Preparation dish fresh-keeping monitoring system and method based on internet of things

Abstract

This invention relates to the field of Internet of Things (IoT) technology and discloses an IoT-based pre-prepared food preservation monitoring system and method. By uniquely encoding each batch of pre-prepared food and establishing a full-link IoT monitoring chain, the system collects environmental temperature, humidity, volatile organic compound (VOC) concentration, carbon dioxide concentration, ammonia concentration, and triaxial vibration acceleration at each node. Temperature and humidity degradation, vibration degradation, and gas degradation are constructed separately, and these three types of degradation are fused to obtain the preservation status value of each node. The preservation status value of each node is recursively accumulated according to a time decay method to form a full-link preservation status curve, identifying rising inflection points and abrupt change inflection points. Finally, the remaining shelf life is predicted based on the preservation status curve, and the preservation status of the pre-prepared food is determined according to the predicted remaining shelf life. This system can identify and comprehensively judge the cumulative degradation of pre-prepared food during multi-node circulation in real time, improving the consistency of quality, the authenticity of shelf life, and consumer safety throughout the entire process of pre-prepared food production.

Description

Technical Field

[0001] This invention relates to the field of Internet of Things (IoT) technology, specifically to an IoT-based pre-prepared vegetable preservation monitoring system. Background Technology

[0002] Pre-prepared dishes typically refer to pre-packaged food products made from one or more edible agricultural products and their derivatives, which undergo pre-processing, seasoning, cooking or semi-cooking, and repackaging. After production, they need to go through multiple circulation stages such as refrigeration, freezing, transportation, sales, and reheating before they can be consumed by consumers. Because pre-prepared dishes generally have high moisture content, sensitive texture, complex flavor composition, and clearly defined shelf-life requirements, their quality stability is highly dependent on factors such as storage temperature, environmental humidity, packaging airtightness, transportation time, and exposure time after opening.

[0003] In existing technologies, the preservation and management of pre-prepared meals largely relies on methods such as cold chain temperature control, manual inspection, and regular sampling to control the storage and transportation process. However, it is difficult to identify and comprehensively judge the quality changes caused by multiple factors such as temperature fluctuations, humidity changes, packaging damage, repeated opening, transportation vibration, and microbial deterioration in real time. Especially in the actual circulation process, pre-prepared meals often undergo transfer at multiple nodes, by multiple entities, and in multiple scenarios, which can easily lead to cold chain interruptions and environmental loss of control. Traditional methods can usually only reflect the environmental state at a local point in time and cannot accurately characterize the cumulative degree of deterioration of pre-prepared meals throughout the entire circulation process. It is also difficult to judge in time whether they are close to the critical state of freshness loss or spoilage, and thus cannot guarantee the consistency of quality, the authenticity of shelf life, and the safety of consumption of pre-prepared meals throughout the entire process. Summary of the Invention

[0004] In response to the problems in related technologies, the present invention provides a pre-prepared vegetable preservation monitoring system based on the Internet of Things to overcome the aforementioned technical problems in existing related technologies.

[0005] To solve the aforementioned technical problem, the present invention is achieved through the following technical solution:

[0006] On the one hand, this invention provides an IoT-based pre-prepared food preservation monitoring system, specifically including:

[0007] The Internet of Things (IoT) data acquisition module is used to deploy IoT data acquisition devices at multiple nodes, including pre-prepared food production and warehousing, cold storage, cold chain transportation, and terminal shelf placement, to obtain information on the circulation status of the same batch of pre-prepared food during the storage period at each node.

[0008] The degradation accumulation module quantifies the temperature and humidity disturbances, mechanical vibration disturbances, and gas metabolism disturbances of pre-cooked food at each node based on the flow status information, and obtains the corresponding temperature and humidity degradation degree, vibration degradation degree, and gas degradation degree.

[0009] The full-node trajectory module performs normalized weighted fusion of the cumulative deterioration of temperature and humidity, vibration, and gas at each node to obtain the freshness status value. The freshness status value of each node is then attenuated in chronological order and recursively accumulated to obtain the cumulative freshness status value. Based on the nodes and their corresponding cumulative freshness status values, the full-link freshness change trajectory of the pre-prepared food is constructed to obtain the freshness status curve. Then, based on the freshness status curve, anomalies are identified and risks are determined. If anomalies are found, the batch of pre-prepared food is in a high-risk state of freshness loss; otherwise, the prediction status module is executed.

[0010] The prediction status module predicts the remaining shelf life based on the freshness status curve and uses this to determine the status of the prepared food.

[0011] Preferred method: Quantification of temperature and humidity disturbances:

[0012] The ambient temperature and humidity are compared with the target storage temperature and humidity, respectively. The absolute values ​​are then normalized and weighted according to preset weights to obtain the storage deviation value. A storage deviation change graph is constructed with time as the horizontal axis and the storage deviation value as the vertical axis. Local peak points and local valley points are extracted as poles. The storage deviation difference and time interval between adjacent poles are calculated to obtain the local fluctuation coefficient. When the local fluctuation coefficient is greater than the effective deviation threshold, the effective fluctuation range is determined. The cumulative deterioration degree of temperature and humidity is calculated based on the effective fluctuation range.

[0013] Preferably, the local fluctuation coefficient is obtained by calculating the difference in preservation deviation between adjacent poles and the time interval:

[0014] In the saved deviation change graph, poles are selected, including local peak points and local valley points. The difference between the saved deviation values ​​between two adjacent poles is calculated to obtain the adjacent change value. The time difference between the two adjacent poles is calculated to obtain the adjacent interval. The adjacent change value is then divided by the adjacent interval to obtain the local fluctuation coefficient. A preset effective deviation threshold is established. All local fluctuation coefficients in the saved deviation change graph are traversed. If any local fluctuation coefficient is greater than the effective deviation threshold, the interval formed by the two adjacent poles corresponding to the local fluctuation coefficient is marked as the effective fluctuation interval, and temperature and humidity deterioration is accumulated for it. Otherwise, the temperature and humidity deterioration is assigned a preset benchmark value.

[0015] Preferably, the cumulative deterioration of temperature and humidity is calculated based on the effective fluctuation range:

[0016] For each effective fluctuation interval, the effective fluctuation intensity is obtained by subtracting the effective deviation threshold from the local fluctuation coefficient and then dividing by the effective deviation threshold. For each effective fluctuation interval, the area of ​​preserved deviation A(i) in the effective fluctuation interval is calculated using the trapezoidal integral method, as follows:

[0017]

[0018] Where i represents the index of the effective fluctuation interval, B(i) and B(i+1) represent the preservation deviation values ​​corresponding to the two extreme points of the interval, and b(i+) and b(i+1) represent the time corresponding to the two extreme points of the interval. The effective fluctuation intensity of the effective fluctuation interval is multiplied by the preservation deviation area to obtain the degradation contribution value of the effective fluctuation interval. The degradation contribution values ​​of all effective fluctuation intervals in the preservation deviation change graph are summed to obtain the temperature and humidity degradation degree.

[0019] Preferably, mechanical vibration disturbance quantification:

[0020] The triaxial vibration acceleration is synthesized into a vibration acceleration, and a vibration variation diagram is constructed with time as the abscissa and vibration acceleration as the ordinate. Local peak points and local valley points are extracted as poles. The vibration fluctuation coefficient is obtained by calculating the vibration acceleration difference and time interval between adjacent poles. When the vibration fluctuation coefficient is greater than the effective vibration threshold, the effective vibration interval is determined, and the vibration deterioration degree is calculated based on the effective vibration interval.

[0021] Preferred method: Quantification of gas metabolism disturbances

[0022] After normalizing the concentrations of volatile organic compounds, carbon dioxide, and ammonia, the gas deviation values ​​were obtained by weighted fusion according to preset weighting coefficients. A gas deviation change map was constructed with time as the x-axis and gas deviation values ​​as the y-axis. Based on the gas deviation values ​​at adjacent acquisition times, the cumulative gas area and gas growth rate were calculated. Continuous deviations were identified when the gas growth rate continuously exceeded a preset positive growth threshold. The continuous deviation contribution value was calculated based on all continuous deviations in the gas deviation change map, and a continuous deviation factor was obtained through a product-based saturation mapping. Finally, the cumulative gas degradation degree was calculated by combining the cumulative gas area and the continuous deviation factor. The formula for the product-based saturation mapping is as follows:

[0023]

[0024] Where φ j This represents the continuous deviation contribution value φ of any consecutive deviation in the gas deviation change graph. j , j represents the index of any consecutive deviation in the gas deviation change graph.

[0025] Preferably, the continuous deviation contribution value is calculated based on all continuous deviations in the gas deviation change graph:

[0026] The gas growth rate between two adjacent sampling times is extracted, reflecting the rate of increase in gas degradation value between adjacent sampling times. The gas growth rates at each adjacent time point are arranged chronologically. If more than n consecutive gas growth rates exceed a preset positive growth threshold, it is recorded as a continuous deviation, and the number of sampling points with consecutive positive growth within the continuous deviation is recorded as m. For each continuous deviation, the cumulative growth amount within that deviation segment is calculated and normalized, recorded as Q. For each continuous deviation, the continuous deviation contribution value is calculated using a formula based on the number of sampling points with consecutive positive growth and the cumulative growth amount within the continuous deviation. The specific formula is as follows:

[0027]

[0028] Where μ is the continuous deviation number adjustment parameter, used to characterize the amplification intensity of the continuous positive growth point number on the continuous deviation degree, and is set to 3 in this embodiment. η is the growth amount adjustment parameter, used to characterize the amplification intensity of the continuous deviation cumulative growth amount on the continuous deviation degree, and is set to 1.0.

[0029] Preferably, anomalies are identified and risks are assessed based on the preservation status curve:

[0030] Arranged chronologically, scatter points are plotted in a two-dimensional rectangular coordinate system with time as the abscissa and cumulative freshness status as the ordinate. Connecting the discrete points forms a continuous curve, resulting in a batch-level freshness status curve. During curve recognition, the first and second derivatives of the curve are calculated. If a node satisfies the condition that the curve slope is greater than a preset rising threshold and the second derivative is positive, it is identified as a rising inflection point. If the slope difference between adjacent intervals exceeds a preset abrupt change threshold and the local curvature exceeds a preset curvature threshold, it is identified as abrupt change inflection point. Rising inflection points and abrupt change inflection points are collectively referred to as abnormal points.

[0031] Preferably, the remaining shelf life is predicted based on the freshness status curve:

[0032] Extract the cumulative freshness status value at the current node and establish a shelf life determination benchmark based on the preset shelf life of this batch of pre-prepared vegetables; use a sliding window to locally fit the freshness status curve to obtain the trend function after the current node; based on the trend function and the correspondence between the cumulative freshness status value and the shelf life consumption progress in historical qualified samples, establish a mapping model from the cumulative freshness status value to the shelf life consumption ratio. The mapping model is used to convert the cumulative freshness status value at the current node into the corresponding shelf life consumption ratio; then calculate the expected time corresponding to the expected end of the preset shelf life based on the shelf life consumption ratio and the preset shelf life. The remaining shelf life is equal to the expected time minus the current time.

[0033] If the remaining shelf life is within a preset timeframe, the prepared food is considered to be in a shelf-life warning state. If the remaining shelf life is less than the lower limit of the timeframe, the prepared food is considered to be in a state of risk of spoilage. If the remaining shelf life is greater than the upper limit of the timeframe, the prepared food is considered to be in a normal preservation state.

[0034] On the other hand, the present invention provides a method for monitoring the freshness of pre-prepared vegetables based on the Internet of Things, specifically including the following steps:

[0035] Step 1 involves deploying IoT data collection nodes at multiple nodes during the production, cold storage, cold chain transportation, and terminal shelf placement of pre-prepared dishes to obtain information on the circulation status of the same batch of pre-prepared dishes during the storage period at each node.

[0036] Step 2: Based on the flow status information, the temperature and humidity disturbances, mechanical vibration disturbances, and gas metabolism disturbances of the pre-cooked food at each node are quantified to obtain the corresponding temperature and humidity deterioration degree, vibration deterioration degree, and gas deterioration degree.

[0037] Step 3: Normalize and weight the cumulative deterioration of temperature and humidity, vibration and gas at each node to obtain the freshness status value; after attenuating the freshness status value of each node in chronological order, recursively accumulate it to obtain the cumulative freshness status value; construct the whole-chain freshness change trajectory of the pre-cooked food based on the nodes and their corresponding cumulative freshness status values ​​to obtain the freshness status curve; then identify abnormal points based on the freshness status curve and make risk judgments. If there are abnormal points, the batch of pre-cooked food is in a high-risk state of freshness loss; otherwise, proceed to step 4.

[0038] Step 4: Based on the preservation status curve, predict the remaining shelf life and determine the status of the prepared food.

[0039] The present invention has the following beneficial effects:

[0040] 1. This invention deploys IoT information collection nodes in multiple stages, including the production and warehousing of pre-prepared dishes, cold storage, cold chain transportation, and terminal shelf placement, to continuously collect data on ambient temperature, ambient humidity, volatile organic compound concentration, carbon dioxide concentration, ammonia concentration, and triaxial vibration acceleration. It also constructs temperature and humidity deterioration, vibration deterioration, and gas deterioration, which can simultaneously characterize the comprehensive deterioration state of pre-prepared dishes during circulation from three dimensions: temperature control instability, mechanical disturbance, and microbial metabolism. This avoids the problems of single monitoring dimensions and one-sided judgment results caused by existing technologies that rely solely on single temperature control or manual inspection.

[0041] 2. This invention recursively accumulates the preservation status values ​​corresponding to each node according to the time decay method to construct a full-link preservation status curve, and identifies rising inflection points and sudden change inflection points in the curve. This allows for accurate capture of the critical moments when pre-cooked food changes from a normal state to a high-risk state of freshness loss during multi-node, multi-subject, and multi-scenario transportation. It enables dynamic identification and trend judgment of the cumulative deterioration process, thereby overcoming the shortcomings of traditional methods that can only reflect the environmental state at a local point in time and cannot characterize the degree of cumulative deterioration throughout the entire process.

[0042] 3. This invention also predicts the remaining shelf life based on the freshness status curve and performs graded early warnings based on the remaining shelf life status. The monitoring results can be directly converted into actionable status warnings such as adjusting the outbound priority, optimizing the sales sequence, isolating products from sales, and reporting losses upon re-inspection. This enables the monitoring of the freshness status of pre-prepared vegetables, improves the consistency of quality throughout the process, the authenticity of the shelf life, and consumer safety, and has strong real-time performance, accuracy, and practicality.

[0043] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of the invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For those skilled in the art, the drawings can be obtained from these drawings without creative effort.

[0045] Figure 1 A block diagram of an IoT-based pre-prepared food preservation monitoring system is provided for this invention.

[0046] Figure 2 The present invention provides a flowchart of a pre-prepared food preservation monitoring system based on the Internet of Things. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] In existing technologies, the preservation and management of pre-prepared meals largely relies on methods such as cold chain temperature control, manual inspection, and regular sampling to control the storage and transportation process. However, it is difficult to identify and comprehensively judge the quality changes caused by multiple factors such as temperature fluctuations, humidity changes, packaging damage, repeated opening, transportation vibration, and microbial deterioration in real time. Especially in the actual circulation process, pre-prepared meals often undergo transfer through multiple nodes, multiple entities, and multiple scenarios, which can easily lead to cold chain interruptions, environmental loss of control, or record lag. Traditional methods can usually only reflect the environmental state at a local point in time and cannot accurately characterize the cumulative degree of deterioration of pre-prepared meals throughout the circulation process. It is also difficult to judge in a timely manner whether they are close to the critical state of loss of freshness or spoilage, and cannot guarantee the consistency of quality, authenticity of shelf life, and safety of consumption of pre-prepared meals throughout the entire process.

[0049] To address the problems mentioned in the background section, such as Figure 1 As shown, this embodiment of the invention provides an IoT-based pre-prepared vegetable preservation monitoring system, which specifically includes: an IoT acquisition module, a deterioration accumulation module, a full-node trajectory module, and a prediction status module;

[0050] The IoT data acquisition module is used to deploy IoT data acquisition nodes at multiple nodes during the production and warehousing of pre-prepared dishes, cold storage, cold chain transportation, and terminal shelf placement to obtain circulation status information of the same batch of pre-prepared dishes during the storage period at each node; circulation status information includes: ambient temperature, ambient humidity, volatile organic compound concentration, carbon dioxide concentration, ammonia concentration, and triaxial vibration acceleration.

[0051] The degradation accumulation module quantifies the temperature and humidity disturbances, mechanical vibration disturbances, and gas metabolism disturbances of the pre-cooked food at each node based on the flow status information, and obtains the corresponding temperature and humidity degradation degree, vibration degradation degree, and gas degradation degree.

[0052] The full-node trajectory module performs normalized weighted fusion of the cumulative deterioration of temperature and humidity, vibration and gas at each node to obtain the node preservation status value. The preservation status value of each node is accumulated and decayed in chronological order to construct the full-link preservation change trajectory and obtain the preservation status curve. Then, based on the preservation status curve, abnormal points are identified and risk is determined. If abnormal points exist, the batch of pre-prepared vegetables is in a high-risk state of freshness loss; otherwise, the prediction status module is executed.

[0053] The prediction status module predicts the remaining shelf life based on the freshness status curve and uses this to determine the status of the prepared food.

[0054] In the specific implementation process of the above embodiments:

[0055] The IoT data acquisition module generates a unique batch code for each batch of pre-prepared food, and binds this batch code with the product name, recipe category, packaging form, initial storage time, target storage temperature, target storage humidity, and preset shelf life to form the basic information of the batch. Simultaneously, IoT information acquisition devices are deployed at each stage of pre-prepared food production and warehousing, cold storage, cold chain transportation, and final shelf placement, enabling each node to form a continuous monitoring link around the same batch. The information collected by each node is stored as the node's circulation status information. This circulation status information includes the storage period of the pre-prepared food at that node, as well as the ambient temperature, humidity, volatile organic compound concentration, carbon dioxide concentration, ammonia concentration, and triaxial vibration acceleration at each acquisition time within that period.

[0056] The degradation accumulation module, based on the flow status information output by the IoT acquisition module, quantifies the temperature and humidity disturbances, mechanical vibration disturbances, and gas metabolism disturbances of the pre-cooked food at each node, obtaining the corresponding temperature and humidity degradation degree, vibration degradation degree, and gas degradation degree; specifically:

[0057] The storage period of the pre-prepared food at a specific node, along with the corresponding ambient temperature and humidity at each sampling time within that period, are extracted. The ambient temperature and humidity are then compared with the target storage temperature and humidity for that batch of pre-prepared food, and the absolute values ​​are taken to calculate the temperature deviation and humidity deviation. These deviations are then normalized. Subsequently, weighting coefficients are assigned to the normalized temperature and humidity deviations, and the deviations are weighted and fused according to these coefficients to obtain the storage deviation value at each sampling time. Finally, a two-dimensional Cartesian coordinate system is constructed with time as the x-axis and the storage deviation value as the y-axis, and the values ​​at each sampling time are plotted. After saving the deviation values ​​as a number of points, they are smoothly connected to obtain a saved deviation change map. Local peak points and local valley points are extracted from the saved deviation change map as poles. The difference in saved deviation and the time interval between two adjacent poles are calculated to obtain the local fluctuation coefficient. When any local fluctuation coefficient is greater than the effective deviation threshold, the corresponding interval is marked as an effective fluctuation interval. The temperature and humidity degradation contribution value is calculated based on the effective fluctuation intensity and the saved deviation area. The temperature and humidity degradation contribution values ​​of all effective fluctuation intervals are accumulated to obtain the temperature and humidity degradation degree. When there is no effective fluctuation interval, the temperature and humidity degradation degree is assigned a preset benchmark value of 0.2.

[0058] The system extracts the storage time period of the pre-prepared food at a certain node and the corresponding triaxial vibration acceleration at each acquisition time within that storage period. The triaxial vibration acceleration is then synthesized into a single vibration acceleration to avoid deviations caused by single-axis fluctuations. A two-dimensional rectangular coordinate system is constructed with time as the x-axis and vibration acceleration as the y-axis. The vibration acceleration at each moment is plotted as several points and then smoothly connected to obtain a vibration variation map. Local peak points and local valley points are extracted from the vibration variation map as poles. The vibration acceleration difference and time interval between adjacent poles are calculated to obtain the vibration fluctuation coefficient. When any vibration fluctuation coefficient is greater than the effective vibration threshold, an effective vibration deviation is determined, and the corresponding interval is marked as an effective vibration interval. The vibration degradation degree is calculated based on the effective vibration interval, and the vibration degradation contribution values ​​of all effective vibration intervals are accumulated to obtain the vibration degradation degree. When no effective vibration interval exists, the vibration degradation degree is assigned a preset benchmark value of 0.2.

[0059] The storage period of pre-cooked food at a certain node and the corresponding concentrations of volatile organic compounds (VOCs), carbon dioxide (COD), and ammonia at each sampling time within that storage period are extracted. The VOC, COD, and ammonia concentrations are normalized to obtain their deviation values. These deviation values ​​are then weighted and fused according to weighting coefficients to obtain the gas deviation values ​​at each sampling time. A gas deviation change graph is constructed with time as the x-axis and gas deviation values ​​as the y-axis, and the cumulative gas area and gas growth rate at two adjacent sampling times are calculated. When the gas growth rate continuously exceeds a preset positive growth threshold, it is recorded as a continuous deviation, and the cumulative growth of continuous deviations within the continuous deviation segment is calculated. After normalizing the cumulative growth of continuous deviations, the continuous deviation contribution value is calculated. Then, the continuous deviation contribution values ​​of any continuous deviation are fused using a product-based saturation mapping to obtain a continuous deviation factor. Finally, the cumulative gas area and the continuous deviation factor are combined to calculate the cumulative gas degradation degree.

[0060] The full-node trajectory module is used to normalize the temperature and humidity degradation, vibration degradation, and gas degradation of pre-prepared food at a certain node, and then perform weighted fusion according to preset weight coefficients to obtain the freshness status value at that node. The freshness status value is used to characterize the overall degree of degradation of pre-prepared food during the storage period at that node. The higher the freshness status value, the higher the risk of freshness loss, spoilage, or quality decline of pre-prepared food at that node. The freshness status values ​​of each node in the process of pre-prepared food going through production and warehousing, cold storage, cold chain transportation, and terminal shelf placement are arranged in the order of the nodes appearing. Combined with the dwell time between nodes and the cumulative impact of the previous node on the next node, a full-link freshness change trajectory curve with a time progression relationship is formed, and rising inflection points and abrupt change inflection points are identified in the curve. Among them, rising inflection points are used to characterize the freshness status from a flat stage to a significant rising stage, and abrupt change inflection points are used to characterize the freshness status undergoing a significant jump in a short period of time. If there is an upward inflection point or a sudden inflection point in the preservation status curve, the batch is determined to be in a high-risk state of freshness loss, and a freshness loss risk warning is generated.

[0061] The prediction status module is used to determine the remaining shelf life of pre-prepared vegetables based on the freshness status curve output by the full node trajectory module.

[0062] like Figure 2 As shown, this embodiment of the invention provides a method for monitoring the freshness of pre-prepared vegetables based on the Internet of Things, specifically including the following steps:

[0063] Step one: First, a unique batch code is generated for each batch of pre-prepared food. This batch code is then linked to the product name, recipe category, packaging form, initial warehousing time, target storage temperature, target storage humidity, and preset shelf life to form the basic information for that batch. IoT information collection nodes are deployed at each stage of pre-prepared food production and warehousing, cold storage, cold chain transportation, and final shelf placement. This allows each node to form a continuous monitoring link around the same batch, and the information collected by each node is stored as the node's circulation status information. The information includes the ambient temperature, humidity, volatile organic compound (VOC) concentration, carbon dioxide concentration, ammonia concentration, and triaxial vibration acceleration at each collection point during the storage period of the prepared food at that node. VOC concentration refers to the total amount of various volatile organic compounds contained in the air, which is usually a comprehensive concentration of volatile organic components such as alcohols, aldehydes, ketones, acids, esters, amines, and sulfides. In the preservation monitoring of prepared food, the higher the VOC concentration, the worse the freshness and the higher the degree of deterioration of the prepared food.

[0064] Step two involves extracting the storage period of the pre-prepared food at a specific point in time, along with the corresponding ambient temperature and humidity at each sampling time within that period. The ambient temperature and humidity are then compared to the target storage temperature and humidity for that batch of pre-prepared food, and the absolute values ​​are used to calculate the temperature deviation and humidity deviation. These deviations characterize the degree of bidirectional deviation between temperature and humidity and the target storage conditions. Both deviations are then normalized, and a weighting coefficient is assigned to each of the normalized humidity and temperature deviations. In this embodiment, considering that during cold chain storage and transportation of pre-prepared food, the influence of temperature on microbial reproduction rate and quality deterioration is generally greater than the influence of humidity on quality, the weighting coefficient corresponding to the temperature deviation is set to... The weighting coefficient for humidity deviation is set to 0.3, and the sum of the two is 1. Based on the assigned weighting coefficient, the temperature deviation and humidity deviation are weighted and fused to obtain the storage deviation value. This yields the storage deviation value of the pre-prepared food at each moment during the storage period at this node. A two-dimensional rectangular coordinate system is constructed with time as the x-axis and the storage deviation value as the y-axis. The storage deviation value at each moment during the storage period is plotted as several points in the coordinate system. Then, a smooth curve is used to connect these points sequentially to obtain a storage deviation variation graph of the pre-prepared food at this node during the storage period. Extreme points are selected in the storage deviation variation graph, including local peak points and local valley points. The difference in storage deviation value between two adjacent extreme points is calculated. The adjacent variation values ​​are calculated, and the time difference between the two adjacent extreme points is used to obtain the adjacent interval. Then, the adjacent variation value is divided by the adjacent interval to obtain the local fluctuation coefficient. The larger the local fluctuation coefficient, the higher the rate of change of storage deviation between adjacent extreme points, indicating that the dynamic disturbance of temperature and humidity at that node deviates from the target storage conditions is more significant, thus reflecting the greater risk of cumulative deterioration of the prepared food in the corresponding period. An effective deviation threshold is preset. The effective deviation threshold is used to distinguish between normal environmental fluctuations and abnormal fluctuations that have a substantial impact on quality. It is usually determined based on the statistical distribution of the local fluctuation coefficient in historical normal storage and transportation data. For example, the 95th percentile value is taken as the threshold to ensure that most normal fluctuations are not misjudged, while also being able to identify abnormal fluctuations. The system maintains high sensitivity to fluctuations. It iterates through all local fluctuation coefficients in the deviation change graph. If any local fluctuation coefficient is greater than the effective deviation threshold, it is considered that there is an effective temperature and humidity deviation. The interval formed by the two adjacent poles corresponding to the local fluctuation coefficient is marked as the effective fluctuation interval, and temperature and humidity degradation is accumulated. Otherwise, it is considered a normal fluctuation, and the temperature and humidity degradation degree is assigned a preset benchmark value. In this embodiment, the preset benchmark value is set to 0.2. The benchmark value is used to represent the background output level of the system when no effective degradation is detected. Its value is preferably determined by the average output level of historical normal samples, so that the output is stable and non-zero under normal conditions, while ensuring that there is obvious distinction when there is abnormal degradation.

[0065] The cumulative temperature and humidity deterioration within the effective fluctuation range is as follows:

[0066] For each effective fluctuation range, calculate its effective fluctuation intensity F(i), where i represents the index of the effective fluctuation range. The calculation formula is as follows:

[0067]

[0068] Where k(i) represents the local fluctuation coefficient of the effective interval, and K represents the effective deviation threshold;

[0069] For each effective fluctuation interval, the area of ​​preserved deviation A(i) within that effective fluctuation interval is calculated using the trapezoidal integral method, as follows:

[0070]

[0071] Where B(i) and B(i+1) represent the storage deviation values ​​corresponding to the poles at both ends of the interval, and b(i+) and b(i+1) represent the time corresponding to the poles at both ends of the interval;

[0072] Multiply the effective fluctuation intensity F(i) of the effective fluctuation range by the area of ​​the preserved deviation A(i) to obtain the deterioration contribution value of the effective fluctuation range. Sum the deterioration contribution values ​​of all effective fluctuation ranges in the preserved deviation change graph to obtain the temperature and humidity deterioration degree.

[0073] Step 3: Extract the three-axis vibration acceleration corresponding to the storage period of the pre-cooked food at a certain node and the collection time within the storage period. To avoid deviation caused by single-axis fluctuations, the three-axis accelerations are synthesized into vibration acceleration. Based on the vibration acceleration of the pre-cooked food at each moment within the storage period at that node, a two-dimensional rectangular coordinate system is constructed with time as the abscissa and vibration acceleration as the ordinate. The vibration acceleration at each moment is plotted as several points and then smoothly connected to obtain a vibration variation diagram. Since the response of the pre-cooked food to vibration is not only related to the instantaneous acceleration but also to the continuity of acceleration change, local peak points and local valley points are extracted from the vibration variation diagram as poles. For two adjacent poles, the difference in vibration acceleration is calculated, and the time interval between the two poles is also calculated. The vibration fluctuation coefficient is obtained by dividing the difference in vibration acceleration by the time interval. The larger the vibration fluctuation coefficient, the more severe the mechanical disturbance change between the two poles, and the more severe the mechanical disturbance change in the pre-cooked food. The more frequent the impact, shaking, or oscillation, the more likely it is to cause a decrease in packaging sealing performance, damage to the food structure, stratification of juices, or localized compression deformation. A preset effective vibration threshold is used to distinguish between normal background vibration during transportation and abnormal mechanical disturbances that may cause quality damage. This threshold is usually determined based on the statistical distribution of vibration fluctuation coefficients in historical transportation data, or calibrated in conjunction with the critical vibration level that causes structural damage in packaging vibration resistance tests. If the vibration fluctuation coefficient between any two extreme points is greater than the effective vibration threshold, it indicates that there is an effective vibration deviation in the vibration change diagram. The interval formed by the two extreme points is taken as the effective vibration interval, and vibration degradation accumulation is performed on the effective vibration interval. Otherwise, it indicates that only normal background vibration exists and does not constitute effective vibration degradation. In this case, the vibration degradation accumulation degree is assigned a preset benchmark value, which is set to 0.2 in this example. Vibration degradation accumulation on the effective vibration interval specifically includes:

[0074] Similarly, for each effective vibration interval, the effective vibration intensity is obtained by subtracting the effective vibration threshold from the vibration fluctuation coefficient and dividing by the effective vibration threshold. Then, the vibration area is obtained by calculating the integral approximation of the vibration change amplitude using the trapezoidal area method. Specifically, the vibration area is the sum of the vibration accelerations corresponding to the extreme values ​​at both ends of the effective vibration interval, divided by 2, and then multiplied by the duration of the effective vibration interval, where the duration refers to the time difference value of the horizontal axis corresponding to the effective vibration interval. The vibration deterioration contribution value of the effective vibration interval is obtained by multiplying the effective vibration intensity by the vibration area. All effective vibration intervals in the vibration change diagram are traversed, and their corresponding vibration deterioration contribution values ​​are summed to obtain the cumulative vibration deterioration degree.

[0075] Step 4: Extract the storage time period of the pre-cooked food at a certain node, and the corresponding volatile organic compound (VOC) concentration, carbon dioxide concentration, and ammonia concentration at each collection time within the storage time period; normalize the VOC, carbon dioxide, and ammonia concentrations to obtain the deviation values ​​for VOC, carbon dioxide, and ammonia. The normalization calculation formula is as follows:

[0076]

[0077]

[0078]

[0079] in, , and These are the concentrations of volatile organic compounds, carbon dioxide, and ammonia, respectively. , and These are the reference concentrations for volatile organic compounds, carbon dioxide, and ammonia, respectively. The reference concentrations are used to characterize the gas reference levels when pre-cooked dishes are in a fresh and stable state. They are obtained by testing fresh samples under standard storage conditions or determined by the statistical mean of historical qualified samples, in order to reflect the gas release characteristics of different types of pre-cooked dishes under normal conditions. , and These are the allowable deviation thresholds for the corresponding gases. These thresholds limit the fluctuation range of gas concentration within a range that does not affect quality, and are determined based on food safety standards, enterprise internal control standards, or the critical level before gas mutation in historical deterioration samples. A weighting coefficient is assigned to the deviation values ​​of volatile organic compounds (VOCs), carbon dioxide, and ammonia, with the sum of these coefficients being one. The deviation values ​​of VOCs, carbon dioxide, and ammonia are weighted and fused according to the assigned weighting coefficients to obtain the gas deviation value. In this embodiment, the weighting coefficients for VOCs, carbon dioxide, and ammonia are set to 0.45, 0.2, and 0.35, respectively. VOCs and ammonia generally reflect the putrefaction and protein decomposition process more directly, so their weights can be appropriately higher than carbon dioxide. A two-dimensional rectangular coordinate system is constructed with time as the x-axis and the gas deviation value as the y-axis to plot the gas deviation change graph. The cumulative area of ​​gas corresponding to two adjacent collection times in the gas deviation change graph is calculated using the following formula. The calculation formula is as follows:

[0080]

[0081] Where t represents the acquisition time index, E t and E t+1These represent the gas deviation values ​​at two adjacent sampling times t and t+1, respectively, and the gas cumulative area characterizes the basic cumulative amount of body deterioration within the time interval of two adjacent sampling times t and t+1.

[0082] Further calculation of gas growth rate The calculation formula is as follows:

[0083]

[0084] Gas growth rate Reflects the rate of increase of gas degradation value between adjacent sampling times; the gas growth rate at each adjacent time point. Arranged chronologically, if there are more than n consecutive gas growth rates exceeding the preset positive growth threshold, it is recorded as a continuous deviation, and the number of sampling points with continuous positive growth within the continuous deviation is recorded as m. The positive growth threshold is used to filter out small random fluctuations; only when the gas growth rate exceeds this threshold is it considered to have a substantial growth trend. It is set based on the mean or standard deviation of historical normal fluctuation data, for example, the mean plus one standard deviation is taken as the threshold. For each continuous deviation, the cumulative growth of the continuous deviation within the continuous deviation segment is calculated, and the cumulative growth of the continuous deviation is normalized and recorded as Q. Usually, the historical maximum continuous growth or the allowable growth threshold is used for normalization. The cumulative growth of the continuous deviation is the cumulative result of the positive increment of the gas deterioration value at each adjacent time point within the continuous deviation segment. For each continuous deviation, its corresponding continuous deviation contribution value φ is calculated, and the calculation formula is as follows:

[0085]

[0086] Where μ is the continuous deviation number adjustment parameter, used to characterize the amplification intensity of the continuous deviation degree by the number of continuous positive growth points. In this embodiment, the value is 3. When the number of continuous positive growth points reaches 3 or more, it indicates that the gas deterioration value has formed a stable and continuous upward trend between adjacent collection times, rather than a single point or short-term fluctuation. η is the growth amount adjustment parameter, used to characterize the amplification intensity of the continuous deviation cumulative growth amount by the continuous deviation degree, and the value is 1.0. Under the combined effect of the continuous deviation number and the continuous deviation cumulative growth amount, the continuous deviation factor achieves the synergistic characterization of the persistence and growth intensity of gas deterioration, thereby improving the accuracy of the characterization of the spoilage development process of pre-prepared vegetables by the cumulative gas deterioration degree.

[0087] From this, we can obtain the continuous deviation contribution value φ for any consecutive deviation in the gas deviation change graph. j , j represents the index of any consecutive deviation in the gas deviation change diagram, and the continuous deviation factor p is obtained by fusion calculation through product saturation mapping. The continuous deviation factor p characterizes the continuity of gas degradation, and the calculation formula is:

[0088]

[0089] By employing the fusion calculation of product-type saturation mapping, the linear distortion caused by simple summation can be avoided, while making the influence of multiple consecutive deviation segments on the final factor more stable and more in line with the cumulative deterioration law.

[0090] The gas degradation area H is calculated by summing the cumulative gas areas at all adjacent time points in the gas deviation change graph. Then, the gas degradation area and the continuous deviation factor are combined to obtain the gas degradation degree D through saturation mapping. The calculation formula is as follows:

[0091]

[0092] Where D0 is the normal fluctuation benchmark value, which is set to 0.2 in this embodiment; the normal fluctuation benchmark value is used to represent the background output level when there is no effective gas deviation, so as to avoid the result being always zero under normal fluctuation conditions, and to facilitate the distinction from the situation where there is effective deviation; through this saturation mapping method, the degree of gas degradation gradually increases with the cumulative degradation over the time interval, while avoiding the result being infinitely amplified under extreme gas conditions, thereby improving the stability and comparability of gas degradation evaluation;

[0093] Step 5: After obtaining the temperature and humidity degradation, vibration degradation, and gas degradation of the pre-cooked food at a certain node, the three are normalized to ensure they are on the same numerical scale. Then, they are weighted and fused according to preset weighting coefficients to obtain the freshness status value at that node. The freshness status value is used to characterize the overall degree of degradation of the pre-cooked food during its storage period at that node. In this embodiment, the weighting coefficients for temperature and humidity degradation, vibration degradation, and gas degradation are set to 0.4, 0.25, and 0.35, respectively, and the sum of the three is 1. Of course, in other embodiments, the weighting coefficients can also be adjusted according to... The weighting coefficients are adjusted based on the type of pre-prepared food, packaging format, and storage and transportation conditions. Using this method, the freshness status value at each node is obtained. A higher freshness status value indicates a higher risk of freshness loss, spoilage, or quality deterioration at that node. After the pre-prepared food passes through multiple nodes—production and warehousing, cold storage, cold chain transportation, and final shelf placement—the freshness status values ​​for each node are obtained and arranged according to their chronological order. The cumulative impact of each node on the next is then integrated to form a time-progressive, end-to-end freshness change trajectory. Specifically:

[0094] Let d = 1, 2, 3...M be the nodes that the pre-prepared food passes through in sequence, where M is a positive integer representing the total number of nodes passed through during the transportation of the pre-prepared food, and d represent the index of any one of these nodes; the preservation status value at each node is denoted as S. dAs time progresses, the influence of the preceding node's state on subsequent nodes gradually diminishes. A time decay coefficient w is assigned to each node d. d Multiplying this by the preservation status value of the corresponding node yields the preservation degradation value R. d Time decay coefficient w d The calculation formula is as follows:

[0095]

[0096] Where Δ d Δ represents the time difference between entering node d and exiting node d, i.e., the storage time of the prepared food at node d; λ is the attenuation adjustment coefficient, λ > 0, used to adjust the degree to which the node interval weakens the transmission intensity of the freshness state; Δ d The larger the value, the weaker the influence of the current node d on the next node (d+1); a cumulative trajectory is formed using a recursive method to construct a continuous end-to-end freshness change trajectory, C1=R1, C2=C1+R2, ..., C d =C d-1 +R d ,Right now S d This represents the cumulative freshness status value up to the d-th node. The meaning of this recursive relationship is that the degradation impact of previous nodes continues to decay with time intervals, but the freshness status value of the current node is re-added to the cumulative result, thus forming a trajectory that reflects the degradation evolution process across the entire link; thus, (d, C) is obtained. d The data points are arranged chronologically and plotted in a two-dimensional rectangular coordinate system with time as the x-axis and cumulative freshness status as the y-axis. Cubic spline interpolation, piecewise polynomial fitting, or smoothing curve connection methods are then used to connect the discrete points into a continuous curve, resulting in a batch-level freshness status curve. During curve identification, the first and second derivatives of the curve are calculated. If a node satisfies the condition that the curve slope changes from a flat state to a significantly increasing state and the second derivative is positive, it is determined to be an upward inflection point. When the curve slope is greater than a preset upward threshold, it indicates that the freshness status has entered a continuous upward phase. This upward threshold is preferably determined based on the upper quantile value of the curve slope in historical normal samples to avoid misjudging ordinary fluctuations as upward inflection points. If the slope difference between adjacent intervals exceeds a preset abrupt change threshold, and the local curvature exceeds a preset curvature threshold, it is determined to be an abrupt change inflection point. The abrupt change threshold is preferably determined based on the smallest characteristic value of the slope surge in historical abnormal samples, and the curvature threshold is preferably calibrated based on the statistical distribution of the second derivatives of normal and abnormal samples. If there is an upward inflection point or a sudden inflection point in the preservation status curve, it indicates that the pre-cooked food has experienced a cold chain interruption, environmental loss of control, or abnormal exposure during the distribution process, and the batch is judged to be in a high-risk state of freshness loss; if there is no upward inflection point or a sudden inflection point, proceed to step six.

[0097] Step 6, based on the preservation status curve (d, C) d To determine the remaining shelf life of pre-prepared meals, the following steps are taken:

[0098] Extract the cumulative freshness status value at the current node and establish a shelf-life judgment benchmark based on the preset shelf life of this batch of pre-prepared vegetables; use a sliding window to locally fit the freshness status curve to obtain the trend function after the current node; the window length and step size of the sliding window are preset and determined according to the fluctuation period of the freshness status value in historical samples; based on the trend function and the correspondence between the cumulative freshness status value and the shelf-life consumption progress in historical qualified samples, establish a mapping model from the cumulative freshness status value to the shelf-life consumption ratio. The mapping model is used to convert the cumulative freshness status value at the current node into the corresponding shelf-life consumption ratio; the mapping model can be any of the piecewise linear function, regression function, or lookup table function; then calculate the expected time corresponding to the expected end of the preset shelf life based on the shelf-life consumption ratio and the preset shelf life, and the remaining shelf life is equal to the expected time minus the current time;

[0099] A preset shelf-life range is established, set according to the standard shelf life, packaging method, and storage and transportation temperature zone of different types of pre-prepared meals. If the remaining shelf life is within the preset range, the pre-prepared meal is considered to be in a shelf-life warning state. If the remaining shelf life is less than the lower limit of the preset range, the pre-prepared meal is considered to be in a state of risk of spoilage. If the remaining shelf life is greater than the upper limit of the preset range, the pre-prepared meal is considered to be in a normal preservation state. When it is determined to be in a normal preservation state, the original outbound and sales plans are maintained. When it is determined to be in a shelf-life warning state, the outbound order is adjusted first and the subsequent circulation time is shortened. When it is determined to be in a state of near-spoilage risk, sales are suspended, isolation is carried out, re-inspection is performed, or priority is given to reporting losses. Through the above methods, local deterioration information in the entire circulation chain can be linked and mapped with the batch-level shelf-life status, thereby realizing the monitoring and control of the consistency of quality, authenticity of shelf life, and safety of consumption of pre-prepared meals throughout the entire process.

[0100] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0101] The preferred embodiments of the invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention.

Claims

1. A pre-prepared dish freshness monitoring system based on the Internet of Things, characterized by, include: The Internet of Things (IoT) data acquisition module is used to deploy IoT data acquisition devices at multiple nodes, including pre-prepared food production and warehousing, cold storage, cold chain transportation, and terminal shelf placement, to obtain information on the circulation status of the same batch of pre-prepared food during the storage period at each node. The degradation accumulation module quantifies the temperature and humidity disturbances, mechanical vibration disturbances, and gas metabolism disturbances of pre-cooked food at each node based on the flow status information, and obtains the corresponding temperature and humidity degradation degree, vibration degradation degree, and gas degradation degree. The full-node trajectory module performs normalized weighted fusion of the cumulative temperature and humidity degradation, vibration degradation, and gas degradation of each node to obtain the preservation status value. The freshness status value of each node is attenuated in chronological order and then accumulated to obtain the cumulative freshness status value. Based on the node and the corresponding cumulative freshness status value, the whole-link freshness change trajectory of the pre-cooked food is constructed to obtain the freshness status curve. Then, based on the freshness status curve, abnormal points are identified and risk is determined. If there are abnormal points, the batch of pre-cooked food is in a high-risk state of freshness loss; otherwise, the prediction status module is executed. The prediction status module predicts the remaining shelf life based on the freshness status curve and uses this to determine the status of the prepared food.

2. The IoT-based pre-prepared food preservation monitoring system according to claim 1, wherein, Quantification of temperature and humidity disturbances: The ambient temperature and humidity are compared with the target storage temperature and humidity, respectively. The absolute values ​​are then normalized and weighted according to preset weights to obtain the storage deviation value. A storage deviation change graph is constructed with time as the horizontal axis and the storage deviation value as the vertical axis. Local peak points and local valley points are extracted as poles. The storage deviation difference and time interval between adjacent poles are calculated to obtain the local fluctuation coefficient. When the local fluctuation coefficient is greater than the effective deviation threshold, the effective fluctuation range is determined. The cumulative deterioration degree of temperature and humidity is calculated based on the effective fluctuation range. 3.The IoT-based pre-prepared dish freshness monitoring system according to claim 2, wherein, The local fluctuation coefficient is obtained by calculating the difference in preservation deviation between adjacent poles and the time interval: In the saved deviation change graph, poles are selected, including local peak points and local valley points. The difference between the saved deviation values ​​between two adjacent poles is calculated to obtain the adjacent change value. The time difference between the two adjacent poles is calculated to obtain the adjacent interval. The adjacent change value is then divided by the adjacent interval to obtain the local fluctuation coefficient. A preset effective deviation threshold is established. All local fluctuation coefficients in the saved deviation change graph are traversed. If any local fluctuation coefficient is greater than the effective deviation threshold, the interval formed by the two adjacent poles corresponding to the local fluctuation coefficient is marked as the effective fluctuation interval, and temperature and humidity deterioration is accumulated for it. Otherwise, the temperature and humidity deterioration is assigned a preset benchmark value. 4.The IoT-based pre-prepared dish freshness monitoring system according to claim 3, wherein, Calculation of cumulative temperature and humidity deterioration based on the effective fluctuation range: For each effective fluctuation interval, the effective fluctuation intensity is obtained by subtracting the effective deviation threshold from the local fluctuation coefficient and then dividing by the effective deviation threshold. For each effective fluctuation interval, the area of ​​the preserved deviation in the effective fluctuation interval is calculated using the trapezoidal integral method. The deterioration contribution value of the effective fluctuation interval is obtained by multiplying the effective fluctuation intensity of the effective fluctuation interval by the area of ​​the preserved deviation. The temperature and humidity deterioration degree is obtained by summing the deterioration contribution values ​​of all effective fluctuation intervals in the preserved deviation change graph. 5.The IoT-based pre-prepared dish freshness monitoring system according to claim 4, wherein, Quantification of mechanical vibration disturbance: The triaxial vibration acceleration is synthesized into a vibration acceleration, and a vibration variation diagram is constructed with time as the abscissa and vibration acceleration as the ordinate. Local peak points and local valley points are extracted as poles. The vibration fluctuation coefficient is obtained by calculating the vibration acceleration difference and time interval between adjacent poles. When the vibration fluctuation coefficient is greater than the effective vibration threshold, the effective vibration interval is determined, and the vibration deterioration degree is calculated based on the effective vibration interval. 6.The IoT-based pre-prepared dish freshness monitoring system according to claim 5, wherein, Quantification of gas metabolism perturbations: After normalizing the concentrations of volatile organic compounds, carbon dioxide, and ammonia, the gas deviation values ​​are obtained by weighted fusion according to preset weighting coefficients. A gas deviation change map is constructed with time as the horizontal axis and gas deviation values ​​as the vertical axis. The cumulative gas area and gas growth rate are calculated based on the gas deviation values ​​at adjacent acquisition times. When the gas growth rate continuously exceeds the preset positive growth threshold, continuous deviation is determined. The continuous deviation contribution value is calculated based on all continuous deviations in the gas deviation change map, and the continuous deviation factor is obtained through product saturation mapping. Finally, the cumulative gas degradation degree is calculated by combining the cumulative gas area and the continuous deviation factor.

7. The pre-prepared vegetable preservation monitoring system based on the Internet of Things according to claim 6, characterized in that, Calculate the continuous deviation contribution value based on all continuous deviations in the gas deviation change graph: Extract the gas growth rate between two adjacent sampling times, which reflects the rate of increase of gas degradation value between adjacent sampling times; arrange the gas growth rates of each adjacent time in chronological order; if there are more than n consecutive gas growth rates greater than the preset positive growth threshold, it is recorded as a continuous deviation, and the number of sampling points with continuous positive growth within the continuous deviation is counted. For each consecutive deviation, the cumulative increase in consecutive deviation within that consecutive deviation segment is calculated and normalized. For each consecutive deviation, the contribution value of this consecutive deviation is calculated by formulating the number of sampling points with consecutive positive growth and the cumulative increase in consecutive deviation within the consecutive deviation.

8. The pre-prepared vegetable preservation monitoring system based on the Internet of Things according to claim 7, characterized in that, Identify anomalies and assess risks based on the preservation status curve: Arranged chronologically, scatter points are plotted in a two-dimensional rectangular coordinate system with time as the abscissa and cumulative freshness status as the ordinate. Connecting the discrete points forms a continuous curve, resulting in a batch-level freshness status curve. During curve recognition, the first and second derivatives of the curve are calculated. If a node satisfies the condition that the curve slope is greater than a preset rising threshold and the second derivative is positive, it is identified as a rising inflection point. If the slope difference between adjacent intervals exceeds a preset abrupt change threshold and the local curvature exceeds a preset curvature threshold, it is identified as abrupt change inflection point. Rising inflection points and abrupt change inflection points are collectively referred to as abnormal points.

9. The pre-prepared vegetable preservation monitoring system based on the Internet of Things according to claim 8, characterized in that, Predicting remaining shelf life based on freshness status curves: Extract the cumulative freshness status value at the current node and establish a shelf life determination benchmark based on the preset shelf life of this batch of pre-prepared vegetables; use a sliding window to locally fit the freshness status curve to obtain the trend function after the current node; based on the trend function and the correspondence between the cumulative freshness status value and the shelf life consumption progress in historical qualified samples, establish a mapping model from the cumulative freshness status value to the shelf life consumption ratio. The mapping model is used to convert the cumulative freshness status value at the current node into the corresponding shelf life consumption ratio; then calculate the expected time corresponding to the expected end of the preset shelf life based on the shelf life consumption ratio and the preset shelf life. The remaining shelf life is equal to the expected time minus the current time. If the remaining shelf life is within a preset timeframe, the prepared food is considered to be in a shelf-life warning state. If the remaining shelf life is less than the lower limit of the timeframe, the prepared food is considered to be in a state of risk of spoilage. If the remaining shelf life is greater than the upper limit of the timeframe, the prepared food is considered to be in a normal preservation state.

10. A method for monitoring the freshness of pre-prepared vegetables based on the Internet of Things, characterized in that... The system applied to the IoT-based pre-prepared vegetable preservation monitoring system as described in any one of claims 1-9 includes the following steps: Step 1 involves deploying IoT data collection nodes at multiple nodes during the production, cold storage, cold chain transportation, and terminal shelf placement of pre-prepared dishes to obtain information on the circulation status of the same batch of pre-prepared dishes during the storage period at each node. Step 2: Based on the flow status information, the temperature and humidity disturbances, mechanical vibration disturbances, and gas metabolism disturbances of the pre-cooked food at each node are quantified to obtain the corresponding temperature and humidity deterioration degree, vibration deterioration degree, and gas deterioration degree. Step 3: Normalize and weight the cumulative deterioration of temperature and humidity, vibration and gas at each node to obtain the freshness status value; after attenuating the freshness status value of each node in chronological order, recursively accumulate it to obtain the cumulative freshness status value; construct the whole-chain freshness change trajectory of the pre-cooked food based on the nodes and their corresponding cumulative freshness status values ​​to obtain the freshness status curve; then identify abnormal points based on the freshness status curve and make risk judgments. If there are abnormal points, the batch of pre-cooked food is in a high-risk state of freshness loss; otherwise, proceed to step 4. Step 4: Based on the preservation status curve, predict the remaining shelf life and determine the status of the prepared food.

Images