical rules. Based on the output of the risk assessment model, combined with the pre-set rules (such as the risk level corresponding to the excessive concentration of different nuclides, the risk sensitivity of different types of seafood, etc.), the corresponding reasoning process is initiated to perform in-depth analysis of real-time data, obtain logical analysis results, and clarify the risk status and key information reflected by the current detection data. In this embodiment, detection data from different modalities are mapped to the same feature space to construct cross-modal aligned feature vectors, enabling collaborative expression of data from various modalities and providing a unified data format for subsequent analysis. Based on a risk assessment model, multi-dimensional information is comprehensively considered to output a quantitative risk assessment score, quantifying the radioactivity risk of the sample. Through logical reasoning algorithms, the data is deeply analyzed according to preset rules to clarify the risk status and key information. This solution not only improves detection efficiency but also enhances the scientific nature and relevance of risk assessment, providing strong support for food safety supervision and nuclear pollution monitoring. It is particularly suitable for complex and diverse detection scenarios, effectively reducing the false alarm rate of reports based on radiation detection information and improving the reliability of decision-making.

[0022] Please see Figure 5 A fifth embodiment of a method for generating a radioactive detection information report according to the present invention includes: 501. Feature extraction of real-time radioactivity detection data based on the covariance method to obtain energy spectrum features; In this embodiment, the real-time radioactivity detection data contains multiple dimensions (such as radioactivity counts in different energy bands). The correlation between data in each dimension can be analyzed by using the covariance method. A data matrix is ​​constructed for the real-time radioactivity detection data, and the covariance between data in each dimension in the matrix is ​​calculated to form a covariance matrix. This matrix reflects the relationship between radioactivity counts in different energy bands. By performing eigenvalue decomposition (such as singular value decomposition SVD) on the covariance matrix, the main eigenvectors and eigenvalues ​​are extracted to obtain the energy spectrum features, thereby realizing the extraction of energy spectrum features from the radioactivity energy spectrum data. 502. Obtain the knowledge information table according to the preset intelligent chart selection algorithm; In this embodiment, the intelligent chart selection algorithm obtains a knowledge information table based on the analysis of the type (numerical, time series, etc.), data scale, and data distribution characteristics of real-time radioactivity detection data. For example, if the data is radioactivity intensity data that changes over time, the algorithm selects a line chart to show the trend; if the data is a comparison of the content of radionuclides in different samples, a bar chart is selected. Based on the selected chart type, combined with knowledge and standards in the field of radioactivity detection, a knowledge information table is generated. 503. Based on the preset cosine similarity algorithm and knowledge information table, feature extraction is performed on the real-time radioactivity detection data to obtain knowledge information features; In this embodiment, by calculating the cosine similarity between the real-time data vector and the standard knowledge information vector, the degree of matching and difference between the real-time data and the standard knowledge can be determined. For example, if the cosine similarity between the real-time data vector and the standard vector when a certain type of radionuclide exceeds the standard is high, it indicates that the current detection data has the characteristic tendency of exceeding the standard for that type of radionuclide. Based on the similarity calculation results, key features related to knowledge information are extracted to obtain knowledge information features, and the radioactive knowledge and risk information hidden behind the data are mined. 504. Based on a pre-defined attention mechanism, knowledge information features and energy spectrum features are aligned to obtain cross-modal aligned feature vectors; In this embodiment, the attention mechanism calculates the weights between different feature dimensions to highlight the features that are important for radioactive risk assessment. It then fuses and aligns the two features through weighted summation and other methods, mapping the features of different modalities (energy spectrum data modality and knowledge information modality) to the same feature space. Finally, it obtains a cross-modal aligned feature vector, providing a unified and information-rich data representation for subsequent risk assessment, logical analysis, etc. In this embodiment, the covariance method is used to analyze the correlation of data across various dimensions and extract energy spectrum features. An intelligent chart selection algorithm matches the optimal visualization format based on data characteristics, and combined with professional knowledge, a knowledge information table is generated, providing a clear framework for data interpretation. A cosine similarity algorithm is used to compare real-time and standard knowledge vectors to capture risk characteristics such as excessive nuclide levels. An attention mechanism dynamically balances the weights of energy spectrum and knowledge information features, achieving precise alignment and fusion of cross-modal features. This solution provides a unified and valuable data foundation for subsequent risk assessment and logical analysis, significantly improving the accuracy, efficiency, and intelligence of radioactive detection, and strongly supporting scientific decision-making in scenarios such as food safety supervision.

[0023] Please see Figure 6 The sixth embodiment of a method for generating a radioactive detection information report in this invention includes: 601. Filter the knowledge information features according to the preset confidence level and the preset wavelet transform denoising algorithm to obtain high-confidence knowledge information features; In this embodiment, a wavelet transform denoising algorithm is used in conjunction with a pre-set confidence level (such as a feature reliability probability threshold) to filter out features with high confidence. For example, abnormal feature vectors with cosine similarity below the threshold are removed, and finally high-confidence knowledge information features are obtained while retaining effective features. 602. Based on the attention mechanism, perform weight analysis on the energy spectrum features and high-confidence knowledge information features to obtain adaptive weight features; 603. Align the high-confidence knowledge information features and energy spectrum features according to the adaptive weight features to obtain cross-modal aligned feature vectors; In this embodiment, wavelet transform denoising algorithm and pre-set confidence level are used to perform dual screening of knowledge information features, effectively eliminating abnormal noise and low-confidence data, while retaining valid high-confidence knowledge information features, thus consolidating the foundation of data reliability. An attention mechanism is introduced to dynamically analyze energy spectrum features and high-confidence knowledge information features, and a unified feature vector is formed through cross-modal alignment. This provides efficient and accurate data support for subsequent risk assessment and logical analysis, enhances the credibility and decision-making value of detection results, and assists in scientific judgment and rapid response in scenarios such as food safety supervision.

[0024] Please see Figure 7 The seventh embodiment of a method for generating a radioactive detection information report in this invention includes: 701. Based on a preset text feature extraction algorithm and a preset occurrence frequency threshold, feature extraction is performed on cross-modal aligned feature vectors to obtain text features and data distribution features; In this embodiment, text feature extraction algorithms (such as TF-IDF, BERT word vectors, etc.) are used to process the text information (such as detection sample descriptions, environmental descriptions, etc.) contained in the cross-modal alignment feature vectors, extracting key text features and mining the semantic information and key terms contained in the text. Simultaneously, by combining frequency thresholds, feature words are filtered to highlight important content in the text, thus obtaining the final text features. For data distribution feature extraction, numerical data (such as radionuclide concentrations, detection time series data, etc.) in the cross-modal alignment feature vectors are analyzed. By calculating statistics such as mean, variance, and quantiles, the central tendency, dispersion, and distribution pattern of the data are obtained. At the same time, the data distribution pattern is determined based on the data frequency thresholds, thereby obtaining comprehensive data distribution characteristics and providing a data-level basis for subsequent report content design. 702. Generate design data items and configure the learning framework based on text features, data distribution characteristics, and preset report display requirements; In this embodiment, based on the extracted text features and data distribution features, and combined with the report display requirements (such as whether the report is aimed at regulatory authorities, researchers, or ordinary consumers, as different audiences have different requirements for the focus and presentation of information), design data items are generated. For example, if the report is aimed at regulatory authorities, the design data items include detailed information on the exceeding of radionuclide limits and risk level assessment results; if it is aimed at ordinary consumers, the focus is on easy-to-understand risk warnings and consumption advice. 703. Based on the logical analysis results, design data items, and configure the learning framework, a radioactive detection report template is constructed. In this embodiment, configuring the learning framework is based on designing data items and logical analysis results to determine the organization, display method, and interaction logic of the report content; for example, selecting appropriate chart types (bar charts, line charts, heatmaps, etc.) to display data distribution characteristics, determining the layout format of text information, and designing the relationships and interactive operations between data items (such as clicking on the chart to view detailed data), providing a structured framework guide for the construction of the report template; In this embodiment, text feature extraction algorithms and frequency thresholds are used to deeply mine text semantics and data distribution features from cross-modal aligned feature vectors. This not only filters out key terms in the descriptions of test samples and environmental descriptions, but also uses statistical analysis to identify the core trends in numerical data, laying a detailed foundation for the report content. Combined with preset report display requirements, customized data items are generated for different audiences such as regulatory authorities, researchers, and ordinary consumers, ensuring the targeted delivery of information. Based on a configuration learning framework, the report structure, visualization methods, and interaction logic are scientifically planned, organically integrating logical analysis results with data features to form a standardized and structured report template. This solution balances data depth and display flexibility, improving report generation efficiency while enhancing information readability and decision-making reference value, providing strong support for applications in multiple scenarios such as food safety supervision and public science education.

[0025] Please see Figure 8 The eighth embodiment of a method for generating a radioactive detection information report according to the present invention includes: 801. Perform a hash operation on the radioactivity detection information report based on a preset hash function to obtain a hash value; In this embodiment, the hash function has one-way and uniqueness. After performing a hash operation on the detection report, the resulting hash value can be used as the "digital fingerprint" of the report to quickly verify the integrity of the report content. For example, if the report is tampered with during transmission or storage, the recalculated hash value will be inconsistent with the original hash value, thereby timely detecting data anomalies. 802. Obtain the acquisition time and acquisition device number of real-time radioactivity detection data; 803. Based on the preset consensus algorithm, the radioactivity detection information report, hash value, collection time, and collection device number are packaged into blocks to obtain blockchain evidence; In this embodiment, the radioactivity detection information report, hash value, collection time, and collection device number are integrated and packaged according to a consensus algorithm (such as Practical Byzantine Fault Tolerance (PBFT) or Proof-of-Stake (PoS). The consensus algorithm ensures that multiple participating nodes agree on the data. Only data that passes the consensus verification can be packaged into a block. In the block, the data items are interconnected. The detection report serves as the core content, the hash value is used for integrity verification, and the collection time and device number provide traceability information. Each block contains the hash value of the previous block, forming a chain structure to ensure the immutability and traceability of the data. 804. Based on the preset consortium blockchain, the blockchain evidence is stored in shards to obtain sharded on-chain evidence. In this embodiment, the consortium blockchain consists of multiple authorized nodes, featuring partial decentralization and controllability. Sharding storage divides data into multiple fragments and distributes them across different nodes, which improves storage efficiency and data processing speed while enhancing data security. Different nodes are responsible for storing and managing different data fragments, and only authorized and verified nodes can access and operate the corresponding fragment data, effectively preventing data leakage and malicious attacks. In this embodiment, hash operations, consensus algorithms, and consortium blockchain sharding storage technology are used to build a complete evidence storage system for radioactivity detection information; the integrity of the report is ensured through "digital fingerprints," the immutability and traceability of data are achieved through blockchain, and security and efficiency are improved through sharding storage, so as to comprehensively ensure that the detection data is authentic, reliable, secure and controllable, and provide a solid basis for supervision and decision-making.

[0026] The above describes a method for generating a radioactive detection information report according to an embodiment of the present invention. The following describes a device for generating a radioactive detection information report according to an embodiment of the present invention. Please refer to [link / reference]. Figure 9 One embodiment of the radioactive detection information report generation device of the present invention includes: The outlier detection module 1 is used to acquire historical radioactivity detection data and perform outlier detection on the historical radioactivity detection data based on the interquartile range method to obtain the lower threshold and upper threshold for outlier judgment. Model building module 2 is used to build a risk assessment model based on the analytic hierarchy process, the lower threshold for outlier detection, and the upper threshold for outlier detection. Data acquisition module 3 is used to acquire real-time radioactivity detection data; The logic analysis module 4 is used to perform logical analysis on real-time radioactivity detection data based on the risk assessment model and the preset logical reasoning algorithm to obtain the logical analysis results. Condition judgment module 5 is used to determine whether the logical analysis result meets the preset logical relationship conditions; Template generation module 6 is used to construct a radioactivity detection report template based on the logical analysis results when the logical analysis results meet the logical relationship conditions; The detection information report generation module 7 is used to populate the radioactivity detection report template with data based on real-time radioactivity detection data to obtain a radioactivity detection information report.

[0027] In this embodiment, the interquartile range method is used to analyze historical data, effectively identifying outliers caused by factors such as equipment failure and sample contamination. Simultaneously, the outlier threshold generated based on this method serves as a real-time monitoring benchmark and is deeply integrated into the risk assessment model, significantly improving the model's sensitivity and adaptability to data fluctuation risks. The risk assessment model, constructed using the analytic hierarchy process (AHP), considers multiple factors such as radionuclide concentration, seafood type, and testing environment. Scientific weighting achieves a comprehensive quantitative assessment of risk, making the assessment results more objective and accurate. Logical reasoning algorithms are used to perform in-depth analysis of real-time data. If logical relationship conditions are met, the system automatically generates targeted and adaptive testing report templates, providing a scientific and efficient decision support tool for food safety supervision and effectively safeguarding the quality and safety of seafood and consumer health.

[0028] Figure 10 This is a schematic diagram of the structure of a radioactive detection information report generation device 900 provided in an embodiment of the present invention. This radioactive detection information report generation device 900 can vary considerably due to different configurations or performance. It may include one or more central processing units (CPUs) 910 (e.g., one or more processors) and a memory 920, and one or more media 930 (e.g., one or more mass storage devices) storing application programs 933 or data 932. The memory 920 and media 930 can be temporary or persistent storage. The program stored in the media 930 may include one or more modules (not shown in the diagram), each module may include a series of instruction operations on the radioactive detection information report generation device 900. Furthermore, the processor 910 may be configured to communicate with the media 930 and execute a series of instruction operations in the media 930 on the radioactive detection information report generation device 900 to implement the steps of the radioactive detection information report generation method provided in the above-described method embodiments.

[0029] A radioactivity detection information report generation device 900 may further include one or more power supplies 940, one or more wired or wireless network interfaces 950, one or more input / output interfaces 960, and / or one or more operating systems 931, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 9 The illustrated structure of a radioactive detection information report generation device does not constitute a limitation on a radioactive detection information report generation device 900. It may include more or fewer components than illustrated, or combine certain components, or have different component arrangements.

[0030] The present invention and its embodiments have been described above. This description is not restrictive. The accompanying drawings are only one embodiment of the present invention. The actual content is not limited thereto. In short, if a person skilled in the art is inspired by this description and designs a similar structure and embodiment without departing from the spirit of the present invention, such design should fall within the protection scope of the present invention.

Claims

1. A radiation detection information report generation method characterized by comprising: The method comprises the following steps: acquiring historical radiation detection data, and performing outlier detection on the historical radiation detection data based on a quartile range method to obtain an outlier determination lower threshold and an outlier determination upper threshold; constructing a risk assessment model based on an analytic hierarchy process method, the outlier determination lower threshold, and the outlier determination upper threshold; acquiring real-time radiation detection data; performing logical analysis on the real-time radiation detection data according to the risk assessment model and a preset logical reasoning algorithm to obtain a logical analysis result; determining whether the logical analysis result meets a preset logical relationship condition; when the logical analysis result meets the logical relationship condition, constructing a radiation detection report template according to the logical analysis result; filling data in the radiation detection report template according to the real-time radiation detection data to obtain a radiation detection information report.

2. The radiation detection information report generation method of claim 1, wherein, The method of performing outlier detection on the historical radiation detection data based on the quartile range method to obtain the outlier determination lower threshold and the outlier determination upper threshold comprises the following steps: sorting the historical radiation detection data according to a preset sorting algorithm and a preset reference quantile to obtain sorted detection data; performing outlier detection on the sorted detection data based on the quartile range method to obtain a first quartile and a second quartile; performing difference calculation on the first quartile according to the second quartile to obtain a quartile range; calculating the outlier determination lower threshold according to a preset outlier determination coefficient, the first quartile, and the quartile range; calculating the outlier determination upper threshold according to the outlier determination coefficient, the second quartile, and the quartile range.

3. The method of claim 1, wherein the radiation detection information report is generated by a radiation detection information report generator. The method of constructing the risk assessment model based on the analytic hierarchy process method, the outlier determination lower threshold, and the outlier determination upper threshold comprises the following steps: constructing a primary assessment model based on the analytic hierarchy process method and a preset fuzzy comprehensive risk evaluation index; performing full-feature extraction on the historical radiation detection data to obtain a sample feature vector; generating kernel function parameters according to the outlier determination lower threshold and the outlier determination upper threshold; training the primary assessment model according to the sample feature vector and the kernel function parameters to obtain the risk assessment model.

4. The radiation detection information report generation method of claim 1, wherein, The method of performing logical analysis on the real-time radiation detection data according to the risk assessment model and the preset logical reasoning algorithm to obtain the logical analysis result comprises the following steps: performing preprocessing on the real-time radiation detection data to obtain a cross-modal alignment feature vector; performing risk assessment on the cross-modal alignment feature vector according to the risk assessment model to obtain a risk assessment score; performing logical analysis on the cross-modal fusion feature vector according to the logical reasoning algorithm and the risk assessment score to obtain the logical analysis result.

5. A radiation detection information report generation method according to claim 4, characterized by, The method of performing preprocessing on the real-time radiation detection data to obtain the cross-modal alignment feature vector comprises the following steps: performing feature extraction on the real-time radiation detection data based on a covariance method to obtain energy spectrum features; acquiring a knowledge information table according to a preset intelligent graph selection algorithm; performing feature extraction on the real-time radiation detection data according to a preset cosine similarity algorithm and the knowledge information table to obtain knowledge information features; The knowledge information features and the energy spectrum features are aligned based on a preset attention mechanism to obtain a cross-modal alignment feature vector.

6. A radiation detection information report generation method according to claim 5, wherein, The knowledge information features and the energy spectrum features are aligned based on a preset attention mechanism to obtain a cross-modal alignment feature vector, including: The knowledge information features are filtered according to a preset confidence and a preset wavelet transform denoising algorithm to obtain high-confidence knowledge information features; The energy spectrum features and the high-confidence knowledge information features are analyzed based on an attention mechanism to obtain adaptive weight features; The high-confidence knowledge information features and the energy spectrum features are aligned according to the adaptive weight features to obtain a cross-modal alignment feature vector.

7. A radiation detection information report generation method according to claim 6, wherein, The radiation detection report template is constructed according to the logical analysis result, including: The cross-modal alignment feature vector is extracted based on a preset text feature extraction algorithm and a preset frequency threshold to obtain text features and data distribution features; Design data items and a configuration learning framework are generated according to the text features, the data distribution features, and a preset report display requirement; The radiation detection report template is constructed according to the logical analysis result, the design data items, and the configuration learning framework.

8. The method of claim 1, wherein the radiation detection information report is generated by a radiation detection information report generator. The step of filling the radiation detection report template with data according to the real-time radiation detection data to obtain a radiation detection information report further includes: The radiation detection information report is subjected to a hash operation based on a preset hash function to obtain a hash value; The collection time and the collection equipment number of the real-time radiation detection data are obtained; The radiation detection information report, the hash value, the collection time, and the collection equipment number are packaged into a block according to a preset consensus algorithm to obtain a blockchain storage; The blockchain storage is stored in fragments based on a preset alliance chain to obtain a fragmented chain storage.

9. A radiation detection information report generating apparatus characterized by comprising: It includes: An outlier detection module for obtaining historical radiation detection data and detecting outliers in the historical radiation detection data based on a quartile range method to obtain an outlier lower limit threshold and an outlier upper limit threshold; A model construction module for constructing a risk assessment model based on an analytic hierarchy process, the outlier lower limit threshold, and the outlier upper limit threshold; A data acquisition module for acquiring real-time radiation detection data; A logical analysis module for logically analyzing the real-time radiation detection data according to a risk assessment model and a preset logical reasoning algorithm to obtain a logical analysis result; A condition judgment module for judging whether the logical analysis result meets a preset logical relationship condition; A template generation module for constructing a radiation detection report template according to a logical analysis result when the logical analysis result meets the logical relationship condition; A detection information report generation module for filling a radiation detection report template with data according to real-time radiation detection data to obtain a radiation detection information report.

10. A radiation detection information report generating apparatus characterized by comprising: It includes: A memory and at least one processor, the memory having instructions stored therein; The at least one processor invokes the instructions in the memory to cause the radiation detection information report generation device to perform the steps of the radiation detection information report generation method of any one of claims 1-8.

Images