Pu'er tea leaf production process monitoring method and system

Abstract

The invention discloses a Pu'er tea leaf production process monitoring method and system, and relates to the field of production monitoring, the Pu'er tea leaf production process monitoring method comprises the following steps: step 1, based on a plurality of production links, deploying a plurality of image acquisition devices and temperature and humidity sensing acquisition devices for each production link correspondingly, presetting an acquisition frequency and an acquisition angle scheme, acquiring tea appearance images and temperature and humidity data, and synchronously performing local storage; step 2, presetting a standard trend of the tea appearance image and the temperature and humidity data under a corresponding production link, acquiring the tea appearance image and the temperature and humidity data which are acquired in real time, performing abnormity capture based on the standard trend, and outputting a captured abnormal point location; the method comprises the following steps: comparing a preset standard trend with data acquired in real time, identifying and capturing abnormal points, establishing association between each abnormal point and a previous link state, constructing an abnormal trend table, and performing reason analysis and corrective measure matching by utilizing traceability tracking.

Description

Technical Field

[0001] The invention relates to the technical field of production monitoring, and in particular to a method and system for monitoring the production process of Pu'er tea. Background Art

[0002] Pu'er tea is a fermented tea with a complex production process and is very sensitive to environmental factors. Therefore, it is very important to maintain its stable quality during the production process. The production process of Pu'er tea includes multiple links such as picking, withering, killing green, rolling, fermentation, and drying. Each link has a direct impact on the quality of tea. Therefore, each link needs to be carefully managed and monitored. Once a problem occurs in a link, it may lead to a decline in the quality of the entire product. With the development of sensor technology, image recognition technology, and big data analysis technology, modern production monitoring systems have begun to introduce more intelligent methods; In the prior art, tea quality monitoring often relies on regular testing, resulting in a long response time for identifying abnormalities. Abnormal problems may not be discovered in the first place, resulting in losses or increased quality fluctuations. In addition, there are usually multiple scattered monitoring points, and the integration and correlation analysis of the obtained monitoring data are insufficient, making it difficult to grasp the overall situation of tea quality monitoring as a whole, and lacking in-depth analysis of the causes of abnormalities. It is difficult to quickly adjust monitoring strategies and parameters according to changes in the production process, and there is a lack of adaptability, which greatly reduces the accuracy and efficiency of monitoring when facing new production processes or market requirements. The lack of an effective feedback mechanism makes it impossible to effectively record and summarize abnormalities in the production process, making it impossible to achieve effective improvement and optimization in subsequent production. Summary of the invention

[0003] 1. Technical issues to be resolved In view of the above-mentioned shortcomings of the prior art, the present invention provides a method and system for monitoring the production process of Pu'er tea, which can effectively solve the problems of the prior art.

[0004] (II) Technical solution To achieve the above objectives, the present invention is implemented through the following technical solutions: The invention discloses a method for monitoring the production process of Pu'er tea, comprising the following steps: Step 1: Based on several production links, deploy several image acquisition devices and temperature and humidity sensor acquisition devices to each production link, preset the acquisition frequency and acquisition angle scheme, collect the appearance image and temperature and humidity data of tea leaves, and store them locally simultaneously; Step 2: preset the standard trend of the tea appearance image and temperature and humidity data under the corresponding production link, obtain the real-time collection of tea appearance image and temperature and humidity data, and capture anomalies based on the standard trend, and output the captured abnormal points; Step 3: Associate each abnormal point with the status of its previous link and build an abnormal trend table; Step 4: Based on the current performance data of different abnormal points and the corresponding abnormal trend table, preset several level adjustment schemes of the acquisition frequency and angle of the image acquisition equipment and the temperature and humidity sensor acquisition equipment; Step 5: Obtain the current abnormal trend table and apply the corresponding level adjustment plan; Step 6: Re-collect and capture the abnormal points of the tea appearance image and temperature and humidity data after the adjustment scheme is applied, preset the standard threshold of the abnormal points, and when the threshold of the abnormal points obtained by re-collection exceeds the standard threshold, trigger the alarm information to be submitted to the management end, and update the abnormal trend table; Step 7: Take the abnormal points and abnormal trend table obtained in step 6 as input, and conduct source analysis on the abnormal causes; Step 8: Build a correction database and match the corrective measures for the output abnormal causes.

[0005] Furthermore, the exception capture process in step 2 is: Step 21: extracting color, texture and shape features from the collected tea appearance images, and converting the extracted image features into numerical values; Step 22: Standardize the collected temperature and humidity data to normalize the data range; Step 23: Merge the image features after numerical conversion and the standardized temperature and humidity data to obtain a feature matrix with samples as rows and features as columns; Step 24: Cluster the merged feature matrix using the K-means algorithm, and determine the center of each cluster and the distribution status of samples outside the center based on the current clustering results; Step 25: construct a distance metric of the samples within the cluster after standard clustering, formulate anomaly detection standards based on the distance metric, and judge whether the distance metric of the samples within the cluster after the current clustering meets the standards based on the anomaly detection standards; Step 26: If the judgment is no, then extract the abnormal data and its distribution in the feature space, construct and output the abnormal points; Step 27: If the judgment is yes, the extraction is terminated.

[0006] Furthermore, the adjustment scheme attributes in step 4 include: the rotation adjustment angle and duration of the image acquisition device generated based on the abnormal point and its associated image acquisition device in all directions, and the acquisition range and acquisition frequency of the temperature and humidity acquisition device generated based on the abnormal point and its associated temperature and humidity acquisition device.

[0007] Furthermore, the calculation formula for the abnormal cause tracing analysis in step 7 is: ; In the formula, Represents the impact of the jth abnormal cause c on all abnormal points, represents the i-th outlier point, represents the lth state record, represents the jth abnormal reason, Represents each outlier For all reasons The posterior probability under represents the correlation measurement between the anomaly point and the state, n represents the number of anomalies, and m represents the number of anomaly causes.

[0008] Furthermore, in the calculation formula of the abnormal cause tracing analysis, The calculation logic is: ; In the formula, represents the probability of observing the i-th abnormal point a under the j-th cause c, represents the prior probability of the jth cause c, Represents the prior probability of the i-th outlier point a.

[0009] Furthermore, in the calculation formula of the abnormal cause tracing analysis, Based on the abnormal trend table, time series data is formed, and the correlation between the trend change and the state of each abnormal point is calculated. The calculation logic is: ; In the formula, Represents the trend information of the i-th abnormal point in the abnormal trend table T.

[0010] Furthermore, the process of corrective measure matching in step 8 is: edit or read a number of corrective measures to build a corrective database, and assign a number of label measures to the stored corrective measures, identify the abnormal causes obtained by tracing, and label the abnormal causes, perform association matching in the corrective database based on the labels obtained by the processing, sort the matched corrective measures according to the number of hit label values, and submit the matched corrective measures sorted to the management end.

[0011] A Pu'er tea production process monitoring system, comprising: The data acquisition module is used to arrange image acquisition equipment and temperature and humidity sensors in various links of Pu'er tea production, operate according to the preset acquisition frequency and angle, and obtain the appearance image and temperature and humidity data of tea in real time; Standard setting module, used to set the standard trend of tea appearance, temperature and humidity at each stage; The abnormality capture module is used to compare the real-time acquired tea appearance images and temperature and humidity data with the established standard trend. When the data deviates from the standard trend, it will automatically capture the abnormal point and record the relevant information of the abnormal event; The trend analysis module is used to associate each captured abnormal point with the status of the previous link in the production process, establish an abnormal trend table, and record the occurrence time, data and corresponding status of the abnormal point; The strategy adjustment module is used to preset several acquisition frequency and angle strategies, and submit the corresponding strategies to the data acquisition module for implementation based on the performance of the currently monitored abnormal points; The threshold monitoring module is used to compare the data collected by the data collection module with the preset standard threshold of the re-collected data according to the adjusted collection strategy, and when the threshold of the re-collected data exceeds the standard threshold, the abnormal trend table is updated; The traceability analysis module is used to capture abnormal points and their corresponding abnormal trend tables, trace the source, and output the causes of the abnormalities; The measure matching module is used to build a database containing various corrective solutions and automatically match the corresponding corrective solutions based on the currently identified abnormal causes.

[0012] Furthermore, the threshold monitoring module is interactively connected to an alarm module via a wireless network, and the alarm module is used to receive a standard threshold judgment result of the threshold monitoring module, and when there is a threshold violation event, an alarm is triggered to report abnormal data to the management end.

[0013] Furthermore, the data acquisition module is interactively connected to the standard setting module and the threshold monitoring module through a wireless network, the standard setting module is interactively connected to the exception capture module through a wireless network, the exception capture module is interactively connected to the trend analysis module through a wireless network, the trend analysis module is interactively connected to the policy adjustment module through a wireless network, the threshold monitoring module is interactively connected to the source tracing analysis module through a wireless network, and the source tracing analysis module is interactively connected to the measure matching module through a wireless network.

[0014] (III) Beneficial effects Compared with the known prior art, the technical solution provided by the present invention has the following beneficial effects: By deploying image acquisition equipment and temperature and humidity sensors in each key production link, regularly collecting tea appearance images and environmental data, and updating and storing data in real time, the preset standard trend is compared with the real-time collected data to identify and capture abnormal points, and obtain accurate abnormal detection results, thereby accurately capturing abnormal situations, improving managers' sensitivity to quality issues, and enabling managers to make quick decisions.

[0015] By associating each abnormal point with the status of its previous link, building an abnormal trend table, and using traceability tracking to perform cause analysis, it is possible to associate abnormal events with production links and status, deeply analyze the causes of abnormalities, and provide a clear direction for subsequent improvements. By building a correction database to match the abnormal causes, the output abnormal causes are matched with corrective measures, forming a closed-loop feedback mechanism, which helps to continuously optimize the production process and improve product quality.

[0016] By analyzing the performance data of different abnormal points and adjusting the level plan according to the set level, the working parameters of the image acquisition equipment and sensors are changed in real time, and the acquisition frequency and angle are automatically adjusted according to the monitoring data, which is more adaptable to the dynamic changes in the production process, thereby improving the monitoring efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A schematic diagram of a process flow of a Pu'er tea production process monitoring method in the present invention; Figure 2 The figure is a schematic diagram of the framework of a Pu'er tea production process monitoring system in the present invention.

[0019] The numbers in the figure represent: 1. Data acquisition module; 2. Standard setting module; 3. Anomaly capture module; 4. Trend analysis module; 5. Strategy adjustment module; 6. Threshold monitoring module; 7. Source tracing analysis module; 8. Measures matching module; 9. Alarm module. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0021] The present invention will be further described below in conjunction with the embodiments.

[0022] Embodiment 1: A method for monitoring the production process of Pu'er tea leaves according to the present embodiment, as follows: Figure 1 As shown, the following steps are included: Step 1: Based on several production links, deploy several image acquisition devices and temperature and humidity sensor acquisition devices to each production link, preset the acquisition frequency and acquisition angle scheme, collect the appearance image and temperature and humidity data of tea leaves, and store them locally simultaneously; Step 2: preset the standard trend of the tea appearance image and temperature and humidity data under the corresponding production link, obtain the real-time collection of tea appearance image and temperature and humidity data, and capture anomalies based on the standard trend, and output the captured abnormal points; Step 3: Associate each abnormal point with the status of the previous link and build an abnormal trend table. The abnormal trend table helps analyze the history of abnormalities and their possible causes, provides data support for subsequent production optimization, and improves the efficiency and refined management of the production process. Step 4: Based on the current performance data of different abnormal points and the corresponding abnormal trend table, preset several levels of adjustment schemes for the acquisition frequency and angle of the image acquisition device and the temperature and humidity sensor acquisition device; the adjustment scheme attributes include: the rotation adjustment angle and duration of the image acquisition device and its associated image acquisition device based on the abnormal point, the acquisition range and acquisition frequency of the temperature and humidity acquisition device and its associated temperature and humidity acquisition device based on the abnormal point; it can implement targeted monitoring measures for the abnormal points to improve the accuracy and effectiveness of monitoring; Step 5: Obtain the current abnormal trend table and apply the corresponding level adjustment plan; implement the level adjustment plan to ensure that abnormal points are continuously monitored and further enhance the ability to respond to problems; Step 6: Re-collect and capture the abnormal points of the tea appearance image and temperature and humidity data after the adjustment scheme is applied, and preset the standard threshold of the abnormal points. When the threshold of the abnormal points obtained by re-collection exceeds the standard threshold, the alarm information is triggered and submitted to the management end, and the abnormal trend table is updated to ensure that important abnormalities will not be ignored, and the monitoring effect is fed back in time to ensure the safety of the production process; Step 7: Take the abnormal points and abnormal trend table obtained in step 6 as input, and conduct source analysis on the abnormal causes; First, calculate each outlier For all reasons The posterior probability under The calculation logic is: ; In the formula, represents the probability of observing the i-th abnormal point a under the j-th cause c, represents the prior probability of the jth cause c, Represents the prior probability of the ith abnormal point a; based on the abnormal trend table T, the time series data is formed, and the correlation between the trend change and the state of each abnormal point can be calculated. The process is: ; In the formula, Represents the trend information of the i-th abnormal point in the abnormal trend table T, which represents the value or characteristics of the abnormal point at a specific time point; finally, the trend data of the abnormal point and the state are combined to comprehensively calculate all the causes Corresponding to the impact of all abnormal points, the calculation formula for the final abnormal cause tracing analysis is: ; In the formula, Represents the impact of the jth abnormal cause c on all abnormal points, represents the i-th outlier point from the collected data set, Represents the lth status record, which is used to indicate the production status of the previous link. Indicates the jth abnormal cause, reflecting the factors that may cause the abnormality. Represents each outlier For all reasons The posterior probability under , reflects the possibility strength of the cause, represents the correlation measurement between the abnormal point and the state. The smaller the value, the higher the probability. n represents the number of abnormal points, which is the total number of all abnormal points captured during the monitoring process. Each abnormal point is identified by the appearance image and temperature and humidity data. m represents the number of abnormal reasons. The above formula can be used to systematically analyze and track the causes of abnormal points in the Pu'er tea production process. By combining historical status and abnormal trends, the potential causes can be quantified and a decision-making basis can be provided so that corresponding corrective measures can be taken. Step 8: Build a correction database and match the output abnormal causes with corrective measures. The corrective measure matching process is as follows: edit or read several corrective measures to build a correction database, and assign several label measures to the stored corrective measures, identify the abnormal causes obtained by tracing, and label the abnormal causes. Based on several processed labels, perform correlation matching in the correction database, sort the matched corrective measures by the number of hit label values, and submit the matched corrective measures to the management end. Through targeted corrective measures, the optimization level of the production process can be improved. Effective corrective measures can prevent the same abnormal events from happening again, thereby achieving continuous improvement.

[0023] Compared with the existing technology, through the comprehensive deployment of collection equipment, multi-dimensional monitoring of each link is realized, the richness and reliability of monitoring data are guaranteed, and the real-time data collection and abnormality capture mechanism enable problems in the production process to be discovered and handled in time, improving the production response capability. The abnormal trend analysis and level adjustment plan can be flexibly adjusted according to actual conditions, which helps to quickly adapt to new situations in the production process. Through mathematical calculations to analyze the causes of abnormalities, data support is provided, making the decision-making process more scientific and refined. By building a correction database, continuous improvement of the production process can be achieved, reducing the recurrence of similar problems, and improving product quality and production efficiency.

[0024] Embodiment 2: In other aspects, this embodiment also provides a Pu'er tea production process monitoring system, such as Figure 2 As shown, including: Data acquisition module 1 is used to arrange image acquisition equipment and temperature and humidity sensors in various links of Pu'er tea production, operate according to the preset acquisition frequency and angle, and obtain the appearance image and temperature and humidity data of tea in real time; Standard setting module 2, used to set the standard trend of tea appearance and temperature and humidity at each stage; The abnormality capture module 3 is used to compare the real-time acquired tea appearance image and temperature and humidity data with the established standard trend. When the data is detected to deviate from the standard trend, the abnormal point will be automatically captured and the relevant information of the abnormal event will be recorded; The trend analysis module 4 is used to associate each captured abnormal point with the status of the previous link in the production process, establish an abnormal trend table, and record the occurrence time, data and corresponding status of the abnormal point; The strategy adjustment module 5 is used to preset a number of acquisition frequency and angle strategies, and submit the corresponding strategies to the data acquisition module 1 for implementation based on the performance of the currently monitored abnormal points; The threshold monitoring module 6 is used to compare the data collected by the data collection module 1 with the preset standard threshold of the re-collected data according to the adjusted collection strategy. When the threshold of the re-collected data exceeds the standard threshold, the abnormal trend table is updated; the threshold monitoring module 6 is interactively connected to the alarm module 9 through a wireless network. The alarm module 9 is used to receive the standard threshold judgment result of the threshold monitoring module 6. When there is a threshold violation event, an alarm is triggered to report the abnormal data to the management end; The source tracing analysis module 7 is used to capture the abnormal points and their corresponding abnormal trend tables, trace the source, and output the causes of the abnormalities; The measure matching module 8 is used to build a database containing various correction solutions, and automatically match the corresponding correction solution based on the currently identified abnormal cause; The data acquisition module 1 is interactively connected to the standard setting module 2 and the threshold monitoring module 6 through a wireless network, the standard setting module 2 is interactively connected to the exception capture module 3 through a wireless network, the exception capture module 3 is interactively connected to the trend analysis module 4 through a wireless network, the trend analysis module 4 is interactively connected to the policy adjustment module 5 through a wireless network, the threshold monitoring module 6 is interactively connected to the source tracing analysis module 7 through a wireless network, and the source tracing analysis module 7 is interactively connected to the measure matching module 8 through a wireless network.

[0025] During the specific implementation of this embodiment, the data acquisition module 1 issues acquisition instructions to each pre-deployed image acquisition and temperature and humidity acquisition device to obtain the acquired data, sets the standard data through the standard setting module 2, compares the acquired data with the standard data through the exception capture module 3, and analyzes the acquired abnormal data through the trend analysis module 4. Based on the analysis results, the acquisition setting instructions of the data acquisition module 1 are adjusted through the strategy adjustment module 5, and the data acquisition module 1 is reused for acquisition. The acquired data is re-analyzed by the exception capture module 3 and the trend analysis module 4, and then submitted to the threshold monitoring module 6 for judgment. If the preset re-acquisition data threshold is exceeded, an alarm is first issued through the alarm module 9, and then the cause of the abnormality is traced back to the target through the source tracing analysis module 7. Based on the cause of the abnormality and the target, corrective measures are matched and provided through the measure matching module 8.

[0026] Embodiment 3: In this embodiment, a measure for abnormal capture is provided, and the specific process is as follows: Step 21: extracting color, texture and shape features from the collected tea appearance images, and converting the extracted image features into numerical values; Step 22: Standardize the collected temperature and humidity data to normalize the data range; Step 23: Merge the image features after numerical conversion and the standardized temperature and humidity data to obtain a feature matrix with samples as rows and features as columns; Step 24: Cluster the merged feature matrix using the K-means algorithm, and determine the center of each cluster and the distribution status of samples outside the center based on the current clustering results; Step 25: construct a distance metric of the samples within the cluster after standard clustering, formulate anomaly detection standards based on the distance metric, and judge whether the distance metric of the samples within the cluster after the current clustering meets the standards based on the anomaly detection standards; Step 26: If the judgment is no, then extract the abnormal data and its distribution in the feature space, construct and output the abnormal points; Step 27: If the judgment is yes, the extraction is terminated.

[0027] Compared with the existing technology, this method simultaneously extracts the color, texture and shape features of the image and combines them with temperature and humidity data to analyze the condition of tea from multiple dimensions, thereby improving the accuracy and comprehensiveness of anomaly detection. The standardized processing of temperature and humidity data eliminates the dimensional influence between different data sources, avoids misjudgment due to inconsistent data, and improves the reliability of the analysis results. Cluster analysis can quickly identify the central trend and abnormal characteristics of data, which improves the speed and efficiency of anomaly detection compared with traditional methods. By establishing a standard distance metric, the standard of anomalies can be scientifically defined, thus providing a more accurate basis for anomaly detection. Compared with simple threshold judgment methods, it is more flexible and accurate, providing the distribution of abnormal data in the feature space, which helps to analyze and understand the causes or trends of anomalies, and facilitates subsequent adjustments and improvements to the decision-making process.

[0028] In summary, the present invention collects the appearance data and production environment information of tea in real time through the deployment of image acquisition equipment and temperature and humidity sensors, realizes real-time monitoring of the production process, promptly discovers and handles abnormal situations, avoids the accumulation of quality defects, and captures abnormal points and trend changes by collecting and analyzing a large amount of historical data and real-time data, providing scientific basis and data support, helping managers make more reasonable decisions, and effectively optimizing the production process; By presetting standard trends and thresholds, comparing the actual collected tea quality data, and adjusting the production process in a timely manner to ensure the consistency of tea quality, the collection plan can be adjusted based on the performance data of abnormal points, which can more flexibly respond to new problems arising in the production process, and then implement targeted monitoring measures on abnormal points to improve the accuracy and effectiveness of monitoring. Through traceability tracking strategies, the relationship between abnormal points and related production status can be analyzed, and the potential causes of the abnormalities can be identified to ensure that the source of the problem can be quickly located, so as to take targeted improvement measures and reduce the risk of similar problems in the future.

[0029] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for monitoring the production process of Pu'er tea, characterized in that: The following steps are involved: Step 1: Based on several production links, deploy several image acquisition devices and temperature and humidity sensor acquisition devices to each production link, preset the acquisition frequency and acquisition angle scheme, collect the appearance image and temperature and humidity data of tea leaves, and store them locally simultaneously; Step 2: preset the standard trend of the tea appearance image and temperature and humidity data under the corresponding production link, obtain the real-time collection of tea appearance image and temperature and humidity data, and capture anomalies based on the standard trend, and output the captured abnormal points; Step 3: Associate each abnormal point with the status of its previous link and build an abnormal trend table; Step 4: Based on the current performance data of different abnormal points and the corresponding abnormal trend table, preset several level adjustment schemes of the acquisition frequency and angle of the image acquisition equipment and the temperature and humidity sensor acquisition equipment; Step 5: Obtain the current abnormal trend table and apply the corresponding level adjustment plan; Step 6: Recollect and capture the abnormal points of the tea appearance image and temperature and humidity data after the adjustment scheme is applied, preset the standard threshold of the abnormal points, and when the threshold of the abnormal points obtained by recollection exceeds the standard threshold, trigger the alarm information to be submitted to the management end, and update the abnormal trend table; Step 7: Take the abnormal points and abnormal trend table obtained in step 6 as input, and conduct source analysis on the abnormal causes; Step 8: Build a correction database and match the corrective measures for the output abnormal causes.

2. A method for monitoring the production process of Pu'er tea according to claim 1, characterized in that: The exception capture process in step 2 is as follows: Step 21: extracting color, texture and shape features from the collected tea appearance images, and converting the extracted image features into numerical values; Step 22: Standardize the collected temperature and humidity data to normalize the data range; Step 23: Merge the image features after numerical conversion and the standardized temperature and humidity data to obtain a feature matrix with samples as rows and features as columns; Step 24: Cluster the merged feature matrix using the K-means algorithm, and determine the center of each cluster and the distribution status of samples outside the center based on the current clustering results; Step 25: construct a distance metric of the samples within the cluster after standard clustering, formulate anomaly detection standards based on the distance metric, and judge whether the distance metric of the samples within the cluster after the current clustering meets the standards based on the anomaly detection standards; Step 26: If the judgment is no, then extract the abnormal data and its distribution in the feature space, construct and output the abnormal points; Step 27: If the judgment is yes, the extraction is terminated.

3. A Pu'er tea production process monitoring method according to claim 1, characterized in that: The adjustment scheme attributes in step 4 include: the rotation adjustment angle and duration of the image acquisition device generated based on the abnormal point and its associated image acquisition device in all directions, and the acquisition range and acquisition frequency of the temperature and humidity acquisition device generated based on the abnormal point and its associated temperature and humidity acquisition device.

4. A Pu'er tea production process monitoring method according to claim 1, characterized in that: The calculation formula for abnormal cause tracing analysis in step 7 is: ; In the formula, Represents the impact of the jth abnormal cause c on all abnormal points, represents the i-th outlier point, represents the lth state record, represents the jth abnormal reason, Represents each outlier For all reasons The posterior probability under represents the correlation measurement between the anomaly point and the state, n represents the number of anomalies, and m represents the number of anomaly causes.

5. A Pu'er tea production process monitoring method according to claim 4, characterized in that: In the calculation formula of the abnormal cause tracing analysis, The calculation logic is: ; In the formula, represents the probability of observing the i-th abnormal point a under the j-th cause c, represents the prior probability of the jth cause c, Represents the prior probability of the i-th outlier point a.

6. A method for monitoring the production process of Pu'er tea according to claim 4, characterized in that: In the calculation formula of the abnormal cause tracing analysis, Based on the abnormal trend table, time series data is formed, and the correlation between the trend change and the state of each abnormal point is calculated. The calculation logic is: ; In the formula, Represents the trend information of the i-th abnormal point in the abnormal trend table T.

7. A method for monitoring the production process of Pu'er tea according to claim 1, characterized in that: The process of corrective measure matching in step 8 is as follows: edit or read a number of corrective measures to construct a corrective database, assign a number of label measures to the stored corrective measures, identify the abnormal causes obtained by tracing, and label the abnormal causes, perform association matching in the corrective database based on the labels obtained by the processing, sort the matched corrective measures by the number of hit label values, and submit the sorted matched corrective measures to the management end.

8. A Pu'er tea production process monitoring system, the system is an implementation system of a Pu'er tea production process monitoring method according to any one of claims 1 to 7, characterized in that: include: The data acquisition module is used to arrange image acquisition equipment and temperature and humidity sensors in various links of Pu'er tea production, operate according to the preset acquisition frequency and angle, and obtain the appearance image and temperature and humidity data of tea in real time; Standard setting module, used to set the standard trend of tea appearance, temperature and humidity at each stage; The abnormality capture module is used to compare the real-time acquired tea appearance images and temperature and humidity data with the established standard trend. When the data is detected to deviate from the standard trend, the abnormal point will be automatically captured and the relevant information of the abnormal event will be recorded; The trend analysis module is used to associate each captured abnormal point with the status of the previous link in the production process, establish an abnormal trend table, and record the occurrence time, data and corresponding status of the abnormal point; The strategy adjustment module is used to preset several acquisition frequency and angle strategies, and submit the corresponding strategies to the data acquisition module for implementation based on the performance of the currently monitored abnormal points; The threshold monitoring module is used to compare the data collected by the data collection module with the preset standard threshold of the re-collected data according to the adjusted collection strategy, and when the threshold of the re-collected data exceeds the standard threshold, the abnormal trend table is updated; The traceability analysis module is used to capture abnormal points and their corresponding abnormal trend tables, trace the source, and output the causes of the abnormalities; The measure matching module is used to build a database containing various corrective solutions and automatically match the corresponding corrective solutions based on the currently identified abnormal causes.

9. A Pu'er tea production process monitoring system according to claim 8, characterized in that: The threshold monitoring module is interactively connected to the alarm module via a wireless network. The alarm module is used to receive the standard threshold judgment result of the threshold monitoring module. When there is a threshold violation event, an alarm is triggered to report abnormal data to the management end.

10. A Pu'er tea production process monitoring system according to claim 8, characterized in that: The data acquisition module is interactively connected to the standard setting module and the threshold monitoring module through a wireless network, the standard setting module is interactively connected to the exception capture module through a wireless network, the exception capture module is interactively connected to the trend analysis module through a wireless network, the trend analysis module is interactively connected to the policy adjustment module through a wireless network, the threshold monitoring module is interactively connected to the source tracing analysis module through a wireless network, and the source tracing analysis module is interactively connected to the measure matching module through a wireless network.

Images