Agricultural product quality safety intelligent grading early warning system and grading method thereof

Abstract

The invention discloses an agricultural product quality safety intelligent grading early warning system and a grading method thereof, and relates to the technical field of agriculture. The method comprises the following steps: acquiring and analyzing production area information of agricultural products to obtain a production area risk value; obtaining biological index data of the agricultural products, and obtaining a production environment risk value based on the production area risk value; performing metabonomics analysis on the agricultural products, and performing pollution early warning scoring on the agricultural products to generate pollution early warning risk values; obtaining and analyzing consumer behavior data of the agricultural product to obtain a consumer feedback score; and integrating the production environment risk value, the pollution early warning risk value and the consumer feedback score, performing analysis to obtain a comprehensive quality safety score of the agricultural product, generating a score level, and triggering score level early warning. According to the invention, a scientific and intelligent agricultural product quality safety dynamic evaluation and early warning system is formed, and the occurrence rate of agricultural product safety accidents is effectively reduced.

Description

Technical Field

[0001] The present application relates to the field of agricultural technology, and in particular to an intelligent grading and early warning system for agricultural product quality and safety and a scoring method thereof. Background Art

[0002] Agricultural product quality and safety refers to the compliance of agricultural products with national quality standards and safety requirements throughout their production, processing, distribution, and consumption, ensuring they are harmless to human health, the ecological environment, and the sustainable development of agricultural production. Its core goal is to ensure consumers can eat and consume products with confidence while promoting high-quality agricultural development. Therefore, agricultural product quality and safety scoring is particularly important. The intelligent agricultural product quality and safety scoring method uses data collection, model calculation, and dynamic assessment to quantitatively classify agricultural product safety risks. It aims to replace traditional manual sampling inspections and achieve efficient and accurate quality control.

[0003] Among related technologies, agricultural product quality and safety testing techniques mostly rely on traditional chemical testing methods, such as pesticide residue testing and heavy metal testing. These methods can usually only be used after contamination occurs, and cannot identify potential contamination risks before or in the early stages of contamination, resulting in late detection of contamination problems. Furthermore, traditional agricultural product quality scoring methods mostly focus on testing physical or chemical properties, such as color, size, sugar content, and pesticide residues, ignoring the biological reactions of agricultural products and consumer feedback, and thus need improvement. Summary of the Invention

[0004] The purpose of the present invention is to provide an agricultural product quality and safety intelligent grading early warning system and a scoring method thereof, so as to solve the problems raised in the above background technology.

[0005] In the first aspect, this application provides an intelligent scoring method for agricultural product quality and safety, which adopts the following technical solutions:

[0006] Acquiring production area information of agricultural products, analyzing the production area information, and obtaining a production area risk value of the agricultural products;

[0007] Obtaining biological indicator data of agricultural products to obtain a growth health index of the agricultural products, and obtaining a production environment risk value of the agricultural products based on the production area risk value;

[0008] Conduct metabolomics analysis on agricultural products to obtain metabolomics analysis results, and then perform pollution warning scores on agricultural products based on the metabolomics analysis results to generate pollution warning risk values;

[0009] Acquiring consumer behavior data of agricultural products, analyzing the consumer behavior data, and obtaining consumer feedback scores of the agricultural products;

[0010] The comprehensive quality and safety score of agricultural products is obtained by analyzing the production environment risk value, pollution warning risk value and consumer feedback score of the agricultural products. A score grade is generated based on the comprehensive quality and safety score, and a score grade warning is triggered.

[0011] Preferably, the steps of obtaining the production area information of the agricultural products, analyzing the production area information, and obtaining the production area risk value of the agricultural products are specifically as follows:

[0012] Acquiring production area information of agricultural products, wherein the production area information includes regional soil information, regional water source information, and regional climate information;

[0013] Based on the regional soil information, according to the soil requirements for agricultural product growth, the matching degree between the regional soil and the soil requirements for agricultural product growth is determined, and the soil matching degree is obtained and marked as P 土壤 ;

[0014] Based on the regional water source information, according to the water source demand of agricultural product growth, the matching degree between the regional water source and the water source demand of agricultural product growth is judged, and the water source matching degree is obtained and marked as P 水源 ;

[0015] Based on the regional climate information, the probability of climate disasters occurring in the agricultural product production area is determined, and the climate disaster probability is obtained and marked as S 气候 ;

[0016] By associating the formula S 区域 =w1×(1-P 土壤 )+w2×(1-P 水源 )+w3×S 气候 Calculate the production area risk value S of agricultural products 区域 .

[0017] Preferably, the steps of obtaining the biological indicator data of the agricultural products, obtaining the growth health index of the agricultural products, and obtaining the production environment risk value of the agricultural products based on the production area risk value are specifically as follows:

[0018] Obtaining biological indicator data of agricultural products, including chlorophyll content B 叶绿素 , leaf area index B 叶面积 , plant stress resistance B 抗逆性 and plant hormone levels B 激素水平 ;

[0019] According to the biological health score formula B 生物 =w4×B 叶绿素 +w5×B 叶面积 +w6×B 抗逆性 +w7×B 激素水平 Calculate the growth health index B of agricultural products生物 , where w4, w5, and w6 are weight coefficients and are not equal to zero;

[0020] Extract the production area risk value S 区域 , through the weighted comprehensive formula S 环境 =n1×S 区域 +n2×1 / B 生物 Calculate the production environment risk value S of agricultural products 环境 , where n1 and n2 are weight coefficients and both are not equal to zero.

[0021] Preferably, the steps of performing metabolomics analysis on agricultural products to obtain metabolomics analysis results, performing pollution warning scores on agricultural products based on the metabolomics analysis results, and generating pollution warning risk values ​​are specifically as follows:

[0022] Collect samples from agricultural products, analyze the metabolite profiles in the samples using high-throughput analytical technology, and obtain information on the types and concentrations of agricultural product metabolites;

[0023] Compare the type and concentration information of agricultural product metabolites obtained from different batches and under different conditions, perform standardization processing, and obtain the target type and concentration data of agricultural product metabolites;

[0024] The target species data and target concentration data are analyzed to obtain the metabolome analysis results. The pollution warning scores of agricultural products are performed based on the metabolome analysis results to generate pollution warning risk values.

[0025] Preferably, the target species data and target concentration data are analyzed to obtain metabolomic analysis results, and the pollution warning scores of agricultural products are scored according to the metabolomic analysis results to generate pollution warning risk values, specifically, the steps of:

[0026] Based on the target species data, obtain the preset standard safety concentration thresholds of the target agricultural product metabolites;

[0027] Based on the target concentration data, the target concentration data is compared with the standard safety concentration threshold to obtain the pollutant concentration level value and generate the pollutant concentration analysis results;

[0028] Screen out agricultural product samples in an uncontaminated and healthy state to establish a normal metabolic profile, and screen out agricultural product samples in a contaminated state to establish a contaminated metabolic profile;

[0029] Perform significant difference analysis on normal metabolic profile and contamination metabolic profile to obtain metabolic profile change analysis results;

[0030] The metabolome analysis results are obtained by integrating the pollutant concentration analysis results and the metabolic spectrum change analysis results. The pollution warning score of agricultural products is performed based on the metabolome analysis results to generate a pollution warning risk value.

[0031] Preferably, the step of performing significant difference analysis on the normal metabolic profile and the contaminated metabolic profile to obtain the metabolic profile change analysis results is specifically:

[0032] Analyze the significant differences between normal metabolic profiles and contamination metabolic profiles to screen out differential metabolites;

[0033] The screened differential metabolites were mapped to the public metabolic pathway database, and the metabolic pathways affected by pollutants were identified through metabolomics network analysis to construct a metabolite-pathway network;

[0034] Based on the metabolite-pathway network, the metabolic cascade changes caused by pollutants are analyzed to obtain the degree of metabolic spectrum change and generate metabolic spectrum change analysis results.

[0035] Preferably, the steps of obtaining metabolome analysis results by integrating the pollutant concentration analysis results and the metabolic spectrum change analysis results, and performing pollution warning scores for agricultural products based on the metabolome analysis results to generate pollution warning risk values ​​are specifically as follows:

[0036] According to the screened differential metabolites, the number of differential metabolites is obtained and marked as D 差异数量， The degree of interference of differential metabolites on metabolic pathways is marked as D 干扰程度 ;

[0037] Based on the pollutant concentration analysis results and metabolic profile change analysis results, the pollutant concentration level value and metabolic profile change degree value were extracted and marked as D 浓度水平 and D 变化程度 ;

[0038] By associating the formula S 污染 =n3×D 差异数量 +n4×D 干扰程度 +n5×D 浓度水平 +n6×D 变化程度 The pollution warning risk value of agricultural products is calculated, where n3, n4, n5, and n6 are weight coefficients and none of them are equal to zero.

[0039] Preferably, the steps of obtaining consumer behavior data of agricultural products, analyzing the consumer behavior data, and obtaining consumer feedback scores of the agricultural products are specifically as follows:

[0040] Acquiring consumer behavior data of agricultural products, wherein the consumer behavior data includes the number of times agricultural products are purchased and agricultural product evaluation information;

[0041] According to the number of times agricultural products are purchased, consumers’ preference for agricultural products is evaluated to obtain the agricultural product preference degree;

[0042] The agricultural product evaluation information includes consumer rating information and evaluation content information of the agricultural products, and sentiment analysis is performed on the rating information and evaluation content information to obtain consumer acceptance of the agricultural products and generate agricultural product acceptance;

[0043] Combining the agricultural product preference and agricultural product acceptance, the consumer feedback score of the agricultural product is obtained and marked as E 消费者 .

[0044] Preferably, the production environment risk value, pollution warning risk value and consumer feedback score of the agricultural products are comprehensively analyzed to obtain a comprehensive quality and safety score of the agricultural products, and a score level is generated according to the comprehensive quality and safety score. The steps of triggering the score level warning are specifically as follows:

[0045] Based on consumer feedback rating E 消费者 , and normalize it to get the consumer feedback score S 消费者 ;

[0046] Extract the production environment risk value S of agricultural products 环境 , pollution warning risk value S 污染 And consumer feedback score S 消费者 , through the formula S 综合 =m1×S 环境 +m2×S 污染 +m3×(1-S 消费者 ) Calculate the comprehensive quality risk value S of agricultural products 综合 , where m1, m2, and m3 are weight coefficients and are not equal to zero;

[0047] By the formula F=1 / S 综合 Calculate the comprehensive quality and safety score F of agricultural products 综合 , the comprehensive quality and safety score F 综合 Compare with the preset standard quality safety threshold range [F1, F2];

[0048] When the comprehensive quality and safety score F 综合 ∈[0, F1), it indicates that there is a serious risk in the quality of agricultural products, and a serious safety level warning signal is generated;

[0049] When the comprehensive quality and safety score F 综合 ∈[F1, F2), it indicates that there is a moderate risk in the quality of agricultural products, and a moderate safety level warning signal is generated;

[0050] When the comprehensive quality and safety score F 综合 ∈[F2, 1), it indicates that there is a slight risk in the quality of agricultural products, and a mild safety level warning signal is generated.

[0051] Secondly, this application provides an intelligent grading and early warning system for agricultural product quality and safety, which adopts the following technical solutions:

[0052] An intelligent grading and early warning system for agricultural product quality and safety, comprising:

[0053] A regional analysis module obtains production area information of agricultural products, analyzes the production area information, and obtains a production area risk value of the agricultural products;

[0054] An environmental analysis module obtains biological indicator data of agricultural products, obtains the growth health index of agricultural products, and obtains the production environment risk value of agricultural products based on the production area risk value;

[0055] The pollution analysis module conducts metabolomics analysis on agricultural products, obtains metabolomics analysis results, and then performs pollution warning scores on agricultural products based on the metabolomics analysis results to generate pollution warning risk values;

[0056] A consumer analysis module, which obtains consumer behavior data of agricultural products, analyzes the consumer behavior data, and obtains consumer feedback scores of agricultural products;

[0057] The scoring and warning module comprehensively analyzes the production environment risk value, pollution warning risk value and consumer feedback score of agricultural products to obtain the comprehensive quality and safety score of agricultural products, generates a scoring level based on the comprehensive quality and safety score, and triggers a scoring level warning.

[0058] In summary, this application includes at least one of the following beneficial technical effects:

[0059] 1. Through regional traceability and risk analysis, assess potential risks in agricultural product growing areas, identify high-risk production areas in advance, reduce systemic quality and safety risks, and provide regional agricultural product cultivation guidance to regulatory authorities. By testing the biological indicators of individual agricultural products to quantitatively reflect their actual health status, combining biological indicators with regional risks, a more targeted and real-time comprehensive risk assessment of the production environment is formed, compensating for the lag and inaccuracy of simple geographical environment assessments and avoiding misjudgments. Metabolomics is used to capture subtle but sensitive metabolite changes in agricultural products, promptly identifying potential contamination or abnormal physiological states. Metabolic changes are quantified into pollution warning scores, thereby quantifying pollution risks and achieving proactive pollution prevention and control. Based on real market feedback and through big data analysis, a quality feedback score at the consumer perception level is generated. Consumer interruption experience data is fed back to the source production quality assessment system, achieving quality tracking and optimization throughout the production-circulation-consumption process, and improving the integrity and reliability of quality and safety assessments. Multi-source heterogeneous data fusion and multi-dimensional comprehensive modeling are used to generate a comprehensive quality and safety score under the final unified standard. According to the comprehensive score, the grades are divided and the corresponding safety warnings and disposal measures are triggered, thus realizing a three-dimensional prevention and control system for agricultural product quality and safety, which can greatly improve the level of agricultural product quality and safety assurance.

[0060] 2. Utilize high-throughput analytical techniques to analyze the metabolite profiles in samples, rapidly, comprehensively, and highly sensitively acquiring molecular fingerprints related to agricultural product physiology and pollution stress. This allows detection of early physiological changes difficult to detect using traditional pesticide residue detection methods, providing a precise, non-destructive data foundation for early warning of pollution and capturing signals of metabolome changes triggered by pollution, disease, and environmental stress. Metabolite data from samples collected from different sampling batches (at different times and locations) and under different treatment conditions are compared to eliminate bias between samples due to differences in experimental batch, sampling time, and treatment, ensuring comparability and consistency in subsequent analyses and improving the accuracy and robustness of the model in identifying pollution risks. Target species and concentration data are analyzed to identify subtle effects of pollutants on agricultural product physiology, enabling modeling of associations between pollutant types and metabolic change patterns, significantly improving the sensitivity and foresight of pollution detection.

[0061] 3. Conduct significant difference analysis between normal and polluted metabolic profiles to accurately identify metabolites that undergo significant changes due to pollutants, providing core molecules for subsequent pathway analysis. Use public databases to connect differential metabolites to corresponding metabolic pathways, conduct metabolic pathway enrichment analysis, identify significantly enriched (significantly affected) metabolic pathways, and construct a metabolite-pathway association network based on the correspondence between metabolites and pathways. Reveal the biochemical mechanism changes caused by pollutants, not just focusing on individual metabolites, but also observing systemic biological pathway changes, enhancing the biological explanatory power of pollution risk analysis and supporting biomarker screening and mechanism modeling for early warning models. Based on the metabolite-pathway network, analyze the metabolic cascade changes caused by pollutants, calculate the degree of metabolic profile change, and output the final metabolic profile change analysis results. Quantify complex metabolic network changes into easily understandable and comparable indicators, providing more sensitive and accurate biological risk signals for pollution warning and intelligent scoring systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a schematic diagram of the specific steps of an embodiment of an intelligent scoring method for agricultural product quality and safety of the present invention.

[0063] Figure 2 The present invention is a module connection diagram of an embodiment of an intelligent grading and early warning system for agricultural product quality and safety. DETAILED DESCRIPTION

[0064] Below is a combination of the embodiments and Figure 1-Figure 2 The present invention will be described in further detail, but the embodiments of the present invention are not limited thereto.

[0065] The present invention discloses an intelligent scoring method for agricultural product quality and safety, which specifically comprises the following steps:

[0066] Step S1, obtaining the production area information of the agricultural products, analyzing the production area information, and obtaining the production area risk value of the agricultural products;

[0067] Step S2, obtaining biological indicator data of agricultural products, obtaining a growth health index of agricultural products, and obtaining a production environment risk value of agricultural products based on the production area risk value;

[0068] Step S3, performing metabolomics analysis on the agricultural products to obtain metabolomics analysis results, and performing pollution warning scores on the agricultural products based on the metabolomics analysis results to generate pollution warning risk values;

[0069] Step S4, obtaining consumer behavior data of agricultural products, analyzing the consumer behavior data, and obtaining consumer feedback scores of the agricultural products;

[0070] Step S5: Comprehensively analyze the production environment risk value, pollution warning risk value and consumer feedback score of the agricultural products to obtain a comprehensive quality and safety score of the agricultural products, generate a score level based on the comprehensive quality and safety score, and trigger a score level warning.

[0071] In practical applications, regional traceability and risk analysis are used to assess potential risks in agricultural production areas, proactively identify high-risk production areas, mitigate systemic quality and safety risks, and provide regional guidance for agricultural production to regulatory authorities. By testing biological indicators of individual agricultural products, a quantitative reflection of their actual health status is provided. Combining these indicators with regional risks creates a more targeted and real-time comprehensive risk assessment of the production environment. This approach can identify complex situations such as "low environmental risk, high biological stress" or "high environmental risk, no obvious biological abnormalities," thereby compensating for the lag and inaccuracy of purely geographical environmental assessments and avoiding misjudgments. Metabolomics is used to capture subtle but sensitive metabolite changes within agricultural products, enabling timely detection of potential contamination (pesticide residues, heavy metals, microbial toxins, etc.) or abnormal physiological states. Metabolic changes are quantified into pollution warning scores, thereby quantifying pollution risk. Metabolic abnormality signals can be used to identify problems before pollutant accumulation exceeds standards, enabling proactive pollution prevention and control. Based on real market feedback, we analyze consumers' subjective and objective experiences after purchase and use. Through big data analysis, we generate quality feedback scores at the consumer perception level. This consumer experience data is fed back into the source production quality assessment system, optimizing quality tracking throughout the production, distribution, and consumption process, and improving the integrity and reliability of quality and safety assessments. By integrating multi-source heterogeneous data and implementing multi-dimensional comprehensive modeling, we generate a comprehensive quality and safety score based on a unified standard. Based on this comprehensive score, we categorize products into different levels and trigger corresponding safety warnings and response measures. This creates a three-dimensional prevention and control system for agricultural product quality and safety, significantly enhancing the quality and safety assurance of agricultural products.

[0072] The steps of obtaining production area information of agricultural products, analyzing the production area information, and obtaining the production area risk value of the agricultural products are specifically as follows:

[0073] Step S11, obtaining production area information of agricultural products, wherein the production area information includes regional soil information, regional water source information, and regional climate information;

[0074] Step S12: Based on the regional soil information and the agricultural product growth soil requirements, the matching degree between the regional soil and the agricultural product growth soil requirements is determined, and the soil matching degree is obtained and marked as P. 土壤 ;

[0075] Step S13: Based on the regional water source information and the water source demand for agricultural product growth, the matching degree between the regional water source and the water source demand for agricultural product growth is determined, and the water source matching degree is obtained and marked as P. 水源 ;

[0076] Step S14: Based on the regional climate information, the probability of climate disasters occurring in the agricultural product production area is determined, and the climate disaster probability is obtained and marked as S. 气候 ;

[0077] Step S15, through the correlation formula S 区域 =w1×(1-P 土壤 )+w2×(1-P 水源 )+w3×S 气候 Calculate the production area risk value S of agricultural products 区域 .

[0078] In practical applications, soil conditions directly impact the growth quality, disease risk, and pesticide residue accumulation of agricultural products. Analyzing the degree to which soil characteristics in a production area match the soil conditions (such as texture and fertility) required for the growth of target agricultural products can provide early warning of potential growth problems or pesticide residue accumulation. Water source safety directly affects risks such as microbial contamination and excessive heavy metal content in agricultural products. Analyzing water source quality (such as heavy metals, pesticide residues, and microbial contamination) and water supply to ensure they meet safe agricultural production requirements is crucial. Monitoring and assessing water source quality, which varies seasonally, can effectively avoid planting during periods of high water risk. A high probability of climate disasters means significant yield fluctuations and a high risk of quality loss. Analyzing historical data and trends on extreme weather conditions (such as heavy rain, drought, frost, and heat waves) in a production area and calculating the probability of disaster occurrence can help plan measures such as adjusting farming schedules and selecting resilient varieties, thereby enhancing risk resilience. A correlation formula is proposed to convert multidimensional heterogeneous factors into standardized risk values, facilitating subsequent modeling with other modules to form a comprehensive scoring system.

[0079] The steps of obtaining biological indicator data of agricultural products, obtaining a growth health index of agricultural products, and obtaining a production environment risk value of agricultural products based on the production area risk value are specifically as follows:

[0080] Step S21, obtaining biological indicator data of agricultural products, said biological indicator data including chlorophyll content B 叶绿素 , leaf area index B 叶面积 , plant stress resistance B 抗逆性 and plant hormone levels B 激素水平 ;

[0081] Step S22, according to the biological health score formula B 生物 =w4×B 叶绿素 +w5×B 叶面积 +w6×B 抗逆性 +w7×B 激素水平 Calculate the growth health index B of agricultural products 生物, where w4, w5, and w6 are weight coefficients and are not equal to zero;

[0082] Step S23, extracting the production area risk value S 区域 , through the weighted comprehensive formula S 环境 =n1×S 区域 +n2×1 / B 生物 Calculate the production environment risk value S of agricultural products 环境 , where n1 and n2 are weight coefficients and both are not equal to zero.

[0083] In practice, chlorophyll content in plant leaves represents photosynthetic efficiency and nutritional status, leaf area index reflects plant growth luxuriance, plant stress tolerance reflects a plant's ability to withstand adverse stresses, such as drought, cold, and disease resistance, and plant hormone levels, such as auxin and abscisic acid, represent physiological regulation. A weighted comprehensive formula integrates these multiple biological indicators into a single, integrated index, providing a visual reflection of the current growth and health of agricultural crops and capturing potential physiological anomalies. This helps identify hidden issues, such as "crop impacts before environmental risks manifest." The integrated production area risk value and biological health index consider both external environmental factors and the crop's intrinsic physiological state. Weighted analysis of these two factors yields a more realistic and dynamic comprehensive risk profile for the production environment, preventing inaccuracies in a single indicator. Even if the baseline environmental risk is low, crop health anomalies can provide timely warnings.

[0084] The steps of performing metabolomics analysis on agricultural products, obtaining metabolomics analysis results, and scoring pollution warnings of agricultural products based on the metabolomics analysis results to generate pollution warning risk values ​​are as follows:

[0085] Step S31, collecting samples from agricultural products, analyzing the metabolite spectrum in the samples using high-throughput analysis technology, and obtaining type information and concentration information of the agricultural product metabolites;

[0086] Step S32, comparing the type information and concentration information of the agricultural product metabolites obtained from different batches and under different conditions, and performing standardization processing to obtain target type data and target concentration data of the agricultural product metabolites;

[0087] Step S33 , analyzing the target species data and the target concentration data to obtain metabolomics analysis results, and performing pollution warning scores on the agricultural products based on the metabolomics analysis results to generate pollution warning risk values.

[0088] In practical applications, high-throughput analytical techniques, such as gas chromatography-mass spectrometry and liquid chromatography-mass spectrometry, are used to analyze the metabolite profiles in samples. These techniques provide rapid, comprehensive, and highly sensitive molecular fingerprints related to agricultural product physiology and pollution stress, detecting early physiological changes that are difficult to detect using traditional pesticide residue detection methods. This provides a precise and non-destructive data foundation for early warning of pollution, capturing metabolomic changes triggered by pollution, disease, and environmental stress. Metabolite data from samples collected from different sampling batches (at different times and locations) and under different treatment conditions are compared to eliminate bias between samples due to differences in experimental batch, sampling time, and treatment. This ensures comparability and consistency in subsequent analyses, and improves the accuracy and robustness of the model's identification of pollution risks. Analysis of target species and concentration data reveals subtle effects of pollutants on agricultural product physiology, enabling modeling of associations between pollutant types and metabolic patterns, significantly enhancing the sensitivity and foresight of pollution detection.

[0089] The target species data and target concentration data are analyzed to obtain metabolomic analysis results. The pollution warning scores of agricultural products are calculated based on the metabolomic analysis results to generate pollution warning risk values. The specific steps are as follows:

[0090] Step S331, obtaining a preset standard safety concentration threshold of a target agricultural product metabolite according to the target category data;

[0091] Step S332: Based on the target concentration data, the target concentration data is compared with the standard safety concentration threshold to obtain a pollutant concentration level value and generate a pollutant concentration analysis result;

[0092] Step S333, screening out agricultural product samples in an uncontaminated and healthy state to establish a normal metabolic profile, and screening out agricultural product samples in a contaminated state to establish a contaminated metabolic profile;

[0093] Step S334, performing a significant difference analysis on the normal metabolic profile and the contaminated metabolic profile to obtain a metabolic profile change analysis result;

[0094] Step S335 , combining the pollutant concentration analysis results and the metabolic spectrum change analysis results to obtain a metabolome analysis result, and performing a pollution warning score for the agricultural product based on the metabolome analysis result to generate a pollution warning risk value.

[0095] In actual application, the deviation between the actual concentration of the target metabolite and the standard threshold is calculated to determine whether each metabolite is in a normal, high or low state, forming a pollutant concentration level matrix for each sample, judging whether there is potential pollution or stress anomaly in the sample, quantitatively assessing the degree of pollution, and providing single indicator evidence for subsequent overall changes in the metabolic spectrum. A metabolite concentration distribution model is established for each, namely the normal metabolic spectrum and the polluted metabolic spectrum, and a significant difference analysis is performed on the normal metabolic spectrum and the polluted metabolic spectrum to identify changes in biological fingerprints in the early asymptomatic stage of pollution, thereby improving the sensitivity and accuracy of the early warning system. The results of the pollutant concentration analysis and the metabolic spectrum change analysis are combined, considering both single indicator abnormalities and systemic metabolic abnormalities, and the risk of agricultural product contamination is assessed from a global perspective to enhance the credibility and scientific nature of the test results.

[0096] The steps for performing significant difference analysis between normal metabolic profile and contamination metabolic profile to obtain metabolic profile change analysis results are as follows:

[0097] Step S3341, performing significant difference analysis on the normal metabolic profile and the contamination metabolic profile to screen out differential metabolites;

[0098] Step S3342: Map the screened differential metabolites to a public metabolic pathway database, identify the metabolic pathways affected by the pollutants through metabolomics network analysis, and construct a metabolite-pathway network;

[0099] Step S3343: Analyze the metabolic cascade changes caused by the pollutants based on the metabolite-pathway network, obtain the metabolic spectrum change degree value, and generate the metabolic spectrum change analysis results.

[0100] In practical applications, significant differences between normal and polluted metabolic profiles are analyzed to accurately identify metabolites that undergo significant changes due to pollutants, providing core molecules for subsequent pathway analysis. Public databases are used to connect differential metabolites to corresponding metabolic pathways, and metabolic pathway enrichment analysis is performed to identify significantly enriched (significantly affected) metabolic pathways. Based on the correspondence between metabolites and pathways, a metabolite-pathway association network is constructed. For example, pesticide pollution may affect antioxidant pathways, and heavy metal pollution may interfere with mineral metabolic pathways. Revealing the changes in biochemical mechanisms caused by pollutants not only focuses on individual metabolites, but also allows observation of systemic biological pathway changes, enhancing the biological explanatory power of pollution risk analysis and supporting biomarker screening and mechanism modeling for early warning models. Based on the metabolite-pathway network, the metabolic cascade changes caused by pollutants are analyzed, the degree of metabolic profile change values ​​are calculated, and the final metabolic profile change analysis results are output. The degree of change value can be calculated using (number of differential metabolites × average difference fold) / total number of detected metabolites, quantifying complex metabolic network changes into easy-to-understand and comparable indicators, providing more sensitive and accurate biological risk signals for pollution warning and intelligent scoring systems.

[0101] The steps for obtaining the metabolome analysis results by integrating the pollutant concentration analysis results and the metabolic spectrum change analysis results, and then performing pollution warning scores for agricultural products based on the metabolome analysis results and generating pollution warning risk values ​​are as follows:

[0102] Step S3351: Based on the screened differential metabolites, the number of differential metabolites is obtained and marked as D. 差异数量， The degree of interference of differential metabolites on metabolic pathways is marked as D 干扰程度 ;

[0103] Step S3352: Based on the pollutant concentration analysis results and the metabolic spectrum change analysis results, extract the pollutant concentration level value and the metabolic spectrum change degree value, marked as D 浓度水平 and D 变化程度 ;

[0104] Step S3353, through the association formula S 污染 =n3×D 差异数量 +n4×D 干扰程度 +n5×D 浓度水平 +n6×D 变化程度 The pollution warning risk value of agricultural products is calculated, where n3, n4, n5, and n6 are weight coefficients and none of them are equal to zero.

[0105] In practical applications, the number of significantly differential metabolites is counted, and the range of metabolite changes caused by pollutants is quantified. The more differential metabolites there are, the more extensive the impact of the pollution on the biological system and the higher the risk. Based on the metabolite-pathway network, the interference intensity of differential metabolites on key metabolic pathways is analyzed to characterize the degree of core damage in the metabolic network. The pollutant concentration level value is extracted to reflect the extent to which the pollutant exceeds the safety standard, and the metabolic spectrum change degree value is extracted to reflect the severity of the abnormal physiological metabolic state. The final draft pollution warning risk value is obtained through weighted synthesis of the correlation formula, realizing early, accurate and intelligent warning of agricultural product pollution. At the same time, the weight coefficient can be dynamically adjusted to flexibly adapt to different agricultural product pollution scenarios.

[0106] The steps of obtaining consumer behavior data of agricultural products, analyzing the consumer behavior data, and obtaining consumer feedback scores of the agricultural products are specifically as follows:

[0107] Step S41, obtaining consumer behavior data of agricultural products, wherein the consumer behavior data includes the number of times agricultural products are purchased and agricultural product evaluation information;

[0108] Step S42, evaluating the consumer's preference for agricultural products based on the number of times the agricultural products are purchased, and obtaining the agricultural product preference degree;

[0109] Step S43: The agricultural product evaluation information includes consumer rating information and evaluation content information of the agricultural products, and sentiment analysis is performed on the rating information and evaluation content information to obtain consumer acceptance of the agricultural products and generate agricultural product acceptance;

[0110] Step S44: Comprehensively consider the preference and acceptance of agricultural products to obtain the consumer feedback score of the agricultural products and mark it as E. 消费者 .

[0111] In practical applications, consumers' initial preference intentions are quantified based on the number of agricultural product purchases per unit time. A high preference score generally indicates good product quality perception and strong brand recognition. Sentiment analysis is conducted on ratings and review content to assess consumers' actual experience and satisfaction after use. A high acceptance score indicates that the agricultural products meet consumer expectations. By integrating preference and acceptance scores into a comprehensive user feedback score, consumer perception of product quality can be quantified and incorporated into the intelligent scoring system, reflecting real market quality, strengthening the system's external validation capabilities, and providing scientific data support for subsequent risk warnings.

[0112] The comprehensive quality and safety scores of agricultural products are analyzed based on the production environment risk value, pollution warning risk value, and consumer feedback score of the agricultural products. The scoring level is generated based on the comprehensive quality and safety score. The steps to trigger the scoring level warning are as follows:

[0113] Step S51: Based on the consumer feedback score E 消费者 , and normalize it to get the consumer feedback score S 消费者 ;

[0114] Step S52: extract the production environment risk value S of the agricultural product 环境 , pollution warning risk value S 污染 And consumer feedback score S 消费者 , through the formula S 综合 =m1×S 环境 +m2×S 污染 +m3×(1-S 消费者 ) Calculate the comprehensive quality risk value S of agricultural products 综合 , where m1, m2, and m3 are weight coefficients and are not equal to zero;

[0115] Step S53, by the formula F=1 / S 综合 Calculate the comprehensive quality and safety score F of agricultural products 综合 , the comprehensive quality and safety score F 综合 Compare with the preset standard quality safety threshold range [F1, F2];

[0116] Step S54: When the comprehensive quality and safety score F 综合 ∈[0, F1), it indicates that there is a serious risk in the quality of agricultural products, and a serious safety level warning signal is generated;

[0117] Step S55: When the comprehensive quality safety score F 综合 ∈[F1, F2), it indicates that there is a moderate risk in the quality of agricultural products, and a moderate safety level warning signal is generated;

[0118] Step S56: When the comprehensive quality safety score F 综合 ∈[F2, 1), it indicates that there is a slight risk in the quality of agricultural products, and a mild safety level warning signal is generated.

[0119] In actual application, the obtained consumer feedback scores will be normalized to make them easy to unify with other risk values, ensure the scientific and comparable scoring system, and avoid scoring distortion caused by data bias. Comprehensively consider the production environment risks, pollution risks and actual consumer feedback to form a comprehensive and dynamic risk assessment, avoid one-sided judgments caused by a single dimension, and achieve accurate control of agricultural product quality risks. Through inverse transformation, an intuitive and positive safety scoring indicator is formed, F 综合The higher the score, the better the safety, facilitating rapid sorting and comparison of the quality levels of agricultural product batches, facilitating management decisions. Based on the score level, corresponding quality warning signals are automatically triggered in a timely manner, improving response speed and enabling early detection, early warning, and early intervention of quality issues, thereby maximizing food safety and enhancing the foresight and response efficiency of agricultural product safety management.

[0120] An agricultural product quality and safety intelligent grading and early warning system, which applies the above-mentioned agricultural product quality and safety intelligent scoring method, includes:

[0121] A regional analysis module obtains production area information of agricultural products, analyzes the production area information, and obtains a production area risk value of the agricultural products;

[0122] An environmental analysis module obtains biological indicator data of agricultural products, obtains the growth health index of agricultural products, and obtains the production environment risk value of agricultural products based on the production area risk value;

[0123] The pollution analysis module conducts metabolomics analysis on agricultural products, obtains metabolomics analysis results, and then performs pollution warning scores on agricultural products based on the metabolomics analysis results to generate pollution warning risk values;

[0124] A consumer analysis module, which obtains consumer behavior data of agricultural products, analyzes the consumer behavior data, and obtains consumer feedback scores of agricultural products;

[0125] The scoring and warning module comprehensively analyzes the production environment risk value, pollution warning risk value and consumer feedback score of agricultural products to obtain the comprehensive quality and safety score of agricultural products, generates a scoring level based on the comprehensive quality and safety score, and triggers a scoring level warning.

[0126] In practical applications, the regional analysis module captures information about agricultural product production areas. Combined with agricultural product growth requirements, data analysis is performed to assess the compatibility of production areas with the healthy growth of agricultural products. This can proactively identify potential risk factors such as soil contamination, abnormal water quality, and frequent climate disasters, providing a scientific basis for subsequent quality and safety testing and enhancing traceability. The environmental analysis module collects biological indicator data for agricultural products, calculates their growth health index, and, combined with production area risk values, further comprehensively analyzes the actual environmental risks of agricultural products. This provides a dynamic and intuitive reflection of the growth health of agricultural products, effectively compensating for the static nature of environmental information and improving the accuracy and real-time nature of risk assessments. The contamination analysis module performs metabolomics testing and analysis on agricultural product samples, identifying changes in metabolite types and concentrations and assessing potential contamination. This module can capture subtle physiological changes before contamination manifests itself or before traditional detection signals are detected, providing early warning of agricultural product contamination risks, increasing sensitivity to identifying hidden contamination, and ensuring agricultural product safety. The consumer analysis module collects consumer behavior data on agricultural products. Through behavioral and sentiment analysis techniques, it assesses consumers' comprehensive experience feedback on agricultural products. This supplements the actual user experience that cannot be captured by laboratory testing, effectively closing the loop of the overall quality assessment system and enhancing the model's market adaptability. The scoring and early warning module integrates production environment risk values, pollution early warning risk values, and consumer feedback scores to calculate a comprehensive quality and safety score for agricultural products. Agricultural products are then graded based on the scoring results, and safety early warning signals of different levels are automatically triggered. This forms a scientific, systematic, and intelligent dynamic evaluation and early warning system for agricultural product quality and safety, effectively reducing the incidence of agricultural product safety accidents and significantly improving the accuracy, efficiency, and foresight of agricultural product quality supervision.

[0127] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. An intelligent scoring method for agricultural product quality and safety, characterized in that: The following steps are involved: Acquiring production area information of agricultural products, analyzing the production area information, and obtaining a production area risk value of the agricultural products; Obtaining biological indicator data of agricultural products to obtain a growth health index of the agricultural products, and obtaining a production environment risk value of the agricultural products based on the production area risk value; Conduct metabolomics analysis on agricultural products to obtain metabolomics analysis results, and then perform pollution warning scores on agricultural products based on the metabolomics analysis results to generate pollution warning risk values; Acquiring consumer behavior data of agricultural products, analyzing the consumer behavior data, and obtaining consumer feedback scores of the agricultural products; The comprehensive quality and safety score of agricultural products is obtained by analyzing the production environment risk value, pollution warning risk value and consumer feedback score of the agricultural products. A score grade is generated based on the comprehensive quality and safety score, and a score grade warning is triggered.

2. The method for intelligently scoring agricultural product quality and safety according to claim 1, characterized in that: The step of obtaining the production area information of the agricultural products, analyzing the production area information, and obtaining the production area risk value of the agricultural products is specifically as follows: Acquiring production area information of agricultural products, wherein the production area information includes regional soil information, regional water source information, and regional climate information; Based on the regional soil information, according to the soil requirements for agricultural product growth, the matching degree between the regional soil and the soil requirements for agricultural product growth is determined, and the soil matching degree is obtained and marked as P 土壤 ; Based on the regional water source information, according to the water source demand of agricultural product growth, the matching degree between the regional water source and the water source demand of agricultural product growth is judged, and the water source matching degree is obtained and marked as P 水源 ; Based on the regional climate information, the probability of climate disasters occurring in the agricultural product production area is determined, and the climate disaster probability is obtained and marked as S 气候 ; By associating the formula S 区域 =w1×(1-P 土壤 )+w2×(1-P 水源 )+w3×S 气候 Calculate the production area risk value S of agricultural products 区域 .

3. The method for intelligently scoring agricultural product quality and safety according to claim 2, characterized in that: The steps of obtaining the biological indicator data of the agricultural products, obtaining the growth health index of the agricultural products, and obtaining the production environment risk value of the agricultural products based on the production area risk value are specifically as follows: Obtaining biological indicator data of agricultural products, including chlorophyll content B 叶绿素 , leaf area index B 叶面积 , plant stress resistance B 抗逆性 and plant hormone levels B 激素水平 ; According to the biological health score formula B 生物 =w4×B 叶绿素 +w5×B 叶面积 +w6×B 抗逆性 +w7×B 激素水平 Calculate the growth health index B of agricultural products 生物 , where w4, w5, and w6 are weight coefficients and are not equal to zero; Extract the production area risk value S 区域 , through the weighted comprehensive formula S 环境 =n1×S 区域 +n2×1 / B 生物 Calculate the production environment risk value S of agricultural products 环境 , where n1 and n2 are weight coefficients and both are not equal to zero.

4. The method for intelligently scoring agricultural product quality and safety according to claim 1, wherein: The steps of performing metabolomics analysis on agricultural products to obtain metabolomics analysis results, performing pollution warning scores on agricultural products based on the metabolomics analysis results, and generating pollution warning risk values ​​are specifically as follows: Collect samples from agricultural products, analyze the metabolite profiles in the samples using high-throughput analytical technology, and obtain information on the types and concentrations of agricultural product metabolites; Compare the type and concentration information of agricultural product metabolites obtained from different batches and under different conditions, perform standardization processing, and obtain the target type and concentration data of agricultural product metabolites; The target species data and target concentration data are analyzed to obtain the metabolome analysis results. The pollution warning scores of agricultural products are performed based on the metabolome analysis results to generate pollution warning risk values.

5. The method for intelligently scoring agricultural product quality and safety according to claim 4, characterized in that: The steps of analyzing the target species data and the target concentration data to obtain metabolome analysis results, performing pollution warning scores on agricultural products based on the metabolome analysis results, and generating pollution warning risk values ​​are specifically as follows: Based on the target species data, obtain the preset standard safety concentration thresholds of the target agricultural product metabolites; Based on the target concentration data, the target concentration data is compared with the standard safety concentration threshold to obtain the pollutant concentration level value and generate the pollutant concentration analysis results; Screen out agricultural product samples in an uncontaminated and healthy state to establish a normal metabolic profile, and screen out agricultural product samples in a contaminated state to establish a contaminated metabolic profile; Perform significant difference analysis on normal metabolic profile and contamination metabolic profile to obtain metabolic profile change analysis results; The metabolome analysis results are obtained by integrating the pollutant concentration analysis results and the metabolic spectrum change analysis results. The pollution warning score of agricultural products is performed based on the metabolome analysis results to generate a pollution warning risk value.

6. The method for intelligently scoring agricultural product quality and safety according to claim 5, characterized in that: The step of performing significant difference analysis on the normal metabolic profile and the contaminated metabolic profile to obtain the metabolic profile change analysis results is specifically as follows: Analyze the significant differences between normal metabolic profiles and contamination metabolic profiles to screen out differential metabolites; The screened differential metabolites were mapped to the public metabolic pathway database, and the metabolic pathways affected by pollutants were identified through metabolomics network analysis to construct a metabolite-pathway network; Based on the metabolite-pathway network, the metabolic cascade changes caused by pollutants are analyzed to obtain the degree of metabolic spectrum change and generate metabolic spectrum change analysis results.

7. The method for intelligently scoring agricultural product quality and safety according to claim 6, characterized in that: The steps of synthesizing the pollutant concentration analysis results and the metabolic spectrum change analysis results to obtain the metabolome analysis results, and performing pollution warning scores on agricultural products based on the metabolome analysis results to generate pollution warning risk values ​​are specifically as follows: According to the screened differential metabolites, the number of differential metabolites is obtained and marked as D 差异数量， The degree of interference of differential metabolites on metabolic pathways is marked as D 干扰程度 ; Based on the pollutant concentration analysis results and metabolic profile change analysis results, the pollutant concentration level value and metabolic profile change degree value were extracted and marked as D 浓度水平 and D 变化程度 ; By associating the formula S 污染 =n3×D 差异数量 +n4×D 干扰程度 +n5×D 浓度水平 +n6×D 变化程度 The pollution warning risk value of agricultural products is calculated, where n3, n4, n5, and n6 are weight coefficients and none of them are equal to zero.

8. The method for intelligently scoring agricultural product quality and safety according to claim 1, wherein: The steps of obtaining consumer behavior data of agricultural products, analyzing the consumer behavior data, and obtaining consumer feedback scores of agricultural products are specifically as follows: Acquiring consumer behavior data of agricultural products, wherein the consumer behavior data includes the number of times agricultural products are purchased and agricultural product evaluation information; According to the number of times agricultural products are purchased, consumers’ preference for agricultural products is evaluated to obtain the agricultural product preference degree; The agricultural product evaluation information includes consumer rating information and evaluation content information of the agricultural products, and sentiment analysis is performed on the rating information and evaluation content information to obtain consumer acceptance of the agricultural products and generate agricultural product acceptance; Combining the agricultural product preference and agricultural product acceptance, the consumer feedback score of the agricultural product is obtained and marked as E 消费者 .

9. The method for intelligently scoring agricultural product quality and safety according to claim 8, characterized in that: The steps of analyzing the comprehensive quality and safety scores of agricultural products by integrating the production environment risk value, pollution warning risk value and consumer feedback score of the agricultural products, generating a rating grade based on the comprehensive quality and safety score, and triggering the rating grade warning are as follows: Based on consumer feedback rating E 消费者 , and normalize it to get the consumer feedback score S 消费者 ; Extract the production environment risk value S of agricultural products 环境 , pollution warning risk value S 污染 And consumer feedback score S 消费者 , through the formula S 综合 =m1×S 环境 +m2×S 污染 +m3×(1-S 消费者 ) Calculate the comprehensive quality risk value S of agricultural products 综合 , where m1, m2, and m3 are weight coefficients and are not equal to zero; By the formula F=1 / S 综合 Calculate the comprehensive quality and safety score F of agricultural products 综合 , the comprehensive quality and safety score F 综合 Compare with the preset standard quality safety threshold range [F1, F2]; When the comprehensive quality and safety score F 综合 ∈[0, F1), it indicates that there is a serious risk in the quality of agricultural products, and a serious safety level warning signal is generated; When the comprehensive quality and safety score F 综合 ∈[F1, F2), it indicates that there is a moderate risk in the quality of agricultural products, and a moderate safety level warning signal is generated; When the comprehensive quality and safety score F 综合 ∈[F2, 1), it indicates that there is a slight risk in the quality of agricultural products, and a mild safety level warning signal is generated.

10. An intelligent grading and early warning system for agricultural product quality and safety, characterized by: Applying the method for intelligent scoring of agricultural product quality and safety according to any one of claims 1 to 9, comprising: A regional analysis module obtains production area information of agricultural products, analyzes the production area information, and obtains a production area risk value of the agricultural products; An environmental analysis module obtains biological indicator data of agricultural products, obtains the growth health index of agricultural products, and obtains the production environment risk value of agricultural products based on the production area risk value; The pollution analysis module conducts metabolomics analysis on agricultural products, obtains metabolomics analysis results, and then performs pollution warning scores on agricultural products based on the metabolomics analysis results to generate pollution warning risk values; A consumer analysis module, which obtains consumer behavior data of agricultural products, analyzes the consumer behavior data, and obtains consumer feedback scores of agricultural products; The scoring and warning module comprehensively analyzes the production environment risk value, pollution warning risk value and consumer feedback score of agricultural products to obtain the comprehensive quality and safety score of agricultural products, generates a scoring level based on the comprehensive quality and safety score, and triggers a scoring level warning.

Images