Method and system for evaluating thermal stability of millet diglyceride oil

Abstract

The present invention relates to the technical field of edible oil quality detection, and discloses a millet diglyceride oil product thermal stability evaluation method, which comprises: grouping oil products to be evaluated according to raw material production places and processing batches, and constructing an evaluation database containing historical thermal stability detection events; screening the sample group subjected to thermal stability improvement treatment as a target sample group, extracting near infrared spectrum data of the target sample group, and determining an abnormal change node; and constructing a detection frequency adjustment response model based on the abnormal nodes, and dynamically optimizing the detection frequency. The system comprises a sample grouping module, an evaluation database construction module, a target sample group analysis module, an abnormal change node determination module and a real-time evaluation module. Through data association analysis and dynamic model adjustment, accurate evaluation of the thermal stability of the oil product is achieved, the problems that a traditional method is low in detection efficiency and insufficient in data utilization are solved, and effective technical support is provided for edible oil quality management and control.

Description

Technical Field

[0001] The present invention relates to the technical field of edible oil quality detection, and in particular to a method and system for evaluating the thermal stability of millet diglyceride oil. Background Art

[0002] The thermal stability of edible oils is one of the most important indicators of their quality, especially for functional oils such as millet diglyceride oil, where thermal stability directly affects their nutritional value and food safety. Traditional thermal stability assessment methods rely primarily on laboratory testing, such as thermogravimetric analysis or oxidative stability testing. While accurate, these methods suffer from long testing cycles and high costs. Furthermore, traditional methods often overlook the correlation between the origin of the oil raw materials, processing batches, and historical test data, making it difficult for the assessment results to fully reflect the actual thermal stability trends of the oil.

[0003] In actual production, the thermal stability of oil products can fluctuate due to raw material differences, processing adjustments, or changes in storage conditions. Existing technologies lack systematic analysis of historical test data, making it difficult to effectively identify points where thermal stability changes abruptly, and even more difficult to dynamically adjust assessment strategies based on testing frequency. For example, certain oil products may experience a significant decrease in thermal stability during a specific storage cycle. However, traditional methods, due to their fixed testing frequency, may miss critical change points, thus affecting the timeliness of subsequent process optimization or composition adjustments.

[0004] Although near-infrared spectroscopy has been applied to oil testing, its application in thermal stability assessment is still limited to the analysis of single test data, failing to integrate historical test events and processing operation records to form a dynamic assessment model. Therefore, developing an assessment method and system that can integrate multi-source data, dynamically adjust testing frequency, and accurately identify thermal stability change nodes is of great significance to improving the efficiency of oil quality management. Summary of the Invention

[0005] The object of the present invention is to provide a method and system for evaluating the thermal stability of millet diglyceride oil to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solution: a method and system for evaluating the thermal stability of millet diglyceride oil, the method comprising: Step A100: The millet diglyceride oil to be evaluated is grouped for sample testing according to the raw material origin and processing batch, with each group of samples representing a unique origin and batch combination; the thermal stability test events of each group of samples within the historical storage period are input into the evaluation system to form an evaluation database; Step A200: Based on the evaluation database, a sample group of the oil products to be evaluated that has a record of overheat stability improvement treatment operations is selected as a first sample group to be analyzed, and a sample group that has no record of such operations is selected as a second sample group to be analyzed; a verification analysis is performed on the first sample group to be analyzed to obtain a target sample group with improvement operations affecting detection frequency; and data of the second sample group to be analyzed is associated and stored; Step A300: extracting near-infrared spectrum detection data recorded in the target sample group, and determining abnormal change nodes for the oil products in the target sample group based on each thermal stability detection event; Step A400: Acquire the detection data and near-infrared spectrum data corresponding to the abnormal change node, and construct a detection frequency adjustment response model for the oil products in the target sample group; Step A500: When there is a real-time detection of the same type of oil that meets the corresponding detection frequency adjustment response model, a frequency adjustment signal is transmitted to the evaluation system.

[0007] Preferably, the step A200 includes the following specific steps: The thermal stability improvement treatment operation refers to the operation that requires ingredient adjustment or process optimization after thermal stability testing; Extract the detection time of the thermal stability detection event recorded in the first sample group to be analyzed, calculate the interval length S1 between two adjacent detection events based on the detection time, and mark the first sample group to be analyzed in which the interval length S1 generated by all detection events is the same as the same frequency sample group to be analyzed, otherwise mark it as the different frequency sample group to be analyzed; All detection events recorded for the same-frequency sample group to be analyzed are generated into a same-frequency time axis sequence in chronological order, and the detection data of each detection event is recorded with the detection time as the coordinate point; the coordinate point corresponding to the detection time of the thermal stability improvement treatment operation is marked as the target coordinate point, and the storage period of the corresponding sample is divided into a pre-treatment period and a post-treatment period with the target coordinate point as the dividing point; The first sample group to be analyzed, which has the same storage period length as the extraction evaluation system record and whose similarity of the ratio of the length of the period before treatment to the length of the period after treatment is greater than the set threshold, is the control sample group; Extracting test data recorded for the same-frequency sample group to be analyzed, wherein the test data refers to thermal decomposition temperature data transmitted by a thermogravimetric analyzer, determining the thermal decomposition temperature data for the same-frequency sample group to be analyzed to be the target data for the thermal stability improvement treatment operation, and calculating the thermal stability fluctuation rate of the same-frequency sample group to be analyzed; The above-mentioned thermal stability fluctuation rate calculation method is used to calculate the test data in the control sample group to obtain the corresponding thermal stability fluctuation rate, and the thermal stability fluctuation discrete value corresponding to the sample group to be analyzed at the same frequency is calculated; Set a discrete value threshold for thermal stability anomalies and extract the same-frequency sample group to be analyzed whose discrete value is less than the threshold as the target sample group.

[0008] Preferably, the step A200 further includes the following specific steps: Generate a hetero-frequency time axis sequence from all detection events recorded in the hetero-frequency sample group to be analyzed in chronological order, and mark the target coordinate points in the hetero-frequency time axis sequence, as well as the pre-processing period and the post-processing period; The interval between the detection time closest to the target coordinate point in the pre-processing period and the target coordinate point is constructed as the first detection period K1, and the adjacent detection times of the first detection period in the pre-processing period constitute the second detection period K2; based on the first detection period, a first detection frequency 1 / K1 is obtained, and a second detection frequency 1 / K2 is obtained in the second detection period; Calculate the thermal stability rate corresponding to the sample group to be analyzed at different frequencies; Set a thermal stability rate threshold, and extract the heterogeneous frequency sample group to be analyzed whose stability rate is less than or equal to the threshold as the target sample group; The data association storage of the second sample group to be analyzed refers to extracting the detection frequency of the second sample group to be analyzed in historical thermal stability detection events, and storing them in association with the raw material production place and the processing batch name.

[0009] Preferably, the step A300 includes the following steps: Step A310: The near-infrared spectrum detection data includes the absorbance value of the oil product at a specific wave number, and the near-infrared spectrum detection data is used to calculate the real-time thermal stability characteristic value of the target sample group in the pre-processing period based on the target coordinate point; Step A320: Using adjacent thermal stability detection events recorded in the target sample group as analysis units, analyzing the thermal stability characteristic value variation range of each analysis unit from the target coordinate point forward, and calculating the real-time thermal stability characteristic value variation amplitude in each analysis unit; Step A330: Select the extraction moment of the corresponding analysis unit whose difference between the change range and the change amplitude is less than the set threshold as the abnormal change node.

[0010] Preferably, the step A400 includes the following steps: Step A410: Constructing a response cycle with the abnormal change node as the end node and the first detection time recorded before the abnormal change node as the start node, extracting the target data value recorded within the response cycle, and calculating the thermal stability fluctuation rate based on the target data value. Using the response cycle as an analysis unit, calculate the variation range of the thermal stability characteristic value within the response cycle. Step A420: Extract the interval duration K0 of the response cycle and calculate the response detection frequency 1 / K0. If 1 / K0 is greater than or equal to the interval duration of the previous adjacent thermal stability detection event, output the detection frequency adjustment response model; if 1 / K0 is less than the interval duration of the previous adjacent detection event, use the interval duration of the previous adjacent detection event as the real-time detection frequency for reminder.

[0011] Preferably, the present invention also includes a millet diglyceride oil product thermal stability assessment system, which applies the above-mentioned millet diglyceride oil product thermal stability assessment method, and the system includes: a sample grouping module, an assessment database construction module, a target sample group analysis module, an association storage module, an abnormal change node determination module, a response model construction module, and a real-time assessment module; The sample grouping module is used to group the millet diglyceride oil to be evaluated according to the raw material origin and processing batch for sample testing; The evaluation database construction module is used to input the thermal stability detection events within the historical storage period of each group of samples into the evaluation system to form an evaluation database; The target sample group analysis module is used to perform verification analysis on the first sample group to be analyzed to obtain a target sample group with detection frequency impact improvement operation; The association storage module is used to perform data association storage on the second sample group to be analyzed; The abnormal change node determination module is used to determine abnormal change nodes for oil products in the target sample group based on each thermal stability detection event; The response model construction module is used to construct a detection frequency adjustment response model for oil products in the target sample group; The real-time evaluation module is used to transmit a frequency adjustment signal to the evaluation system when there is a real-time detection of the same type of oil that meets the corresponding detection frequency adjustment response model.

[0012] Preferably, the target sample group analysis module includes a sample group generation unit to be analyzed, a sample group labeling unit to be analyzed, a control sample group output unit, a thermal stability fluctuation rate calculation unit, a fluctuation discrete value calculation unit and a thermal stability stabilization rate calculation unit; The to-be-analyzed sample group generating unit is configured to generate a first to-be-analyzed sample group and a second to-be-analyzed sample group; The to-be-analyzed sample group marking unit is used to analyze the interval duration of the first to-be-analyzed sample group and mark it as a same-frequency to-be-analyzed sample group and a different-frequency to-be-analyzed sample group; The control sample group output unit is used to extract the first to-be-analyzed sample group having the same storage period length as the evaluation system record and whose similarity of the ratio of the length of the period before treatment to the length of the period after treatment is greater than a set threshold as the control sample group; The thermal stability fluctuation rate calculation unit is used to calculate the thermal stability fluctuation rate of the same frequency sample group to be analyzed; The anomaly discrete value calculation unit is used to calculate the thermal stability anomaly discrete values ​​corresponding to the sample groups to be analyzed at the same frequency, and extract the sample groups to be analyzed at the same frequency whose discrete values ​​are less than a threshold value as the target sample group; The thermal stability rate calculation unit is used to calculate the thermal stability rate corresponding to the heterogeneous frequency sample group to be analyzed, and extract the heterogeneous frequency sample group to be analyzed whose stability rate is less than or equal to a threshold as the target sample group.

[0013] Preferably, the abnormal change node determination module includes a characteristic value calculation unit, a characteristic value change analysis unit and a node output unit; The characteristic value calculation unit is used to calculate the real-time thermal stability characteristic value of the target sample group in the pre-processing period based on the target coordinate point using the near-infrared spectrum detection data; The characteristic value change analysis unit is used to analyze the thermal stability characteristic value change range of each analysis unit from the target coordinate point forward, and calculate the real-time thermal stability characteristic value change amplitude in each analysis unit; The node output unit is used to select the extraction moment of the corresponding analysis unit whose difference between the change range and the change amplitude is less than a set threshold as the abnormal change node.

[0014] Preferably, the response model construction module includes a response cycle construction unit and a response model output unit; The response cycle construction unit is used to construct a response cycle with the abnormal change node as the end node and the first detection time recorded before the abnormal change node as the start node; The response model output unit is used to extract the interval duration of the response cycle and calculate the response detection frequency. If the response detection frequency is greater than or equal to the interval duration of the previous adjacent thermal stability detection event, the corresponding detection frequency adjustment response model is output.

[0015] Preferably, the real-time evaluation module includes a signal transmission unit, and the signal transmission unit is used to transmit a frequency adjustment signal to the evaluation system when there is a real-time detection of the same type of oil that meets the corresponding detection frequency adjustment response model.

[0016] Compared with the prior art, the present invention has the following beneficial effects: This invention achieves a systematic analysis of the thermal stability of multi-source oil products by grouping them by raw material origin and processing batch, and constructing an evaluation database based on historical test data. By screening target sample groups based on the evaluation database, the impact of test frequency on thermal stability improvements can be accurately identified, providing reliable data support for subsequent analysis. By extracting near-infrared spectral data and identifying nodes with abnormal changes, the thermal stability of oil products over the storage cycle can be dynamically captured, avoiding the evaluation lag caused by traditional fixed test frequencies.

[0017] By building a detection frequency adjustment response model, the system dynamically optimizes detection strategies based on actual oil product changes, improving detection efficiency and assessment accuracy. When real-time detection of similar oil products meets the model's conditions, the system automatically transmits a frequency adjustment signal, enabling intelligent and precise adjustment of detection frequency and reducing errors caused by human intervention.

[0018] This invention also provides historical detection frequency references for oils that haven't undergone thermal stability enhancement treatment by associating data with the second sample group to be analyzed, further expanding the system's applicability and practicality. The system's modules work together to automate the entire process, from data acquisition to model building and real-time evaluation, significantly reducing labor and time costs.

[0019] Furthermore, this invention quantifies the changing trends of oil thermal stability by calculating the thermal stability fluctuation rate and stability rate, providing a scientific basis for process optimization and composition adjustment. The precise identification of abnormal change nodes and the dynamic construction of response cycles further enhance the system's flexibility and adaptability, enabling it to meet the assessment requirements of different storage conditions and oil characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a working principle diagram of the method for evaluating the thermal stability of millet diglyceride oil according to the present invention; Figure 2 A flowchart for processing a sample group to be analyzed at different frequencies; Figure 3 Flowchart for adjusting response model construction for detection frequency; Figure 4 This is the structural diagram of the evaluation system. DETAILED DESCRIPTION

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0022] See also Figures 1-4 The present invention relates to a method and system for evaluating the thermal stability of millet diglyceride oil, and the specific implementation steps are as follows: Step A100: The millet diglyceride oil to be evaluated is grouped for sample testing according to the raw material origin and processing batch, with each group of samples representing a unique origin and batch combination; the thermal stability test events of each group of samples within the historical storage period are input into the evaluation system to form an evaluation database.

[0023] Step A200: Based on the evaluation database, a sample group of the oil products to be evaluated that has recorded the operation of improving the overheat stability is selected as the first sample group to be analyzed, and a sample group that has not recorded such operation is selected as the second sample group to be analyzed; a verification analysis is performed on the first sample group to be analyzed to obtain a target sample group with the operation of improving the detection frequency; and data association is performed on the second sample group to be analyzed.

[0024] Step A300: extracting near-infrared spectrum detection data recorded in the target sample group, and determining abnormal change nodes for the oil products in the target sample group based on various thermal stability detection events.

[0025] Step A400: Acquire the detection data and near-infrared spectrum data corresponding to the abnormal change node, and construct a detection frequency adjustment response model for the oil products in the target sample group.

[0026] Step A500: When there is a real-time detection of the same type of oil that meets the corresponding detection frequency adjustment response model, a frequency adjustment signal is transmitted to the evaluation system.

[0027] Example 1: During step A200, it's important to clearly define the thermal stability improvement treatment, which refers to any ingredient adjustments or process optimizations required after thermal stability testing. This step is fundamental to all subsequent analysis; only by clearly defining these treatments can we accurately identify eligible samples.

[0028] For the first sample group to be analyzed, it is necessary to extract the detection time of the thermal stability detection event recorded therein. Based on these detection times, calculate the interval duration S1 between two adjacent detection events. Here, such a calculation operation needs to be performed for each first sample group to be analyzed. Then, the sample group is marked according to the calculated interval duration S1. If the interval duration S1 generated by all detection events recorded in a first sample group to be analyzed is the same, then it is marked as a same-frequency sample group to be analyzed; conversely, if there are different interval durations S1, it is marked as a different-frequency sample group to be analyzed.

[0029] For the sample group marked as being analyzed at the same frequency, it is necessary to generate a same-frequency timeline sequence for all the detection events recorded in chronological order. In this sequence, the detection data of each detection event is recorded with the detection time as the coordinate point. The detection data here will be used for further analysis and calculations later. At the same time, the coordinate point recording the detection time corresponding to the thermal stability improvement treatment operation should be marked and determined as the target coordinate point. Using this target coordinate point as the dividing point, the storage period of the corresponding sample is divided into the pre-processing period and the post-processing period. This division facilitates comparative analysis of the situation before and after the treatment.

[0030] A control sample group is extracted from the evaluation system, containing samples that meet specific criteria within the first set of samples to be analyzed. These criteria require that the storage cycles recorded in the evaluation system have the same length, and that the similarity between the lengths of the pre-processing cycle and the lengths of the post-processing cycle exceeds a predetermined threshold. The threshold is a pre-determined criterion used to determine whether the similarity meets the requirements.

[0031] After extracting the control sample group, the test data recorded for the sample group to be analyzed at the same frequency must be extracted. Here, the test data refers to the thermal decomposition temperature data transmitted by the thermogravimetric analyzer. Using this data, the thermal decomposition temperature data for the sample group to be analyzed at the same frequency is determined as the target data for the thermal stability improvement treatment. Then, based on the target data, the thermal stability fluctuation rate of the sample group to be analyzed at the same frequency is calculated. Calculating the thermal stability fluctuation rate requires a specific calculation method. Although no formula is involved here, it is important to clarify that the relevant calculation is based on the target data.

[0032] Using the same method used to calculate the thermal stability fluctuation rate for the same-frequency sample group, the test data within the control sample group is calculated to obtain the corresponding thermal stability fluctuation rate for the control sample group. With the thermal stability fluctuation rates of the same-frequency sample group and the control sample group, the thermal stability fluctuation discrete value corresponding to the same-frequency sample group can be calculated.

[0033] A pre-set threshold for the discrete value of thermal stability fluctuations is set. Based on this threshold, sample groups with discrete values ​​below the threshold are extracted from the sample group with the same frequency to be analyzed. These sample groups are identified as target sample groups. This completes the processing of the sample groups with the same frequency to be analyzed in the first sample group, resulting in a target sample group that meets the criteria. The entire process must be carried out strictly according to the above steps, and accurate data processing is required at every stage to ensure that the final target sample group accurately reflects the impact of the testing frequency on operational improvements.

[0034] Example 2: Step A200 also includes processing of the heterogeneous frequency sample group to be analyzed and data association storage operations for the second sample group to be analyzed. For the heterogeneous frequency sample group to be analyzed, it is necessary to generate a heterogeneous frequency time axis sequence for all the detection events recorded in it in chronological order. In the process of generating the time axis sequence, the target coordinate points therein must be marked at the same time. The target coordinate points here are consistent with the definition in the same frequency sample group to be analyzed, that is, the coordinate points that record the detection time corresponding to the thermal stability improvement treatment operation. In addition, it is necessary to use the target coordinate point as the dividing point to divide the storage period of the corresponding sample into a pre-processing period and a post-processing period, which is the same as the period division method of the same frequency sample group to be analyzed.

[0035] After completing the generation of the heterofrequency time axis sequence and marking the target coordinate points and periods, it is necessary to construct the detection period in the pre-processing period. Specifically, the interval between the detection time closest to the target coordinate point and the target coordinate point in the pre-processing period is constructed and defined as the first detection period K1. At the same time, the adjacent detection times of the first detection period in the pre-processing period constitute the second detection period K2. It should be clarified here that the adjacent detection time refers to the interval between the detection times adjacent to the first detection period in the pre-processing period. Based on the first detection period K1, the first detection frequency can be obtained, and the calculation method is 1 / K1; similarly, based on the second detection period K2, the second detection frequency, that is, 1 / K2, can be obtained.

[0036] The thermal stability rate for the heterogeneous sample group to be analyzed must be calculated. This calculation relies on the test data from the heterogeneous sample group, as well as the previously established test cycle and frequency information. While a specific formula is not required, it is important to clarify that this is a comprehensive calculation based on these data and parameters. After the calculation is complete, a predetermined threshold for the thermal stability rate is set. Based on this threshold, sample groups with stability rates equal to or less than the threshold are extracted from the heterogeneous sample group to be analyzed, and these sample groups are designated as target sample groups.

[0037] After completing the processing of the heterofrequency sample group to be analyzed in the first sample group to be analyzed, it is necessary to perform data association storage operations on the second sample group to be analyzed. Specifically, data association storage refers to extracting the detection frequency of the second sample group to be analyzed in the historical thermal stability detection events. The detection frequency here refers to the reciprocal of the time interval between adjacent detection events in the previous thermal stability detection process of the second sample group to be analyzed, that is, detection frequency = 1 / the length of the interval between adjacent detection events. After extracting the detection frequency, it is associated and stored according to the origin of the raw materials and the name of the processing batch. In other words, the detection frequency of each second sample group to be analyzed is associated with the origin of the raw materials and the name of the processing batch corresponding to the sample group, and stored in the evaluation system for subsequent query and use.

[0038] Throughout the entire implementation process, every step must be strictly followed according to the prescribed procedures. For the processing of heterogeneous sample groups to be analyzed, accurate data processing is required at every stage, from generating timeline sequences to marking target coordinate points and periods, to constructing detection cycles and calculating detection frequencies. When calculating the thermal stability rate, it is important to ensure that the underlying data is accurate and the calculation process is logical. When setting thresholds and extracting target sample groups, strict screening must be carried out according to the threshold criteria to ensure that the selected target sample groups accurately reflect the impact of detection frequency on thermal stability.

[0039] For the associated storage of data from the second sample group to be analyzed, ensure that the extracted test frequencies are accurate and correctly associated with the raw material origin and processing batch name. Ensure data integrity and queryability during storage to ensure rapid and accurate access to the required historical test frequency data for subsequent evaluations.

[0040] Example 3: Step A300 primarily involves identifying nodes within the target sample group indicating abnormal oil product changes, specifically through analysis of near-infrared spectroscopy data. The composition of near-infrared spectroscopy data must be clearly defined, as it includes the absorbance values ​​of the oil at specific wavenumbers. These absorbance values, obtained through near-infrared spectroscopy, reflect the composition and structural characteristics of the oil and, in turn, correlate with its thermal stability.

[0041] After acquiring the near-infrared spectral test data recorded for the target sample group, this data is used to calculate the real-time thermal stability characteristic value of the target sample group during the pre-treatment period based on the target coordinate point. The target coordinate point here is the coordinate point corresponding to the test time of the recorded thermal stability improvement treatment operation, as marked in step A200. The pre-treatment period is the first half of the storage period divided by this target coordinate point. The calculation of the real-time thermal stability characteristic value requires comprehensive consideration of the near-infrared spectral absorbance values ​​at each test time point during the pre-treatment period. These absorbance values ​​are converted into characteristic values ​​that represent thermal stability through a specific calculation method (not a formal description). This characteristic value can reflect the thermal stability state of the oil product at different times during the pre-treatment period.

[0042] Adjacent thermal stability detection events recorded in the target sample group are used as analysis units. Adjacent thermal stability detection events here refer to two detection events that are consecutive in time sequence, and the time period between each two adjacent detection events constitutes an analysis unit. The variation range of the thermal stability characteristic value of each analysis unit is analyzed from the target coordinate point forward. In other words, starting from the detection time corresponding to the target coordinate point, each analysis unit is traced back in sequence, and the maximum and minimum thermal stability characteristic values ​​within each analysis unit are calculated to determine the variation range of the characteristic value within the analysis unit.

[0043] In addition to determining the range of thermal stability characteristic value variation for each analysis unit, the real-time thermal stability characteristic value variation amplitude must also be calculated for each analysis unit. This variation amplitude is calculated based on the differences in thermal stability characteristic values ​​between adjacent detection time points within the analysis unit. By comparing the characteristic values ​​at adjacent detection time points, the variation in the characteristic values ​​is calculated, and the amplitude of the characteristic value variation within the analysis unit is derived. This amplitude reflects the rate of change or degree of fluctuation in the thermal stability of the oil product within that analysis unit's time period.

[0044] After completing the calculation of the variation range and variation amplitude of the thermal stability characteristic value of each analysis unit, it is necessary to select the extraction moment of the corresponding analysis unit whose difference between the variation range and the variation amplitude is less than the set threshold as the abnormal change node. The set threshold here is a predetermined standard value used to judge whether the difference between the variation range and the variation amplitude is within a reasonable range. For each analysis unit, the difference between the variation range and the variation amplitude of its thermal stability characteristic value is calculated. If the difference is less than the set threshold, it is considered that there is an abnormality in the thermal stability change within the analysis unit, and the extraction moment of the analysis unit, that is, the end time point of the analysis unit or a specific time point, is determined as the abnormal change node.

[0045] Throughout the entire process, every step must closely align with near-infrared spectroscopy data and thermal stability characteristic values. When acquiring near-infrared spectroscopy data, the accuracy of the instrument and the standardization of the testing process must be ensured to guarantee data reliability. When calculating thermal stability characteristic values, the actual state of the oil product during the pre-treatment cycle must be fully considered to ensure that the characteristic values ​​accurately reflect the thermal stability state. When analyzing the range and magnitude of change within each analytical unit, strict chronological order must be followed to avoid missing or incorrectly assigning analytical units.

[0046] Thresholds must be determined based on the oil's characteristics and actual testing experience to ensure reasonable thresholds and accurately screen for abnormal change nodes. Once identified, these nodes serve as a crucial basis for constructing a response model to adjust the testing frequency. Therefore, the entire process must ensure accurate data processing and logical rigor. By identifying abnormal change nodes within the target sample group, we can provide key time references for subsequent analysis of the causes and patterns of thermal stability changes, laying the foundation for optimizing testing frequency and improving the accuracy of thermal stability assessments.

[0047] Example 4: Step A400 is mainly to construct a detection frequency adjustment response model for oil products in the target sample group, which is specifically achieved through the extraction and analysis of data related to abnormal change nodes. First, the abnormal change node is used as the end node, and the first detection time recorded before the abnormal change node is used as the starting node to construct a response cycle. For example, assuming that the detection time corresponding to the abnormal change node of a certain oil product in the target sample group is T3, and the first detection time recorded before T3 is T1, then the time period from T1 to T3 constitutes a response cycle. The first detection time here refers to the detection time point before the abnormal change node that is farthest from the abnormal change node, ensuring that the response cycle can cover a complete detection interval before the abnormal change occurs.

[0048] After establishing a response cycle, it's necessary to extract the target data values ​​recorded within that cycle. These target data values ​​refer to the thermal decomposition temperature data for the thermal stability improvement treatment performed on the same-frequency sample group to be analyzed, as determined in step A200. These data, transmitted by the thermogravimetric analyzer, reflect the thermal decomposition temperatures of the oil product at different testing time points. The thermal stability fluctuation rate is calculated based on these target data values. This calculation comprehensively considers the changes in the target data values ​​within the response cycle, such as the fluctuation amplitude of the data at different time points and comparisons with historical data. Using specific calculation logic (not formalized), a numerical value representing the degree of abnormal thermal stability changes is derived.

[0049] At the same time, the response cycle is used as an analysis unit to calculate the range of the thermal stability characteristic value within that cycle. The thermal stability characteristic value, calculated in step A300 using near-infrared spectroscopy data, reflects the thermal stability of the oil product. Within this analysis unit, the maximum and minimum values ​​of the thermal stability characteristic value must be determined to determine the range of variation within that cycle. This range can reflect the fluctuation of the oil product's thermal stability within the response cycle.

[0050] After calculating the relevant data within the response cycle, extract the interval duration K0 of the response cycle, which is the time length between the start node and the end node. The response detection frequency is calculated based on the interval duration K0, using the calculation method of 1 / K0. Next, the response detection frequency needs to be compared with the interval duration of the previous adjacent thermal stability detection event. The previous adjacent thermal stability detection event refers to the detection event that is adjacent to the starting node before the start node of the response cycle. Its interval duration refers to the time length between this adjacent detection event and the starting node.

[0051] If the calculated response detection frequency 1 / K0 is greater than or equal to the interval between the previous thermal stability detection events, a detection frequency adjustment response model is output. For example, assuming the interval between the previous thermal stability detection events is 10 days and the response cycle interval K0 is 8 days, then the response detection frequency 1 / K0 is 1 / 8 day⁻¹, which translates to an interval of 8 days. In this case, 8 days is shorter than 10 days, and the corresponding response detection frequency is higher. This indicates that the higher detection frequency may have played a role in capturing thermal stability changes before the abnormal change node appeared. Therefore, a corresponding detection frequency adjustment response model can be output, which may indicate the detection frequency that should be adopted in similar situations.

[0052] If 1 / K0 is less than the interval between the previous detection events, the interval between the previous detection events is used as the real-time detection frequency for reminders. For example, if the interval between the previous detection events is 5 days, the response cycle interval K0 is 6 days, and the response detection frequency 1 / K0 is 1 / 6 day⁻¹, the corresponding interval of 6 days is greater than 5 days, and the response detection frequency is lower. This may mean that the detection frequency was low before the abnormal change node appeared, and the change in thermal stability was not captured in time. Therefore, it is necessary to use the interval of 5 days between the previous detection events as the real-time detection frequency for reminders, prompting the user to maintain a high detection frequency in subsequent tests to avoid similar situations.

[0053] Throughout the implementation process, the construction of the response cycle requires accurate determination of the start and end points to ensure that the cycle fully encompasses the detection data related to abnormal changes. When extracting target data values ​​and thermal stability characteristic values, data accuracy and completeness must be ensured to avoid data omissions or errors. When calculating the thermal stability fluctuation rate and variation range, established logic and methods must be followed to ensure that the results truly reflect the changes in the thermal stability of the oil product.

[0054] When comparing the response detection frequency with the duration of the previous adjacent detection event, the comparison criteria and logic must be clarified, and reasonable judgments and processing must be made based on the comparison results. When outputting the detection frequency adjustment response model or issuing reminders, it is necessary to ensure that the model and reminder content can provide effective guidance for subsequent real-time detection, help optimize the detection frequency, and improve the accuracy of the thermal stability assessment of oil products. The entire process needs to focus on the relevance and logic of the data. Each step of the operation must have a clear basis and purpose to ensure the scientific and practical nature of the detection frequency adjustment response model and provide strong support for the assessment and management of oil thermal stability. During the implementation process, data recording and archiving must also be done to facilitate subsequent verification and optimization of the model's effectiveness.

[0055] Example 5: In this example, a system for evaluating the thermal stability of millet diglyceride oil applies the above-mentioned evaluation method. The system is composed of multiple functional modules, and each module works together to achieve the evaluation of the thermal stability of the oil.

[0056] The sample grouping module groups the millet diglyceride oil samples for testing by origin and processing batch. For example, if the oil to be evaluated comes from three different origins, each with two processing batches, the sample grouping module will divide these oils into 3 x 2 = 6 sample groups. Each group of samples corresponds to a unique origin and batch combination, such as Batch 1 from Origin A, Batch 2 from Origin A, Batch 1 from Origin B, and so on. This ensures that the source information of each sample group is clear and unambiguous.

[0057] The evaluation database construction module is responsible for inputting thermal stability test events within the historical storage period of each sample group into the evaluation system to form an evaluation database. For example, for the sample group of Batch 1 from Origin A, its historical storage period is 12 months. During this period, four thermal stability tests were conducted, with the test time being the first, third, sixth, and twelfth months of storage. The thermal decomposition temperature data, near-infrared spectrum test data, and other related information of each test are input into the evaluation system to form the evaluation data record of the sample group. These data of all sample groups together constitute the evaluation database.

[0058] The target sample group analysis module is used to perform a verification analysis on the first sample group to be analyzed to obtain a target sample group with improvement operations that are affected by the detection frequency. Assuming that 10 first sample groups to be analyzed that record overheating stability improvement operations are screened out from the evaluation database, the target sample group analysis module will first process these sample groups. Among them, the sample group generation unit to be analyzed generates a first sample group to be analyzed and a second sample group to be analyzed that does not record such operations; the sample group marking unit to be analyzed analyzes the interval length of the first sample group to be analyzed and marks it as a same-frequency or different-frequency sample group to be analyzed. For example, if the interval length between adjacent detection events in one sample group is 3 months, it is marked as a same-frequency sample group to be analyzed. If the interval length of another sample group is 2 months, 4 months, and 3 months, respectively, it is marked as a different-frequency sample group to be analyzed.

[0059] The control sample group output unit extracts, from the same-frequency sample group to be analyzed, a sample group with the same storage period length and a similarity ratio between the pre-processing period and the post-processing period length that exceeds a set threshold as the control sample group. The thermal stability fluctuation rate calculation unit calculates the thermal stability fluctuation rate of the same-frequency sample group to be analyzed, and the fluctuation discrete value calculation unit calculates the fluctuation discrete value, extracting the sample group with a discrete value less than the threshold as the target sample group. Simultaneously, the thermal stability stability rate calculation unit calculates the stability rate of the different-frequency sample group to be analyzed, extracting the sample group with a stability rate less than or equal to the threshold as the target sample group.

[0060] The associated storage module associates and stores data for the second set of samples to be analyzed. For example, if the second set of samples to be analyzed includes samples from Batch 2 of Origin B, and its historical thermal stability testing frequency is once every two months, the associated storage module will associate this testing frequency with the name "Origin B, Batch 2" and store it in the system for subsequent querying.

[0061] The Abnormal Change Node Determination Module identifies abnormal change nodes for oil products within a target sample group based on each thermal stability test event. For a specific target sample group, the eigenvalue calculation unit uses its near-infrared spectral test data to calculate the real-time thermal stability eigenvalue for that sample group during the pre-processing period based on the target coordinate point. Assuming the target coordinate point corresponds to the sixth month of storage and the pre-processing period is the previous six months, the eigenvalue calculation unit calculates the corresponding thermal stability eigenvalue based on the near-infrared spectral absorbance values ​​at each test time point within the previous six months.

[0062] The eigenvalue change analysis unit analyzes the range and magnitude of thermal stability eigenvalue changes for each adjacent detection event, starting from the target coordinate point (month 6). For example, if the adjacent detection events are months 3 and 6, the maximum and minimum eigenvalues ​​within the unit are analyzed to obtain the range of change, and the difference in eigenvalues ​​between adjacent detection time points is calculated to obtain the magnitude of change. The node output unit selects the extraction time of the corresponding analysis unit where the difference between the range and magnitude of change is less than a set threshold as the abnormal change node. If the difference between the analysis unit from month 3 to month 6 meets the criteria, month 6 is identified as the abnormal change node.

[0063] The response model construction module is used to construct a response model for adjusting the detection frequency of oil products within the target sample group. Taking the identified abnormal change node (month 6) as an example, the response cycle construction unit uses this node as the end node and the first detection time recorded before it (month 1) as the start node to construct the response cycle (months 1 to 6). The response model output unit extracts the interval duration K0 of this response cycle as 5 months, calculates the response detection frequency 1 / K0 as 1 / 5 months⁻¹, and compares it with the interval duration of the previous adjacent detection event (assuming there were no detection events before month 1, the previous adjacent detection event has an initial detection time of 0 months and an interval duration of 1 month). If 1 / K0 is greater than or equal to this interval duration, the detection frequency adjustment response model is output; if it is less than this interval duration, the interval duration of the previous adjacent detection event is used as the real-time detection frequency reminder.

[0064] The real-time assessment module is designed to transmit a frequency adjustment signal to the assessment system when real-time testing of similar oil products meets the corresponding detection frequency adjustment response model. For example, if the real-time testing of similar oil products from Origin A, Batch 1, meets the previously constructed detection frequency adjustment response model, the signal transmission unit in the real-time assessment module will transmit a frequency adjustment signal to the assessment system, prompting it to adjust the detection frequency.

[0065] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0066] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for evaluating the thermal stability of millet diglyceride oil, characterized in that: The evaluation method comprises the following steps: Step A100: The millet diglyceride oil to be evaluated is grouped for sample testing according to the raw material origin and processing batch, with each group of samples representing a unique origin and batch combination; the thermal stability test events of each group of samples within the historical storage period are input into the evaluation system to form an evaluation database; Step A200: Based on the evaluation database, a sample group of the oil products to be evaluated that has a record of overheat stability improvement treatment operations is selected as a first sample group to be analyzed, and a sample group that has no record of such operations is selected as a second sample group to be analyzed; a verification analysis is performed on the first sample group to be analyzed to obtain a target sample group with improvement operations affecting detection frequency; and data of the second sample group to be analyzed is associated and stored; Step A300: extracting near-infrared spectrum detection data recorded in the target sample group, and determining abnormal change nodes for the oil products in the target sample group based on each thermal stability detection event; Step A400: Acquire the detection data and near-infrared spectrum data corresponding to the abnormal change node, and construct a detection frequency adjustment response model for the oil products in the target sample group; Step A500: When there is a real-time detection of the same type of oil that meets the corresponding detection frequency adjustment response model, a frequency adjustment signal is transmitted to the evaluation system.

2. The method for evaluating the thermal stability of millet diglyceride oil according to claim 1, wherein: The step A200 includes the following specific steps: The thermal stability improvement treatment operation refers to the operation that requires ingredient adjustment or process optimization after thermal stability testing; Extract the detection time of the thermal stability detection event recorded in the first sample group to be analyzed, calculate the interval length S1 between two adjacent detection events based on the detection time, and mark the first sample group to be analyzed in which the interval length S1 generated by all detection events is the same as the same frequency sample group to be analyzed, otherwise mark it as the different frequency sample group to be analyzed; All detection events recorded for the same-frequency sample group to be analyzed are generated into a same-frequency time axis sequence in chronological order, and the detection data of each detection event is recorded with the detection time as the coordinate point; the coordinate point corresponding to the detection time of the thermal stability improvement treatment operation is marked as the target coordinate point, and the storage period of the corresponding sample is divided into a pre-treatment period and a post-treatment period with the target coordinate point as the dividing point; The first sample group to be analyzed, which has the same storage period length as the extraction evaluation system record and whose similarity of the ratio of the length of the period before treatment to the length of the period after treatment is greater than the set threshold, is the control sample group; Extracting test data recorded for the same-frequency sample group to be analyzed, wherein the test data refers to thermal decomposition temperature data transmitted by a thermogravimetric analyzer, determining the thermal decomposition temperature data for the same-frequency sample group to be analyzed to be the target data for the thermal stability improvement treatment operation, and calculating the thermal stability fluctuation rate of the same-frequency sample group to be analyzed; The above-mentioned thermal stability fluctuation rate calculation method is used to calculate the test data in the control sample group to obtain the corresponding thermal stability fluctuation rate, and the thermal stability fluctuation discrete value corresponding to the sample group to be analyzed at the same frequency is calculated; Set a discrete value threshold for thermal stability anomalies and extract the same-frequency sample group to be analyzed whose discrete value is less than the threshold as the target sample group.

3. The method for evaluating the thermal stability of millet diglyceride oil according to claim 2, wherein: The step A200 further includes the following specific steps: Generate a hetero-frequency time axis sequence from all detection events recorded in the hetero-frequency sample group to be analyzed in chronological order, and mark the target coordinate points in the hetero-frequency time axis sequence, as well as the pre-processing period and the post-processing period; The interval between the detection time closest to the target coordinate point in the pre-processing period and the target coordinate point is constructed as the first detection period K1, and the adjacent detection times of the first detection period in the pre-processing period constitute the second detection period K2; based on the first detection period, a first detection frequency 1 / K1 is obtained, and a second detection frequency 1 / K2 is obtained in the second detection period; Calculate the thermal stability rate corresponding to the sample group to be analyzed at different frequencies; Set a thermal stability rate threshold, and extract the heterogeneous frequency sample group to be analyzed whose stability rate is less than or equal to the threshold as the target sample group; The data association storage of the second sample group to be analyzed refers to extracting the detection frequency of the second sample group to be analyzed in historical thermal stability detection events, and storing them in association with the raw material production place and the processing batch name.

4. The method for evaluating the thermal stability of millet diglyceride oil according to claim 1, wherein: The step A300 includes the following steps: Step A310: The near-infrared spectrum detection data includes the absorbance value of the oil product at a specific wave number, and the near-infrared spectrum detection data is used to calculate the real-time thermal stability characteristic value of the target sample group in the pre-processing period based on the target coordinate point; Step A320: Using adjacent thermal stability detection events recorded in the target sample group as analysis units, analyzing the thermal stability characteristic value variation range of each analysis unit from the target coordinate point forward, and calculating the real-time thermal stability characteristic value variation amplitude in each analysis unit; Step A330: Select the extraction moment of the corresponding analysis unit whose difference between the change range and the change amplitude is less than the set threshold as the abnormal change node.

5. The method for evaluating the thermal stability of millet diglyceride oil according to claim 1, wherein: The step A400 includes the following steps: Step A410: Constructing a response cycle with the abnormal change node as the end node and the first detection time recorded before the abnormal change node as the start node, extracting the target data value recorded within the response cycle, and calculating the thermal stability fluctuation rate based on the target data value. Using the response cycle as an analysis unit, calculate the variation range of the thermal stability characteristic value within the response cycle. Step A420: Extract the interval duration K0 of the response cycle and calculate the response detection frequency 1 / K0. If 1 / K0 is greater than or equal to the interval duration of the previous adjacent thermal stability detection event, output the detection frequency adjustment response model; if 1 / K0 is less than the interval duration of the previous adjacent detection event, use the interval duration of the previous adjacent detection event as the real-time detection frequency for reminder.

6. A system for evaluating the thermal stability of millet diglyceride oil, using the method for evaluating the thermal stability of millet diglyceride oil according to any one of claims 1 to 5, characterized in that: The system includes a sample grouping module, an evaluation database construction module, a target sample group analysis module, an association storage module, an abnormal change node determination module, a response model construction module and a real-time evaluation module; The sample grouping module is used to group the millet diglyceride oil to be evaluated according to the raw material origin and processing batch for sample testing; The evaluation database construction module is used to input the thermal stability detection events within the historical storage period of each group of samples into the evaluation system to form an evaluation database; The target sample group analysis module is used to perform verification analysis on the first sample group to be analyzed to obtain a target sample group with detection frequency impact improvement operation; The association storage module is used to perform data association storage on the second sample group to be analyzed; The abnormal change node determination module is used to determine abnormal change nodes for oil products in the target sample group based on each thermal stability detection event; The response model construction module is used to construct a detection frequency adjustment response model for oil products in the target sample group; The real-time evaluation module is used to transmit a frequency adjustment signal to the evaluation system when there is a real-time detection of the same type of oil that meets the corresponding detection frequency adjustment response model.

7. The oil product thermal stability evaluation system of millet diglyceride oil according to claim 6, characterized in that: The target sample group analysis module includes a sample group generation unit to be analyzed, a sample group marking unit to be analyzed, a control sample group output unit, a thermal stability fluctuation rate calculation unit, a fluctuation discrete value calculation unit and a thermal stability stability rate calculation unit; The to-be-analyzed sample group generating unit is configured to generate a first to-be-analyzed sample group and a second to-be-analyzed sample group; The to-be-analyzed sample group marking unit is used to analyze the interval duration of the first to-be-analyzed sample group and mark it as a same-frequency to-be-analyzed sample group and a different-frequency to-be-analyzed sample group; The control sample group output unit is used to extract the first to-be-analyzed sample group having the same storage period length as the evaluation system record and whose similarity of the ratio of the length of the period before treatment to the length of the period after treatment is greater than a set threshold as the control sample group; The thermal stability fluctuation rate calculation unit is used to calculate the thermal stability fluctuation rate of the same frequency sample group to be analyzed; The anomaly discrete value calculation unit is used to calculate the thermal stability anomaly discrete values ​​corresponding to the sample groups to be analyzed at the same frequency, and extract the sample groups to be analyzed at the same frequency whose discrete values ​​are less than a threshold value as the target sample group; The thermal stability rate calculation unit is used to calculate the thermal stability rate corresponding to the heterogeneous frequency sample group to be analyzed, and extract the heterogeneous frequency sample group to be analyzed whose stability rate is less than or equal to a threshold as the target sample group.

8. The oil product thermal stability evaluation system of millet diglyceride oil according to claim 6, characterized in that: The abnormal change node determination module includes a characteristic value calculation unit, a characteristic value change analysis unit and a node output unit; The characteristic value calculation unit is used to calculate the real-time thermal stability characteristic value of the target sample group in the pre-processing period based on the target coordinate point using the near-infrared spectrum detection data; The characteristic value change analysis unit is used to analyze the thermal stability characteristic value change range of each analysis unit from the target coordinate point forward, and calculate the real-time thermal stability characteristic value change amplitude in each analysis unit; The node output unit is used to select the extraction moment of the corresponding analysis unit whose difference between the change range and the change amplitude is less than a set threshold as the abnormal change node.

9. The oil product thermal stability evaluation system of millet diglyceride oil according to claim 6, characterized in that: The response model construction module includes a response cycle construction unit and a response model output unit; The response cycle construction unit is used to construct a response cycle with the abnormal change node as the end node and the first detection time recorded before the abnormal change node as the start node; The response model output unit is used to extract the interval duration of the response cycle and calculate the response detection frequency. If the response detection frequency is greater than or equal to the interval duration of the previous adjacent thermal stability detection event, the corresponding detection frequency adjustment response model is output.

10. The oil product thermal stability evaluation system of millet diglyceride oil according to claim 6, characterized in that: The real-time evaluation module includes a signal transmission unit, which is used to transmit a frequency adjustment signal to the evaluation system when there is a real-time detection of the same type of oil that meets the corresponding detection frequency adjustment response model.

Images