Composite production line management method and system for food processing
Through dough image recognition and fermentation parameter optimization, combined with composite consistency analysis, the problem of difficulty in bread production consistency analysis is solved, the fine adjustment of the bread production line is achieved, and the production efficiency and quality are improved.
Patent Information
- Application Number
- CN202411446805.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-16
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-10-16
AI Technical Summary
Due to the large production process of existing bread composite production management methods, it is difficult to analyze the consistency of bread production, and it is impossible to carry out targeted adjustment and management of the bread production line, which affects production efficiency and production quality.
By collecting dough images for identification, the dough specification information and preliminary fermentation information are obtained, and the fermentation parameters are optimized; after fermentation, the fermentation images are collected, the fermentation specification information is identified, and the decorative placement parameters are optimized; after baking, the composite consistency analysis is performed, Combining the consistent parameters of cutting, fermentation and decorative placement, the production line management parameters are analyzed to achieve fine adjustment of the bread production line.
It improves the precision and accuracy of bread production consistency analysis, realizes accurate adjustment of bread production line, and improves production efficiency and production quality.
Smart Images

Figure CN119026893A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of production management, and in particular to a composite production line management method and system for food processing. Background Art
[0002] The bread compound production line is a bread production system that integrates multiple functions. By integrating the equipment and systems of the above processes, the continuity and automation of the bread production process can be realized. It has many advantages such as high degree of automation, high production efficiency and stable product quality.
[0003] At present, the existing composite bread production management methods make it difficult to analyze the consistency of bread production due to the large number of production processes, resulting in the inability to carry out targeted adjustment and management of the bread production line, which affects the technical problems of bread production efficiency and production quality. Summary of the invention
[0004] The purpose of this application is to provide a composite production line management method and system for food processing, in order to solve the technical problem that the existing composite production management method of bread has many production processes, making it difficult to analyze the consistency of bread production, resulting in the inability to carry out targeted adjustment and management of the bread production line, affecting the bread production efficiency and production quality.
[0005] In view of the above problems, the present application provides a composite production line management method and system for food processing.
[0006] In a first aspect, the present application provides a composite production line management method for food processing, which is implemented by a composite production line management system for food processing, wherein the method comprises: collecting multiple dough images of multiple doughs that have been divided, identifying them, and obtaining multiple dough specification information and multiple preliminary fermentation information; performing fermentation parameter analysis of final fermentation of dough according to the multiple dough specification information and multiple preliminary fermentation information, obtaining multiple final fermentation parameters, and performing fermentation; after fermentation is completed, collecting multiple fermented dough images of multiple doughs, identifying them, obtaining multiple fermentation specification information and multiple final fermentation information; performing decoration placement parameter analysis of fermented dough according to the multiple fermentation specification information and multiple final fermentation information, obtaining multiple decoration placement parameters, and performing decoration placement and baking; after baking is completed, collecting multiple baking images, performing production composite consistency analysis, and obtaining composite consistency parameters; performing consistency analysis of cutting, fermentation, and decoration placement according to multiple dough images, multiple final fermentation parameters, and multiple decoration placement parameters, respectively, obtaining cutting consistency parameters, fermentation consistency parameters, and decoration consistency parameters, combining the composite consistency parameters, analyzing to obtain production line management parameters, and performing discrimination management.
[0007] In a second aspect, the present application further provides a composite production line management system for food processing, which is used to execute a composite production line management method for food processing as described in the first aspect, wherein the system comprises: a dough image recognition module, which is used to collect multiple dough images of multiple doughs that have been divided, perform recognition, and obtain multiple dough specification information and multiple preliminary fermentation information; a fermentation parameter analysis module, which is used to perform fermentation parameter analysis of the final fermentation of the dough according to the multiple dough specification information and the multiple preliminary fermentation information, obtain multiple final fermentation parameters, and perform fermentation; a fermentation dough image recognition module, which is used to collect multiple fermentation dough images of multiple doughs after fermentation is completed, perform recognition, and obtain multiple fermentation specification information and multiple final fermentation parameters. Fermentation information; a decoration placement parameter analysis module, which is used to perform decoration placement parameter analysis of the fermented dough according to the multiple fermentation specification information and the multiple final fermentation information, obtain multiple decoration placement parameters, and perform decoration placement and baking; a production composite consistency analysis module, which is used to collect multiple baking images after baking is completed, perform production composite consistency analysis, and obtain composite consistency parameters; a production line management parameter discrimination module, which is used to perform consistency analysis of cutting, fermentation and decoration placement according to multiple dough images, multiple final fermentation parameters, and multiple decoration placement parameters, obtain cutting consistency parameters, fermentation consistency parameters and decoration consistency parameters, and combine the composite consistency parameters to analyze and obtain production line management parameters for discrimination management.
[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0009] By collecting multiple dough images of multiple doughs that have been segmented and identifying them, multiple dough specification information and multiple preliminary fermentation information are obtained; according to the multiple dough specification information and multiple preliminary fermentation information, fermentation parameter analysis of the final fermentation of the dough is performed respectively to obtain multiple final fermentation parameters, and fermentation is performed; after the fermentation is completed, multiple fermented dough images of the multiple doughs are collected and identified to obtain multiple fermentation specification information and multiple final fermentation information; according to the multiple fermentation specification information and multiple final fermentation information, decoration placement parameter analysis of the fermented dough is performed respectively to obtain multiple decoration placement parameters, and decoration placement and baking are performed; after the baking is completed, multiple baking images are collected to perform production composite consistency analysis to obtain composite consistency parameters; consistency analysis of cutting, fermentation and decoration placement is performed respectively according to the multiple dough images, multiple final fermentation parameters and multiple decoration placement parameters to obtain cutting consistency parameters, fermentation consistency parameters and decoration consistency parameters, and production line management parameters are obtained by analysis in combination with the composite consistency parameters, and discrimination management is performed. That is to say, by identifying the multiple dough images that have been segmented, multiple dough specification information and multiple preliminary fermentation information are obtained; then, the dough fermentation parameters are optimized and analyzed according to the multiple dough specification information and the multiple preliminary fermentation information, and multiple optimized final fermentation parameters are obtained for fermentation, which can improve the accuracy of the final fermentation parameter setting, thereby improving the dough fermentation quality; then, after the fermentation is completed, the multiple fermented dough images are identified to obtain multiple fermentation specification information and multiple final fermentation information; further, according to the multiple fermentation specification information and the multiple final fermentation information, the decoration placement parameters of the fermented dough are optimized and analyzed, and multiple optimized decoration placement parameters are obtained for decoration placement and baking, which can improve the consistency of the distribution of decorations on the bread, thereby improving the appearance quality of the bread; and after baking, the production composite consistency analysis is performed based on the multiple baking images to obtain the composite consistency parameters; on the other hand, the consistency analysis of cutting, fermentation and decoration placement is performed according to the multiple dough images, the multiple final fermentation parameters, and the multiple decoration placement parameters, and the production line management parameters are obtained based on the cutting consistency parameters, the fermentation consistency parameters, the decoration consistency parameters and the composite consistency parameters. When the production line management parameters do not meet the preset management parameter thresholds, the bread composite production line is adjusted and optimized. It can improve the precision and accuracy of bread production consistency analysis, achieve the technical goal of accurately adjusting the bread production line, and achieve the technical effect of improving bread production efficiency and production quality.
[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.
[0012] Figure 1 A schematic diagram of a process flow of a composite production line management method for food processing in this application;
[0013] Figure 2 A schematic diagram of a process for obtaining multiple dough specification information and multiple preliminary fermentation information in a composite production line management method for food processing in the present application;
[0014] Figure 3 This is a structural schematic diagram of a composite production line management system for food processing in the present application.
[0015] Description of reference numerals:
[0016] Dough image recognition module 11, fermentation parameter analysis module 12, fermented dough image recognition module 13, decoration placement parameter analysis module 14, production composite consistency analysis module 15, production line management parameter determination module 16. DETAILED DESCRIPTION
[0017] This application provides a composite production line management method and system for food processing, which solves the technical problem that the existing composite bread production management method has many production processes, making it difficult to analyze the consistency of bread production, resulting in the inability to conduct targeted regulation and management of the bread production line, affecting the efficiency and quality of bread production. The precision and accuracy of bread production consistency analysis can be improved, the technical goal of accurately adjusting the bread production line can be achieved, and the technical effect of improving the efficiency and quality of bread production can be achieved.
[0018] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.
[0019] Embodiment 1
[0020] Please refer to the attached Figure 1 The present application provides a composite production line management method for food processing, wherein the method is applied to a composite production line management system for food processing, and the method specifically comprises the following steps:
[0021] Step 1: collecting multiple dough images of the multiple doughs that have been segmented, performing recognition, and obtaining multiple dough specification information and multiple preliminary fermentation information;
[0022] Specifically, firstly, an image sensing device is used to collect images of multiple doughs that have been divided, wherein dough division refers to dividing the dough into multiple small doughs by a divider after the initial fermentation of the dough is completed; then a dough image recognizer is constructed based on a convolutional neural network to perform feature recognition on multiple dough images to obtain multiple dough specification information and multiple preliminary fermentation information, wherein the dough specification information includes data such as dough shape and size, and the preliminary fermentation information includes the degree of preliminary fermentation of the dough, which can be obtained by analyzing the number and size of pores in the dough image, wherein the number and size of pores are proportional to the degree of preliminary fermentation of the dough, that is, the more pores on the dough and the larger the pore size, the better the degree of preliminary fermentation of the dough.
[0023] Step 2: According to the plurality of dough specification information and the plurality of preliminary fermentation information, respectively analyze the fermentation parameters of the final fermentation of the dough, obtain a plurality of final fermentation parameters, and perform fermentation;
[0024] Specifically, based on the multiple dough specification information and multiple preliminary fermentation information, the fermentation parameters for the final fermentation of the dough are optimized respectively, wherein the final fermentation is a key step in the bread-making process. After the dough has undergone the initial fermentation, it will be divided, shaped, and placed in a baking tray or other baking container, and then undergo final fermentation. The purpose of the final fermentation is to allow the dough to expand again, increase the volume of the bread, and improve its taste and texture. During the final fermentation process, the yeast will continue to consume the sugar in the dough, produce carbon dioxide and alcohol, thereby causing the dough to expand and produce a unique aroma; wherein the fermentation parameters include fermentation control parameters such as fermentation temperature, fermentation humidity, and fermentation time; based on the optimization results, multiple final fermentation parameters are obtained, and fermentation control of the corresponding dough is performed according to the multiple final fermentation parameters.
[0025] By optimizing the final fermentation parameters of the dough based on the dough specification information and the preliminary fermentation information, the optimized final fermentation parameters are obtained to control the fermentation of the dough. This can improve the adaptability of the final fermentation parameters to the current state of the dough, making the final fermentation parameter settings more accurate, thereby improving the dough fermentation quality.
[0026] Step 3: After fermentation is completed, multiple fermented dough images of multiple doughs are collected and identified to obtain multiple fermentation specification information and multiple final fermentation information;
[0027] Specifically, after the dough fermentation is completed, an image sensor device is used to collect images of multiple doughs after fermentation to obtain multiple fermented dough images; then a fermented dough image recognizer is constructed based on a convolutional neural network, and the fermented dough image recognizer includes an input layer, a convolutional layer, a pooling layer, a fully connected layer and an output layer, wherein the input data of the input layer is the fermented dough image, and the output data is the fermentation specification information (fermentation shape, fermentation size, etc.) and the final fermentation information (volume expansion ratio, dough elasticity, dough softness); then a sample fermented dough image set, a sample fermentation specification information set and a sample final fermentation information set are collected to perform supervised training on the fermented dough image recognizer, wherein the training process is as follows: first, the sample fermented dough image is used as input and input into the input layer of the convolutional neural network; then the output layer of the model outputs the predicted value of the fermentation specification information and the preliminary fermentation information; further, the mean square error is used to evaluate the error between the predicted value of the model and the true value (sample fermentation specification information, sample final fermentation information); then the network weight is updated by a back propagation algorithm to continuously reduce the error between the predicted value and the true value, and the model is continuously trained using the same method to obtain a fermented dough image recognizer that meets the expected convergence conditions.
[0028] Then, the fermented dough image recognizer performs feature recognition on multiple fermented dough images to obtain multiple fermentation specification information and multiple final fermentation information, wherein the fermentation specification information includes information such as the shape and size of the dough after final fermentation; and the final fermentation information includes information such as the fermentation degree of the dough, volume expansion ratio, dough elasticity, and dough softness.
[0029] Step 4: Analyze the decoration placement parameters of the fermented dough according to the multiple fermentation specification information and the multiple final fermentation information, obtain multiple decoration placement parameters, and place and bake the decorations;
[0030] Specifically, according to the multiple fermentation specification information and the multiple final fermentation information, the decoration placement parameters of the fermented dough are optimized respectively, wherein the decoration refers to the decorative food on the bread, such as sesame, nuts, egg liquid, meat floss, etc., which can be set according to actual conditions, and the decoration placement parameters include the decoration placement area on the bread, the decoration spacing distance, etc., wherein, due to the different shapes, sizes and fermentation degrees of the fermented dough, the distribution area of the expanded bread decorations after baking may also be different, affecting the consistency of the bread product. Therefore, the decoration placement parameters are optimized according to the fermentation specification information and the final fermentation information, so that the distribution of the decorations on the baked bread is as consistent as possible, for example: the decorations are placed in the same position area on the bread; multiple decoration placement parameters are obtained, and the decorations are placed on the fermented dough according to the multiple decoration placement parameters, and the fermented dough with the decorations placed is baked.
[0031] By optimizing the decoration placement parameters of the fermented dough according to the fermentation specification information and the final fermentation information, the adaptability of the decoration placement parameters to the fermented dough state can be improved, making the decoration placement parameters more accurate, thereby improving the distribution consistency and appearance quality of the decorations on the bread after baking.
[0032] Step 5: After baking is completed, multiple baking images are collected to perform production composite consistency analysis to obtain composite consistency parameters;
[0033] Specifically, after the bread baking operation is completed, the image sensing equipment is used to capture images of the baked bread to obtain multiple baking images; then, baking quality recognition is performed based on the multiple baking images, and production composite consistency analysis is performed based on the multiple baking quality recognition results, that is, the overall uniformity and consistency of the baking quality of multiple breads are calculated, and a composite consistency parameter is obtained to characterize the uniformity of the bread baking quality in the production line, wherein the higher the baking quality uniformity, the larger the composite consistency parameter.
[0034] Step six: Perform consistency analysis of cutting, fermentation and decoration placement based on multiple dough images, multiple final fermentation parameters and multiple decoration placement parameters to obtain cutting consistency parameters, fermentation consistency parameters and decoration consistency parameters. Combine the composite consistency parameters to obtain production line management parameters for discrimination management.
[0035] Specifically, dough cutting consistency analysis is then performed based on multiple dough images, dough final fermentation consistency analysis is performed based on multiple final fermentation parameters, and decoration placement consistency analysis is performed based on multiple decoration placement parameters to obtain cutting consistency parameters, fermentation consistency parameters, and decoration consistency parameters, wherein the cutting consistency parameter represents the overall uniformity of dough cutting, the fermentation consistency parameter represents the overall uniformity of dough final fermentation, and the decoration consistency parameter represents the overall uniformity of decoration placement on bread. If the consistency is low, it means that each dough needs different processing parameters for each production, or the consistency of the produced bread is low, which affects the production efficiency and production quality of bread. Then, the production line production consistency evaluation is performed based on the cutting consistency parameter, the fermentation consistency parameter, the decoration consistency parameter, and the composite consistency parameter to obtain the production line management parameter; further determine whether the production line management parameter meets the preset management parameter threshold. If not, the bread production line is optimized and adjusted, for example, the control parameters of the equipment in the bread production line are optimized. The above method can improve the precision and accuracy of bread production consistency analysis, achieve the technical goal of accurately adjusting the bread production line, and achieve the technical effect of improving bread production efficiency and production quality.
[0036] The composite production line management method for food processing is applied to a composite production line management system for food processing, which can solve the technical problem that the existing composite production management method for bread has more production processes, making it difficult to analyze the consistency of bread production, resulting in the inability to carry out targeted adjustment and management of the bread production line, affecting the bread production efficiency and production quality. First, multiple dough images of multiple doughs that have been segmented are collected and identified to obtain multiple dough specification information and multiple preliminary fermentation information; then, according to the multiple dough specification information and multiple preliminary fermentation information, fermentation parameter analysis of final fermentation of dough is performed respectively to obtain multiple final fermentation parameters and fermentation is performed; then, after fermentation is completed, multiple fermentation dough images of multiple doughs are collected and identified to obtain multiple fermentation specification information and multiple final fermentation information; next, according to the multiple fermentation specification information and multiple final fermentation information, decoration placement parameter analysis of fermented dough is performed respectively to obtain multiple decoration placement parameters, and decoration placement and baking are performed; in addition, after baking is completed, multiple baking images are collected to perform production composite consistency analysis to obtain composite consistency parameters; finally, according to multiple dough images, multiple final fermentation parameters, and multiple decoration placement parameters, consistency analysis of cutting, fermentation, and decoration placement is performed to obtain cutting consistency parameters, fermentation consistency parameters, and decoration consistency parameters, and combined with the composite consistency parameters, production line management parameters are analyzed to perform discrimination management. The precision and accuracy of bread production consistency analysis can be improved, the technical goal of accurately adjusting the bread production line can be achieved, and the technical effect of improving bread production efficiency and production quality can be achieved.
[0037] Further, multiple dough images of the multiple doughs that have been segmented are collected and identified to obtain multiple dough specification information and multiple preliminary fermentation information, as shown in the attached figure. Figure 2 As shown, step one of this application includes:
[0038] collecting multiple dough images of the multiple doughs that have been segmented;
[0039] According to the historical production data of the composite production line, a sample dough image set is obtained;
[0040] Collecting and marking dough size specification information and preliminary fermentation information of the sample dough image to obtain a sample dough specification information set and a sample preliminary fermentation information set, wherein the preliminary fermentation information includes a preliminary fermentation degree;
[0041] The sample dough image set, the sample dough specification information set and the sample preliminary fermentation information set are used to construct a dough image recognizer, and the multiple dough images are recognized to obtain multiple dough specification information and multiple preliminary fermentation information.
[0042] Specifically, firstly, image acquisition is performed on multiple doughs that have been segmented by an image sensing device to obtain multiple dough images; on the other hand, historical production data of the composite production line is called to obtain a set of sample dough images based on the historical production data; then, information acquisition and identification are performed on the dough size specification information and preliminary fermentation information of multiple sample dough images in the sample dough image set to obtain a set of sample dough specification information and a set of sample preliminary fermentation information, wherein the sample dough images and the sample dough specification information and the sample preliminary fermentation information have a mapping relationship, wherein the preliminary fermentation information includes a degree of preliminary fermentation, and the degree of preliminary fermentation can be obtained by analyzing the number and size of pores in the dough image, wherein the number of pores, the size of pores and the degree of preliminary fermentation of the dough are proportional, that is, the more the number of pores on the dough and the larger the size of the pores, the better the degree of preliminary fermentation of the dough.
[0043] A dough image recognizer is constructed based on a convolutional neural network. The dough image recognizer is a convolutional neural network model that can be iteratively optimized in machine learning and is obtained through supervised training. The dough image recognizer includes an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer. The input data of the input layer is a dough image, and the output data is dough specification information and preliminary fermentation information. Then, the sample dough image set, the sample dough specification information set, and the sample preliminary fermentation information set are used as training data sets to perform supervised training on the dough image recognizer. The training method of the dough image recognizer is the same as the training method of the fermented dough image recognizer. Those skilled in the art can refer to the training process of the fermented dough image recognizer until a dough image recognizer that meets the preset accuracy constraint is obtained. Finally, the dough image recognizer recognizes the multiple dough images and outputs multiple dough specification information and multiple preliminary fermentation information. By constructing a dough image recognizer based on a convolutional neural network to perform dough feature analysis, the efficiency and accuracy of dough feature analysis can be improved, providing support for the next step of optimizing the final fermentation parameters.
[0044] Further, according to the plurality of dough specification information and the plurality of preliminary fermentation information, fermentation parameter analysis of final fermentation of the dough is performed respectively, a plurality of final fermentation parameters are obtained, and fermentation is performed. Step 2 of the present application includes:
[0045] Obtain standard fermentation specification information and standard final fermentation information;
[0046] Based on the standard fermentation specification information and the standard final fermentation information, a final fermentation optimization function is constructed as follows:
[0047]
[0048] Among them, FER is fermentation fitness, w1 is specification weight, w2 is fermentation weight, D y is the standard fermentation specification information, F y is the standard final fermentation information, D f The actual fermentation specification information after fermentation according to the final fermentation parameters, F f The actual final fermentation information after fermentation according to the final fermentation parameters;
[0049] According to the final fermentation optimization function, the fermentation parameters of the final fermentation of the dough are optimized to obtain the final fermentation parameters.
[0050] Specifically, first, obtain standard fermentation specification information and standard final fermentation information, wherein the standard fermentation specification information refers to the expected fermentation specification requirements, including standard fermentation shape, standard fermentation size (thickness, diameter, etc.), etc., and the standard final fermentation information refers to the expected final state of the dough after fermentation, including volume expansion ratio, dough elasticity, dough softness and other indicators, wherein the standard fermentation specification information and standard final fermentation information can be set by technical personnel in this field according to actual bread production requirements.
[0051] Then, based on the standard fermentation specification information and the standard final fermentation information, a final fermentation optimization function is constructed, wherein the final fermentation optimization function expression is:
[0052] In the final fermentation optimization function, FER is the fermentation fitness, wherein the greater the fermentation fitness, the better the fermentation quality; w1 is the specification weight, which represents the influence of the fermentation specification information on the fermentation quality; w2 is the fermentation weight, which represents the influence of the final fermentation information on the fermentation quality, wherein the sum of the specification weight and the fermentation weight is 1, and it can be set according to the influence of the fermentation specification information and the final fermentation information on the overall fermentation quality. The greater the influence of any indicator, the greater the weight corresponding to the indicator. Weighting can be performed according to the existing coefficient of variation method. For example, firstly, the mean of all indicators in the fermentation specification information (fermentation shape, fermentation size, etc.) and the mean of all indicators in the final fermentation information (volume expansion ratio, dough elasticity, etc.) are calculated; then the standard deviation of each indicator in the fermentation specification information and the final fermentation information is calculated respectively; further according to the fermentation specification The coefficient of variation of the fermentation specification information is calculated based on the indicator mean and indicator standard deviation of the information, where the coefficient of variation is the ratio of the indicator standard deviation to the indicator mean. The coefficient of variation reflects the degree of data dispersion. The smaller the coefficient of variation, the more stable the indicator and the greater the impact on the fermentation fitness. The coefficient of variation of the final fermentation information is calculated based on the indicator mean and indicator standard deviation of the final fermentation information. Then, the corresponding weight is assigned based on the coefficient of variation of the fermentation specification information and the coefficient of variation of the final fermentation information. The coefficient of variation and the weight are negatively correlated, that is, the larger the coefficient of variation, the smaller the weight, and the smaller the coefficient of variation, the larger the weight. For example, the ratio of the coefficient of variation of the fermentation specification information to the sum of the coefficients of variation of the two can be set as the fermentation weight of the final fermentation information, and the ratio of the coefficient of variation of the final fermentation information to the sum of the coefficients of variation of the two can be set as the fermentation weight of the fermentation specification information, and the specification weight and fermentation weight can be obtained. y is the standard fermentation specification information, F y is the standard final fermentation information, D f The actual fermentation specification information after fermentation according to the final fermentation parameters, including the actual fermentation shape, actual fermentation size, etc.; F f The actual final fermentation information after fermentation according to the final fermentation parameters includes the actual volume expansion ratio, the actual dough elasticity, the actual dough softness, etc. Further based on the final fermentation optimization function, the fermentation parameters of the final fermentation of the dough are optimized, and the fermentation parameters with the maximum fermentation fitness in the optimization process are output and set as the final fermentation parameters.
[0053] Furthermore, the present application also includes the following steps:
[0054] According to the historical production data of final fermentation of dough, a sample dough specification information set and a sample preliminary fermentation information set are obtained, and a sample final fermentation parameter set, a sample fermentation specification information set and a sample final fermentation information set are obtained to construct a final fermentation predictor;
[0055] Randomly generate a first final fermentation parameter, and based on the final fermentation predictor, perform a final fermentation prediction according to the first final fermentation parameter to obtain first actual fermentation specification information and first actual final fermentation information;
[0056] According to the final fermentation optimization function, combined with the first actual fermentation specification information and the first actual final fermentation information, a first fermentation fitness is calculated;
[0057] Continue the optimization until convergence, and output the final fermentation parameters with the maximum fermentation fitness.
[0058] Specifically, according to the final fermentation optimization function, the fermentation parameters of the final fermentation of the dough are optimized, and the method for obtaining the final fermentation parameters is as follows: first, the historical production data of the final fermentation of the dough is called, and a sample dough specification information set and a sample preliminary fermentation information set are obtained based on the historical production data, and a sample final fermentation parameter set, a sample fermentation specification information set and a sample final fermentation information set are obtained, wherein the sample dough specification information, the sample preliminary fermentation information and the sample final fermentation parameters, the sample fermentation specification information and the sample final fermentation information have a mapping relationship.
[0059] A final fermentation predictor is constructed based on a BP neural network. The final fermentation predictor is a neural network model that can be iteratively optimized in a feedforward neural network, and is obtained by supervised training through a training data set; wherein the final fermentation predictor includes an input layer, a plurality of hidden layers, and an output layer, wherein the input data of the input layer are dough specification information, preliminary fermentation information, and final fermentation parameters, and the output data of the output layer are fermentation specification information and final fermentation information. Then, the sample dough specification information set, the sample preliminary fermentation information set, the sample final fermentation parameter set, the sample fermentation specification information set, and the sample final fermentation information set are used as training data sets to supervise the final fermentation predictor; first, the collected training data set is divided into a training set, a validation set, and a test set, such as 70% of the training set is used for model training, and 15% of the validation set is used for validation during the training process to evaluate the performance of the model during training to prevent overfitting; 15% of the test set is used to finally evaluate the generalization ability of the model. Then, the training set is used to conduct supervised training on the final fermentation predictor. First, the input data (sample dough specification information, sample preliminary fermentation information and sample final fermentation parameters) are input into the neural network. After calculation of the hidden layer, the prediction result of the output layer (fermentation specification information and final fermentation information) is obtained. Then, the error between the predicted value and the actual target value is calculated through the loss function. Further, the gradient descent algorithm is used to backpropagate the gradient information of the error to update the weight parameters of the neural network. Finally, the model is iteratively optimized through multiple rounds of training, so that the loss function gradually converges, the prediction accuracy of the model is gradually improved, and the final fermentation predictor that tends to converge is obtained.
[0060] Then, the validation set and test set were used to perform performance verification and model testing on the final fermentation predictor, respectively, to obtain the trained final fermentation predictor.
[0061] Obtain fermentation parameter rated thresholds, including temperature rated thresholds, humidity rated thresholds, etc.; then randomly select any parameter from multiple parameter rated thresholds of the fermentation parameter rated thresholds to form a first final fermentation parameter; then input the first final fermentation parameter into the final fermentation predictor for final fermentation prediction, and output first actual fermentation specification information and first actual final fermentation information. Further, according to the final fermentation optimization function, perform fermentation fitness calculation on the first actual fermentation specification information and the first actual final fermentation information to obtain a first fermentation fitness.
[0062] Then, any parameter is randomly selected again within the multiple parameter rated thresholds of the fermentation parameter rated threshold to form the second final fermentation parameter, wherein the second final fermentation parameter is not completely the same as the first final fermentation parameter, and the second fermentation fitness of the second final fermentation parameter is calculated; then the first fermentation fitness and the second fermentation fitness are compared, if the first fermentation fitness is greater than the second fermentation fitness, the first final fermentation parameter is set as the current optimal fermentation parameter; conversely, if the first fermentation fitness is less than or equal to the second fermentation fitness, the second final fermentation parameter is set as the current optimal fermentation parameter; continue to iterate and optimize until the preset number of optimizations is reached, and output the final fermentation parameter with the maximum fermentation fitness, and the preset number of optimizations can be set according to actual conditions. By optimizing the final fermentation parameters of the dough based on the dough specification information and the preliminary fermentation information, the optimized final fermentation parameters are obtained to control the fermentation of the dough, which can improve the adaptability of the final fermentation parameters to the current state of the dough, make the final fermentation parameter setting more accurate, and thus improve the dough fermentation quality.
[0063] Further, according to the plurality of fermentation specification information and the plurality of final fermentation information, the decoration placement parameter analysis of the fermented dough is performed respectively. Step 4 of the present application includes:
[0064] Acquire standard decoration distribution information, wherein the standard decoration distribution information includes a standard distribution area of the decoration after baking;
[0065] Based on the standard decoration distribution information, a decoration placement optimization function is constructed as follows:
[0066]
[0067] Among them, ORN is the decoration fitness, Z y is the standard decoration distribution information, Z k The actual distribution information of the decorations after they are placed and baked according to the decoration placement parameters;
[0068] According to the decoration placement optimization function, the decoration placement parameters of the fermented dough are optimized and analyzed to obtain the multiple decoration placement parameters, wherein in the decoration placement parameter optimization analysis process, the decoration distribution information after baking is predicted based on the decoration placement parameters, fermentation specification information, final fermentation information and baking parameters.
[0069] Specifically, first, obtain standard decoration distribution information, wherein the standard decoration distribution information includes a standard distribution area of decorations after baking, such as a standard distribution area of the decoration distribution area, which can be set by a person skilled in the art according to actual needs of bread products, such as by extracting information from a bread product design drawing; then, based on the standard decoration distribution information, construct a decoration placement optimization function, and the decoration placement optimization function expression is: In the decoration placement optimization function, ORN is the decoration fitness, where the larger the decoration fitness, the better the decoration quality; Z y is the standard decoration distribution information, Z k The actual decoration distribution information after the bread is placed and baked according to the decoration placement parameters includes the actual distribution area of the decoration distribution region. The actual decoration distribution information can be obtained by collecting the baked bread image using an image sensor and extracting the decoration distribution area of the bread image.
[0070] A decoration distribution prediction model is constructed based on a BP neural network, wherein the decoration distribution prediction model is a neural network model that can be iteratively optimized in machine learning, including an input layer, multiple hidden layers and an output layer, wherein the input data of the input layer are decoration placement parameters, fermentation specification information, final fermentation information and baking parameters, and the output data of the output layer are decoration distribution prediction results; then, the historical production data of dough baking is called to obtain a sample training data set, wherein the sample training data includes sample decoration placement parameters, sample fermentation specification information, sample final fermentation information, sample baking parameters and sample decoration distribution; then, the sample decoration placement parameters, sample fermentation specification information, sample final fermentation information, sample baking parameters and sample decoration distribution are used to predict the decoration distribution of the dough; , sample final fermentation information, and sample baking parameters are used as inputs, and the sample decoration distribution is used as supervision. The sample training data set is used to perform supervised training on the decoration distribution prediction model, that is, the input data (sample decoration placement parameters, sample fermentation specification information, sample fermentation information, and sample baking parameters) are forward propagated through the neural network to obtain the predicted output of the model (decoration distribution); then the loss (error) between the model prediction value and the true value is calculated, and the weight is adjusted through the back propagation algorithm; then iterative training is performed, and forward propagation, error calculation, and back propagation are repeated, and the model weights are continuously updated until the loss function converges to obtain a trained decoration distribution prediction model.
[0071] Then, according to the decoration placement optimization function, the decoration placement parameters of the fermented dough are optimized based on the decoration distribution prediction model, and the multiple decoration placement parameters are obtained based on the optimization results, wherein the optimization method is the same as the optimization method for obtaining the final fermentation parameters mentioned above, and will not be described in detail here. By optimizing the decoration placement parameters of the fermented dough according to the fermentation specification information and the final fermentation information, the adaptability of the decoration placement parameters to the fermented dough state can be improved, making the decoration placement parameters more accurate, thereby improving the distribution consistency and appearance quality of the decorations on the bread after baking.
[0072] Further, multiple baking images are collected to perform production composite consistency analysis to obtain composite consistency parameters. Step five of this application includes:
[0073] Collect multiple baking images after baking is completed;
[0074] Based on the historical data of bread production in the composite production line, a set of sample baking images and a set of sample baking quality are obtained to build a baking quality identifier;
[0075] Based on the baking quality identifier, the plurality of baking images are identified to obtain a plurality of baking quality information;
[0076] The composite consistency parameter is obtained by calculation according to the plurality of baking quality information.
[0077] Specifically, an image sensor device is used to collect images of multiple breads after baking to obtain multiple baking images; the historical data of bread production in the composite production line is called, and a sample baking image set and a sample baking quality set are obtained based on the historical data, wherein the sample baking image and the sample baking quality correspond one to one, and the baking quality can be represented by a baking quality coefficient. The larger the quality coefficient, the better the baking quality. A baking quality identifier is constructed based on a convolutional neural network, and the baking quality identifier includes an input layer, a convolution layer, a pooling layer, a fully connected layer, and an output layer, wherein the input data of the input layer is a baking image, and the output data of the output layer is a baking quality coefficient; then, the sample baking image is used as input data, and the sample baking quality is used as supervision, and the sample baking image set and the sample baking quality set are used to supervise the baking quality identifier. First, the input baking image is passed through the convolution layer, the pooling layer, and the fully connected layer to generate the predicted output (baking quality coefficient) of the model; then, according to the value of the loss function, the weights in the network are adjusted using the back propagation algorithm to minimize the prediction error; then iterative training is performed until the loss function converges, and the trained baking quality identifier is output.
[0078] Then, the baking quality identifier identifies the multiple baking images and outputs multiple baking quality information. Finally, a composite consistency calculation is performed based on the multiple baking quality information to obtain the composite consistency parameter, wherein the composite consistency parameter represents the uniformity of the baking quality of the bread. For example, the variance of the multiple baking quality information can be calculated, and the composite consistency parameter can be calculated according to the variance calculation result, wherein the variance of the quality information is inversely proportional to the composite consistency parameter, that is, the smaller the variance of the quality information is, the greater the uniformity of the baking quality is, and the larger the composite consistency parameter is, such as taking the inverse of the variance of the quality information as the composite consistency parameter.
[0079] Further, the consistency analysis of cutting, fermentation and decoration placement is performed according to multiple dough images, multiple final fermentation parameters and multiple decoration placement parameters to obtain cutting consistency parameters, fermentation consistency parameters and decoration consistency parameters. The production line management parameters are obtained by analysis in combination with the composite consistency parameters. Step 6 of this application includes:
[0080] Randomly combining and similarity identifying the multiple dough images to obtain a dough similarity set, wherein the similarity identification is performed by a twin neural network;
[0081] Calculating and obtaining a cutting consistency parameter according to the dough similarity set;
[0082] According to the multiple final fermentation parameters and the multiple decoration placement parameters, respectively calculating and obtaining a fermentation consistency parameter and a decoration consistency parameter;
[0083] Obtain a sample cutting consistency parameter set, a sample fermentation consistency parameter set, a sample decoration consistency parameter set, and a sample composite consistency parameter set, and obtain a sample production line management parameter set;
[0084] Using the sample cutting consistency parameter set, sample fermentation consistency parameter set, sample decoration consistency parameter set, sample compound consistency parameter set and sample production line management parameter set, based on a decision tree, a production line management decision maker is constructed;
[0085] Based on the production line management decision maker, the composite consistency parameter, the cutting consistency parameter, the fermentation consistency parameter and the decoration consistency parameter are analyzed and decided to obtain the production line management parameter;
[0086] Determine whether the production line management parameter is greater than the management parameter threshold. If so, adjust and optimize the production line management.
[0087] Specifically, first, the plurality of dough images are randomly combined in pairs, and the dough image combination results are similarity identified to obtain a dough similarity set, wherein the dough image combination results can be similarity identified by a twin neural network, and the twin neural network is a special neural network structure, which is composed of two or more sub-networks with shared weights, each of which can be a convolutional neural network or other types of neural networks, used to extract features from input images and calculate the similarity of these features, and the output of the twin neural network is a similarity score, which represents the distance between the two images in the feature space, wherein the smaller the distance between the two images in the feature space, the greater the similarity score. The twin neural network includes an input layer, a shared convolutional network layer and an output layer, wherein the input data of the input layer is two dough images, and the output data of the output layer is the similarity score of the two dough images, wherein the shared convolutional network layer is used to extract similar features through the same convolutional network (shared parameters). By inputting pairs of dough images and corresponding similarity labels and optimizing the model by calculating the loss function, the model is gradually optimized through multiple rounds of training iterations using an optimization algorithm (such as Adam or SGD) until the loss function converges to obtain a twin neural network that meets the expected requirements.
[0088] Then, a similarity variance calculation is performed based on the dough similarity set, and a cutting consistency parameter is obtained based on the similarity variance calculation result, wherein the cutting consistency parameter is negatively correlated with the similarity variance calculation result, that is, the larger the similarity variance calculation result is, the smaller the cutting consistency parameter is, and the smaller the similarity variance calculation result is, the larger the cutting consistency parameter is. For example, the inverse of the similarity variance calculation result can be set as the cutting consistency parameter.
[0089] Then, according to the multiple final fermentation parameters and multiple decoration placement parameters, variance calculations are performed respectively, and fermentation consistency parameters and decoration consistency parameters are obtained based on the variance calculation results, wherein the fermentation consistency parameter is the inverse of the fermentation parameter variance calculation result, and the decoration consistency parameter is the inverse of the decoration placement parameter variance calculation result; then, a sample cutting consistency parameter set, a sample fermentation consistency parameter set, a sample decoration consistency parameter set, and a sample composite consistency parameter set are obtained, and a sample production line management parameter set is obtained, wherein the sample production line management parameter represents the overall management quality of the production line, and the larger the sample production line management parameter, the better the overall management quality of the production line, wherein the sample cutting consistency parameter, the sample fermentation consistency parameter, the sample decoration consistency parameter, the sample composite consistency parameter, and the sample production line management parameter have a mapping relationship. The decision tree model is to split the features of the data and gradually refine them until the optimal classification or regression result is achieved. In the production line management scenario, the decision tree will make management decisions based on the consistency parameters. Then, based on the decision tree principle, the cutting consistency parameter, fermentation consistency parameter, decoration consistency parameter and composite consistency parameter are used as input features, and the production line management parameter is used as the output target variable to construct an initial production line management decision maker. The initial production line management decision maker is further trained using the sample cutting consistency parameter set, sample fermentation consistency parameter set, sample decoration consistency parameter set, sample composite consistency parameter set and sample production line management parameter set. By recursively splitting the feature space, the optimal splitting condition of each node is found, and the node splitting method is continuously optimized to minimize the prediction error, so as to obtain a production line management decision maker that meets the expected output accuracy. Then, the composite consistency parameter, cutting consistency parameter, fermentation consistency parameter and decoration consistency parameter are input into the production line management decision maker for matching decision analysis, and the matching output production line management parameter is obtained; the management parameter threshold is obtained, and the management parameter threshold can be set according to the management quality requirement, wherein the higher the quality requirement, the smaller the management parameter threshold; then, it is determined whether the production line management parameter is greater than the management parameter threshold. If so, the production line is adjusted and optimized, such as optimizing the control parameters of the equipment in the bread production line.
[0090] In summary, the composite production line management method for food processing provided by the present application has the following technical effects:
[0091] 1. By identifying multiple dough images that have been segmented, multiple dough specification information and multiple preliminary fermentation information are obtained; then, dough fermentation parameters are optimized and analyzed according to multiple dough specification information and multiple preliminary fermentation information, and multiple optimized final fermentation parameters are obtained for fermentation; then, after fermentation is completed, multiple fermented dough images are identified to obtain multiple fermentation specification information and multiple final fermentation information; further, according to multiple fermentation specification information and multiple final fermentation information, the decoration placement parameters of the fermented dough are optimized and analyzed to obtain multiple optimized decoration placement parameters for decoration placement and baking; and after baking, production composite consistency analysis is performed based on multiple baking images to obtain composite consistency parameters; on the other hand, consistency analysis of cutting, fermentation and decoration placement is performed according to multiple dough images, multiple final fermentation parameters, and multiple decoration placement parameters, and production line management parameters are obtained based on cutting consistency parameters, fermentation consistency parameters, decoration consistency parameters and composite consistency parameters. When the production line management parameters do not meet the preset management parameter threshold, the bread composite production line is adjusted and optimized. The precision and accuracy of bread production consistency analysis can be improved, the technical goal of accurately adjusting the bread production line can be achieved, and the technical effect of improving bread production efficiency and production quality can be achieved.
[0092] 2. By optimizing the final fermentation parameters of the dough based on the dough specification information and the preliminary fermentation information, the optimized final fermentation parameters are obtained to control the fermentation of the dough. This can improve the adaptability of the final fermentation parameters to the current state of the dough, making the final fermentation parameter settings more accurate, thereby improving the dough fermentation quality.
[0093] 3. By optimizing the decoration placement parameters of the fermented dough according to the fermentation specification information and the final fermentation information, the adaptability of the decoration placement parameters to the fermented dough state can be improved, making the decoration placement parameters more accurate, thereby improving the distribution consistency and appearance quality of the decorations on the bread after baking.
[0094] Embodiment 2
[0095] Based on the same inventive concept as the composite production line management method for food processing in the aforementioned embodiment, the present application also provides a composite production line management system for food processing, please refer to the attached Figure 3 , the system comprising:
[0096] The dough image recognition module 11 is used to collect multiple dough images of the multiple doughs that have been divided, perform recognition, and obtain multiple dough specification information and multiple preliminary fermentation information;
[0097] A fermentation parameter analysis module 12 is used to analyze the fermentation parameters of the final fermentation of the dough according to the plurality of dough specification information and the plurality of preliminary fermentation information, obtain a plurality of final fermentation parameters, and perform fermentation;
[0098] The fermented dough image recognition module 13 is used to collect multiple fermented dough images of multiple doughs after fermentation is completed, perform recognition, and obtain multiple fermentation specification information and multiple final fermentation information;
[0099] A decoration placement parameter analysis module 14 is used to analyze the decoration placement parameters of the fermented dough according to the plurality of fermentation specification information and the plurality of final fermentation information, obtain a plurality of decoration placement parameters, and perform decoration placement and baking;
[0100] The production composite consistency analysis module 15 is used to collect multiple baking images after baking is completed, perform production composite consistency analysis, and obtain composite consistency parameters;
[0101] The production line management parameter identification module 16 is used to perform consistency analysis of cutting, fermentation and decoration placement according to multiple dough images, multiple final fermentation parameters and multiple decoration placement parameters, obtain cutting consistency parameters, fermentation consistency parameters and decoration consistency parameters, and combine the composite consistency parameters to analyze and obtain production line management parameters for identification management.
[0102] Furthermore, the dough image recognition module 11 in the system is also used for:
[0103] collecting multiple dough images of the multiple doughs that have been segmented;
[0104] According to the historical production data of the composite production line, a sample dough image set is obtained;
[0105] Collecting and marking dough size specification information and preliminary fermentation information of the sample dough image to obtain a sample dough specification information set and a sample preliminary fermentation information set, wherein the preliminary fermentation information includes a preliminary fermentation degree;
[0106] The sample dough image set, the sample dough specification information set and the sample preliminary fermentation information set are used to construct a dough image recognizer, and the multiple dough images are recognized to obtain multiple dough specification information and multiple preliminary fermentation information.
[0107] Further, the fermentation parameter analysis module 12 in the system is also used for:
[0108] Obtain standard fermentation specification information and standard final fermentation information;
[0109] Based on the standard fermentation specification information and the standard final fermentation information, a final fermentation optimization function is constructed as follows:
[0110]
[0111] Among them, FER is fermentation fitness, w1 is specification weight, w2 is fermentation weight, D y is the standard fermentation specification information, F y is the standard final fermentation information, D f The actual fermentation specification information after fermentation according to the final fermentation parameters, F f The actual final fermentation information after fermentation according to the final fermentation parameters;
[0112] According to the final fermentation optimization function, the fermentation parameters of the final fermentation of the dough are optimized to obtain the final fermentation parameters.
[0113] Further, the fermentation parameter analysis module 12 in the system is also used for:
[0114] According to the historical production data of final fermentation of dough, a sample dough specification information set and a sample preliminary fermentation information set are obtained, and a sample final fermentation parameter set, a sample fermentation specification information set and a sample final fermentation information set are obtained to construct a final fermentation predictor;
[0115] Randomly generate a first final fermentation parameter, and based on the final fermentation predictor, perform a final fermentation prediction according to the first final fermentation parameter to obtain first actual fermentation specification information and first actual final fermentation information;
[0116] According to the final fermentation optimization function, combined with the first actual fermentation specification information and the first actual final fermentation information, a first fermentation fitness is calculated;
[0117] Continue the optimization until convergence, and output the final fermentation parameters with the maximum fermentation fitness.
[0118] Furthermore, the decoration placement parameter analysis module 14 in the system is also used for:
[0119] Acquire standard decoration distribution information, wherein the standard decoration distribution information includes a standard distribution area of the decoration after baking;
[0120] Based on the standard decoration distribution information, a decoration placement optimization function is constructed as follows:
[0121]
[0122] Among them, ORN is the decoration fitness, Z y is the standard decoration distribution information, Zk The actual distribution information of the decorations after they are placed and baked according to the decoration placement parameters;
[0123] According to the decoration placement optimization function, the decoration placement parameters of the fermented dough are optimized and analyzed to obtain the multiple decoration placement parameters, wherein in the decoration placement parameter optimization analysis process, the decoration distribution information after baking is predicted based on the decoration placement parameters, fermentation specification information, final fermentation information and baking parameters.
[0124] Further, the production composite consistency analysis module 15 in the system is also used for:
[0125] Collect multiple baking images after baking is completed;
[0126] Based on the historical data of bread production in the composite production line, a set of sample baking images and a set of sample baking quality are obtained to build a baking quality identifier;
[0127] Based on the baking quality identifier, the plurality of baking images are identified to obtain a plurality of baking quality information;
[0128] The composite consistency parameter is obtained by calculation according to the plurality of baking quality information.
[0129] Furthermore, the production line management parameter determination module 16 in the system is also used for:
[0130] Randomly combining and similarity identifying the multiple dough images to obtain a dough similarity set, wherein the similarity identification is performed by a twin neural network;
[0131] Calculating and obtaining a cutting consistency parameter according to the dough similarity set;
[0132] According to the multiple final fermentation parameters and the multiple decoration placement parameters, respectively calculating and obtaining a fermentation consistency parameter and a decoration consistency parameter;
[0133] Obtain a sample cutting consistency parameter set, a sample fermentation consistency parameter set, a sample decoration consistency parameter set, and a sample composite consistency parameter set, and obtain a sample production line management parameter set;
[0134] Using the sample cutting consistency parameter set, sample fermentation consistency parameter set, sample decoration consistency parameter set, sample compound consistency parameter set and sample production line management parameter set, based on a decision tree, a production line management decision maker is constructed;
[0135] Based on the production line management decision maker, the composite consistency parameter, the cutting consistency parameter, the fermentation consistency parameter and the decoration consistency parameter are analyzed and decided to obtain the production line management parameter;
[0136] Determine whether the production line management parameter is greater than the management parameter threshold. If so, adjust and optimize the production line management.
[0137] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The composite production line management method for food processing and the specific examples in the aforementioned embodiment 1 are also applicable to a composite production line management system for food processing in this embodiment. Through the aforementioned detailed description of a composite production line management method for food processing, those skilled in the art can clearly know a composite production line management system for food processing in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be described in the method part.
[0138] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.
[0139] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.
Claims
1. A composite production line management method for food processing, characterized in that: The method comprises: Collecting multiple dough images of the multiple doughs that have been segmented, performing recognition, and obtaining multiple dough specification information and multiple preliminary fermentation information; According to the plurality of dough specification information and the plurality of preliminary fermentation information, respectively, fermentation parameter analysis of final fermentation of the dough is performed to obtain a plurality of final fermentation parameters and perform fermentation; After fermentation is completed, a plurality of fermented dough images of the plurality of doughs are collected and identified to obtain a plurality of fermentation specification information and a plurality of final fermentation information; According to the plurality of fermentation specification information and the plurality of final fermentation information, respectively analyze the decoration placement parameters of the fermented dough, obtain a plurality of decoration placement parameters, and perform decoration placement and baking; After baking is completed, multiple baking images are collected to perform production composite consistency analysis and obtain composite consistency parameters; The consistency analysis of cutting, fermentation and decoration placement is performed according to multiple dough images, multiple final fermentation parameters and multiple decoration placement parameters to obtain cutting consistency parameters, fermentation consistency parameters and decoration consistency parameters. The production line management parameters are analyzed and obtained in combination with the composite consistency parameters to perform discrimination management.
2. The method according to claim 1, characterized in that Collect multiple dough images of the multiple doughs that have been segmented, perform recognition, and obtain multiple dough specification information and multiple preliminary fermentation information, including: collecting multiple dough images of the multiple doughs that have been segmented; According to the historical production data of the composite production line, a sample dough image set is obtained; Collecting and marking dough size specification information and preliminary fermentation information of the sample dough image to obtain a sample dough specification information set and a sample preliminary fermentation information set, wherein the preliminary fermentation information includes a preliminary fermentation degree; The sample dough image set, the sample dough specification information set and the sample preliminary fermentation information set are used to construct a dough image recognizer, and the multiple dough images are recognized to obtain multiple dough specification information and multiple preliminary fermentation information.
3. The method according to claim 1, characterized in that: According to the plurality of dough specification information and the plurality of preliminary fermentation information, respectively, fermentation parameter analysis of final fermentation of the dough is performed to obtain a plurality of final fermentation parameters and perform fermentation, including: Obtain standard fermentation specification information and standard final fermentation information; Based on the standard fermentation specification information and the standard final fermentation information, a final fermentation optimization function is constructed as follows: Among them, FER is fermentation fitness, w1 is specification weight, w2 is fermentation weight, D y is the standard fermentation specification information, F y is the standard final fermentation information, D f The actual fermentation specification information after fermentation according to the final fermentation parameters, F f The actual final fermentation information after fermentation according to the final fermentation parameters; According to the final fermentation optimization function, the fermentation parameters of the final fermentation of the dough are optimized to obtain the final fermentation parameters.
4. The method according to claim 3, characterized in that According to the final fermentation optimization function, the fermentation parameters of the final fermentation of the dough are optimized to obtain the final fermentation parameters, including: According to the historical production data of final fermentation of dough, a sample dough specification information set and a sample preliminary fermentation information set are obtained, and a sample final fermentation parameter set, a sample fermentation specification information set and a sample final fermentation information set are obtained to construct a final fermentation predictor; Randomly generate a first final fermentation parameter, and based on the final fermentation predictor, perform a final fermentation prediction according to the first final fermentation parameter to obtain first actual fermentation specification information and first actual final fermentation information; According to the final fermentation optimization function, combined with the first actual fermentation specification information and the first actual final fermentation information, a first fermentation fitness is calculated; Continue the optimization until convergence, and output the final fermentation parameters with the maximum fermentation fitness.
5. The method according to claim 1, characterized in that According to the plurality of fermentation specification information and the plurality of final fermentation information, respectively performing decoration placement parameter analysis of the fermented dough, including: Acquire standard decoration distribution information, wherein the standard decoration distribution information includes a standard distribution area of the decoration after baking; Based on the standard decoration distribution information, a decoration placement optimization function is constructed as follows: Among them, ORN is the decoration fitness, Z y is the standard decoration distribution information, Z k The actual distribution information of the decorations after they are placed and baked according to the decoration placement parameters; According to the decoration placement optimization function, the decoration placement parameters of the fermented dough are optimized and analyzed to obtain the multiple decoration placement parameters, wherein in the decoration placement parameter optimization analysis process, the decoration distribution information after baking is predicted based on the decoration placement parameters, fermentation specification information, final fermentation information and baking parameters.
6. The method according to claim 1, characterized in that Collect multiple baking images, perform production composite consistency analysis, and obtain composite consistency parameters, including: Collect multiple baking images after baking is completed; Based on the historical data of bread production in the composite production line, a set of sample baking images and a set of sample baking quality are obtained to build a baking quality identifier; Based on the baking quality identifier, the plurality of baking images are identified to obtain a plurality of baking quality information; The composite consistency parameter is obtained by calculation according to the plurality of baking quality information.
7. The method according to claim 1, characterized in that According to multiple dough images, multiple final fermentation parameters, and multiple decoration placement parameters, consistency analysis of cutting, fermentation, and decoration placement is performed to obtain cutting consistency parameters, fermentation consistency parameters, and decoration consistency parameters. Combined with the composite consistency parameters, production line management parameters are analyzed and obtained, including: Randomly combining and similarity identifying the multiple dough images to obtain a dough similarity set, wherein the similarity identification is performed by a twin neural network; Calculating and obtaining a cutting consistency parameter according to the dough similarity set; According to the multiple final fermentation parameters and the multiple decoration placement parameters, respectively calculating and obtaining a fermentation consistency parameter and a decoration consistency parameter; Obtain a sample cutting consistency parameter set, a sample fermentation consistency parameter set, a sample decoration consistency parameter set, and a sample composite consistency parameter set, and obtain a sample production line management parameter set; Using the sample cutting consistency parameter set, sample fermentation consistency parameter set, sample decoration consistency parameter set, sample compound consistency parameter set and sample production line management parameter set, based on a decision tree, a production line management decision maker is constructed; Based on the production line management decision maker, the composite consistency parameter, the cutting consistency parameter, the fermentation consistency parameter and the decoration consistency parameter are analyzed and decided to obtain the production line management parameter; Determine whether the production line management parameter is greater than the management parameter threshold. If so, adjust and optimize the production line management.
8. A composite production line management system for food processing, characterized in that: For implementing the steps of the method according to any one of claims 1 to 7, the system comprises: A dough image recognition module is used to collect multiple dough images of multiple doughs that have been divided, perform recognition, and obtain multiple dough specification information and multiple preliminary fermentation information; A fermentation parameter analysis module, used to analyze fermentation parameters of final fermentation of dough according to the plurality of dough specification information and the plurality of preliminary fermentation information, obtain a plurality of final fermentation parameters, and perform fermentation; A fermented dough image recognition module is used to collect multiple fermented dough images of multiple doughs after fermentation is completed, perform recognition, and obtain multiple fermentation specification information and multiple final fermentation information; A decoration placement parameter analysis module, for performing decoration placement parameter analysis of the fermented dough according to the plurality of fermentation specification information and the plurality of final fermentation information, obtaining a plurality of decoration placement parameters, and performing decoration placement and baking; The production composite consistency analysis module is used to collect multiple baking images after baking is completed, perform production composite consistency analysis, and obtain composite consistency parameters; The production line management parameter discrimination module is used to perform consistency analysis of cutting, fermentation and decoration placement according to multiple dough images, multiple final fermentation parameters and multiple decoration placement parameters, obtain cutting consistency parameters, fermentation consistency parameters and decoration consistency parameters, and combine the composite consistency parameters to analyze and obtain production line management parameters for discrimination management.
Citation Information
Patent Citations
Bread baking simulation method, device and equipment and readable storage medium
CN115935704A
Intelligent unmanned factory production method, system and equipment
CN117784730A
Bread baking simulation method and device based on deep learning model and electronic equipment
CN118313278A
Optimization control method for fermented glutinous rice steamed bun production process based on artificial intelligence
CN118466426A
Image sample processing method and apparatus and data processing method and apparatus
WO2022205158A1
Cited By
Production line management method and system for mixed production of konjak
CN121329039A
Industrial product quality tracing method based on industrial big data
CN121329230A
