A method and system for dynamically adjusting the temperature gradient of pepper freezing

Abstract

The present invention discloses a method and system for dynamically adjusting the temperature gradient of pepper freezing, which specifically relates to the technical field of industrial freezing temperature control. The method is used to solve the problem of temperature gradient inaccuracy caused by the static model failing to perceive the dynamic changes of material thermodynamic parameters in real time in the existing freezing control method. The method collects temperature distribution and gas flow rate data in the freezing process in real time, screens a high-confidence data subset in the low-flow rate area based on spatial correlation, uses a recursive algorithm to identify the dynamic thermal conductivity coefficient of the material online, and combines a dual-threshold judgment mechanism of energy release rate and phase lag angle to generate a stable parameter sequence through segmented fitting. The method further dynamically adjusts the preset temperature gradient model through a nonlinear correction strategy to generate an adaptive gradient control curve, ultimately driving the heat exchange rate adjustment to achieve closed-loop tracking of the target curve by the actual temperature gradient. The method significantly improves the freezing uniformity and energy efficiency of high-moisture materials, and improves the control accuracy caused by thermal response mismatch.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial freezing temperature control, and more particularly to a method and system for dynamically adjusting the freezing temperature gradient of peppers. Background Art

[0002] In the temperature control of industrial freezing production lines, dynamic gradient adjustment is the key to ensuring uniform freezing of materials. Existing control methods generally adopt fixed temperature curves or linear cooling strategies based on static models. For example, basic temperature control is achieved through preset PID parameters or PLC programmed logic. That is, the default material thermodynamic parameters (such as thermal conductivity and phase change rate) are constant during the freezing process, which can meet conventional freezing requirements. However, when processing high-moisture materials such as peppers, the temperature gradient tracking accuracy will gradually decay with the freezing process.

[0003] In the existing technology, the existing freezing control system fails to perceive the changes in dynamic physical parameters such as thermal conductivity and moisture phase change rate when peppers are frozen in real time, resulting in a mismatch between the preset temperature gradient model and the actual thermal response of the material, thereby causing local supercooling damage and abnormal fluctuations in energy consumption. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a method and system for dynamically adjusting the temperature gradient of pepper freezing to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A method for dynamically adjusting the temperature gradient of pepper freezing, comprising the following steps:

[0007] S1. Real-time collection of temperature distribution data of target materials during freezing and gas flow rate distribution in the freezing area;

[0008] S2. Perform dynamic confidence scoring on the temperature distribution data based on the spatial correlation between the gas velocity distribution and the temperature distribution data, and select a high-confidence data subset in the low velocity area;

[0009] S3. Based on the high-confidence data subset, the real-time thermal conductivity coefficient of the target material is identified online through a recursive algorithm to generate a thermal conductivity coefficient sequence, including:

[0010] Based on the high-confidence data subset, the real-time thermal conductivity coefficient of the target material is identified online through the recursive least squares method. The forgetting factor of the recursive least squares method is adjusted according to the dynamic confidence score of the high-confidence data subset. The lower the dynamic confidence score, the larger the forgetting factor value.

[0011] The thermal conductivity coefficients identified at each moment are arranged in chronological order to generate a thermal conductivity coefficient sequence. During the generation of the thermal conductivity coefficient sequence, the change rate of the thermal conductivity coefficients at adjacent moments is smoothed. If the change rate of the thermal conductivity coefficient exceeds a preset change threshold, the thermal conductivity coefficient at the current moment is corrected using a sliding average algorithm.

[0012] S4. Calculating the energy release rate and the phase lag angle of the adjacent regions based on the thermal conductivity coefficient sequence. If the energy release rate exceeds the critical threshold of the phase change and the phase lag angle exceeds the dynamic tolerance, a stable parameter sequence is generated by piecewise fitting.

[0013] S5. Performing nonlinear correction on the preset temperature gradient model according to the stable parameter sequence to generate an updated gradient control curve;

[0014] S6. Generate a temperature set value deviation for each freezing zone based on the gradient control curve, and generate a drive instruction to adjust the heat exchange rate so that the actual temperature gradient tracks the gradient control curve.

[0015] In a preferred embodiment, real-time acquisition of temperature distribution data of the target material during freezing and gas flow rate distribution in the freezing area includes:

[0016] The temperature data of each monitoring point during the freezing process of the target material is collected in real time through a matrix temperature sensor array to generate temperature distribution data;

[0017] The ultrasonic flow velocity sensor is used to collect the gas flow velocity data of each monitoring point in the freezing area and generate the gas flow velocity distribution;

[0018] The temperature distribution data and gas flow rate distribution are synchronized and aligned, and the temperature data is subjected to signal noise reduction based on the sliding average filtering algorithm, and the gas flow rate data is subjected to outlier filtering based on the median filtering algorithm.

[0019] In a preferred embodiment, the temperature distribution data is dynamically scored based on the spatial correlation between the gas velocity distribution and the temperature distribution data, and a high confidence data subset in the low velocity area is screened, including:

[0020] Marking the area in the gas flow rate distribution where the gas flow rate is lower than a preset flow rate threshold as a low flow rate area, otherwise marking it as a high flow rate area; and extracting the temperature distribution data corresponding to the low flow rate area;

[0021] Based on the trend consistency of the temperature change rate of adjacent freezing areas within the time window, the temperature data stability coefficient of the low flow rate area is calculated. The temperature data stability coefficient is the standard deviation of the temperature data of the corresponding freezing area.

[0022] Generate a dynamic confidence score for the temperature distribution data based on the ratio of the temperature data stability coefficient of the low flow rate area to the temperature change rate of the adjacent high flow rate area;

[0023] If the dynamic confidence score is higher than the dynamic confidence threshold, the corresponding temperature data point is determined to be a high-confidence data point;

[0024] All high-confidence data points are density clustered according to spatial coordinates. If the distance between a high-confidence data point and its nearest neighbor exceeds the preset clustering radius, it is removed to generate a high-confidence data subset.

[0025] In a preferred embodiment, the energy release rate and the phase lag angle of the adjacent region are calculated based on the thermal conductivity coefficient sequence. If the energy release rate exceeds the phase change critical threshold and the phase lag angle exceeds the dynamic tolerance, a stable parameter sequence is generated by piecewise fitting, including:

[0026] The energy release rate is calculated based on the thermal conductivity series. The energy release rate is the product of the thermal conductivity change rate and the temperature gradient. The temperature gradient is calculated based on the difference of the temperature distribution data of adjacent frozen areas.

[0027] Based on the heat transfer coefficient sequence, a cross-correlation analysis is performed on the heat transfer coefficient time series of adjacent freezing areas, the time delay corresponding to the maximum cross-correlation coefficient is extracted, and the time delay is converted into a phase lag angle.

[0028] If the energy release rate exceeds the critical threshold of phase change and the phase lag angle exceeds the dynamic tolerance, a mutation point is marked in the thermal conductivity coefficient sequence, and the thermal conductivity coefficient sequence is divided into multiple sub-segments based on the mutation point;

[0029] The least squares fitting is performed on the heat conduction coefficient in each sub-segment to generate a stable parameter sequence; the order of the least squares fitting is dynamically adjusted according to the length of the sub-segment. The longer the sub-segment, the higher the fitting order.

[0030] In a preferred embodiment, a nonlinear correction is performed on the preset temperature gradient model according to the stable parameter sequence to generate an updated gradient control curve, including:

[0031] Based on the baseline value and change trend of the heat transfer coefficient in the stable parameter sequence, the target temperature value of each freezing area of the preset temperature gradient model is corrected by nonlinear mapping;

[0032] The corrected target temperature values of each freezing zone are interpolated piecewise according to the spatial distribution to generate an updated gradient control curve;

[0033] The time axis of the gradient control curve is smoothed and filtered to eliminate the sudden step change of the temperature setting value caused by the parameter correction, and the final updated gradient control curve is generated.

[0034] In a preferred embodiment, the order of the segmented interpolation is adaptively adjusted according to the temperature difference between adjacent regions. The higher the temperature difference between adjacent regions, the higher the interpolation order.

[0035] In a preferred embodiment, the temperature set value deviation of each freezing zone is generated based on the gradient control curve, and the drive instruction is generated to adjust the heat exchange rate so that the actual temperature gradient tracks the gradient control curve, including:

[0036] Calculate the temperature set value deviation of each freezing zone based on the gradient control curve. The temperature set value deviation is the difference between the current actual temperature and the target temperature at the corresponding moment in the gradient control curve.

[0037] When generating drive instructions to adjust the heat exchange rate based on the temperature set value deviation, a proportional-integral controller is used to calculate the heat exchange rate adjustment amount for each freezing zone;

[0038] Dynamic damping compensation is performed on the heat exchange rate adjustment based on the temperature change rate difference between adjacent freezing zones. The greater the temperature change rate difference, the higher the damping compensation coefficient. Nonlinear compensation is enabled when the absolute value difference of the temperature change rate between adjacent zones exceeds a preset difference threshold.

[0039] The compensated heat exchange rate adjustment amount is converted into a compressor frequency adjustment instruction and an air valve opening instruction, so that the actual temperature gradient tracks the gradient control curve.

[0040] In another aspect, the present invention provides a system for dynamically adjusting the temperature gradient of pepper freezing, comprising:

[0041] Temperature and flow synchronous acquisition module: real-time acquisition of temperature distribution data of target materials during freezing and gas flow rate distribution in the freezing area;

[0042] Dynamic confidence screening module: Dynamic confidence scoring of temperature distribution data is performed based on the spatial correlation between gas velocity distribution and temperature distribution data, and a high-confidence data subset in the low velocity area is screened;

[0043] Thermal conductivity recursive identification module: Based on a high-confidence data subset, the module uses a recursive algorithm to online identify the real-time thermal conductivity coefficient of the target material and generate a thermal conductivity coefficient sequence;

[0044] Parameter segmented fitting module: Calculates the energy release rate and the phase lag angle of adjacent regions based on the thermal conductivity coefficient sequence. If the energy release rate exceeds the critical threshold of phase change and the phase lag angle exceeds the dynamic tolerance, segmented fitting is performed to generate a stable parameter sequence.

[0045] Gradient curve generation module: performs nonlinear correction on the preset temperature gradient model according to the stable parameter sequence to generate an updated gradient control curve;

[0046] Thermal control drive generation module: Generates the temperature set value deviation of each freezing area based on the gradient control curve, generates drive instructions to adjust the heat exchange rate, and makes the actual temperature gradient track the gradient control curve.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] 1. By real-time sensing of dynamic physical parameter changes during the freezing process and coupling a multi-source data fusion mechanism, the temperature gradient control accuracy and system stability during the freezing of high-moisture materials are significantly improved; based on the spatial correlation of temperature and flow velocity distribution, high-confidence data in low-flow velocity areas are dynamically screened to effectively avoid measurement noise caused by airflow disturbances and provide reliable input for subsequent parameter identification; through online identification of the heat conductivity coefficient through a recursive algorithm, the drift of thermodynamic properties during the phase change of the material is captured in real time, solving the gradient mismatch problem caused by parameter solidification in traditional static models; combining the dual-threshold judgment mechanism of energy release rate and phase lag angle, it accurately identifies abnormal sudden changes in heat conduction and triggers segmented fitting, generating a stable parameter sequence with clear physical meaning, ensuring the timeliness and robustness of model correction; based on the nonlinear correction strategy of stable parameters, the local slope and curvature of the gradient control curve are dynamically adjusted, so that the preset model can adaptively fit the actual thermal response characteristics of the material, greatly reducing the risk of local overcooling or undercooling.

[0049] 2. Dynamic tracking and control of temperature gradients are achieved through a closed-loop feedback mechanism, which optimizes the system's energy efficiency while ensuring freezing uniformity. The real-time update of the gradient control curve and the coordinated adjustment of the heat exchange rate overcome the defects in handling complex heat transfer hysteresis effects. The dynamic damping compensation mechanism based on the difference in temperature change rates between adjacent regions suppresses the gradient tearing problem caused by sudden changes in control instructions, ensuring a smooth transition during the freezing process. At the same time, the parameter-driven instruction generation logic converts abstract thermodynamic responses into executable device control signals, forming a complete technical closed loop from data perception, model iteration to execution optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is a flow chart of a method for dynamically adjusting the temperature gradient of pepper freezing according to the present invention;

[0051] Figure 2 The present invention is a schematic structural diagram of a system for dynamically adjusting the temperature gradient of pepper freezing. DETAILED DESCRIPTION

[0052] The following will provide a clear and complete description of 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. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0053] Example 1: Figure 1 The present invention provides a method for dynamically adjusting the temperature gradient of pepper freezing, comprising the following steps:

[0054] S1. Real-time collection of temperature distribution data of target materials during freezing and gas flow rate distribution in the freezing area;

[0055] S2. Perform dynamic confidence scoring on the temperature distribution data based on the spatial correlation between the gas velocity distribution and the temperature distribution data, and select a high-confidence data subset in the low velocity area;

[0056] S3. Based on the high-confidence data subset, the real-time thermal conductivity coefficient of the target material is identified online through a recursive algorithm to generate a thermal conductivity coefficient sequence;

[0057] S4. Calculating the energy release rate and the phase lag angle of the adjacent regions based on the thermal conductivity coefficient sequence. If the energy release rate exceeds the critical threshold of the phase change and the phase lag angle exceeds the dynamic tolerance, a stable parameter sequence is generated by piecewise fitting.

[0058] S5. Performing nonlinear correction on the preset temperature gradient model according to the stable parameter sequence to generate an updated gradient control curve;

[0059] S6. Generate a temperature set value deviation for each freezing zone based on the gradient control curve, and generate a drive instruction to adjust the heat exchange rate so that the actual temperature gradient tracks the gradient control curve.

[0060] S1. Real-time collection of temperature distribution data of the target material during freezing and gas flow rate distribution in the freezing area. Specifically:

[0061] The temperature data of each monitoring point during the freezing process of the target material is collected in real time through a matrix temperature sensor array to generate temperature distribution data. The matrix temperature sensor array consists of multiple temperature sensors arranged in rows and columns with equal spacing. The row spacing and column spacing between the temperature sensors are set to 50 mm to 100 mm according to the size of the target material and the freezing uniformity requirements. For example, for a larger stack of peppers, the row spacing and column spacing are set to 100 mm to balance data density and cost. The installation positions of the temperature sensors cover the surface of the target material and the key areas of the freezing cavity. For example, sensors are arranged in the center and edge areas of the material surface. Each temperature sensor collects temperature data in real time at a sampling frequency of once per second, and the data of all temperature sensors are aggregated to generate temperature distribution data. The temperature distribution data is stored in the form of a two-dimensional matrix, and the row and column coordinates of the matrix correspond one-to-one to the physical layout of the temperature sensors. For example, the first row and the first column correspond to the temperature value of the first sensor in the upper left corner.

[0062] The ultrasonic flow rate sensors are used to synchronously collect gas flow rate data at each monitoring point in the freezing area to generate a gas flow rate distribution. The number of ultrasonic flow rate sensors is consistent with the number of temperature sensors, and the installation positions correspond one-to-one to the monitoring point positions of the temperature sensors. For example, an ultrasonic flow rate sensor is installed directly above each temperature sensor. The ultrasonic flow rate sensors synchronously collect gas flow rate data at the same sampling frequency as the temperature sensors, for example, the sampling frequency is also 1 time per second, and the data from all flow rate sensors are aggregated to generate a gas flow rate distribution. The gas flow rate distribution is stored in the form of a two-dimensional matrix, and the row and column coordinates of the matrix correspond one-to-one to the physical layout of the ultrasonic flow rate sensors. For example, the first row and the first column correspond to the gas flow rate value of the first flow rate sensor in the upper left corner.

[0063] The temperature distribution data and the gas flow rate distribution are synchronized and time-scaled. A unified clock source is used to synchronize the data acquisition process of the temperature sensor array and the ultrasonic flow rate sensor. For example, time synchronization is achieved through a GPS module or an internal crystal oscillator clock to ensure that the timestamp error between each temperature data point and the corresponding gas flow rate data point is less than 10 milliseconds. For time deviations caused by sensor response delays, the temperature data and gas flow rate data are aligned on the time axis through an interpolation algorithm. The interpolation algorithm uses a linear interpolation method, specifically: if the temperature data at a certain moment is missing, the arithmetic mean of the temperature data at two adjacent moments is taken as the temperature value at that moment; if the gas flow rate data at a certain moment is missing, the arithmetic mean of the gas flow rate data at two adjacent moments is taken as the gas flow rate value at that moment.

[0064] Signal denoising is performed on the temperature data based on a sliding average filtering algorithm. The window size of the sliding average filtering algorithm is set to 5 sampling points, and the sliding window moves forward by 1 sampling point each time. For each temperature data point, the arithmetic mean of the 5 temperature data in the current sliding window is taken to replace the original temperature value. The arithmetic mean is calculated by adding all the temperature values in the window and dividing by 5. The sliding window size is set based on the typical cycle of temperature fluctuations during the freezing process. For example, the optimal value is selected after testing the filtering ability of different window sizes for high-frequency noise in preliminary experiments. A window with 5 sampling points can effectively suppress high-frequency noise without masking the true temperature change trend.

[0065] Outlier filtering is performed on the gas flow rate data based on the median filtering algorithm; the window size of the median filtering algorithm is set to 3 sampling points, and the window moves forward 1 sampling point each time; for each gas flow rate data point, the median value of the 3 gas flow rate data in the current window is taken to replace the original gas flow rate value, and the median value is selected by sorting the 3 gas flow rate data in the window by size and then taking the middle value; if the gas flow rate data is detected to deviate from the flow rate value of the adjacent monitoring point by more than 30%, for example, the flow rate value of a certain monitoring point is 1.5 m / s and the flow rate value of the adjacent point is 1.0 m / s, it is determined to be an outlier, and the outlier is replaced by the median value in the window; the setting basis of the median filtering window size is the typical duration of the gas flow rate mutation, for example, in the preliminary experiment, a window of 3 sampling points is selected after testing the ability of different window sizes to suppress sudden interference.

[0066] S2. Perform dynamic confidence scoring on the temperature distribution data based on the spatial correlation between the gas velocity distribution and the temperature distribution data, and select a high-confidence data subset in the low velocity area. Specifically:

[0067] The area in the gas flow rate distribution where the gas flow rate is lower than the preset flow rate threshold is marked as a low flow rate area, otherwise it is marked as a high flow rate area, and the temperature distribution data corresponding to the low flow rate area is extracted; the preset flow rate threshold is set according to the air supply efficiency of the refrigeration equipment and the characteristics of the target material. For example, in the pepper freezing scenario, the preset flow rate threshold is set to 0.5 m / s. When the gas flow rate is lower than 0.5 m / s, it is marked as a low flow rate area; the temperature distribution data of the low flow rate area is extracted in the following way: the temperature data points corresponding to the coordinates of the low flow rate area are filtered out from the temperature distribution data matrix to generate a subset of the temperature data of the low flow rate area; the coordinate correspondence method is a one-to-one mapping of the matrix row and column indices and the physical position of the sensor.

[0068] The temperature data stability coefficient of the low-flow-velocity area is calculated based on the trend consistency of the temperature change rate of adjacent freezing areas within the time window. The duration of the time window is set according to the dynamic characteristics of the freezing process. For example, the time window is set to 5 seconds to capture short-term trends in temperature changes. The adjacent freezing area is defined as a set of monitoring points that are physically adjacent to the low-flow-velocity area and marked as a high-flow-velocity area. The temperature data stability coefficient is calculated by statistically analyzing the standard deviation of all temperature data in the low-flow-velocity area within the time window. The smaller the standard deviation, the higher the temperature data stability. For example, the temperature data of the low-flow-velocity area within the 5-second time window is [-18.5°C, -18.6°C, -18.4°C, -18.5°C, -18.5°C], and the standard deviation is calculated to be 0.07°C, indicating a high stability coefficient. The standard deviation is calculated by taking the data mean, calculating the average of the squared differences between each data point and the mean, and then square rooting it.

[0069] The temperature change rate of adjacent high-flow rate areas within the same time window is extracted. The temperature change rate is calculated by taking the difference between the initial and final temperatures of adjacent high-flow rate areas within the time window and dividing it by the time window duration to obtain the temperature change per unit time. The absolute value of the temperature change rate reflects the cooling rate and is used to quantify the heat exchange intensity of adjacent areas.

[0070] A dynamic confidence score for the temperature distribution data is generated based on the temperature data stability coefficient of the low-flow region and the absolute value of the temperature change rate of the adjacent high-flow region. The dynamic confidence score is calculated as follows: Dynamic Confidence Score = Temperature Data Stability Coefficient / Absolute Value of the Temperature Change Rate of the Adjacent High-Flow Region. For example, if the temperature data stability coefficient of the low-flow region is 0.07°C and the temperature change rate of the adjacent high-flow region is -0.2°C / second, the dynamic confidence score is 0.07 / 0.2 = 0.35. This ratio represents the reliability of the temperature data in the low-flow region. When the cooling rate of the adjacent region is higher (increased denominator) or the temperature fluctuation in the low-flow region is larger (increased numerator), the score decreases, and the data credibility decreases. The dynamic confidence threshold is set based on historical data statistics. For example, by analyzing 100 sets of freezing data, a threshold of 0.5 is set. A data point with a dynamic confidence score above 0.5 is considered high-confidence.

[0071] If the dynamic confidence score is higher than the dynamic confidence threshold, the corresponding temperature data point is determined to be a high-confidence data point; for example, when the dynamic confidence score is 0.6, which is higher than the threshold of 0.5, the temperature data point is marked as a high-confidence data point; the spatial coordinates of all high-confidence data points are extracted from the low-flow area temperature data subset to form an initial high-confidence data set.

[0072] All high-confidence data points are density-clusters based on their spatial coordinates. The density clustering rule is as follows: the Euclidean distance between each high-confidence data point and its nearest neighbor is calculated. If the distance exceeds the preset cluster radius, the point is considered an outlier. The preset cluster radius is set based on the spatial scale of the freezing chamber and the sensor layout density. For example, in the pepper freezing scenario, the preset cluster radius is set to 0.3 meters. The Euclidean distance is calculated by taking the square root of the sum of the squares of the coordinate differences between two data points. For example, if the coordinates of a high-confidence data point are (1.0, 2.0) and its nearest neighbor are (1.2, 2.1), the Euclidean distance is approximately 0.22 meters, which is less than 0.3 meters, so the data point is retained. If the coordinates of a data point are (3.0, 4.0) and its nearest neighbor are (3.5, 4.6), the Euclidean distance is approximately 0.78 meters, which is greater than 0.3 meters, so the data point is removed. After removing the outliers, the final high-confidence data subset is generated.

[0073] S3. Based on the high-confidence data subset, the real-time thermal conductivity coefficient of the target material is identified online through a recursive algorithm to generate a thermal conductivity coefficient sequence. Specifically:

[0074] Based on the high-confidence data subset, the real-time thermal conductivity coefficient of the target material is identified online through the recursive least squares method; the forgetting factor of the recursive least squares method is adjusted according to the dynamic confidence score of the high-confidence data subset. The lower the dynamic confidence score, the larger the forgetting factor value; the dynamic confidence score ranges from 0 to 1, and the adjustment rule of the forgetting factor is: forgetting factor = 1-dynamic confidence score; for example, when the dynamic confidence score is 0.3, the forgetting factor is set to 0.7, indicating that the current data credibility is low, and the algorithm accelerates the forgetting of historical data to suppress the influence of noise; when the dynamic confidence score is 0.8, the forgetting factor is set to 0.2, indicating that the current data credibility is high, and the algorithm retains more historical information to improve the stability of parameter identification; the specific value of the forgetting factor is calibrated through preliminary experiments. For example, in the pepper freezing scenario, by comparing the thermal conductivity coefficient identification error under different forgetting factors, the mapping relationship with the smallest error is selected.

[0075] The thermal conductivity coefficients identified at each moment are arranged in chronological order to generate a thermal conductivity coefficient sequence. During the generation of the thermal conductivity coefficient sequence, the rate of change of the thermal conductivity coefficients at adjacent moments is smoothed. The rate of change of the thermal conductivity coefficient is calculated by taking the difference between the thermal conductivity coefficient at the current moment and the thermal conductivity coefficient at the previous time period and dividing it by the time interval. For example, if the thermal conductivity coefficient at the current moment is 1.2 W / (m·K), the thermal conductivity coefficient at the previous time period is 1.0 W / (m·K), and the time interval is 1 second, the rate of change is 0.2 W / (m·K) / second. The preset change threshold is set based on the normal fluctuation range of the thermal conductivity coefficient in historical data. For example, during the freezing process of peppers, the distribution of the rate of change of the thermal conductivity coefficient of 100 sets of experimental data is statistically analyzed, and the 95% quantile of 0.15 W / (m·K) / second is used as the threshold. When the rate of change exceeds this value, it is determined to be an abnormal jump.

[0076] If the rate of change of the thermal conductivity coefficient exceeds a preset change threshold, the thermal conductivity coefficient at the current moment is corrected using a sliding average algorithm. The window size of the sliding average algorithm is set according to the typical period of thermal conductivity change, for example, the window size is set to five time points. The sliding average algorithm is calculated as follows: the thermal conductivity coefficients at the current moment and the four previous moments are taken and the arithmetic mean is calculated as the corrected current thermal conductivity coefficient. For example, if the thermal conductivity coefficients at the current moment and the four previous moments are [1.2, 1.1, 1.15, 1.05, 1.0] W / (m·K), the corrected value is (1.2+1.1+1.15+1.05+1.0) / 5= 1.1 W / (m·K). The sliding average window size is set based on the smoothing effect verified by preliminary experiments. For example, in the pepper freezing scenario, the noise suppression ability of different window sizes was tested, and a window of five time points was ultimately selected to achieve the best balance.

[0077] S4. Calculate the energy release rate and the phase lag angle of the adjacent region based on the thermal conductivity coefficient sequence. If the energy release rate exceeds the critical threshold of the phase change and the phase lag angle exceeds the dynamic tolerance, a stable parameter sequence is generated by piecewise fitting. Specifically:

[0078] The energy release rate is calculated based on the thermal conductivity sequence. The energy release rate is the product of the rate of change of the thermal conductivity and the temperature gradient. The rate of change of the thermal conductivity is calculated by taking the difference between the thermal conductivity at the current moment and the thermal conductivity at the previous moment, divided by the time interval. The temperature gradient is calculated by taking the difference between the temperature distribution data of adjacent frozen areas and dividing it by the physical distance. For example, if the temperature of adjacent area A is -20°C and the temperature of area B is -21°C, and the physical distance is 0.1 meter, the temperature gradient is (-21 - (-20)) / 0.1 = -10°C / meter. The energy release rate is calculated as: rate of change of thermal conductivity × temperature gradient. For example, if the rate of change of thermal conductivity is 0.2 W / (m·K) / second and the temperature gradient is -10°C / meter, the energy release rate is 0.2 × (-10) = -2 W / (m·K·second).

[0079] Based on the heat conductivity coefficient sequence, cross-correlation analysis was performed on the heat conductivity coefficient time series of adjacent freezing zones. The time delay corresponding to the maximum cross-correlation coefficient was extracted and converted into a phase lag angle. The specific steps of the cross-correlation analysis are: the heat conductivity coefficient sequences of two adjacent zones are aligned according to the time window, the cross-correlation coefficient is calculated by sliding, and the time offset with the maximum cross-correlation coefficient is found. For example, the heat conductivity coefficient sequence of zone A is [1.0, 1.1, 1.2, 1.15, 1.1], and the sequence of zone B is [1.0, 1.05, 1.1, 1.15, 1.2]. After sliding calculation, the time delay corresponding to the maximum cross-correlation coefficient is 2 seconds. The phase lag angle is converted by dividing the time delay by the total time window length of the current freezing stage and multiplying it by 360 degrees. For example, if the time delay is 2 seconds and the total time window is 10 seconds, the phase lag angle is (2 / 10) × 360 = 72 degrees.

[0080] If the energy release rate exceeds the critical phase change threshold and the phase lag angle exceeds the dynamic tolerance, a mutation point is marked in the thermal conductivity coefficient sequence, and the thermal conductivity coefficient sequence is divided into multiple subsegments based on the mutation point. The critical phase change threshold is set according to the phase change latent heat characteristics of the target material. For example, the critical energy release rate for ice crystal formation during pepper freezing, measured by differential scanning calorimetry (DSC), is 1.5 W / (m·K·s). When the energy release rate exceeds this value, a phase change mutation is determined. The dynamic tolerance is set based on the normal fluctuation range of the phase lag angle in historical data. For example, after statistically analyzing 100 sets of pepper freezing data, the frequency distribution of the phase lag angle is calculated, and the dynamic tolerance is 60 degrees corresponding to the 90th percentile. The mutation point marking rule is: when both the energy release rate exceeds the threshold and the phase lag angle exceeds the tolerance, for example, if the energy release rate is 2.0 W / (m·K·s) and the phase lag angle is 75 degrees at t = 5 seconds, t = 5 seconds is marked as the mutation point.

[0081] Least squares fitting is performed on the thermal conductivity coefficient within each subsegment to generate a stable parameter sequence. The order of the least squares fitting is dynamically adjusted according to the subsegment length. The longer the subsegment, the higher the fitting order. For example, through preliminary experiments to test the fitting errors under different subsegment lengths, the following mapping rules are determined: first-order linear fitting is used when the subsegment length is less than 10 data points, second-order polynomial fitting is used when the subsegment length is 10 to 20 data points, and third-order polynomial fitting is used when the subsegment length exceeds 20 data points. The fitting parameters include the baseline value and change trend of the thermal conductivity coefficient. For example, a second-order fitting is performed on the subsegment [1.1, 1.2, 1.15, 1.18, 1.2], and the stable parameter sequence is [1.12, 0.05, -0.002], which represent the baseline value, linear growth rate, and quadratic term coefficient, respectively.

[0082] S5. Perform nonlinear correction on the preset temperature gradient model according to the stable parameter sequence to generate an updated gradient control curve. Specifically:

[0083] Based on the baseline value and change trend of the heat conductivity coefficient in the stable parameter sequence, the target temperature value of each freezing area of the preset temperature gradient model is corrected by nonlinear mapping; the baseline value of the heat conductivity coefficient in the stable parameter sequence is the mean value of the steady-state heat conductivity coefficient after segmented fitting, and the change trend is the first-order derivative of the fitting polynomial; the correction weight of the nonlinear mapping correction is dynamically adjusted according to the variance of the stable parameter. The smaller the variance of the stable parameter, the higher the correction weight; for example, if the variance of the stable parameter of a freezing area is 0.1, the correction weight is set to 0.9; if the variance is 0.5, the correction weight is set to 0.5; the adjustment rule of the correction weight is: correction weight = 1 - variance / maximum variance threshold, where the maximum variance threshold is set according to the variance distribution of the stable parameter in the historical data. For example, after counting 100 sets of pepper freezing data, the 95% quantile of the variance distribution is 1.0 as the maximum variance threshold; the calculation method of the corrected target temperature value is: original target temperature value + correction weight × heat conductivity change trend × Temperature compensation coefficient. The temperature compensation coefficient is set based on the product of the material's specific heat and density. For example, if the specific heat of chili peppers is 3.8 kJ / (kg·K) and their density is 500 kg / m³, the temperature compensation coefficient is 3.8 × 500 = 1900 J / (m³·K). The temperature compensation coefficient is set based on calibration through material thermodynamic experiments. For example, the temperature compensation coefficient is obtained by measuring the energy required to raise the temperature of chili peppers by 1°C per unit volume in a sealed, insulated chamber.

[0084] The corrected target temperature values for each freezing zone are piecewise interpolated according to their spatial distribution to generate an updated gradient control curve. The order of the piecewise interpolation is adaptively adjusted based on the temperature difference between adjacent zones. The greater the temperature difference between adjacent zones, the higher the interpolation order. The temperature difference is calculated by taking the absolute difference between the corrected target temperatures of two adjacent freezing zones. For example, if the target temperature of zone A is -20°C and the target temperature of zone B is -22°C, the temperature difference is 2°C. The interpolation order is adjusted as follows: linear interpolation is used when the temperature difference is less than 1°C, quadratic polynomial interpolation is used when the difference is between 1°C and 3°C, and cubic spline interpolation is used when the difference is greater than 3°C. For example, if the temperature difference between zones A and B is 2°C, quadratic polynomial interpolation is used to calculate the temperature value of the intermediate zone. The interpolation curve ensures continuity of the temperature and first-order derivative of zones A and B. The boundary condition of the cubic spline interpolation is set to natural spline, that is, the second derivative is zero at the endpoints to ensure a smooth interpolation curve.

[0085] The time axis of the gradient control curve is smoothed and filtered to eliminate sudden temperature setpoint jumps caused by parameter correction, generating the final updated gradient control curve. The smoothing filter uses a sliding average algorithm, with the sliding window size set based on the frequency of temperature setpoint adjustment. For example, during the pepper freezing process, the temperature setpoint is updated every 5 seconds, and the sliding window size is set to three time points. The sliding average algorithm calculates the arithmetic mean of the temperature setpoints at the current moment and the two previous moments, providing the smoothed temperature setpoint value. For example, if the temperature setpoints at the current moment and the two previous moments are [-20°C, -20.5°C, -21°C], the smoothed value is (-20 + (-20.5) + (-21)) / 3 = -20.5°C. The sliding window size is determined based on noise suppression verified in preliminary experiments. For example, a temperature step signal (e.g., from -20°C to -25°C) was artificially injected into the pepper freezing equipment to test the ability of different window sizes to suppress sudden changes. A window size of three time points was ultimately selected to balance smoothness and response speed.

[0086] It is worth noting that the method for constructing the preset temperature gradient model is: based on the freezing process requirements of the target material and historical freezing data statistics, an initial temperature gradient curve is generated, and the curve parameters are optimized through experimental calibration; the specific process includes the following steps:

[0087] Under the typical freezing process conditions of the target material, multiple sets of temperature distribution data of the normal freezing process are collected; for example, in the pepper freezing scenario, 100 sets of temperature distribution data at each time point in the freezing process are collected, and after eliminating abnormal data, they are aligned on the time axis and the average temperature value of each freezing area is calculated; based on the average temperature value, an initial temperature gradient curve is generated, and the initial temperature gradient curve represents a theoretical curve of the target temperature of each freezing area under the standard freezing process changing with time; the parameters of the initial temperature gradient curve are optimized through experimental calibration, specifically: the initial temperature gradient curve is executed in the freezing equipment, and the material freezing uniformity index is monitored in real time. If overcooling or undercooling occurs in a local area, the target temperature value of the corresponding area is adjusted; for example, when the temperature in the center area of the pepper stack lags behind the curve set value, the target temperature value of the area is reduced by 1°C based on the original curve to accelerate cooling; the optimized temperature gradient curve is stored as a preset temperature gradient model for reference in subsequent control processes.

[0088] S6. Generate temperature setpoint deviations for each freezing zone based on the gradient control curve, and generate drive instructions to adjust the heat exchange rate so that the actual temperature gradient tracks the gradient control curve. Specifically:

[0089] The temperature set point deviation of each freezing zone is calculated based on the gradient control curve. The temperature set point deviation is the difference between the current actual temperature and the target temperature at the corresponding time in the gradient control curve. The target temperature in the gradient control curve is obtained by extracting the corresponding timestamp and the value of the freezing zone coordinates from the stored gradient control curve matrix. For example, if the current time is 300 seconds after the start of freezing and the freezing zone coordinates are (2,3), the target temperature value corresponding to the coordinates at 300 seconds is extracted from the gradient control curve matrix. The actual temperature is collected in real time through the temperature sensor array, and the layout of the temperature sensor array corresponds one-to-one with the coordinates of the gradient control curve matrix. The positive and negative signs of the temperature set point deviation are defined as follows: when the actual temperature is higher than the target temperature, the deviation is positive, and the heat exchange rate needs to be reduced; when the actual temperature is lower than the target temperature, the deviation is negative, and the heat exchange rate needs to be increased.

[0090] When generating drive commands to adjust the heat exchange rate based on the temperature setpoint deviation, a proportional-integral controller is used to calculate the heat exchange rate adjustment for each freezing zone. The proportional gain and integral time are dynamically adjusted based on the absolute value of the temperature setpoint deviation. The larger the absolute value of the temperature setpoint deviation, the smaller the proportional gain and the shorter the integral time. The proportional gain adjustment rule is: Proportional gain = Basic proportional gain / (1 + Absolute value of deviation × Attenuation coefficient). The basic proportional gain is calibrated through step response experiments, and the attenuation coefficient is set based on historical control stability data. For example, if the basic proportional gain is 1.0 and the attenuation coefficient is 0.2, when the absolute value of the deviation is 1°C, the proportional gain = 1.0 / (1 + 1 × 0.2) ≈ 0.83. The integral time adjustment rule is: Integral time = Basic integral time × (1 + Absolute value of deviation × Acceleration coefficient). The basic integral time is determined using the critical proportional method, and the acceleration coefficient is set based on the system response speed requirements. For example, if the basic integral time is 10 seconds and the acceleration coefficient is 0.5, when the absolute value of the deviation is 2°C, the integral time = 10 × (1 + 2×0.5) = 20 seconds.

[0091] Dynamic damping compensation is performed on the heat exchange rate adjustment based on the temperature change rate difference between adjacent freezing zones. The greater the temperature change rate difference, the higher the damping compensation coefficient. The temperature change rate difference is calculated by taking the absolute value difference of the temperature change rate of two adjacent freezing zones. The temperature change rate is calculated using a sliding window difference with a sliding window size of three time points. For example, the temperature sequence of zone A is [-20.0°C, -20.5°C, -21.0°C], and the temperature change rate is (-21.0 - (-20.0)) / 2 = -0.5°C / second; the temperature sequence of zone B is [-19.5°C, -20.0°C, -20.3°C], and the temperature change rate is (-20.3 - (-19.5)) / 2 = -0.4°C / second. The difference is |-0.5 - (-0.4)| = 0.1°C / second. The segmented setting rule for the damping compensation coefficient is: when the difference value ≤ When the difference threshold is preset, the damping compensation coefficient = difference value × linear compensation coefficient. When the difference value is greater than the preset difference threshold, the damping compensation coefficient = difference value² × nonlinear compensation coefficient. The linear compensation coefficient and nonlinear compensation coefficient are calibrated through control stability experiments. For example, disturbance signals with different difference values are injected into the chili freezing equipment, and the compensation coefficient combination that makes the overshoot less than 5% is selected.

[0092] The compensated heat exchange rate adjustment is converted into a compressor frequency control command and a damper opening command, ensuring that the actual temperature gradient tracks the gradient control curve. The conversion logic for the compressor frequency control command is as follows: the mapping between the heat exchange rate adjustment and the compressor frequency is determined by the equipment performance curve. The equipment performance curve is calibrated using the following method: under steady-state operating conditions, the heat exchange rates corresponding to different compressor frequencies are recorded and fitted with a frequency-rate linear equation. For example, if each 1 Hz increase in frequency increases the heat exchange rate by 50 W / (m²·K), then an adjustment of 10 W / (m²·K) corresponds to a frequency increase of 0.2 Hz. The conversion logic for the damper opening command is a piecewise linear relationship between the heat exchange rate adjustment and the damper opening, with the segmented threshold set according to the damper characteristic curve. For example, from 0% to 50% opening, each 10% opening corresponds to a rate increase of 20 W / (m²·K); from 50% to 100% opening, each 10% opening corresponds to a rate increase of 10 W / (m²·K), to avoid efficiency saturation in the high-opening range.

[0093] In this embodiment, the target material is chili pepper.

[0094] It is worth noting that the present invention synchronously collects the temperature distribution and fluid dynamics data of the freezing process through a matrix temperature sensor array and a gas flow rate sensor, constructs a dynamic confidence scoring mechanism to screen high-reliability data, and updates key physical parameters such as the thermal conductivity coefficient online based on the recursive identification algorithm; further, through the dual-threshold judgment mechanism of energy release rate and phase lag angle, the thermal conductivity coefficient sequence is dynamically segmented and a stable parameter set is generated to drive the nonlinear correction of the temperature gradient model and the update of the gradient control curve; finally, the proportional-integral control and difference compensation logic are combined to generate a heat exchange rate adjustment instruction to achieve closed-loop tracking control of the temperature gradient; it is through deep coupling of dynamic physical parameter identification, multi-source data fusion and adaptive control algorithm in the freezing process, and improving the temperature gradient tracking accuracy through real-time feedback and parameter iterative optimization, focusing on signal processing, model correction and actuator drive logic optimization of industrial control systems, and belonging to the field of automatic control technology.

[0095] Example 2: Figure 2 A schematic structural diagram of a pepper freezing temperature gradient dynamic adjustment system of the present invention is provided. The pepper freezing temperature gradient dynamic adjustment system comprises:

[0096] Temperature and flow synchronous acquisition module: real-time acquisition of temperature distribution data of target materials during freezing and gas flow rate distribution in the freezing area;

[0097] Dynamic confidence screening module: Dynamic confidence scoring of temperature distribution data is performed based on the spatial correlation between gas velocity distribution and temperature distribution data, and a high-confidence data subset in the low velocity area is screened;

[0098] Thermal conductivity recursive identification module: Based on a high-confidence data subset, the module uses a recursive algorithm to online identify the real-time thermal conductivity coefficient of the target material and generate a thermal conductivity coefficient sequence;

[0099] Parameter segmented fitting module: Calculates the energy release rate and the phase lag angle of adjacent regions based on the thermal conductivity coefficient sequence. If the energy release rate exceeds the critical threshold of phase change and the phase lag angle exceeds the dynamic tolerance, segmented fitting is performed to generate a stable parameter sequence.

[0100] Gradient curve generation module: performs nonlinear correction on the preset temperature gradient model according to the stable parameter sequence to generate an updated gradient control curve;

[0101] Thermal control drive generation module: Generates the temperature set value deviation of each freezing area based on the gradient control curve, generates drive instructions to adjust the heat exchange rate, and makes the actual temperature gradient track the gradient control curve.

[0102] The above formulas are all dimensionless and numerical calculations. The formula is a formula that is closest to the actual situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0103] It should be noted that the present invention can be deployed on the device itself to implement embedded applications, and can also be run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0104] The above embodiments can be implemented in whole or in part via software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in the embodiments of this application are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired means (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0105] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0106] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0107] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of this embodiment according to actual needs.

[0108] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0109] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0110] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

[0111] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for dynamically adjusting the temperature gradient of pepper freezing, characterized in that: The steps include: S1. Real-time collection of temperature distribution data of target materials during freezing and gas flow rate distribution in the freezing area; S2. Perform dynamic confidence scoring on the temperature distribution data based on the spatial correlation between the gas velocity distribution and the temperature distribution data, and select a high-confidence data subset in the low velocity area; S3. Based on the high-confidence data subset, the real-time thermal conductivity coefficient of the target material is identified online through a recursive algorithm to generate a thermal conductivity coefficient sequence, including: Based on the high-confidence data subset, the real-time thermal conductivity coefficient of the target material is identified online through the recursive least squares method. The forgetting factor of the recursive least squares method is adjusted according to the dynamic confidence score of the high-confidence data subset. The lower the dynamic confidence score, the larger the forgetting factor value. The thermal conductivity coefficients identified at each moment are arranged in chronological order to generate a thermal conductivity coefficient sequence. During the generation of the thermal conductivity coefficient sequence, the change rate of the thermal conductivity coefficients at adjacent moments is smoothed. If the change rate of the thermal conductivity coefficient exceeds a preset change threshold, the thermal conductivity coefficient at the current moment is corrected using a sliding average algorithm. S4. Calculating the energy release rate and the phase lag angle of adjacent regions based on the heat conductivity coefficient sequence, including: The energy release rate is calculated based on the thermal conductivity coefficient sequence, which is the product of the thermal conductivity coefficient change rate and the temperature gradient. The temperature gradient is calculated by differentially calculating the temperature distribution data of adjacent frozen areas. Based on the thermal conductivity coefficient sequence, a cross-correlation analysis is performed on the thermal conductivity time series of adjacent frozen areas. The time delay corresponding to the maximum cross-correlation coefficient is extracted and converted into a phase lag angle. If the energy release rate exceeds the critical threshold of phase transition and the phase lag angle exceeds the dynamic tolerance, a stable parameter sequence is generated by piecewise fitting; S5. Performing nonlinear correction on the preset temperature gradient model according to the stable parameter sequence to generate an updated gradient control curve; S6. Generate a temperature set value deviation for each freezing zone based on the gradient control curve, and generate a drive instruction to adjust the heat exchange rate so that the actual temperature gradient tracks the gradient control curve.

2. The method for dynamically adjusting the temperature gradient of pepper freezing according to claim 1, wherein: Real-time collection of temperature distribution data of target materials during freezing and gas velocity distribution in the freezing area, including: The temperature data of each monitoring point during the freezing process of the target material is collected in real time through a matrix temperature sensor array to generate temperature distribution data; The ultrasonic flow velocity sensor is used to collect the gas flow velocity data of each monitoring point in the freezing area and generate the gas flow velocity distribution; The temperature distribution data and gas flow rate distribution are synchronized and aligned, and the temperature data is subjected to signal noise reduction based on the sliding average filtering algorithm, and the gas flow rate data is subjected to outlier filtering based on the median filtering algorithm.

3. The method for dynamically adjusting the temperature gradient of pepper freezing according to claim 1, wherein: Based on the spatial correlation between the gas velocity distribution and the temperature distribution data, a dynamic confidence score is performed on the temperature distribution data to filter out a high-confidence data subset in the low velocity area, including: Marking the area in the gas flow rate distribution where the gas flow rate is lower than a preset flow rate threshold as a low flow rate area, otherwise marking it as a high flow rate area; and extracting the temperature distribution data corresponding to the low flow rate area; Based on the trend consistency of the temperature change rate of adjacent freezing areas within the time window, the temperature data stability coefficient of the low flow rate area is calculated. The temperature data stability coefficient is the standard deviation of the temperature data of the corresponding freezing area. Generate a dynamic confidence score for the temperature distribution data based on the ratio of the temperature data stability coefficient of the low flow rate area to the temperature change rate of the adjacent high flow rate area; If the dynamic confidence score is higher than the dynamic confidence threshold, the corresponding temperature data point is determined to be a high-confidence data point; All high-confidence data points are density clustered according to spatial coordinates. If the distance between a high-confidence data point and its nearest neighbor exceeds the preset clustering radius, it is removed to generate a high-confidence data subset.

4. The method for dynamically adjusting the temperature gradient of pepper freezing according to claim 1, wherein: If the energy release rate exceeds the critical threshold of the phase transition and the phase lag angle exceeds the dynamic tolerance, the piecewise fitting generates a stable parameter sequence, including: If the energy release rate exceeds the critical threshold of phase change and the phase lag angle exceeds the dynamic tolerance, a mutation point is marked in the thermal conductivity coefficient sequence, and the thermal conductivity coefficient sequence is divided into multiple sub-segments based on the mutation point; The least squares fitting is performed on the heat conduction coefficient in each sub-segment to generate a stable parameter sequence; the order of the least squares fitting is dynamically adjusted according to the length of the sub-segment. The longer the sub-segment, the higher the fitting order.

5. The method for dynamically adjusting the temperature gradient of pepper freezing according to claim 1, wherein: The preset temperature gradient model is nonlinearly modified according to the stable parameter sequence to generate an updated gradient control curve, including: Based on the baseline value and change trend of the heat transfer coefficient in the stable parameter sequence, the target temperature value of each freezing area of the preset temperature gradient model is corrected by nonlinear mapping; The corrected target temperature values of each freezing zone are interpolated piecewise according to the spatial distribution to generate an updated gradient control curve; The time axis of the gradient control curve is smoothed and filtered to eliminate the sudden step change of the temperature setting value caused by the parameter correction, and the final updated gradient control curve is generated.

6. The method for dynamically adjusting the temperature gradient of pepper freezing according to claim 5, characterized in that: The order of the segmented interpolation is adaptively adjusted according to the temperature difference between adjacent areas. The greater the temperature difference between adjacent areas, the higher the interpolation order.

7. The method for dynamically adjusting the temperature gradient of pepper freezing according to claim 1, wherein: Based on the gradient control curve, the temperature set value deviation of each freezing zone is generated, and the drive instruction is generated to adjust the heat exchange rate so that the actual temperature gradient tracks the gradient control curve, including: Calculate the temperature set value deviation of each freezing zone based on the gradient control curve. The temperature set value deviation is the difference between the current actual temperature and the target temperature at the corresponding moment in the gradient control curve. When generating drive instructions to adjust the heat exchange rate based on the temperature set value deviation, a proportional-integral controller is used to calculate the heat exchange rate adjustment amount for each freezing zone; Dynamic damping compensation is performed on the heat exchange rate adjustment based on the temperature change rate difference between adjacent freezing zones. The greater the temperature change rate difference, the higher the damping compensation coefficient. Nonlinear compensation is enabled when the absolute value difference of the temperature change rate between adjacent zones exceeds a preset difference threshold. The compensated heat exchange rate adjustment amount is converted into a compressor frequency adjustment instruction and an air valve opening instruction, so that the actual temperature gradient tracks the gradient control curve.

8. A system for dynamically adjusting the temperature gradient of pepper freezing, used to implement the method for dynamically adjusting the temperature gradient of pepper freezing according to any one of claims 1 to 7, characterized in that: include: Temperature and flow synchronous acquisition module: real-time acquisition of temperature distribution data of target materials during freezing and gas flow rate distribution in the freezing area; Dynamic confidence screening module: Dynamic confidence scoring of temperature distribution data is performed based on the spatial correlation between gas velocity distribution and temperature distribution data, and a high-confidence data subset in the low velocity area is screened; Thermal conductivity recursive identification module: Based on a high-confidence data subset, the module uses a recursive algorithm to online identify the real-time thermal conductivity coefficient of the target material and generate a thermal conductivity coefficient sequence; Parameter segmented fitting module: Calculates the energy release rate and the phase lag angle of adjacent regions based on the thermal conductivity coefficient sequence. If the energy release rate exceeds the critical threshold of phase change and the phase lag angle exceeds the dynamic tolerance, segmented fitting is performed to generate a stable parameter sequence. Gradient curve generation module: performs nonlinear correction on the preset temperature gradient model according to the stable parameter sequence to generate an updated gradient control curve; Thermal control drive generation module: Generates the temperature set value deviation of each freezing area based on the gradient control curve, generates drive instructions to adjust the heat exchange rate, and makes the actual temperature gradient track the gradient control curve.

Images