An intelligent tea quality traceability system based on Internet of Things technology

Abstract

The invention relates to the technical field of dynamic monitoring of tea quality, and specifically discloses an intelligent tea quality tracing system based on the Internet of Things technology. The system collects environmental humidity, tea moisture content and tea polyphenol content in real time respectively, transmits the data to a central processing system, uses an autoencoder and a Bayesian network to perform deep feature extraction and probability dependency relationship modeling on the collected data, generates an abnormal characteristic index of moisture content and an abnormal characteristic index of tea polyphenol content, and finally obtains a comprehensive abnormal characteristic index to evaluate whether the tea quality is qualified. The system calculates a humidity stability index through a wavelet decomposition method, identifies and quantifies abnormal fluctuations in environmental humidity. The invention constructs an interference analysis and dynamic control module, uses a random forest model to evaluate the relationship between humidity stability and tea quality, and realizes dynamic control of environmental humidity.

Description

Technical Field

[0001] The present invention relates to the technical field of dynamic monitoring of tea quality, and in particular to an intelligent tea quality tracing system based on the Internet of Things technology. Background Art

[0002] As a widely consumed beverage around the world, the quality of tea directly affects consumers' experience and market competitiveness. In the traditional tea production process, key parameters such as environmental humidity, moisture content and tea polyphenol content have a vital impact on the quality of tea. However, the changes in these parameters are often difficult to accurately control, resulting in unstable tea quality. In order to improve the quality and efficiency of tea production, modern tea production companies are increasingly relying on advanced sensor technology and data analysis methods. By real-time monitoring and analysis of environmental conditions and tea composition data, companies can better optimize the production process and ensure the consistency of quality of each batch of tea. Therefore, the development of a system that can achieve accurate data collection, intelligent analysis and dynamic regulation is of great significance to improving tea production efficiency and product quality.

[0003] The prior art has the following deficiencies:

[0004] Traditional methods mainly rely on manual experience to monitor environmental humidity and tea ingredients, which lacks real-time and accuracy, and cannot detect and correct potential problems in time, resulting in large fluctuations in tea quality. Secondly, the existing technology usually uses a single type of sensor for data collection, which cannot fully cover all key parameters, such as environmental humidity, moisture content and tea polyphenols content, limiting the ability to comprehensively evaluate tea quality. In addition, although some systems are capable of data collection, they are still insufficient in data analysis and processing, and fail to effectively use deep learning algorithms and probabilistic models for feature extraction and relationship modeling, making it difficult to generate accurate abnormal feature indexes and comprehensive quality evaluation results. Finally, the existing technology lacks an effective dynamic control mechanism and cannot automatically adjust the environmental humidity according to real-time data, resulting in the failure to effectively control the impact of changes in environmental conditions on tea quality during the production process. These problems urgently require an integrated and intelligent solution to comprehensively improve the quality and efficiency of tea production. Summary of the invention

[0005] The purpose of the present invention is to provide an intelligent tea quality traceability system based on Internet of Things technology to solve the above-mentioned problems.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] An intelligent tea quality traceability system based on Internet of Things technology, comprising:

[0008] A data collection module, which collects environmental humidity data in real time during the tea production process, and collects moisture content data and tea polyphenol content data of all tea leaves in real time;

[0009] A data accuracy evaluation module, which compares and analyzes the real-time collected humidity data, water content data and tea polyphenol content data with historical data to evaluate the accuracy of the collected real-time data;

[0010] A humidity data stability evaluation module, wherein the humidity data stability evaluation module extracts accurate environmental humidity data according to the evaluation result, obtains a time series of environmental humidity data within a monitoring period based on the accurate environmental humidity data, and evaluates the stability of the environmental humidity data by analyzing the fluctuation range and change rate of the humidity data;

[0011] A tea quality judgment module, which is used to perform deep processing on accurate moisture content data and tea polyphenol content data, generate moisture content abnormality characteristics and tea polyphenol content abnormality characteristics according to the deep processing results, and perform comprehensive analysis on the generated moisture content abnormality characteristics and tea polyphenol content abnormality characteristics, and judge whether the tea quality is qualified according to the analysis results;

[0012] The interference analysis and dynamic control module comprehensively analyzes the stability of the environmental humidity data and the quality of the tea leaves, determines the degree of interference of the abnormal environmental humidity data on the quality of the tea leaves, and dynamically controls the environmental humidity according to the degree of interference.

[0013] As a further solution of the present invention: the real-time collected humidity data, water content data and tea polyphenol content data are compared and analyzed with historical data to evaluate the accuracy of the collected real-time data, specifically including:

[0014] Collect humidity data, water content data and tea polyphenol content data in real time to form a real-time data sequence, and extract the corresponding historical reference data sequence from the historical database;

[0015] The distance between each element in the real-time data sequence and the historical reference data sequence is calculated using the Euclidean distance formula to generate a distance matrix , where the matrix element calculation formula is:

[0016] ;

[0017] in, Indicates the first The value of the data point, represents the data collection point, Indicates the number of historical reference data sequences The value of the data point, Indicates the first The data point is compared with the historical reference data sequence The Euclidean distance between data points is Indicates the length of the real-time data sequence, Indicates the length of the historical data sequence;

[0018] Constructing cumulative distance matrix based on dynamic programming , indicating the distance from the starting point to the position The minimum matching distance, the recursive formula of the cumulative distance matrix is:

[0019] ;

[0020] In the formula, From the starting point to the location The minimum cumulative distance, Indicates the minimum value;

[0021] From the cumulative distance matrix, the backtracking algorithm Backtrack to , get the optimal matching path ;

[0022] in, represents the set of matching points of the optimal path, represents the length of the optimal path, is the index value in the optimal path, indicating the Matching points, Represents the first The index of the matching points, Represents the first historical data sequence in the optimal path The index of the matching points;

[0023] According to the optimal path, the total distance on the path is calculated. The calculation expression is:

[0024] ;

[0025] In the formula, Indicates the total distance;

[0026] Calculate the normalized maximum distance between the real-time data series and the historical data series. The calculation expression is:

[0027] ;

[0028] In the formula, represents the average value of the historical data series, represents the normalized maximum distance;

[0029] According to the total distance of the optimal path and the normalized maximum distance, the accuracy coefficient is calculated, and the calculation expression is:

[0030] ;

[0031] In the formula, represents the accuracy coefficient;

[0032] Compare the accuracy coefficient of real-time data with the preset threshold and extract accurate data.

[0033] As a further solution of the present invention: the extraction of accurate environmental humidity data specifically includes:

[0034] Determine whether the accuracy coefficient of each real-time humidity data is greater than or equal to a preset threshold. If so, it means that the corresponding humidity data is accurate, and the humidity data is extracted. If not, it means that the corresponding data is inaccurate, and the humidity data is corrected.

[0035] As a further solution of the present invention: the stability of the environmental humidity data is evaluated by analyzing the fluctuation range and change rate of the humidity data, specifically including:

[0036] Obtain accurate environmental humidity data, analyze the abnormal fluctuation degree of humidity data according to the fluctuation range of humidity data, and calculate the humidity abnormal fluctuation coefficient;

[0037] Obtain accurate environmental humidity data, analyze the abnormal change rate of humidity data according to the change rate of humidity data, and calculate the humidity change rate abnormality coefficient;

[0038] The humidity abnormal fluctuation coefficient and the humidity change rate abnormal coefficient are normalized and calculated to obtain the humidity stability index.

[0039] It is determined whether the humidity stability index is greater than or equal to a preset threshold value. If so, it indicates that the corresponding environmental humidity data is stable; if not, it indicates that the corresponding environmental humidity data is unstable.

[0040] As a further solution of the present invention: the acquisition logic of the humidity stability index is:

[0041] Obtain the humidity time series data during the tea production process and use the wavelet decomposition method to transform the humidity data series , decomposed into low-frequency components and high-frequency components, the calculation expression is:

[0042] ;

[0043] in, Indicates The low-frequency component of each time acquisition point, Indicates The high-frequency component of each time acquisition point, Indicates Humidity data collected at each time point, Indicates the time collection point;

[0044] Based on high frequency components Calculate the wave energy, the calculation expression is:

[0045] ;

[0046] In the formula, Represents the energy of the high-frequency component of humidity data, Indicates the total number of time collection points;

[0047] The energy of the high frequency component of humidity data Normalized with the total energy of the humidity series, the humidity anomaly fluctuation coefficient is obtained, and the calculation expression is:

[0048] ;

[0049] In the formula, Indicates the humidity abnormal fluctuation coefficient;

[0050] Calculate the rate of change of humidity , calculate the mean humidity change rate and standard deviation ;

[0051] Indicates Humidity change rate at each time collection point;

[0052] Calculate the humidity change rate anomaly coefficient, the calculation expression is:

[0053] ;

[0054] In the formula, Represents the humidity change rate anomaly coefficient, Indicates the maximum absolute value of the humidity change rate;

[0055] The humidity abnormal fluctuation coefficient and the humidity change rate abnormal coefficient are comprehensively normalized and calculated to obtain the humidity stability index. The calculation expression is:

[0056] ;

[0057] In the formula, Represents the humidity stability index, and represents a preset scale factor, and and are greater than 0, Represents the logarithm of the base of a natural number.

[0058] As a further solution of the present invention: the accurate moisture content data and tea polyphenol content data are deeply processed, and according to the deep processing results, moisture content abnormality characteristics and tea polyphenol content abnormality characteristics are generated, and the generated moisture content abnormality characteristics and tea polyphenol content abnormality characteristics are comprehensively analyzed, specifically including:

[0059] Acquire the moisture content data and tea polyphenol content data of each tea leaf, and perform standardization on the moisture content data and tea polyphenol content data of each tea leaf to obtain standardized data;

[0060] The standardized data include: water content data and tea polyphenol content data;

[0061] The water content data and tea polyphenol content data are deeply extracted through the autoencoder to obtain low-dimensional feature representation;

[0062] A Bayesian network is used to establish a probabilistic dependency relationship between moisture content and tea polyphenols content, and the network is trained based on historical data. The structure of the Bayesian network is expressed as follows:

[0063] ;

[0064] In the formula, represents the prior distribution of moisture content, represents the conditional probability distribution of tea polyphenols content under given moisture content, represents the joint probability distribution of water content and tea polyphenols content, Indicates moisture content data, Indicates the content of tea polyphenols;

[0065] The posterior probability of moisture content data and tea polyphenol content data is calculated by Bayesian inference, and the calculation expression is:

[0066] ;

[0067] ;

[0068] in, Indicates the data collected in real time. Indicates the moisture content in a given data The posterior probability of Indicates that the content of tea polyphenols is in the given data The posterior probability of represents the prior distribution of real-time data, represents the conditional probability distribution of moisture content under given real-time data conditions, Represents the conditional probability distribution of tea polyphenols content under given real-time data conditions;

[0069] The moisture content abnormal characteristic index is calculated based on the posterior probability, and the calculation expression is:

[0070] ;

[0071] In the formula, Indicates the characteristic index of abnormal water content;

[0072] The abnormal characteristic index of tea polyphenols content is calculated according to the posterior probability, and the calculation expression is:

[0073] ;

[0074] In the formula, Indicates the abnormal characteristic index of tea polyphenols content;

[0075] The moisture content abnormal characteristic index and the tea polyphenols content abnormal characteristic index are jointly analyzed to obtain the comprehensive abnormal characteristic index, and the calculation expression is:

[0076] ;

[0077] In the formula, represents the comprehensive abnormal characteristic index;

[0078] The comprehensive abnormal characteristic index is compared with the preset threshold value to evaluate the quality of tea leaves.

[0079] As a further solution of the present invention: the method of judging whether the quality of tea leaves is qualified specifically includes:

[0080] It is determined whether the comprehensive abnormal characteristic index of the tea is greater than or equal to a preset threshold value. If so, it means that the quality of the corresponding tea is unqualified; if not, it means that the quality of the corresponding tea is qualified.

[0081] As a further solution of the present invention: the stability of the environmental humidity data and the eligibility of the tea quality are comprehensively analyzed to determine the degree of interference of the abnormal environmental humidity data on the tea quality, specifically including:

[0082] Obtain the humidity stability index and comprehensive abnormal characteristic index during the production process;

[0083] The humidity stability index and the comprehensive abnormal characteristic index are used as input features, and the random forest model is trained using historical data. The interference relationship between humidity anomaly and tea quality is learned through multiple decision trees.

[0084] Through the trained random forest model, the input feature humidity stability index and the comprehensive abnormal feature index are inferred to calculate the interference coefficient. The calculation expression is:

[0085] ;

[0086] In the formula, represents the number of decision trees, represents the total number of decision trees, Indicates The output of a decision tree, Represents the humidity stability index, represents the comprehensive abnormal characteristic index, represents the interference coefficient;

[0087] The construction process of the random forest model is as follows:

[0088] Based on the historical data set, construct a A random forest model of decision trees, where each decision tree Use different training subsets for training, and each decision tree randomly selects a part of the features for node splitting;

[0089] The construction process of each decision tree includes:

[0090] A subset is randomly selected from the training data set. When each node is split, a part of the features is randomly selected for splitting. The output of each decision tree is predicted using the average method, and the interference coefficient is calculated according to the calculation expression of the interference coefficient.

[0091] As a further solution of the present invention: the dynamic control of the ambient humidity according to the interference degree specifically includes:

[0092] Determine whether the interference coefficient of tea leaves in the production process is greater than or equal to a preset threshold. If so, it means that the corresponding humidity data affects the production of tea leaves, and humidity data control is performed, specifically including:

[0093] According to the error between the current interference coefficient and the preset threshold, the proportional control term is calculated, and the calculation expression is:

[0094] ;

[0095] In the formula, represents the proportional control term, represents the proportional gain, represents the interference coefficient, Indicates the preset interference coefficient threshold;

[0096] According to the historical accumulated error, the integral control value is calculated, and the calculation expression is:

[0097] ;

[0098] In the formula, represents the integral control value, represents the integral gain, A time variable representing historical data;

[0099] According to the system change rate, the differential control value is calculated, and the calculation expression is:

[0100] ;

[0101] in, represents the differential control value, represents the differential gain;

[0102] The proportional control item, integral control value and differential control value are summed to obtain the humidity data control coefficient in the production process, the mean humidity data of the current production process is multiplied by the control coefficient to obtain the humidity data correction value, and the humidity data is corrected according to the humidity data correction value.

[0103] Beneficial effects of the present invention:

[0104] (1) The present invention collects environmental humidity, tea moisture content and tea polyphenol content in real time. These devices can not only obtain data in a non-invasive manner, but also efficiently convert physical signals into electrical signals, ensuring the accuracy and timeliness of data transmission. By extracting features from the collected data, a low-dimensional representation is obtained. Further, a Bayesian network is used to construct a probabilistic dependency relationship between moisture content and tea polyphenol content. The model is trained based on historical data to generate an abnormal moisture content characteristic index and an abnormal tea polyphenol content characteristic index, and finally a comprehensive abnormal characteristic index that fully reflects the quality status of tea is formed. In order to ensure the best environmental conditions for tea production, the present invention also introduces a humidity data stability evaluation mechanism. By performing wavelet decomposition on the time series data of environmental humidity within the monitoring period, the system can separate low-frequency and high-frequency components, and calculate the humidity abnormal fluctuation coefficient and the change rate abnormal coefficient based on this, and then obtain the humidity stability index. It can not only effectively identify the abnormal fluctuation of environmental humidity, but also quantify its impact on tea growth and processing. Based on this, the system can quickly take measures to adjust the environmental humidity to the optimal level and minimize the negative impact caused by environmental changes. In summary, the present invention not only significantly improves the accuracy and efficiency of data collection, but also optimizes each link of tea production through intelligent data analysis and dynamic control mechanism, ensuring the quality consistency and reliability of each batch of tea.

[0105] (2) The present invention innovatively constructs an interference analysis and dynamic control module. The dynamic control module uses a random forest model to model the relationship between humidity stability and tea quality, and evaluates the potential impact of abnormal environmental humidity on tea quality by calculating the interference coefficient. Based on this intelligent evaluation result, the system can respond to environmental changes in real time and take rapid measures to adjust humidity levels to minimize adverse effects and ensure that tea is produced under optimal growth and processing conditions. In addition, the system combines a proportional-integral-differential control algorithm to accurately calculate the humidity data control coefficient and achieve refined management of humidity in the production environment. This intelligent dynamic control mechanism not only significantly reduces the need for manual intervention and reduces production costs, but also improves production efficiency. More importantly, with its powerful data analysis capabilities, the system provides decision support for tea growers and processing plants. BRIEF DESCRIPTION OF THE DRAWINGS

[0106] The present invention will be further described below in conjunction with the accompanying drawings.

[0107] Figure 1 It is a flow chart of an intelligent tea quality traceability system based on Internet of Things technology of the present invention. DETAILED DESCRIPTION

[0108] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0109] See also Figure 1 As shown, the present invention is an intelligent tea quality traceability system based on Internet of Things technology, comprising:

[0110] A data collection module, which collects environmental humidity data in real time during the tea production process, and collects moisture content data and tea polyphenol content data of all tea leaves in real time;

[0111] A data accuracy evaluation module, which compares and analyzes the real-time collected humidity data, water content data and tea polyphenol content data with historical data to evaluate the accuracy of the collected real-time data;

[0112] A humidity data stability evaluation module, wherein the humidity data stability evaluation module extracts accurate environmental humidity data according to the evaluation result, obtains a time series of environmental humidity data within a monitoring period based on the accurate environmental humidity data, and evaluates the stability of the environmental humidity data by analyzing the fluctuation range and change rate of the humidity data;

[0113] A tea quality judgment module, which is used to perform deep processing on accurate moisture content data and tea polyphenol content data, generate moisture content abnormality characteristics and tea polyphenol content abnormality characteristics according to the deep processing results, and perform comprehensive analysis on the generated moisture content abnormality characteristics and tea polyphenol content abnormality characteristics, and judge whether the tea quality is qualified according to the analysis results;

[0114] The interference analysis and dynamic control module comprehensively analyzes the stability of the environmental humidity data and the quality of the tea leaves, determines the degree of interference of the abnormal environmental humidity data on the quality of the tea leaves, and dynamically controls the environmental humidity according to the degree of interference.

[0115] In the data acquisition module, the environmental humidity data is collected in real time, and the moisture content data and tea polyphenol content data of all tea leaves are collected in real time, including:

[0116] During the tea production process, humidity sensors are deployed in the tea processing area to monitor the environmental humidity data of the tea production process in real time, and the environmental humidity data is converted into electrical signals and transmitted to the central data processing system;

[0117] The moisture content data of tea leaves is collected through a near-infrared spectrometer. The near-infrared spectrometer can quickly and accurately measure the moisture content of tea leaves without destroying the sample. It irradiates near-infrared light onto the tea sample and analyzes the reflected spectrum to determine the specific moisture content. The moisture content data is then converted into an electrical signal and transmitted to the central data processing system.

[0118] High performance liquid chromatography is usually used to monitor the tea polyphenol content of tea in real time. The tea extract is separated and tested to accurately determine the tea polyphenol content, and the tea polyphenol content data is converted into electrical signals and transmitted to the central data processing system.

[0119] In the data accuracy assessment module, the real-time collected humidity data, water content data and tea polyphenol content data are compared and analyzed with historical data to assess the accuracy of the collected real-time data, including:

[0120] Collect humidity data, water content data and tea polyphenol content data in real time to form a real-time data sequence, and extract the corresponding historical reference data sequence from the historical database;

[0121] The distance between each element in the real-time data sequence and the historical reference data sequence is calculated using the Euclidean distance formula to generate a distance matrix , where the matrix element calculation formula is:

[0122] ;

[0123] in, Indicates the first The value of the data point, represents the data collection point, Indicates the number of historical reference data sequences The value of the data point, Indicates the first The data point is compared with the historical reference data sequence The Euclidean distance between data points is Indicates the length of the real-time data sequence, Indicates the length of the historical data sequence;

[0124] Constructing cumulative distance matrix based on dynamic programming , indicating the distance from the starting point to the position The minimum matching distance, the recursive formula of the cumulative distance matrix is:

[0125] ;

[0126] In the formula, From the starting point to the location The minimum cumulative distance, Indicates the minimum value;

[0127] From the cumulative distance matrix, the backtracking algorithm Backtrack to , get the optimal matching path ;

[0128] in, represents the set of matching points of the optimal path, represents the length of the optimal path, is the index value in the optimal path, indicating the Matching points, Represents the first The index of the matching points, Represents the first historical data sequence in the optimal path The index of the matching points;

[0129] According to the optimal path, the total distance on the path is calculated. The calculation expression is:

[0130] ;

[0131] In the formula, Indicates the total distance;

[0132] Calculate the normalized maximum distance between the real-time data series and the historical data series. The calculation expression is:

[0133] ;

[0134] In the formula, represents the average value of the historical data series, represents the normalized maximum distance;

[0135] According to the total distance of the optimal path and the normalized maximum distance, the accuracy coefficient is calculated, and the calculation expression is:

[0136] ;

[0137] In the formula, represents the accuracy coefficient;

[0138] Compare the accuracy coefficient of real-time data with the preset threshold and extract accurate data;

[0139] The accuracy of the real-time data collected by the evaluation specifically includes:

[0140] Determine whether the accuracy coefficient of each real-time data is greater than or equal to a preset threshold. If so, it means that the corresponding data is accurate, and the data is extracted. If not, it means that the corresponding data is inaccurate, and the data is corrected.

[0141] It should be noted that the accuracy coefficient is used to evaluate the accuracy of the data being collected, and the larger the value of the accuracy coefficient is, the more accurate the corresponding data collection is.

[0142] In the humidity data stability evaluation module, accurate environmental humidity data is extracted according to the evaluation results. Based on the accurate environmental humidity data, a time series of environmental humidity data within a monitoring period is obtained. By analyzing the fluctuation range and change rate of the humidity data, the stability of the environmental humidity data is evaluated, including:

[0143] Obtain accurate environmental humidity data, analyze the abnormal fluctuation degree of humidity data according to the fluctuation range of humidity data, and calculate the humidity abnormal fluctuation coefficient;

[0144] Obtain accurate environmental humidity data, analyze the abnormal change rate of humidity data according to the change rate of humidity data, and calculate the humidity change rate abnormality coefficient;

[0145] The humidity abnormal fluctuation coefficient and the humidity change rate abnormal coefficient are normalized and calculated to obtain the humidity stability index.

[0146] Determine whether the humidity stability index is greater than or equal to a preset threshold value. If so, it means that the corresponding environmental humidity data is stable. If not, it means that the corresponding environmental humidity data is unstable.

[0147] The acquisition logic of the humidity stability index is:

[0148] Obtain the humidity time series data during the tea production process and use the wavelet decomposition method to transform the humidity data series , decomposed into low-frequency components and high-frequency components, the calculation expression is:

[0149] ;

[0150] in, Indicates The low-frequency component of each time acquisition point, Indicates The high-frequency component of each time acquisition point, Indicates Humidity data collected at each time point, Indicates the time collection point;

[0151] Based on high frequency components Calculate the wave energy, the calculation expression is:

[0152] ;

[0153] In the formula, Represents the energy of the high-frequency component of humidity data, Indicates the total number of time collection points;

[0154] The energy of the high frequency component of humidity data Normalized with the total energy of the humidity series, the humidity anomaly fluctuation coefficient is obtained, and the calculation expression is:

[0155] ;

[0156] In the formula, Indicates the humidity abnormal fluctuation coefficient;

[0157] Calculate the rate of change of humidity , calculate the mean humidity change rate and standard deviation ;

[0158] Indicates Humidity change rate at each time collection point;

[0159] Calculate the humidity change rate anomaly coefficient, the calculation expression is:

[0160] ;

[0161] In the formula, Represents the humidity change rate anomaly coefficient, Indicates the maximum absolute value of the humidity change rate;

[0162] The humidity abnormal fluctuation coefficient and the humidity change rate abnormal coefficient are comprehensively normalized and calculated to obtain the humidity stability index. The calculation expression is:

[0163] ;

[0164] In the formula, Represents the humidity stability index, and represents a preset scale factor, and and are greater than 0, Represents the logarithm of the base of a natural number.

[0165] It should be noted that the humidity stability index reflects whether the humidity control of the tea production environment is stable during the tea production process. Stable humidity control can improve the quality of tea, and the larger the value of the humidity stability index, the smaller the corresponding impact on the quality of tea.

[0166] In the tea quality judgment module, the accurate moisture content data and tea polyphenol content data are deeply processed, and according to the deep processing results, the moisture content abnormality characteristics and tea polyphenol content abnormality characteristics are generated, and the generated moisture content abnormality characteristics and tea polyphenol content abnormality characteristics are comprehensively analyzed. According to the analysis results, it is judged whether the tea quality is qualified, which specifically includes:

[0167] Acquire the moisture content data and tea polyphenol content data of each tea leaf, and perform standardization on the moisture content data and tea polyphenol content data of each tea leaf to obtain standardized data;

[0168] The standardized data include: water content data and tea polyphenol content data;

[0169] The water content data and tea polyphenol content data are deeply extracted through the autoencoder to obtain low-dimensional feature representation;

[0170] A Bayesian network is used to establish a probabilistic dependency relationship between moisture content and tea polyphenols content, and the network is trained based on historical data. The structure of the Bayesian network is expressed as follows:

[0171] ;

[0172] In the formula, represents the prior distribution of moisture content, represents the conditional probability distribution of tea polyphenols content under given moisture content, represents the joint probability distribution of water content and tea polyphenols content, Indicates moisture content data, Indicates the content of tea polyphenols;

[0173] The posterior probability of moisture content data and tea polyphenol content data is calculated by Bayesian inference, and the calculation expression is:

[0174] ;

[0175] ;

[0176] in, Indicates the data collected in real time. Indicates the moisture content in a given data The posterior probability of Indicates that the content of tea polyphenols is in the given data The posterior probability of represents the prior distribution of real-time data, represents the conditional probability distribution of moisture content under given real-time data conditions, Represents the conditional probability distribution of tea polyphenols content under given real-time data conditions;

[0177] The moisture content abnormal characteristic index is calculated based on the posterior probability, and the calculation expression is:

[0178] ;

[0179] In the formula, Indicates the characteristic index of abnormal water content;

[0180] The abnormal characteristic index of tea polyphenols content is calculated according to the posterior probability, and the calculation expression is:

[0181] ;

[0182] In the formula, Indicates the abnormal characteristic index of tea polyphenols content;

[0183] The moisture content abnormal characteristic index and the tea polyphenols content abnormal characteristic index are jointly analyzed to obtain the comprehensive abnormal characteristic index, and the calculation expression is:

[0184] ;

[0185] In the formula, represents the comprehensive abnormal characteristic index;

[0186] The comprehensive abnormal characteristic index is compared with the preset threshold value to evaluate the quality of tea leaves;

[0187] Determine whether the comprehensive abnormal characteristic index of the tea is greater than or equal to a preset threshold value. If so, it means that the quality of the corresponding tea is unqualified; if not, it means that the quality of the corresponding tea is qualified;

[0188] It should be noted that the comprehensive abnormal characteristic index reflects the quality of tea, and the larger the value of the comprehensive abnormal characteristic index is, the worse the corresponding tea quality is.

[0189] In the interference analysis and dynamic control module, the stability of the environmental humidity data and the quality of the tea are comprehensively analyzed to determine the degree of interference of the abnormal environmental humidity data on the quality of the tea. According to the degree of interference, the environmental humidity is dynamically controlled, including:

[0190] Obtain the humidity stability index and comprehensive abnormal characteristic index during the production process;

[0191] The humidity stability index and the comprehensive abnormal characteristic index are used as input features, and the random forest model is trained using historical data. The interference relationship between humidity anomaly and tea quality is learned through multiple decision trees.

[0192] Through the trained random forest model, the input feature humidity stability index and the comprehensive abnormal feature index are inferred to calculate the interference coefficient. The calculation expression is:

[0193] ;

[0194] In the formula, represents the number of decision trees, represents the total number of decision trees, Indicates The output of a decision tree, Represents the humidity stability index, represents the comprehensive abnormal characteristic index, represents the interference coefficient;

[0195] The construction process of the random forest model is as follows:

[0196] Based on the historical data set, construct a A random forest model of decision trees, where each decision tree Use different training subsets for training, and each decision tree randomly selects a part of the features for node splitting;

[0197] The construction process of each decision tree includes:

[0198] A subset is randomly selected from the training data set. When each node is split, a portion of the features are randomly selected for splitting. The output of each decision tree is predicted using the average method. The interference is calculated based on the calculation expression of the interference coefficient.

[0199] Determine whether the interference coefficient of tea leaves in the production process is greater than or equal to a preset threshold. If so, it means that the corresponding humidity data affects the production of tea leaves, and humidity data control is performed, specifically including:

[0200] According to the error between the current interference coefficient and the preset threshold, the proportional control term is calculated, and the calculation expression is:

[0201] ;

[0202] In the formula, represents the proportional control term, represents the proportional gain, represents the interference coefficient, Indicates the preset interference coefficient threshold;

[0203] According to the historical accumulated error, the integral control value is calculated, and the calculation expression is:

[0204] ;

[0205] In the formula, represents the integral control value, represents the integral gain, A time variable representing historical data;

[0206] According to the system change rate, the differential control value is calculated, and the calculation expression is:

[0207] ;

[0208] in, represents the differential control value, represents the differential gain;

[0209] The proportional control item, integral control value and differential control value are summed to obtain the humidity data control coefficient in the production process, the mean humidity data of the current production process is multiplied by the control coefficient to obtain the humidity data correction value, and the humidity data is corrected according to the humidity data correction value.

[0210] It should be noted that the interference coefficient reflects the degree of influence of humidity data in the tea production process on the quality of tea, and the larger the value of the interference coefficient, the greater the influence of the corresponding tea quality on the environmental humidity data.

[0211] The working principle of the present invention is as follows: through humidity sensors and near-infrared spectrometers, environmental humidity data, tea moisture content and tea polyphenol content are collected in real time, and these information are converted into electrical signals and transmitted to the central processing system for further processing. In the data accuracy evaluation module, the system uses the Euclidean distance formula to calculate the difference between the real-time data sequence and the historical reference data sequence, so as to evaluate the accuracy of the data and extract accurate data; in the humidity data stability evaluation module, the humidity data fluctuation range and change rate are analyzed by wavelet decomposition method, and the humidity stability index is calculated to evaluate the stability of environmental humidity. The tea quality judgment module uses Bayesian network to perform deep feature extraction and probability dependency modeling on the standardized moisture content and tea polyphenol content data, and generates a comprehensive abnormal feature index for judging the quality of tea. In the interference analysis and dynamic control module, the system constructs a random forest model, calculates the interference coefficient according to the comprehensive analysis results of humidity stability and tea quality, and then determines whether it is necessary to dynamically adjust the environmental humidity to reduce adverse effects and ensure environmental humidity control in tea production. The present invention can not only effectively monitor and optimize environmental factors in the tea production process, but also improve the quality consistency of the final product, providing an efficient and intelligent quality management and traceability solution for the tea industry.

[0212] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0213] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. 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 computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0214] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0215] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0216] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An intelligent tea quality traceability system based on Internet of Things technology, characterized in that: include: A data collection module, which collects environmental humidity data in real time during the tea production process, and collects moisture content data and tea polyphenol content data of all tea leaves in real time; A data accuracy evaluation module, which compares and analyzes the real-time collected humidity data, water content data and tea polyphenol content data with historical data to evaluate the accuracy of the collected real-time data; A humidity data stability evaluation module, wherein the humidity data stability evaluation module extracts accurate environmental humidity data according to the evaluation result, obtains a time series of environmental humidity data within a monitoring period based on the accurate environmental humidity data, and evaluates the stability of the environmental humidity data by analyzing the fluctuation range and change rate of the humidity data; The stability of the environmental humidity data is evaluated by analyzing the fluctuation range and change rate of the humidity data, specifically including: Obtain accurate environmental humidity data, analyze the abnormal fluctuation degree of humidity data according to the fluctuation range of humidity data, and calculate the humidity abnormal fluctuation coefficient; Obtain accurate environmental humidity data, analyze the abnormal change rate of humidity data according to the change rate of humidity data, and calculate the humidity change rate abnormality coefficient; The humidity abnormal fluctuation coefficient and the humidity change rate abnormal coefficient are normalized and calculated to obtain the humidity stability index. Determine whether the humidity stability index is greater than or equal to a preset threshold value. If so, it means that the corresponding environmental humidity data is stable. If not, it means that the corresponding environmental humidity data is unstable. A tea quality judgment module, which is used to perform deep processing on accurate moisture content data and tea polyphenol content data, generate moisture content abnormality characteristics and tea polyphenol content abnormality characteristics according to the deep processing results, and perform comprehensive analysis on the generated moisture content abnormality characteristics and tea polyphenol content abnormality characteristics, and judge whether the tea quality is qualified according to the analysis results; The interference analysis and dynamic control module comprehensively analyzes the stability of the environmental humidity data and the quality of the tea leaves, determines the degree of interference of the abnormal environmental humidity data on the quality of the tea leaves, and dynamically controls the environmental humidity according to the degree of interference.

2. According to claim 1, the intelligent tea quality traceability system based on Internet of Things technology is characterized in that: The real-time collected humidity data, water content data and tea polyphenol content data are compared and analyzed with historical data to evaluate the accuracy of the collected real-time data, specifically including: Collect humidity data, water content data and tea polyphenol content data in real time to form a real-time data sequence, and extract the corresponding historical reference data sequence from the historical database; The distance between each element in the real-time data sequence and the historical reference data sequence is calculated using the Euclidean distance formula to generate a distance matrix , where the matrix element calculation formula is: ； in, Indicates the first The value of the data point, represents the data collection point, Indicates the number of historical reference data sequences The value of the data point, Indicates the first The data point is compared with the historical reference data sequence The Euclidean distance between data points, Indicates the length of the real-time data sequence, Indicates the length of the historical data sequence; Constructing cumulative distance matrix based on dynamic programming , indicating the distance from the starting point to the position The minimum matching distance, the recursive formula of the cumulative distance matrix is: ； In the formula, From the starting point to the location The minimum cumulative distance, Indicates the minimum value; From the cumulative distance matrix, the backtracking algorithm Backtrack to , get the optimal matching path ; in, represents the set of matching points of the optimal path, represents the length of the optimal path, is the index value in the optimal path, indicating the Matching points, Represents the first The index of the matching points, Represents the first historical data sequence in the optimal path The index of the matching points; According to the optimal path, the total distance on the path is calculated. The calculation expression is: ； In the formula, Indicates the total distance; Calculate the normalized maximum distance between the real-time data series and the historical data series. The calculation expression is: ； In the formula, represents the average value of the historical data series, represents the normalized maximum distance; According to the total distance of the optimal path and the normalized maximum distance, the accuracy coefficient is calculated, and the calculation expression is: ； In the formula, represents the accuracy coefficient; Compare the accuracy coefficient of real-time data with the preset threshold and extract accurate data.

3. According to the intelligent tea quality traceability system of the Internet of Things technology of claim 1, it is characterized in that: The accuracy of the real-time data collected by the evaluation specifically includes: Determine whether the accuracy coefficient of each real-time data is greater than or equal to a preset threshold. If so, it means that the corresponding data is accurate, and then the data is extracted. If not, it means that the corresponding data is inaccurate, and then the data is corrected; Wherein, the real-time data includes: humidity data, water content data and tea polyphenols content data.

4. The intelligent tea quality traceability system based on Internet of Things technology according to claim 1 is characterized in that: The acquisition logic of the humidity stability index is: Obtain the humidity time series data during the tea production process and use the wavelet decomposition method to transform the humidity data series , decomposed into low-frequency components and high-frequency components, the calculation expression is: ； in, Indicates The low-frequency component of each time acquisition point, Indicates The high-frequency component of each time acquisition point, Indicates Humidity data collected at each time point, Indicates the time collection point; Based on high frequency components Calculate the wave energy, the calculation expression is: ； In the formula, Represents the energy of the high-frequency component of humidity data, Indicates the total number of time collection points; The energy of the high frequency component of humidity data Normalized with the total energy of the humidity series, the humidity anomaly fluctuation coefficient is obtained, and the calculation expression is: ； In the formula, It indicates the humidity abnormal fluctuation coefficient; Calculate the rate of change of humidity , calculate the mean humidity change rate and standard deviation ; Indicates Humidity change rate at each time collection point; Calculate the humidity change rate anomaly coefficient, the calculation expression is: ； In the formula, Represents the humidity change rate anomaly coefficient, Indicates the maximum absolute value of the humidity change rate; The humidity abnormal fluctuation coefficient and the humidity change rate abnormal coefficient are comprehensively normalized and calculated to obtain the humidity stability index. The calculation expression is: ； In the formula, Represents the humidity stability index, and represents a preset scale factor, and and are greater than 0, Represents the logarithm of the base of a natural number.

5. The intelligent tea quality traceability system based on Internet of Things technology according to claim 1 is characterized in that: The accurate moisture content data and tea polyphenol content data are deeply processed, and according to the deep processing results, moisture content abnormality characteristics and tea polyphenol content abnormality characteristics are generated, and the generated moisture content abnormality characteristics and tea polyphenol content abnormality characteristics are comprehensively analyzed, specifically including: Acquire the moisture content data and tea polyphenol content data of each tea leaf, and perform standardization on the moisture content data and tea polyphenol content data of each tea leaf to obtain standardized data; The standardized data include: water content data and tea polyphenol content data; The water content data and tea polyphenol content data are deeply extracted through the autoencoder to obtain low-dimensional feature representation; A Bayesian network is used to establish a probabilistic dependency relationship between moisture content and tea polyphenols content, and the network is trained based on historical data. The structure of the Bayesian network is expressed as follows: ； In the formula, represents the prior distribution of moisture content, represents the conditional probability distribution of tea polyphenols content under given moisture content, represents the joint probability distribution of water content and tea polyphenols content, Indicates moisture content data, Indicates the content of tea polyphenols; The posterior probability of moisture content data and tea polyphenol content data is calculated by Bayesian inference, and the calculation expression is: ； ； in, Indicates the data collected in real time. Indicates the moisture content in a given data The posterior probability of Indicates that the content of tea polyphenols is in the given data The posterior probability of represents the prior distribution of real-time data, represents the conditional probability distribution of moisture content under given real-time data conditions, Represents the conditional probability distribution of tea polyphenols content under given real-time data conditions; The moisture content abnormal characteristic index is calculated based on the posterior probability, and the calculation expression is: ； In the formula, Indicates the characteristic index of abnormal water content; The abnormal characteristic index of tea polyphenols content is calculated according to the posterior probability, and the calculation expression is: ； In the formula, Indicates the abnormal characteristic index of tea polyphenols content; The moisture content abnormal characteristic index and the tea polyphenols content abnormal characteristic index are jointly analyzed to obtain the comprehensive abnormal characteristic index, and the calculation expression is: ； In the formula, represents the comprehensive abnormal characteristic index; The comprehensive abnormal characteristic index is compared with the preset threshold value to evaluate the quality of tea leaves.

6. The intelligent tea quality traceability system based on Internet of Things technology according to claim 1 is characterized in that: The determination of whether the tea quality is qualified specifically includes: It is determined whether the comprehensive abnormal characteristic index of the tea is greater than or equal to a preset threshold value. If so, it means that the quality of the corresponding tea is unqualified; if not, it means that the quality of the corresponding tea is qualified.

7. The intelligent tea quality traceability system based on Internet of Things technology according to claim 1 is characterized in that: The stability of the environmental humidity data and the eligibility of the tea quality are comprehensively analyzed to determine the degree of interference of the abnormal environmental humidity data on the tea quality, specifically including: Obtain the humidity stability index and comprehensive abnormal characteristic index during the production process; The humidity stability index and the comprehensive abnormal characteristic index are used as input features, and the random forest model is trained using historical data. The interference relationship between humidity anomaly and tea quality is learned through multiple decision trees. Through the trained random forest model, the input feature humidity stability index and the comprehensive abnormal feature index are inferred to calculate the interference coefficient. The calculation expression is: ； In the formula, represents the number of decision trees, represents the total number of decision trees, Indicates The output of a decision tree, Represents the humidity stability index, represents the comprehensive abnormal characteristic index, represents the interference coefficient; The construction process of the random forest model is as follows: Based on the historical data set, construct a A random forest model of decision trees, where each decision tree Use different training subsets for training, and each decision tree randomly selects a part of the features for node splitting; The construction process of each decision tree includes: A subset is randomly selected from the training data set. When each node is split, a part of the features is randomly selected for splitting. The output of each decision tree is predicted using the average method, and the interference coefficient is calculated according to the calculation expression of the interference coefficient.

8. The intelligent tea quality traceability system based on Internet of Things technology according to claim 1 is characterized in that: The dynamic control of the ambient humidity according to the interference degree specifically includes: Determine whether the interference coefficient of tea leaves in the production process is greater than or equal to a preset threshold. If so, it means that the corresponding humidity data affects the production of tea leaves, and humidity data control is performed, specifically including: According to the error between the current interference coefficient and the preset threshold, the proportional control term is calculated, and the calculation expression is: ； In the formula, represents the proportional control term, represents the proportional gain, represents the interference coefficient, Indicates the preset interference coefficient threshold; According to the historical accumulated error, the integral control value is calculated, and the calculation expression is: ； In the formula, represents the integral control value, represents the integral gain, A time variable representing historical data; According to the system change rate, the differential control value is calculated, and the calculation expression is: ； in, represents the differential control value, represents the differential gain; The proportional control item, integral control value and differential control value are summed to obtain the humidity data control coefficient in the production process, the mean humidity data of the current production process is multiplied by the control coefficient to obtain the humidity data correction value, and the humidity data is corrected according to the humidity data correction value.

Images