Cooking processing method and system of self-learning cooking robot

Abstract

The invention relates to the technical field of data identification, and provides a cooking processing method and system of a self-learning cooking robot, the method is applied to the cooking robot, and the method comprises the steps: obtaining the type, shape and weight information of a to-be-processed food material, querying the type information and the shape information, and obtaining a recipe candidate list; selecting a cooking recipe and obtaining cooking steps according to the recipe candidate list and the cuisine preference of the user, obtaining condiment dosage based on the cuisine preference and food material weight information, inputting the condiment dosage and the cooking steps into a cooking robot to pre-cook food materials, obtaining feedback information of the user on a pre-cooking result, and sending the feedback information to the user. And after adjusting the dosage of the flavoring agent and cooking steps, performing formal cooking. By introducing the self-learning mechanism, the cooking robot can select the cooking recipe according to the type, shape and weight information, adjustment is carried out in combination with the cuisine preference of the user, and the problem that adjustment cannot be carried out according to changes of different food materials and instant feedback of the user in practical application is solved.

Description

Technical Field

[0001] The present application relates to the technical field of data recognition, and more particularly to a cooking processing method and system for a self-learning cooking robot. Background Art

[0002] In recent years, with the rapid development of smart technology, smart homes have gradually entered ordinary people's homes. As an important part of smart homes, smart cooking robots have attracted much attention from consumers. This kind of robot can not only help users automatically complete complex cooking operations, but also customize personalized cooking plans according to users' dietary preferences and health needs, thus greatly improving the convenience and fun of home cooking.

[0003] In the related technical means, cooking robots usually use preset cooking programs and steps to cook. Specifically, the cooking robot uses the built-in recipe database to perform cooking operations such as cutting, stirring, and heating according to the preset recipes. This method enables the cooking robot to complete various complex cooking tasks in a shorter time and ensure a certain cooking effect.

[0004] Regarding the above technical solution, although a certain degree of automated cooking can be achieved through preset cooking programs and steps, in actual application, there is a problem that it cannot be adjusted according to changes in different ingredients and instant feedback from users. Summary of the invention

[0005] In order to improve the problem of being unable to make adjustments according to changes in different ingredients and instant feedback from users in actual applications, the present application provides a cooking processing method and system for a self-learning cooking robot, which continuously adjusts and optimizes the cooking steps and the amount of seasonings by obtaining feedback from users, thereby achieving a more intelligent cooking experience.

[0006] The present invention provides a cooking processing method of a self-learning cooking robot, which is applied to the cooking robot. The cooking processing method includes: obtaining food type information, food shape information and food weight information of food to be processed, inputting the food type information and the food shape information into a preset recipe database for query, and obtaining a recipe candidate list; selecting a cooking recipe according to the recipe candidate list and the user's cuisine preference, obtaining cooking steps based on the cooking recipe, obtaining a seasoning amount based on the cuisine preference and the food weight information, inputting the seasoning amount and the cooking steps into the cooking robot to pre-cook the food, and obtaining a pre-cooking result; obtaining user feedback information on the pre-cooking result, adjusting the seasoning amount and the cooking steps based on the feedback information, and then performing formal cooking.

[0007] As a preferred embodiment, the step of obtaining the food type information, food shape information and food weight information of the food to be processed includes: using a camera to capture images of the food to be processed to obtain food image data, inputting the food image data into a preset food recognition model for recognition, and obtaining preliminary food information; performing feature extraction on the preliminary food information through an edge detection algorithm to obtain shape features and type features of the food, using a convolutional neural network to perform shape analysis on the shape features to obtain food shape information; using a support vector machine to classify the type features to obtain food type information; and weighing the food through a weight sensor to obtain food weight information.

[0008] As a preferred scheme, the step of inputting the ingredient type information and the ingredient shape information into a preset recipe database for query to obtain a recipe candidate list includes: matching the ingredient type information and the ingredient shape information using a cosine similarity algorithm to obtain type matching data and shape matching data, processing the type matching data using a dynamic time warping algorithm to obtain processed type data; processing the shape matching data based on principal component analysis to obtain processed shape data, standardizing the processed type data and the processed shape data using a Z-score normalization process to obtain a standardized query index; inputting the standardized query index into a preset recipe database for search, in the recipe database, processing the standardized query index using a K-nearest neighbor algorithm to obtain a candidate type index and a candidate shape index, processing the candidate type index using a binary search to obtain the row where the recipe is located, processing the candidate shape index through a hash table to obtain the column where the recipe is located; cross-referencing the recipe row and the recipe column through a cross-reference technology to obtain a recipe candidate list.

[0009] As a preferred scheme, the step of selecting a cooking recipe based on the recipe candidate list and the user's cuisine preference includes: generating cuisine preference data and cuisine preference priority based on the user's historical selection data, matching the cuisine preference data with the recipe candidate list to generate a preference matching list; sorting the preference matching list based on the cuisine preference priority, and selecting the optimal recipe based on the sorting result to obtain a cooking recipe.

[0010] As a preferred scheme, the step of generating cuisine preference data and cuisine preference priority based on the user's historical selection data includes: analyzing the user's historical selection data through a decision tree algorithm to obtain the dish types and preference strengths frequently selected by the user; inputting the dish types and preference strengths into a preset database for storage, and then probabilistically classifying the stored data through a naive Bayes classifier to obtain cuisine preference data and cuisine preference priorities.

[0011] As a preferred embodiment, the step of obtaining the amount of seasoning based on the cuisine preference and the ingredient weight information includes: analyzing the relationship between the cuisine preference and the seasoning type through a linear regression model to obtain the seasoning ratio, applying matrix operations in linear algebra to quantify the ingredient weight information, and inputting the quantified ingredient weight information and the seasoning ratio into a preset linear optimization model to adjust the seasoning ratio to obtain the seasoning amount.

[0012] As a preferred embodiment, the step of obtaining user feedback on the pre-cooking result, adjusting the amount of seasoning and the cooking steps based on the feedback information, and then performing formal cooking includes: obtaining user feedback data on the taste, texture and appearance of the pre-cooking result, inputting the feedback data into a preset feedback model, and outputting feedback analysis results; adjusting the amount of seasoning according to the feedback analysis results to generate an adjusted seasoning formula; adjusting the cooking steps according to the feedback analysis results to generate adjusted cooking steps; and inputting the adjusted seasoning formula and the adjusted cooking steps into the cooking robot for cooking.

[0013] The present application also provides a cooking processing system of a self-learning cooking robot, which is applied to the cooking robot, and the cooking processing system includes: an acquisition unit, which is used to acquire ingredient type information, ingredient shape information and ingredient weight information of the ingredients to be processed, and input the ingredient type information and the ingredient shape information into a preset recipe database for query to obtain a recipe candidate list; a pre-cooking unit, which is used to select a cooking recipe according to the recipe candidate list and the user's cuisine preference, obtain cooking steps based on the cooking recipe, obtain a seasoning amount based on the cuisine preference and the ingredient weight information, input the seasoning amount and the cooking steps into the cooking robot to pre-cook the ingredients, and obtain a pre-cooking result; a feedback unit, which is used to acquire user feedback information on the pre-cooking result, and perform formal cooking after adjusting the seasoning amount and the cooking steps based on the feedback information.

[0014] Compared with the prior art, the present application has the following beneficial effects: timely feedback and high flexibility. By introducing a self-learning mechanism, the cooking robot can automatically select a suitable cooking recipe based on the information of the type, shape and weight of the ingredients, and make personalized adjustments based on the user's cuisine preferences, including not only user feedback in the pre-cooking stage, but also dynamic optimization of the amount of seasoning and cooking steps; through such a self-learning and optimization process, the cooking robot can provide a more intelligent and personalized cooking experience to meet the diverse needs of users; after each cooking, the cooking robot will continuously improve and adjust based on the user's feedback, thereby gradually improving its cooking skills and effects, and improving the problem that in actual applications, it cannot be adjusted according to the changes in different ingredients and the user's immediate feedback. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in 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 described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0016] The structures, proportions, sizes, etc. illustrated in the drawings of this specification are only used to match the contents disclosed in the specification so as to facilitate understanding and reading by persons familiar with this technology. They are not used to limit the conditions under which the present invention can be implemented, and therefore have no substantive technical significance. Any structural modification, change in proportion or adjustment of size, without affecting the effects and purposes that can be achieved by the present invention, should still fall within the scope of the technical contents disclosed by the present invention.

[0017] Figure 1 is a flow chart of a cooking method of a self-learning cooking robot provided in an embodiment of the present invention; Figure 2 It is a schematic block diagram of the structure of a cooking processing system of a self-learning cooking robot provided in an embodiment of the present invention.

[0018] Description of reference numerals: 10. A cooking processing system of a self-learning cooking robot; 11. An acquisition unit; 12. A pre-cooking unit; 13. A feedback unit. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] The flowcharts shown in the accompanying drawings are only examples and do not necessarily include all the contents and operations / steps, nor must they be executed in the order described. For example, some operations / steps may also be decomposed, combined or partially merged, so the actual execution order may change according to actual conditions.

[0021] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0022] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0023] The technical solution of the present invention is further described below with reference to the accompanying drawings and through specific implementation methods.

[0024] Embodiment 1: like Figure 1 As shown, the cooking processing method of the self-learning cooking robot provided in the embodiment of the present application is applied to the cooking robot, and the cooking processing method includes steps S100 to S300.

[0025] Step S100: Obtain the food type information, food shape information and food weight information of the food to be processed, input the food type information and food shape information into a preset recipe database for query, and obtain a recipe candidate list.

[0026] In this step, the cooking robot first uses cameras and sensors to identify the type, shape and weight of the ingredients to be processed. Specifically, the cooking robot uses image recognition technology to obtain the appearance characteristics of the ingredients, measures the actual weight of the ingredients through weight sensors, and inputs this information into a preset recipe database for query to obtain a list of recipe candidates that match these ingredient characteristics.

[0027] For example, if the ingredient to be processed is potato, the cooking robot will identify it as a tuber vegetable, measure its weight and determine that its shape is oval, and then look for recipes containing potatoes in the recipe database, such as potato goulash, French fries, etc.

[0028] Step S200: select a cooking recipe based on the recipe candidate list and the user's cuisine preference, obtain cooking steps based on the cooking recipe, obtain the amount of seasoning based on the cuisine preference and ingredient weight information, input the amount of seasoning and cooking steps into the cooking robot to pre-cook the ingredients, and obtain the pre-cooking results.

[0029] In this step, a suitable cooking recipe is selected from the recipe candidate list according to the user's preset cuisine preference. Specifically, the corresponding seasoning dosage is calculated based on the user's preference for different cuisines such as Chinese, Japanese, and French, as well as the weight information of the ingredients. Then, these cooking steps and seasoning dosage information are input into the cooking robot, and the ingredients are pre-cooked so that the user can preview the cooking effect.

[0030] For example, if the user prefers Chinese cuisine and chooses beef stewed with potatoes, the required amount of soy sauce, salt, sugar and other seasonings will be calculated based on the weight information of the potatoes and beef, and the cooking robot will be guided to pre-cook and output the pre-cooking results for the user's reference.

[0031] Step S300: Obtain user feedback on the pre-cooking result, adjust the amount of seasoning and cooking steps based on the feedback information, and then perform formal cooking.

[0032] In this step, the cooking robot will display the pre-cooking results and obtain user feedback. Specifically, the user can provide voice feedback to the cooking robot through the mobile phone application or directly, pointing out the deficiencies in the pre-cooking results, such as whether the taste is appropriate, whether the cooking time is reasonable, etc. The cooking robot adjusts the amount of seasoning and cooking steps based on user feedback to ensure the best results during formal cooking.

[0033] For example, if the user feels that the pre-cooked potato beef stew tastes bland, the cooking robot will increase the amount of seasoning, adjust the cooking time, and cook it again based on the user's feedback until the user is satisfied.

[0034] In this embodiment, by obtaining the food type information, food shape information and food weight information of the food to be processed, and inputting them into a preset recipe database for query, a recipe candidate list is obtained. Then, a suitable cooking recipe is selected according to the recipe candidate list and the user's cuisine preference, and detailed cooking steps are obtained based on the cooking recipe. Next, the amount of seasoning is obtained according to the cuisine preference and food weight information, and the amount of seasoning and cooking steps are input into the cooking robot to pre-cook the food to obtain the pre-cooking result. Finally, the user's feedback information on the pre-cooking result is obtained, and the amount of seasoning and cooking steps are adjusted based on the feedback information, and finally the formal cooking is performed. Through the above workflow, the cooking steps and the amount of seasoning can be self-learned and optimized according to the characteristics of different food materials and the personalized needs of users, thereby improving the flexibility and adaptability of cooking. At the same time, through iterative optimization of user feedback, higher cooking quality and user satisfaction are achieved, thereby greatly improving the practical value of the cooking robot, and improving the problem that in actual application, it is impossible to adjust according to the changes of different food materials and the instant feedback of users.

[0035] Embodiment 2: In step S100, the step of obtaining the food type information, food shape information and food weight information of the food to be processed includes: using a camera to capture images of the food to be processed to obtain food image data, inputting the food image data into a preset food recognition model for recognition, and obtaining preliminary food information.

[0036] The image of the food to be processed is obtained through the camera. Specifically, the multi-angle image of the food is captured by the high-definition camera to ensure the integrity and clarity of the image. Then, the collected food image data is transmitted to the preset food recognition model, which is processed based on the convolutional neural network (CNN) to extract the characteristic information of the food, thereby obtaining the preliminary information of the food.

[0037] For example, a camera captures an image of an apple and inputs the image data of the apple from multiple angles into the recognition model. The model then outputs preliminary information that the ingredient is an apple.

[0038] The edge detection algorithm is used to extract features from the preliminary information of the ingredients to obtain the shape features and type features of the ingredients. The convolutional neural network is used to perform shape analysis on the shape features to obtain the shape information of the ingredients. The support vector machine is used to classify the type features to obtain the type information of the ingredients.

[0039] By applying edge detection algorithms, such as Canny edge detection, the edges in the food image are detected. Specifically, by extracting edge information, the outline and background of the food are distinguished, and the shape characteristics of the food are obtained. After that, these shape features are further analyzed using a pre-trained convolutional neural network to determine the specific shape information of the food. At the same time, the type characteristics of the food are classified using a support vector machine (SVM) to accurately identify the type information of the food.

[0040] For example, the outline of a potato is identified through an edge detection algorithm, its elliptical shape is obtained using a convolutional neural network analysis, and it is determined to be a tuberous vegetable after classification using a support vector machine.

[0041] The food is weighed by a weight sensor to obtain the food weight information.

[0042] Weighing is performed through the high-precision weight sensor built into the cooking robot. Specifically, the ingredients are placed on the sensor, which measures and records the weight of the ingredients in real time. These weight data are matched with the ingredient information to obtain accurate ingredient weight information.

[0043] For example, a potato to be processed is placed on the weight sensor, which displays that the weight of the potato is 200 grams.

[0044] In step S100, the ingredient type information and the ingredient shape information are input into a preset recipe database for query, and the step of obtaining a recipe candidate list includes: using a cosine similarity algorithm to match the ingredient type information and the ingredient shape information to obtain type matching data and shape matching data, processing the type matching data through a dynamic time warping algorithm to obtain processed type data; processing the shape matching data based on principal component analysis to obtain processed shape data, and standardizing the processed type data and processed shape data using Z-score normalization to obtain a standardized query index.

[0045] The cosine similarity algorithm is applied to calculate the similarity between the input ingredient type information and shape information and the data in the recipe database. Specifically, this information is converted into vector form and the cosine similarity between them is calculated to obtain preliminary type matching data and shape matching data. Then, the dynamic time warping algorithm is used to refine these matching data to obtain more accurate type data. At the same time, the principal component analysis (PCA) is used to reduce the dimension of the shape matching data to obtain simplified and representative shape data. Finally, the processed type and shape data are standardized into a unified query index through Z-score normalization.

[0046] For example, several possible recipes are matched through the cosine similarity algorithm, and then further processed through the dynamic time warping algorithm and principal component analysis to finally obtain a standardized query index suitable for the current ingredients.

[0047] The standardized query index is input into the preset recipe database for search. In the recipe database, the standardized query index is processed using the K-nearest neighbor algorithm to obtain the candidate category index and the candidate shape index. The candidate category index is processed using binary search to obtain the row where the recipe is located. The candidate shape index is processed through a hash table to obtain the column where the recipe is located.

[0048] The standardized query index is input into the preset recipe database for searching. Specifically, the K-nearest neighbor (KNN) algorithm is applied to calculate the distance between the query index and each data point in the database, and the most similar candidate category index and candidate shape index are determined. Then, the binary search algorithm is used to quickly locate the specific location of the candidate category index and determine the row where the recipe is located. At the same time, the candidate shape index is quickly processed through the hash table index to determine the column where the recipe is located.

[0049] For example, the K-nearest neighbor algorithm is used to find several candidate types and shapes closest to the standardized query index, and then binary search and hash tables are used to quickly locate the specific recipe location.

[0050] The recipe row and recipe column are cross-referenced through the cross-reference technology to obtain a recipe candidate list.

[0051] By applying the cross-reference technology, the rows where the previously determined recipes are located and the columns where the recipes are located are cross-processed. Specifically, the two are combined and cross-matched to obtain a final recipe candidate list.

[0052] For example, by cross-referencing, we can find all recipes that contain potatoes and match the shape feature, generating a candidate list of various potato dishes.

[0053] In step S200, the step of selecting a cooking recipe according to the recipe candidate list and the user's cuisine preference includes: generating cuisine preference data and cuisine preference priority based on the user's historical selection data, matching the cuisine preference data with the recipe candidate list, and generating a preference matching list.

[0054] By analyzing the user's historical selection data, specifically, the user's past cooking choices and preferences are counted, and the cuisine preference data and priorities reflecting the user's preferences are generated. Then, these preference data are matched with the recipe candidate list to generate a preference matching list.

[0055] For example, if a user has frequently chosen Chinese dishes in the past, Chinese recipes that meet the user's preferences will be matched and recommended first.

[0056] The preference matching list is sorted based on the cuisine preference priority, and the optimal recipe is selected based on the sorting result to obtain the cooking recipe.

[0057] By sorting the preference matching list, specifically, arranging the recipes in the list according to the cuisine preference priority, the recipes that best meet the user's preferences are preferentially selected to determine the optimal cooking recipe.

[0058] For example, based on the strength of user preference, recipes that match the Chinese style are placed at the top of the list, and ultimately the best one is selected as the cooking recipe.

[0059] Among them, the step of generating cuisine preference data and cuisine preference priority based on the user's historical selection data includes: analyzing the user's historical selection data through a decision tree algorithm to obtain the dish types and preference strengths frequently selected by the user.

[0060] By applying the decision tree algorithm to analyze the user's historical selection data, specifically, building a decision tree model, analyzing the user's choices in different situations, and identifying the types of dishes that users frequently choose and the strength of their preference.

[0061] For example, through decision tree analysis, it was found that users tend to choose light Chinese dishes for dinner and heavy Western dishes for lunch.

[0062] The dish type and preference strength are input into a preset database for storage, and then the stored data is probabilistically classified through a naive Bayes classifier to obtain the dish preference data and dish preference priority.

[0063] The dish type and preference strength data obtained by decision tree analysis are stored in a preset database. Specifically, a naive Bayes classifier is applied to perform probability classification on these stored data, the probability and priority of each cuisine are calculated, and the cuisine preference data and priority are obtained.

[0064] For example, the user's preference data can be probabilistically classified to generate a preference strength and priority list for each cuisine for subsequent queries.

[0065] In step S200, the step of obtaining the amount of seasoning based on the cuisine preference and the ingredient weight information includes: analyzing the relationship between the cuisine preference and the seasoning type through a linear regression model to obtain the seasoning ratio, applying matrix operations in linear algebra to quantify the ingredient weight information, and inputting the quantified ingredient weight information and the seasoning ratio into a preset linear optimization model to adjust the seasoning ratio to obtain the seasoning amount.

[0066] By establishing a linear regression model, specifically, analyzing the linear relationship between user cuisine preferences and seasoning types, the applicable proportions of various seasonings in different cuisines are calculated. Then, matrix operations in linear algebra are applied to standardize and quantify the weight information of the ingredients to ensure that all data are comparable in the same dimension. Finally, the quantified ingredient weight information and seasoning ratio are input into the preset linear optimization model to accurately adjust the seasoning dosage to obtain the seasoning dosage that meets cooking needs.

[0067] For example, for the user's preferred Chinese-style stewed beef with potatoes, the specific dosage ratio of soy sauce, salt, sugar and other seasonings required is calculated based on the weight of the ingredients and the cuisine preference, and adjusted to the optimal ratio.

[0068] In step S300, the user's feedback information on the pre-cooking results is obtained, and after adjusting the amount of seasoning and the cooking steps based on the feedback information, the formal cooking steps are performed, including: obtaining the user's feedback data on the taste, texture and appearance of the pre-cooking results, inputting the feedback data into a preset feedback model, and outputting the feedback analysis results.

[0069] The user interface collects detailed feedback from users on the pre-cooking results. Specifically, users can provide feedback data on the taste, texture and appearance of dishes through a mobile phone application or by voice input directly to the cooking robot. These data are then input into a preset feedback model, analyzed, and specific feedback analysis results are output.

[0070] For example, if a user reports that the pre-cooked goulash is too bland, the cooking robot will record this information and adjust the amount of seasoning in subsequent steps.

[0071] According to the feedback analysis results, the amount of flavoring agent is adjusted to generate an adjusted flavoring agent formula.

[0072] By analyzing the results of the feedback model, specifically, based on the user's taste preferences and feedback, the dosage ratio of the seasoning is accurately adjusted to generate an optimized seasoning formula to ensure that the taste during actual cooking is more in line with user expectations.

[0073] For example, based on user feedback on the taste of potato beef stew, the ratio of soy sauce and salt is increased to generate a new seasoning recipe.

[0074] According to the feedback analysis results, the cooking steps are adjusted to generate adjusted cooking steps.

[0075] By analyzing user feedback data, specifically, based on user opinions on pre-cooking steps, key steps such as cooking time and heat are adjusted, and new cooking steps are generated to improve the taste and appearance of the finished product.

[0076] For example, if a user reports that the cooking time of potato stewed with beef is too long and the potatoes are overcooked, the cooking time will be shortened, the heat will be adjusted, and new cooking steps will be generated.

[0077] The adjusted seasoning formula and the adjusted cooking steps are input into the cooking robot for cooking.

[0078] By inputting the above-mentioned optimized seasoning formula and cooking steps into the cooking robot, specifically, the robot performs formal cooking according to the updated instructions, ensuring that the final product meets the user's taste and expectations.

[0079] For example, the cooking robot cooks goulash according to new recipes and steps to ensure the best taste and texture.

[0080] In this embodiment, by using a camera to collect images of the food to be processed, using algorithms such as convolutional neural networks and support vector machines to identify and classify the food, obtaining accurate food type information and food shape information, and obtaining food weight information through a weight sensor. At the same time, the food information is processed through a series of algorithms to generate a standardized query index, and a preset recipe database and multiple algorithms are used for matching and querying to obtain a recipe candidate list. The cuisine preference data and priority are generated according to the user's historical selection data, and the optimal cooking recipe is obtained through matching and sorting. In the pre-cooking stage, the amount of seasoning is obtained through a linear regression model and a linear optimization model, and the detailed feedback data of the user is collected after pre-cooking. The feedback model is used to adjust and optimize the amount of seasoning and cooking steps, and finally input into the cooking robot for formal cooking. This method not only realizes the automation and intelligence of the cooking process, but also greatly improves the personalization of cooking and user satisfaction, ensuring that each cooking can be adjusted and optimized in real time according to actual needs to provide the best cooking experience.

[0081] Embodiment 3: like Figure 2 As shown, the present application also provides a cooking processing system 10 of a self-learning cooking robot, which is applied to the cooking robot. The cooking processing system 10 of the self-learning cooking robot includes an acquisition unit 11, a pre-cooking unit 12 and a feedback unit 13.

[0082] The acquisition unit 11 is used to acquire the food type information, food shape information and food weight information of the food to be processed, input the food type information and food shape information into a preset recipe database for query, and obtain a recipe candidate list.

[0083] The pre-cooking unit 12 is used to select a cooking recipe based on the recipe candidate list and the user's cuisine preference, obtain cooking steps based on the cooking recipe, obtain the amount of seasoning based on the cuisine preference and ingredient weight information, input the amount of seasoning and the cooking steps into the cooking robot to pre-cook the ingredients, and obtain the pre-cooking results.

[0084] The feedback unit 13 is used to obtain the user's feedback information on the pre-cooking result, and perform formal cooking after adjusting the amount of seasoning and cooking steps based on the feedback information.

[0085] In this embodiment, a highly intelligent and personalized cooking processing system is realized by integrating the acquisition unit 11, the pre-cooking unit 12 and the feedback unit 13. The acquisition unit 11 accurately acquires the information of the type, shape and weight of the ingredients through the camera and the weight sensor, and uses advanced algorithms such as convolutional neural networks and support vector machines to ensure the accuracy and reliability of the information. This information is input into a preset recipe database, and a standardized query index is generated through the cosine similarity algorithm, the dynamic time warping algorithm and the principal component analysis technology, and then a matching recipe candidate list is obtained from the recipe database.

[0086] The pre-cooking unit 12 selects the best cooking recipe from the recipe candidate list according to the user's cuisine preference, and generates cuisine preference data and priority by analyzing the user's historical selection data to ensure that the recommended recipe meets the user's preferences. The unit also calculates and adjusts the amount of seasoning based on the cuisine preference and ingredient weight information through a linear regression model and a linear optimization model, performs pre-cooking, and obtains preliminary cooking results. The feedback unit 13 collects detailed feedback from the user on the pre-cooking results, including data on taste, mouthfeel, and appearance, analyzes the feedback information through a preset feedback model, and adjusts and optimizes the amount of seasoning and cooking steps accordingly.

[0087] Finally, the adjusted seasoning formula and optimized cooking steps are input into the cooking robot for formal cooking. Through the process of self-learning and continuous optimization, the entire system realizes intelligent adaptation to various ingredients and user preferences, providing a highly personalized and high-quality cooking experience. The system not only improves the flexibility and precision of cooking, meets the diverse needs of users, but also significantly improves user satisfaction, so that every cooking can achieve the desired effect.

[0088] It should be noted that technical personnel in the relevant technical field can clearly understand that, for the convenience and conciseness of description, the specific working process of the system and each unit described above can refer to the corresponding process in the aforementioned cooking processing method embodiment of the self-learning cooking robot, and will not be repeated here.

[0089] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A cooking processing method for a self-learning cooking robot, applied to the cooking robot, characterized in that: The cooking method comprises: Obtaining food type information, food shape information, and food weight information of the food to be processed, inputting the food type information and the food shape information into a preset recipe database for query, and obtaining a recipe candidate list; selecting a cooking recipe according to the recipe candidate list and the cuisine preference of the user, obtaining cooking steps based on the cooking recipe, obtaining a seasoning amount based on the cuisine preference and the food weight information, inputting the seasoning amount and the cooking steps into the cooking robot to pre-cook the food, and obtaining a pre-cooking result; The user's feedback information on the pre-cooking result is obtained, and the amount of the seasoning and the cooking steps are adjusted based on the feedback information before formal cooking is performed.

2. The cooking method of the self-learning cooking robot according to claim 1, characterized in that: The step of obtaining the food type information, food shape information and food weight information of the food to be processed includes: Using a camera to collect images of the food to be processed to obtain food image data, inputting the food image data into a preset food recognition model for recognition to obtain preliminary food information; Extracting features from the preliminary food information using an edge detection algorithm to obtain shape features and type features of the food, performing shape analysis on the shape features using a convolutional neural network to obtain food shape information; and classifying the type features using a support vector machine to obtain food type information; The food is weighed by a weight sensor to obtain the food weight information.

3. The cooking method of the self-learning cooking robot according to claim 1, characterized in that: The step of inputting the food type information and the food shape information into a preset recipe database for query to obtain a recipe candidate list includes: The food type information and the food shape information are matched by using a cosine similarity algorithm to obtain type matching data and shape matching data, and the type matching data is processed by a dynamic time warping algorithm to obtain processed type data; the shape matching data is processed based on principal component analysis to obtain processed shape data, and the processed type data and the processed shape data are standardized by using a Z-score standardization process to obtain a standardized query index; The standardized query index is input into a preset recipe database for searching. In the recipe database, the standardized query index is processed using a K-nearest neighbor algorithm to obtain a candidate category index and a candidate shape index. The candidate category index is processed using a binary search to obtain a row where the recipe is located. The candidate shape index is processed using a hash table to obtain a column where the recipe is located. The row where the recipe is located and the column where the recipe is located are cross-referenced by cross-reference technology to obtain a recipe candidate list.

4. The cooking method of the self-learning cooking robot according to claim 1, characterized in that: The step of selecting a cooking recipe according to the recipe candidate list and the user's cuisine preference comprises: generating cuisine preference data and cuisine preference priority based on the user's historical selection data, matching the cuisine preference data with the recipe candidate list, and generating a preference matching list; The preference matching list is sorted based on the cuisine preference priority, and an optimal recipe is selected based on the sorting result to obtain a cooking recipe.

5. The cooking method of the self-learning cooking robot according to claim 4, characterized in that: The step of generating cuisine preference data and cuisine preference priority based on the user's historical selection data includes: Analyze the user's historical selection data through the decision tree algorithm to obtain the types of dishes that users frequently choose and the strength of their preference; The dish type and preference strength are input into a preset database for storage, and then the stored data is probabilistically classified by a naive Bayes classifier to obtain dish preference data and dish preference priority.

6. The cooking method of the self-learning cooking robot according to claim 1, characterized in that: The step of obtaining the amount of seasoning based on the cuisine preference and the food weight information includes: The relationship between the cuisine preference and the seasoning type is analyzed through a linear regression model to obtain the seasoning ratio, and the matrix operation in linear algebra is used to quantify the food weight information. The quantified food weight information and the seasoning ratio are input into a preset linear optimization model to adjust the seasoning ratio and obtain the seasoning dosage.

7. The cooking method of the self-learning cooking robot according to claim 1, characterized in that: The step of obtaining user feedback on the pre-cooking result, adjusting the amount of the seasoning and the cooking steps based on the feedback information, and then performing formal cooking comprises: Acquiring user feedback data on the taste, texture and appearance of the pre-cooking result, inputting the feedback data into a preset feedback model, and outputting feedback analysis results; According to the feedback analysis result, the amount of the flavoring agent is adjusted to generate an adjusted flavoring agent formula; According to the feedback analysis result, the cooking steps are adjusted to generate adjusted cooking steps; The adjusted seasoning formula and the adjusted cooking steps are input into the cooking robot for cooking.

8. A cooking processing system for a self-learning cooking robot, applied to the cooking robot, characterized in that: The cooking processing system comprises: an acquisition unit, used to acquire food type information, food shape information and food weight information of the food to be processed, input the food type information and the food shape information into a preset recipe database for query, and obtain a recipe candidate list; a pre-cooking unit, configured to select a cooking recipe according to the recipe candidate list and the cuisine preference of the user, obtain cooking steps based on the cooking recipe, obtain a seasoning amount based on the cuisine preference and the food weight information, input the seasoning amount and the cooking steps into the cooking robot to pre-cook the food, and obtain a pre-cooking result; The feedback unit is used to obtain user feedback information on the pre-cooking result, and perform formal cooking after adjusting the amount of the seasoning and the cooking steps based on the feedback information.

Images