Food production information traceability management method and system

Abstract

The invention provides a food production information traceability management method and system, and relates to the technical field of food production monitoring. The method comprises the steps of obtaining multiple groups of multi-process production record data of target baked food, performing production process characteristic linkage analysis on the target baked food, and generating a plurality of process joint characteristic matrixes; performing mode clustering based on production process distance perception on the multiple groups of multi-process production record data to generate a plurality of abnormal production mode class clusters; performing process anomaly analysis and back propagation traceability on the to-be-analyzed batch to generate a process anomaly detection result of the to-be-analyzed batch; according to the plurality of production abnormity mode class clusters, mode abnormity tracing is carried out on the to-be-analyzed batch, a production abnormity mode matching result of the to-be-analyzed batch is generated, and a production abnormity tracing analysis result of the to-be-analyzed batch is generated through fusion. According to the invention, the positioning precision of the abnormal source process of the baked food is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of food production monitoring, and in particular to a food production information traceability management method and system. Background Art

[0002] In the food production industry, especially in the production process of baked goods such as bread and biscuits, tracing quality anomalies is a key link in ensuring food quality and reducing losses.

[0003] The production process of baked goods involves multiple links and processes, such as raw material preparation, fermentation, mixing, baking, cooling and packaging. The quality control of each link may have a significant impact on the subsequent processes. If the mutual influence between the processes is ignored in the process of tracing the production information, and the scheme that only relies on parameter threshold judgment, such as temperature over-limit alarm, is used for abnormal analysis, although it can well identify the dominant abnormality, it cannot track the propagation path of the abnormality in the process chain. For example, when a batch of bread has a problem of insufficient volume, this method may only detect the abnormal temperature in the baking stage, but ignore the indirect impact that changes in temperature and humidity during the fermentation process may have on the control of baking temperature, resulting in failure to identify the fundamental cause of insufficient gluten formation in the previous mixing process.

[0004] Moreover, the causes of abnormalities between different batches have certain similarities and trends. If the common characteristics between different production batches can be comprehensively analyzed, it will help to identify potential quality problems and better locate the source factors based on the trend changes in the overall production process parameters, which will facilitate the optimization of the production process and thus improve the production quality of baked goods. Summary of the invention

[0005] The present invention proposes a food production information traceability management method and system, aiming to solve at least one technical problem existing in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions: A food production information traceability management method, comprising: Obtain the multi-process production record data corresponding to multiple historical production batches of the target baked food, perform deviation discretization and trend discretization processing respectively, obtain the deviation characteristic data and trend characteristic data of each group of multi-process production record data, and construct the deviation trend joint features of multiple groups of multi-process production record data for each production process; Divide multiple groups of multi-process production record data into normal sample data and abnormal sample data, conduct linkage analysis of production process characteristics of target baked goods based on the joint characteristics of deviation trends of normal sample data, and generate a process joint feature matrix between any two production processes; According to the deviation trend joint features, multiple groups of multi-process production record data are clustered based on production process distance perception to generate multiple production abnormal pattern clusters; According to the deviation trend joint feature and process joint feature matrix, process anomaly analysis and back propagation traceability are performed on the target baked food batch to be analyzed, and process anomaly detection results of the batch to be analyzed are generated; The pattern anomaly traceability of the batch to be analyzed is performed based on multiple production anomaly pattern clusters, and the production anomaly pattern matching results of the batch to be analyzed are generated. The results are then integrated with the process anomaly detection results to generate the production anomaly traceability analysis results of the batch to be analyzed.

[0007] Preferably, the production process characteristics of the target baked food are linked and analyzed to generate a process joint feature matrix between any two production processes, including: Determine multiple local combination features of any production process according to the deviation trend joint feature, the local combination feature includes the deviation discrete feature and trend discrete feature of the production parameters in the deviation characteristic data and the trend characteristic data; The frequency parameters corresponding to multiple local combination features of each production process are counted, and the process joint parameters for each joint feature pair between any two production processes are calculated, including determining the joint co-occurrence parameter of each joint feature pair, and determining the downstream reference frequency of the joint feature pair according to the production sequence between the two production processes associated with the joint feature pair, recording the ratio of the joint co-occurrence parameter of the joint feature pair to the downstream reference frequency as the process joint parameter of the joint feature pair, and generating a process joint feature matrix between the two production processes according to the process joint parameters corresponding to multiple joint feature pairs.

[0008] Preferably, multiple groups of multi-process production record data are clustered based on production process distance perception according to the deviation trend joint features to generate multiple production abnormality pattern clusters, including: Vectorized processing is performed on each group of multi-process production record data according to the deviation trend joint feature, including feature scarcity analysis of multiple local combination features of the multi-process production record data in each production process, calculating the global scarcity parameter of each local combination feature, generating the joint weight of the local combination feature according to the global scarcity parameter and the frequency parameter, and constructing a joint weight feature vector for each group of multi-process production record data; According to multiple joint weight feature vectors, multiple groups of multi-process production record data are clustered based on production process distance perception, including introducing production process distance perception weights for correction in the process of calculating the distance between any two groups of multi-process production record data according to the joint weight feature vector to calculate the comprehensive distance parameter between any two groups of multi-process production record data, and using a target clustering algorithm to cluster the multiple groups of multi-process production record data according to the comprehensive distance parameter to obtain multiple production abnormality pattern clusters.

[0009] Preferably, according to the deviation trend joint feature and the process joint feature matrix, the process anomaly analysis and back propagation traceability are performed on the target baked food batch to be analyzed, including: Determine the target production record data of the batch to be analyzed, perform process anomaly analysis on the target production record data according to the frequency parameters corresponding to each local combination feature in the deviation trend joint feature, mark the abnormal process of the target production record data, determine the multiple process joint feature matrices associated with the abnormal process, perform back propagation tracing on the abnormal process, calculate the back propagation matching parameters corresponding to the abnormal process with respect to multiple upstream processes, and generate the process anomaly tracing analysis results of the batch to be analyzed.

[0010] Preferably, back propagation tracing is performed on the abnormal process to calculate the back propagation matching parameters of the abnormal process corresponding to multiple upstream processes, including: According to the deviation trend joint characteristics, trend transfer analysis is performed on multiple local combination features in the abnormal process to determine the abnormal features in the abnormal process. According to the process joint feature matrix between the abnormal process and the upstream process, the process joint parameters between the abnormal features and multiple local combination features of the upstream process are determined, and the back propagation matching parameters of the abnormal process with respect to the upstream process are calculated.

[0011] Preferably, a DBSCAN clustering algorithm is used to cluster multiple groups of multi-process production record data according to the comprehensive distance parameter to obtain multiple production abnormality pattern clusters.

[0012] A food production information traceability management system, comprising: The data joint feature extraction module is used to obtain the multi-process production record data corresponding to multiple historical production batches of the target baked food, perform deviation discretization and trend discretization processing respectively, obtain the deviation characteristic data and trend characteristic data of each group of multi-process production record data, and construct the deviation trend joint features of multiple groups of multi-process production record data for each production process; The process joint analysis module is used to divide multiple groups of multi-process production record data into normal sample data and abnormal sample data, and conduct production process characteristic linkage analysis on the target baked food according to the deviation trend joint characteristics of the normal sample data, and generate the process joint characteristic matrix between any two production processes; The production abnormality pattern clustering module is used to perform pattern clustering based on production process distance perception on multiple groups of multi-process production record data according to the joint characteristics of deviation trends, and generate multiple production abnormality pattern clusters; The process anomaly detection module is used to perform process anomaly analysis and back-propagation traceability on the target baked food batch to be analyzed based on the deviation trend joint feature and the process joint feature matrix, and generate the process anomaly detection result of the batch to be analyzed; The production anomaly traceability analysis module is used to trace the pattern anomaly of the batch to be analyzed based on multiple production anomaly pattern clusters, generate the production anomaly pattern matching results of the batch to be analyzed, and integrate them with the process anomaly detection results to generate the production anomaly traceability analysis results of the batch to be analyzed.

[0013] The present invention has the following beneficial effects: The present invention constructs a deviation trend joint feature and a process joint feature matrix by discretely processing multi-process production record data, quantifies the parameter linkage relationship between upstream and downstream production processes, and realizes implicit anomaly detection through back-propagation calculation of anomaly contribution, so as to identify weak coordinated offsets of parameters between multiple processes; performs pattern recognition on multiple groups of multi-process production record data from the perspective of feature scarcity, and amplifies the contribution of key process differences through process perception distance correction, extracts the changing relationship of implicit anomalies in the entire production process from a large amount of sample data, integrates anomaly tracing detection from different angles, and combines individual anomaly conduction paths with group common features to identify complex anomalies, which can well capture phenomena such as process drift and sudden anomalies in production data, thereby improving the accuracy of anomaly tracing in baked food production. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 The figure is a flow chart of an exemplary method for food production information traceability management according to the present invention.

[0015] Figure 2 The figure is a schematic diagram of the structure of an exemplary food production information traceability management system of the present invention. DETAILED DESCRIPTION

[0016] In order to make those skilled in the art better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0017] See also Figure 1 , a food production information traceability management method provided by an embodiment of the present invention specifically includes the following contents: Step S10: Obtain multi-process production record data corresponding to multiple historical production batches of the target baked food, perform deviation discretization and trend discretization processing respectively, obtain deviation characteristic data and trend characteristic data for each group of multi-process production record data, and construct deviation trend joint features of multiple groups of multi-process production record data for each production process.

[0018] It should be noted that the original data involves the multi-process production record data corresponding to different production batches of a baked food, including the production-related information involved in different production processes, such as the relevant time series record data of production parameters such as temperature, humidity, and speed in the mixing stage. Deviation discretization and trend discretization processing are performed on these data to extract feature information of different granularities. Deviation discretization can calculate the mean level of production parameters in different time periods in each production process, and the deviation between it and the expected value such as a preset reference value or a mean determined by historical data, and discretize it into deviation discrete features such as higher than normal level, lower than normal level and in line with normal level, reflecting the degree of fluctuation of parameters in each process. Trend discretization can calculate the change trend of production parameters in different time periods in each production process, and discretize it into trend discrete features such as rising, falling or stable trends, reflecting the stability of parameter changes in each process during the production process. Thus, the deviation characteristic data and trend characteristic data of each group of multi-process production record data are obtained, and the deviation discrete features and trend discrete features corresponding to the production parameters in each process are integrated to obtain the deviation trend joint features of each production process.

[0019] Step S20: divide the multiple groups of multi-process production record data into normal sample data and abnormal sample data, perform production process characteristic linkage analysis on the target baked food according to the deviation trend joint characteristics of the normal sample data, and generate a process joint feature matrix between any two production processes.

[0020] It should be noted that, based on the above extraction of the deviation trend joint features, the data of multiple production batches are divided into two categories, namely normal sample data and abnormal sample data, to represent the difference in data changes between the normal production process and the abnormal production process. Then, the joint features of the deviation trend of the normal sample data are used to conduct linkage analysis of the production process characteristics, which is specifically used to identify the relationship between the production processes and to achieve modeling in a way that the characteristics of these processes jointly affect the product quality.

[0021] In one of the implementation processes, the process of generating a process joint feature matrix by conducting a linkage analysis of the production process characteristics of the target baked food includes: According to the deviation trend joint feature, multiple local combination features of any production process are determined, and the local combination features include the deviation discrete features and trend discrete features of the production parameters in the deviation characteristic data and trend characteristic data. For example, for the mixing stage, the deviation discrete features and trend discrete features of parameters such as mixing speed, material humidity, and material temperature in a specific period of time are combined to obtain the local combination features corresponding to the overall multiple production parameters in the period, such as the characteristic levels of the temperature of the ingredients at the initial mixing stage and the mixing speed of the equipment.

[0022] Then, the frequency parameters corresponding to multiple local combination features of each production process are counted, and the process joint parameters for each joint feature pair between any two production processes are calculated based on the frequency parameters corresponding to different local combination features.

[0023] In this process, the joint co-occurrence parameter of each joint feature pair is first determined. Taking any two processes as an example, such as the stirring stage and the fermentation stage, a local combination feature corresponding to each of them, such as the parameter-related features of the late stirring stage and the middle fermentation stage, constitutes a feature pair. Then, according to the frequency parameter of the local combination feature, the joint co-occurrence parameter of the feature combination pair, that is, the frequency of co-occurrence, is counted in the multi-process production record data corresponding to multiple production batches. According to the production sequence between the two production processes associated with the joint feature pair, the downstream reference frequency of the joint feature pair is determined, that is, the frequency parameter corresponding to the local combination feature belonging to the downstream process in the two production processes involved in the joint feature pair. The ratio of the joint co-occurrence parameter of the joint feature pair to the downstream reference frequency is recorded as the process joint parameter of the joint feature pair, which characterizes whether the downstream process is affected by the upstream process. Through the process joint parameters of multiple joint feature pairs, the relationship strength and interaction between the processes are quantified. Finally, the process joint feature matrix between the two production processes is constructed to reflect the correlation characteristics between the upstream and downstream processes.

[0024] Step S30: performing pattern clustering based on production process distance perception on multiple groups of multi-process production record data according to the deviation trend joint features to generate multiple production abnormality pattern clusters.

[0025] It should be noted that in the process of using the deviation trend joint feature to cluster the patterns of multiple groups of production record data, the normal samples and abnormal samples are specifically analyzed jointly. The normal samples constitute the main distribution of the data, and the abnormal samples exist as outliers or subclusters that deviate from the main distribution. Considering that some abnormalities may appear as a combination of deviations from normal parameters, such as normal temperature but abnormal humidity and speed, it is necessary to identify them by comparing with normal samples to discover hidden abnormal models. At the same time, the distribution changes of normal samples, such as gradual shifts, may indicate subsequent production risks. Therefore, they are observed in a unified data space to achieve drift detection of the production process. In the process of analyzing the distance between different production batches, the distance perception of the production process is introduced to amplify the impact of the key production process on the whole. At the same time, the matching between different batches is analyzed considering the timing characteristics of the process flow, and finally the clustering of multiple groups of multi-process production record data is realized to generate multiple production abnormality pattern clusters.

[0026] In one implementation process, the above-mentioned pattern clustering based on production process distance perception is performed on multiple groups of multi-process production record data according to the deviation trend joint features to generate multiple production abnormal pattern clusters, including the following: Each group of multi-process production record data is vectorized according to the joint characteristics of the deviation trend, and a joint weight feature vector of each group of multi-process production record data is constructed.

[0027] Specifically, we first perform feature scarcity analysis on multiple local combination features of multi-process production record data in each production process, and calculate the global scarcity parameter of each local combination feature: In the formula, Represents a local combination feature in one of the production processes. Indicates the total number of batches, that is, the total number of production batches involved in multiple groups of multi-process production record data. Indicates that local combination features appear in multiple production batches in this production process The number of batches.

[0028] After the global scarce parameters of the local combination features are calculated in the above manner, the joint weights of the local combination features are generated according to the global scarce parameters and the frequency parameters. Specifically, the characteristic frequency of the local combination features is calculated by the frequency parameters, that is, the ratio between the frequency parameters of the local combination features and the total frequency of multiple local combination features in the production process, and then the product between the characteristic frequency of the local combination features and the global scarce parameters is used as the joint weight of the local combination features. In the process of calculating the joint weight, a large number of conventional combination features are suppressed and some key combination features are enhanced. Specifically, the normal samples in the multi-process production record data are taken as the data subject, and some of the hidden anomalies and local drift scenarios of the parameters are mined. Finally, the local combination features with low joint weights represent some hidden abnormal trend features in the data, which are more likely to be key indicators involving abnormal production. Low joint weights indicate that the combination features are common and conform to the overall normal change law in the actual production process. The joint weight feature vector constructed in this way can well highlight the hidden abnormal patterns in a large number of historical sample data, such as highlighting the abnormal parameter deviation phenomenon and abnormal change trend in the production process.

[0029] Finally, multiple groups of multi-process production record data are clustered based on production process distance perception according to multiple joint weight feature vectors.

[0030] Specifically, in the process of calculating the distance between any two sets of multi-process production record data according to the joint weight feature vector, the production process distance perception weight is introduced for correction to calculate the comprehensive distance parameter between any two sets of multi-process production record data. For the production process distance perception weight: In the formula Indicates production batch and production batches The distance perception weight between production processes, Indicates production batch and production batches In the The matching parameters between the joint weight feature vectors under each production process are: Indicates The maximum length of the joint weight feature vector under production processes, Indicates The importance weight of each production process, Indicates the number of production operations.

[0031] It is worth mentioning that the distance-aware weight of the production process is used to measure the differences between different production batches, including amplifying the contribution of key production processes to the whole. It is specifically optimized through the importance weight of the production process. The importance weight can be reasonably set based on the frequency of anomalies caused by specific production processes in historical data or combined with expert experience. The matching parameter is used to measure the matching degree of combined features between joint weight feature vectors. For example, due to the difference in production batches, one of the batches has additional local combined features in a certain production process, resulting in a difference in the number of combined features in the corresponding joint weight feature vectors. Data filling is required to maintain a consistent sequence degree. The matching parameter can specifically be the number of local combined features shared by the two parameters. The conventional distance between the two joint weight feature vectors, such as the Euclidean distance, is corrected through the distance-aware weight of the production process. The calculated comprehensive distance parameter between the multi-process production record data can better fit the actual production and more comprehensively characterize the differences between different production batches.

[0032] Finally, a target clustering algorithm is adopted according to the comprehensive distance parameter, such as a density-based or distance-based algorithm. In this embodiment, the DBSCAN clustering algorithm is taken as an example, and the comprehensive distance parameter is used as the distance between the multi-process production record data. The DBSCAN clustering algorithm is used to cluster multiple groups of multi-process production record data to obtain multiple production abnormality pattern clusters, each cluster representing one or more similar production abnormality patterns. For example, a production abnormality pattern cluster corresponds to a related production batch where production abnormalities occur due to too low stirring humidity, and by suppressing common signals and amplifying implicit abnormal signals, normal patterns with similar change trends to abnormal patterns are integrated into the cluster, which can assist in locating the root process and identifying the abnormal change pattern of the overall production process caused by a certain root abnormal point.

[0033] Step S40: Perform process anomaly analysis and back propagation tracing on the target baked food batch to be analyzed according to the deviation trend joint feature and the process joint feature matrix, and generate a process anomaly detection result for the batch to be analyzed.

[0034] It should be noted that by analyzing the target production record data of the target batch of baked goods to be analyzed, including first performing process anomaly analysis based on the joint characteristics of the deviation trend, identifying the deviation or change trend of the data from the normal level, and determining the abnormal process in the batch to be analyzed, and then analyzing the correlation between it and multiple upstream production processes based on the process joint feature matrix, tracing the source through back propagation, locating the root cause of the anomaly, and generating the process anomaly detection result of the batch to be analyzed.

[0035] In this process, according to the frequency parameters corresponding to each local combination feature in the deviation trend joint feature, the process anomaly analysis is performed on the target production record data, and some special local combination features in the target production record data are identified, that is, the combination features that appear with a small probability in the normal sample data, and the corresponding production processes are marked as abnormal processes.

[0036] Then, multiple process joint feature matrices associated with the abnormal process are determined, that is, the process joint feature matrix between the abnormal process and multiple upstream processes, and the abnormal process is back-propagated and traced according to the process joint feature matrix, including calculating the back-propagation matching parameters between the abnormal process and each upstream process. The degree of influence of the upstream process on the abnormal process is measured by the back-propagation matching parameters, thereby generating the process abnormality traceability analysis results of the batch to be analyzed.

[0037] In this process, multiple local combination features in the abnormal process are first determined, and trend transfer analysis is performed on each local combination feature according to the deviation trend joint feature to determine the abnormal features in the abnormal process. Specifically, the frequency of occurrence of each local combination feature in the abnormal process is counted, and the frequency of occurrence is compared with the reference frequency determined by the deviation trend joint feature of the normal sample data. If it is significantly higher than the normal level, it means that the local combination feature is an abnormal feature, that is, there is a more significant abnormality compared with the historical law. Then, the back propagation matching parameters between the abnormal process and each upstream process are calculated according to the abnormal features. Specifically, through the corresponding process joint feature matrix, the cumulative value corresponding to the process joint parameters of the abnormal feature and the multiple combination features in the upstream process is counted as the back propagation matching parameter between the abnormal feature and one of the upstream processes, which is used to characterize the overall correlation between the abnormal feature and the upstream process. If in a certain upstream process, some local combination features and the abnormal feature have a high frequency of occurrence, that is, these local combination features in the upstream process will have a greater probability of causing the abnormal feature to appear in the abnormal process with a greater probability, then it means that the upstream process is the source process of the abnormal feature, and the abnormal feature in the abnormal process is abnormal, and there is a greater correlation with a certain upstream process. In this way, the back propagation matching parameters between the abnormal process and each upstream process are calculated to quantitatively characterize the impact of different upstream processes on the abnormal process.

[0038] Step S50: trace the pattern anomaly of the batch to be analyzed according to multiple production anomaly pattern clusters, generate production anomaly pattern matching results for the batch to be analyzed, and merge them with the process anomaly detection results to generate production anomaly tracing analysis results for the batch to be analyzed.

[0039] It should be noted that, through the target production record data of the batch to be analyzed and the results of vectorizing multiple groups of multi-process production record data in the aforementioned steps, a joint weight feature vector corresponding to the target production record data of the batch to be analyzed is generated, and then the joint weight feature vector corresponding to the target production record data is matched with multiple production anomaly pattern clusters, including calculating the distance between the joint weight feature vector corresponding to the target production record data and the cluster center of each production anomaly pattern cluster to determine the production anomaly pattern cluster with the closest distance, and obtain the production anomaly pattern matching result of the batch to be analyzed.

[0040] In the above process, the process anomaly detection results are generated from the perspective of upstream and downstream influences between production processes. This process may result in the deviation degree or change trend of some parameters not being very significant due to weak signals, resulting in the possibility that the source process may not be accurately locked, but a production process between the real source process and the abnormal process is targeted, or the abnormal process may appear due to the combined effect of minor anomalies in multiple processes. The production anomaly pattern cluster analyzes the parameter change law between the entire process in the entire production process, and at the same time integrates the influence law brought by the source process in the abnormal sample, which can better lock the root cause process, and can integrate the production anomaly pattern matching results with the process anomaly detection results. Combined with the historical group law, the joint influence characteristics between independent processes are corrected, and the matching between each production process and the abnormal process is comprehensively considered. It can more comprehensively locate and explain the quality problems in production, realize high-precision source process positioning, and provide the overall correlation and influence relationship data of multiple processes between the source and the current process, which is convenient for optimizing the production process and improving the production quality of baked goods.

[0041] Based on the same inventive concept, the embodiment of the present invention provides a food production information traceability management system, see Figure 2 , the system comprises: The data joint feature extraction module is used to obtain the multi-process production record data corresponding to multiple historical production batches of the target baked food, perform deviation discretization and trend discretization processing respectively, obtain the deviation characteristic data and trend characteristic data of each group of multi-process production record data, and construct the deviation trend joint features of multiple groups of multi-process production record data for each production process; The process joint analysis module is used to divide multiple groups of multi-process production record data into normal sample data and abnormal sample data, and conduct production process characteristic linkage analysis on the target baked food according to the deviation trend joint characteristics of the normal sample data, and generate the process joint characteristic matrix between any two production processes; It includes determining a plurality of local combination features of any production process according to the deviation trend joint feature, the local combination feature including the deviation discrete feature and trend discrete feature of the production parameter in the deviation characteristic data and the trend characteristic data; The frequency parameters corresponding to multiple local combination features of each production process are counted, and the process joint parameters for each joint feature pair between any two production processes are calculated, including determining the joint co-occurrence parameter of each joint feature pair, and determining the downstream reference frequency of the joint feature pair according to the production sequence between the two production processes associated with the joint feature pair, recording the ratio of the joint co-occurrence parameter of the joint feature pair to the downstream reference frequency as the process joint parameter of the joint feature pair, and generating a process joint feature matrix between the two production processes according to the process joint parameters corresponding to multiple joint feature pairs.

[0042] The production abnormality pattern clustering module is used to perform pattern clustering based on production process distance perception on multiple groups of multi-process production record data according to the joint characteristics of deviation trends, and generate multiple production abnormality pattern clusters; The method includes performing vectorization processing on each group of multi-process production record data according to the deviation trend joint feature, performing feature scarcity analysis on multiple local combination features of the multi-process production record data in each production process, calculating the global scarcity parameter of each local combination feature, generating the joint weight of the local combination feature according to the global scarcity parameter and the frequency parameter, and constructing a joint weight feature vector for each group of multi-process production record data; According to multiple joint weight feature vectors, multiple groups of multi-process production record data are clustered based on production process distance perception, including introducing production process distance perception weights for correction in the process of calculating the distance between any two groups of multi-process production record data according to the joint weight feature vector to calculate the comprehensive distance parameter between any two groups of multi-process production record data, and using a target clustering algorithm to cluster the multiple groups of multi-process production record data according to the comprehensive distance parameter to obtain multiple production abnormality pattern clusters.

[0043] The process anomaly detection module is used to perform process anomaly analysis and back-propagation traceability on the target baked food batch to be analyzed based on the deviation trend joint feature and the process joint feature matrix, and generate the process anomaly detection result of the batch to be analyzed; The production anomaly traceability analysis module is used to trace the pattern anomaly of the batch to be analyzed based on multiple production anomaly pattern clusters, generate the production anomaly pattern matching results of the batch to be analyzed, and integrate them with the process anomaly detection results to generate the production anomaly traceability analysis results of the batch to be analyzed.

[0044] The above is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. Parts not described in detail in this specification belong to the prior art known to those skilled in the art.

Claims

1. A food production information traceability management method, characterized in that: include: Obtain the multi-process production record data corresponding to multiple historical production batches of the target baked food, perform deviation discretization and trend discretization processing respectively, obtain the deviation characteristic data and trend characteristic data of each group of multi-process production record data, and construct the deviation trend joint features of multiple groups of multi-process production record data for each production process; Divide multiple groups of multi-process production record data into normal sample data and abnormal sample data, conduct linkage analysis of production process characteristics of target baked goods based on the joint characteristics of deviation trends of normal sample data, and generate a process joint feature matrix between any two production processes; According to the deviation trend joint features, multiple groups of multi-process production record data are clustered based on production process distance perception to generate multiple production abnormal pattern clusters; According to the deviation trend joint feature and process joint feature matrix, process anomaly analysis and back propagation traceability are performed on the target baked food batch to be analyzed, and process anomaly detection results of the batch to be analyzed are generated; The pattern anomaly traceability of the batch to be analyzed is performed based on multiple production anomaly pattern clusters, and the production anomaly pattern matching results of the batch to be analyzed are generated. The results are then integrated with the process anomaly detection results to generate the production anomaly traceability analysis results of the batch to be analyzed.

2. A food production information traceability management method according to claim 1, characterized in that: Perform linkage analysis on the production process characteristics of the target baked goods and generate a joint feature matrix between any two production processes, including: Determine multiple local combination features of any production process according to the deviation trend joint feature, the local combination feature includes the deviation discrete feature and trend discrete feature of the production parameters in the deviation characteristic data and the trend characteristic data; The frequency parameters corresponding to multiple local combination features of each production process are counted, and the process joint parameters for each joint feature pair between any two production processes are calculated, including determining the joint co-occurrence parameter of each joint feature pair, and determining the downstream reference frequency of the joint feature pair according to the production sequence between the two production processes associated with the joint feature pair, recording the ratio of the joint co-occurrence parameter of the joint feature pair to the downstream reference frequency as the process joint parameter of the joint feature pair, and generating a process joint feature matrix between the two production processes according to the process joint parameters corresponding to multiple joint feature pairs.

3. A food production information traceability management method according to claim 1, characterized in that: According to the deviation trend joint features, multiple groups of multi-process production record data are clustered based on production process distance perception to generate multiple production abnormality pattern clusters, including: Vectorized processing is performed on each group of multi-process production record data according to the deviation trend joint feature, including feature scarcity analysis of multiple local combination features of the multi-process production record data in each production process, calculating the global scarcity parameter of each local combination feature, generating the joint weight of the local combination feature according to the global scarcity parameter and the frequency parameter, and constructing a joint weight feature vector for each group of multi-process production record data; According to multiple joint weight feature vectors, multiple groups of multi-process production record data are clustered based on production process distance perception, including introducing production process distance perception weights for correction in the process of calculating the distance between any two groups of multi-process production record data according to the joint weight feature vector to calculate the comprehensive distance parameter between any two groups of multi-process production record data, and using a target clustering algorithm to cluster the multiple groups of multi-process production record data according to the comprehensive distance parameter to obtain multiple production abnormality pattern clusters.

4. A food production information traceability management method according to claim 3, characterized in that: According to the deviation trend joint feature and process joint feature matrix, process anomaly analysis and back propagation traceability are performed on the target baked food batch to be analyzed, including: Determine the target production record data of the batch to be analyzed, perform process anomaly analysis on the target production record data according to the frequency parameters corresponding to each local combination feature in the deviation trend joint feature, mark the abnormal process of the target production record data, determine the multiple process joint feature matrices associated with the abnormal process, perform back propagation tracing on the abnormal process, calculate the back propagation matching parameters corresponding to the abnormal process with respect to multiple upstream processes, and generate the process anomaly tracing analysis results of the batch to be analyzed.

5. A food production information traceability management method according to claim 4, characterized in that: Back propagation is performed on the abnormal process to trace its source, and the back propagation matching parameters corresponding to multiple upstream processes of the abnormal process are calculated, including: According to the deviation trend joint characteristics, trend transfer analysis is performed on multiple local combination features in the abnormal process to determine the abnormal features in the abnormal process. According to the process joint feature matrix between the abnormal process and the upstream process, the process joint parameters between the abnormal features and multiple local combination features of the upstream process are determined, and the back propagation matching parameters of the abnormal process with respect to the upstream process are calculated.

6. A food production information traceability management method according to claim 3, characterized in that: According to the comprehensive distance parameter, the DBSCAN clustering algorithm is used to cluster multiple groups of multi-process production record data to obtain multiple production abnormality pattern clusters.

7. A food production information traceability management system, characterized in that: The system is used to implement a food production information traceability management method as described in any one of claims 1 to 6, comprising: The data joint feature extraction module is used to obtain the multi-process production record data corresponding to multiple historical production batches of the target baked food, perform deviation discretization and trend discretization processing respectively, obtain the deviation characteristic data and trend characteristic data of each group of multi-process production record data, and construct the deviation trend joint features of multiple groups of multi-process production record data for each production process; The process joint analysis module is used to divide multiple groups of multi-process production record data into normal sample data and abnormal sample data, and conduct production process characteristic linkage analysis on the target baked food according to the deviation trend joint characteristics of the normal sample data, and generate the process joint characteristic matrix between any two production processes; The production abnormality pattern clustering module is used to perform pattern clustering based on production process distance perception on multiple groups of multi-process production record data according to the joint characteristics of deviation trends, and generate multiple production abnormality pattern clusters; The process anomaly detection module is used to perform process anomaly analysis and back-propagation traceability on the target baked food batch to be analyzed based on the deviation trend joint feature and the process joint feature matrix, and generate the process anomaly detection result of the batch to be analyzed; The production anomaly traceability analysis module is used to trace the pattern anomaly of the batch to be analyzed based on multiple production anomaly pattern clusters, generate the production anomaly pattern matching results of the batch to be analyzed, and integrate them with the process anomaly detection results to generate the production anomaly traceability analysis results of the batch to be analyzed.

Images