Intelligent intervention method and device of kitchen electrical equipment and electronic equipment

Abstract

The application relates to an intelligent intervention method and device of a kitchen electrical equipment and an electronic device. The method can comprise the following steps: acquiring multi-dimensional scene information of a preset space in a preset time period; matching the multi-dimensional scene information with a plurality of historical cooking events to obtain a candidate cooking event set; performing health filtering and confidence evaluation processing on historical cooking dish information in the matched historical cooking events to obtain confidence information; and in the case that the confidence information of target historical cooking dish information is greater than or equal to a first preset confidence threshold, controlling the kitchen electrical equipment in the preset space to perform active cooking preparation based on the target historical cooking dish information. According to the technical scheme provided by the application, energy waste and invalid disturbance to the user caused by frequent misoperation can be avoided, active intelligent service can be provided for the user in a high-confidence scene, and the cooking preparation efficiency, equipment operation reliability and overall user experience are improved.

Description

Technical Field

[0001] This application relates to the field of intelligent kitchen appliance control technology, and in particular to an intelligent intervention method, device and electronic device for kitchen appliances. Background Technology

[0002] With the rapid development of the Internet of Things, artificial intelligence, and big data technologies, smart homes have become an inevitable trend in modern family life. As the core of family life, the kitchen's level of intelligence directly affects the user's cooking experience and quality of life. Traditional kitchen appliances, such as smart stoves, range hoods, and ovens, mainly rely on manual operation or simple preset programs, lacking the ability to proactively perceive and make decisions regarding the cooking environment.

[0003] Currently, some smart kitchen appliances on the market use timer or reminder schemes based on fixed rules. Users preset cooking plans via an app or device panel, and the device performs scheduled operations at the designated time, such as preheating or reminding users to defrost. Some products even support setting recurring rules based on the day of the week. This scheme is essentially a time-triggered rule engine, and the rules need to be entered and maintained by the user. It cannot be automatically updated when family structure or daily routines change, resulting in high maintenance costs, and users eventually abandon its use. Furthermore, it is only based on time triggering and cannot detect dynamic changes such as food inventory or the status of family members.

[0004] Some devices or accompanying apps record users' historical usage records, such as frequently used recipes and functions. When users actively open the app or inquire, recipes or functions are recommended based on simple usage frequency statistics. The recommendations are presented in the form of a list, and all subsequent actions still require manual triggering by the user. This solution lacks the ability to pre-heat; and it relies on only a single dimension, time or frequency, failing to integrate multi-dimensional contexts such as taste preferences, ingredient inventory, number of people at home, and weather, resulting in low prediction accuracy; furthermore, it lacks a health constraint mechanism, unable to filter recommendations based on dietary restrictions or allergies, posing a safety risk. Summary of the Invention

[0005] This application provides a method, device, and electronic device for intelligent intervention in kitchen appliances, aiming to at least solve problems in related technologies such as how to accurately predict a user's cooking intentions to achieve proactive cooking preparation. The technical solution of this application is as follows: According to a first aspect of the embodiments of this application, a method for intelligent intervention of kitchen appliances is provided, comprising: Obtain multi-dimensional scene information of a preset space within a preset time period; the preset space refers to the cooking-related area where the kitchen appliances are installed. The multidimensional scene information is matched with multiple historical cooking events in the multidimensional historical cooking memory database to obtain a candidate cooking event set; the candidate cooking event set includes the historical cooking events matched with the multidimensional scene information. The historical cooking dish information in the matched historical cooking events is subjected to health filtering and confidence evaluation to obtain the confidence information of the target historical cooking dish information after health filtering. If the confidence level of the target historical cooking dish information is greater than or equal to a first preset confidence threshold, the kitchen appliances in the preset space are controlled to perform active cooking preparation based on the target historical cooking dish information.

[0006] According to a second aspect of the embodiments of this application, a smart intervention device for kitchen appliances is provided, comprising: The multi-dimensional scene information acquisition module is used to acquire multi-dimensional scene information of a preset space within a preset time period; the preset space refers to the cooking-related area where the kitchen appliance is installed. The candidate cooking event set matching module is used to match the multi-dimensional scene information with multiple historical cooking events in the multi-dimensional historical cooking memory bank to obtain a candidate cooking event set; the candidate cooking event set includes the historical cooking events matched by the multi-dimensional scene information. The confidence assessment module is used to perform health filtering and confidence assessment on the historical cooking information in the matched historical cooking events to obtain the confidence information of the target historical cooking information after health filtering. An active cooking module is used to control kitchen appliances in a preset space to perform active cooking preparation based on the target historical cooking information when the confidence information of the target historical cooking information is greater than or equal to a first preset confidence threshold.

[0007] According to a third aspect of the embodiments of this application, an electronic device is provided, comprising: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the method as described in any one of the first aspects above.

[0008] According to a fourth aspect of the present application, a computer-readable storage medium is provided, wherein when instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform any of the methods described in the first aspect of the present application. According to a fifth aspect of the embodiments of this application, a computer program product is provided, including computer instructions that, when executed by a processor, cause a computer to perform the method described in any one of the first aspects of the embodiments of this application.

[0009] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this application.

[0010] The technical solutions provided by the embodiments of this application bring at least the following beneficial effects: By acquiring multi-dimensional scene information of a preset space within a preset time period and automatically matching it with historical cooking events in a multi-dimensional historical cooking memory bank, a set of candidate cooking events is obtained. This avoids the limitation of traditional kitchen appliances relying solely on a single current instruction or fixed timing rules to work, and realizes the transformation from passive response to active perception, significantly improving the device's ability to perceive and understand potential user needs. By performing health filtering and confidence assessment on historical cooking information in matched historical cooking events, the confidence information of the target historical cooking information after health filtering is obtained. This can effectively remove dishes that do not meet the user's health constraints, ensuring food health and safety. At the same time, it can also effectively suppress low-quality recommendations and avoid interference caused by erroneous recommendations to users. When the confidence level of the target historical cooking dish information is greater than or equal to the first preset confidence threshold, the kitchen appliances in the preset space are controlled to perform proactive cooking preparation based on the target historical cooking dish information. This can avoid energy waste and unnecessary disturbance to users caused by frequent malfunctions, and provide users with seamless proactive intelligent services in high confidence scenarios, significantly improving cooking preparation efficiency, equipment operation reliability and overall user experience.

[0011] Other features and aspects of this application will become clear from the following detailed description of exemplary embodiments with reference to the accompanying drawings. Attached Figure Description

[0012] To more clearly illustrate the technical solutions and advantages in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart illustrating an intelligent intervention method for kitchen appliances according to an exemplary embodiment.

[0014] Figure 2 This is a flowchart illustrating a candidate cooking event matching method according to an exemplary embodiment.

[0015] Figure 3 This is a flowchart illustrating a health filtering method according to an exemplary embodiment.

[0016] Figure 4 This is a block diagram of an intelligent intervention device for a kitchen appliance according to an exemplary embodiment.

[0017] Figure 5 This is a block diagram of an electronic device for intelligent intervention of kitchen appliances, illustrated according to an exemplary embodiment. Figure 1 .

[0018] Figure 6 This is a block diagram of an electronic device for intelligent intervention of kitchen appliances, illustrated according to an exemplary embodiment. Figure 2 . Detailed Implementation

[0019] To enable those skilled in the art to better understand the technical solutions of this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments in the specification, and not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0020] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0021] Various exemplary embodiments, features, and aspects of this application will now be described in detail with reference to the accompanying drawings. The same reference numerals in the drawings denote elements that have the same or similar functions. Although various aspects of the embodiments are shown in the drawings, they are not necessarily drawn to scale unless specifically indicated otherwise.

[0022] The term "exemplary" as used herein means "serving as an example, embodiment, or illustration." Any embodiment illustrated herein as "exemplary" is not necessarily to be construed as superior to or better than other embodiments. The term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships may exist, for example, A and / or B, which can represent: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more of a plurality, for example, including at least one of A, B, and C, which can represent including any one or more elements selected from the set consisting of A, B, and C.

[0023] Unless otherwise specified, the directions in this article should be understood as follows: the direction closer to the user is forward, and the direction farther from the user is backward.

[0024] Furthermore, to better illustrate this application, numerous specific details are provided in the following detailed description. Those skilled in the art should understand that this application can be implemented without certain specific details. In some instances, methods, means, components, and circuits well-known to those skilled in the art have not been described in detail in order to highlight the main points of this application.

[0025] Existing kitchen appliance control methods mainly rely on manual user commands or fixed timing rules. The equipment only performs corresponding actions when the user explicitly requests them, lacking the ability to proactively perceive the cooking environment and predict user intentions. When family members' structures or daily routines change, fixed rules cannot be automatically updated, requiring repeated manual maintenance by the user, resulting in high operating costs. Recommendation systems based on historical records can only provide suggestions when the user actively initiates interaction, failing to intervene before the user has a cooking intention. This means that the preheating time of appliances such as ovens directly affects the dining experience. Furthermore, existing solutions only make inferences based on a single dimension of time or frequency, failing to integrate multi-dimensional contextual information such as food inventory, family members' status, and the external environment, resulting in low prediction accuracy. At the same time, they lack health constraint mechanisms and cannot filter recommendation results based on dietary restrictions, allergens, or chronic diseases, posing health and safety risks.

[0026] To address the aforementioned technical problems, this application provides an intelligent intervention method, device, and electronic device for kitchen appliances. This method represents a technological leap from passive response to active perception, ensuring food safety and health while eliminating preheating time, preventing frequent malfunctions from disturbing users, and improving the efficiency and level of intelligent service in cooking preparation. Specifically, as illustrated in the following embodiments.

[0027] Figure 1 This is a flowchart illustrating an intelligent intervention method for kitchen appliances according to an exemplary embodiment. Figure 1 As shown, it may include the following steps.

[0028] In step S101, multi-dimensional scene information of a preset space within a preset time period is obtained.

[0029] In the embodiments of this specification, the preset event segment can refer to a pre-set time window or time interval. For example, it can be "the past 30 minutes", "the past hour", "from 0:00 on the same day to the present moment", or "the previous complete meal period", etc., and this application does not limit it.

[0030] The preset space can refer to a cooking-related area equipped with kitchen appliances, such as the entire kitchen space, the area where the cooking counter is located, or a specific area equipped with smart kitchen appliances. This application does not limit this. The cooking-related area can refer to a spatial range that is directly or indirectly related to cooking activities, and can include the kitchen, the cooking area of ​​an open kitchen, or the operating area where kitchen appliances are installed.

[0031] Multidimensional scene information can refer to a collection of scene information describing the current or preset time period within a preset control from multiple dimensions.

[0032] In one possible implementation, the system acquires multi-dimensional scene information of a preset space within a preset time period to comprehensively perceive the user's current cooking scenario. Specifically, the preset space can be the entire kitchen space, the area where the cooking counter is located, or a specific area where smart kitchen appliances are installed, with at least one or more kitchen appliances deployed within its area; this application does not limit this.

[0033] By setting a preset space, the collection range of multi-dimensional scene information can be focused on the area directly related to cooking activities, avoiding interference from information in irrelevant areas.

[0034] Multidimensional scene information includes at least one of the following dimensions: Food inventory information is used to characterize the current availability of food supplies within a preset space. This information may include the food name, remaining weight, storage method (e.g., refrigerated, frozen, or room temperature), and freshness rating. For example, food inventory information can be automatically collected via weight sensors, image acquisition devices, or RFID tag readers installed in refrigerators or storage cabinets, or it can be obtained through manual input or voice input by the user.

[0035] Object information is used to characterize the status information of members within a preset space or the household to which the preset space belongs. This information may include the current number of people at home, the roles of the members, such as adults, children, the elderly, and whether there are any special dietary needs. For example, the number of people at home can be determined through the location information of smart door locks, human body sensors, or user mobile terminals, and the roles of members can be determined through user registration information or historical data.

[0036] External environmental information is used to characterize the environmental state outside or inside the preset space. This information may include weather type (sunny, cloudy, rainy, snowy, foggy, etc.), outdoor temperature, indoor temperature, and indoor humidity. For example, weather and outdoor temperature data can be obtained from a meteorological service platform via a network interface, and indoor temperature and humidity data can be collected by temperature and humidity sensors installed in the preset space. This application does not limit the specific content of the multi-dimensional scene information.

[0037] In step S103, the multi-dimensional scene information is matched with multiple historical cooking events in the multi-dimensional historical cooking memory bank to obtain a candidate cooking event set; In the embodiments of this specification, a multidimensional historical cooking memory bank can refer to a database or data structure that stores historical cooking event data. A historical cooking event can refer to a complete record of cooking-related activities that occurred before the current moment, such as cooking time, dishes, ingredients, cooking methods, etc.

[0038] A candidate cooking event set can refer to a collection of historical cooking events that have a high degree of matching with the current multidimensional scene information, including historical cooking events that match the multidimensional scene information.

[0039] Figure 2 This is a flowchart illustrating a candidate cooking event matching method according to an exemplary embodiment. Figure 2 As shown, it includes the following steps.

[0040] Step S201: Input the multi-dimensional scene information into the preset vector encoding model for vector encoding processing to obtain multiple scene feature vectors within a preset time period; Step S203: Perform similarity calculation on multiple scene feature vectors and each historical cooking event in the preset multi-dimensional historical cooking memory bank to obtain multiple initial similarity information. Step S205: The multiple initial similarity information and multiple preset scene weights are weighted and summed to obtain the target similarity information of each historical cooking event and multi-dimensional scene information. Step S207: Combine historical cooking events with target similarity information greater than a preset scene similarity threshold to obtain a candidate cooking event set.

[0041] Specifically, the preset vector encoding model is a pre-trained or configured encoder used to convert raw multi-dimensional scene information, such as food inventory information, object information, and external environment information, into numerical feature vector representations. Since multi-dimensional scene information contains various types of data—for example, food names are textual data, ambient temperature is numerical data, and weather type is categorical data—different data types cannot be directly calculated and compared uniformly. Therefore, the vector encoding model maps them to the same vector space. For example, for food inventory information, the model can encode it as a food category distribution vector and a quantity feature vector; for object information, it can be encoded as a vector representing the number of members and their roles; and for external environment information, it can be encoded as a weather type vector and temperature and humidity numerical vectors. Through the above processing, scene information within a preset time period is converted into multiple scene feature vectors.

[0042] Then, for each dimension of the multi-dimensional scene information, such as the ingredient dimension, member dimension, and environment dimension, the similarity between the scene feature vector at the current moment and the historical vectors of the corresponding dimensions of each historical cooking event in the memory bank is calculated. For example, the cosine similarity algorithm can be used to calculate the cosine value of the angle between two vectors to obtain the initial similarity in that dimension. The closer the similarity value is to 1, the more similar the current scene and the historical event are in that dimension; the closer it is to 0, the greater the difference. This application does not impose any restrictions on this.

[0043] Next, different dimensions of scene information contribute differently to the prediction of cooking intentions, and corresponding weight coefficients are assigned to each dimension for weighted fusion processing. In one example, the ingredient dimension has the highest weight, such as 0.45, the member dimension, such as 0.25, the time dimension, such as 0.2, and the environment dimension, such as 0.1. By weighting and summing the initial similarity of each dimension with the corresponding preset weights, the target similarity information is obtained. This information can comprehensively reflect the overall matching degree between the historical cooking event and the current multi-dimensional scene.

[0044] Finally, a preset scene similarity threshold, such as 0.6, is used to filter historical events with a high degree of matching with the current scene. When the target similarity information of a historical cooking event is greater than the preset scene similarity threshold, the event is considered to have a high similarity with the current scene and is included in the candidate cooking event set; conversely, events below or equal to the preset scene similarity threshold are filtered out and do not enter the subsequent processing flow. Through the above filtering method, the events most relevant to the current scene can be efficiently filtered from a large number of historical cooking events, effectively narrowing the data processing scope of subsequent health filtering and confidence assessment, reducing computational resource consumption, and improving system operating efficiency.

[0045] In step S105, the historical cooking dish information in the matched historical cooking events is subjected to health filtering and confidence evaluation to obtain the confidence information of the target historical cooking dish information after health filtering. In the embodiments of this specification, confidence information can refer to numerical information that quantitatively describes the degree of matching between the candidate cooking dish and the current scene and the degree to which the target object may accept it.

[0046] In one possible implementation, based on the health data of the object information within a preset space, the historical cooking dish information in each matched historical cooking event is subjected to health filtering processing to obtain a target cooking event set; the target cooking dataset includes the target historical cooking dish information after health filtering. Based on the confidence assessment conditions, the confidence assessment process is performed on the target historical cooking dish information after health filtering to obtain the confidence information of each target historical cooking dish information after health filtering.

[0047] Specifically, object information can refer to personnel information related to cooking activities within the preset space, which may include information on family members currently at home, information on diners, or information on specific users, such as information on family members with chronic diseases. Health data refers to restrictive data related to the object's health status, including but not limited to: dietary restrictions (such as religious restrictions, personal dietary restrictions), allergen lists (such as peanut allergy, seafood allergy, gluten intolerance, etc.), chronic disease restrictions (such as the daily sodium intake limit for hypertensive patients, the glycemic index limit for diabetic patients, the purine intake limit for gout patients, etc.), and currently taken medications and their corresponding dietary restrictions (such as avoiding dark green leafy vegetables while taking warfarin). This application does not limit these.

[0048] Figure 3 This is a flowchart illustrating a health filtering method according to an exemplary embodiment. Figure 3 As shown, it includes the following steps.

[0049] Step S301: Determine the health constraints corresponding to the health data based on the health data of the object information; Step S303: Extract the nutritional component characteristic data of historical cooked dishes; Step S305: If the nutritional component characteristic data of historical cooking dish information meets the health constraints, retain the historical cooking dish information, and determine the target historical cooking dish information after health filtering based on the historical cooking dish information that meets the health constraints. Step S307: Construct a target cooking event set based on the target historical cooking dish information; Alternatively, in step S309, if the nutritional component characteristic data of the historical cooking dish information does not meet the health constraints, the historical cooking dish information is removed.

[0050] Specifically, health constraints include, but are not limited to: dietary restrictions (such as the range of prohibited ingredients defined by religious or personal dietary restrictions), allergen exclusion conditions (such as the prohibition of specific allergens such as peanuts, seafood, and gluten), chronic disease restrictions (such as the upper limit of sodium content per meal for hypertensive patients, the upper limit of glycemic index for diabetic patients, and the upper limit of purine content per meal for gout patients), and medication restrictions (such as a list of prohibited ingredients for currently taken medications). These health constraints collectively constitute the criteria for assessing the safety and usability of candidate dishes.

[0051] For each candidate dish in a matched historical cooking event, the system extracts its corresponding nutritional component characteristic data. Nutritional component characteristic data can refer to a set of data describing the nutritional composition and health-related attributes of the dish, and may include, but is not limited to: a list of ingredients contained in the dish, used for comparison with lists of dietary restrictions and allergens; sodium content, used to assess whether it meets hypertension restrictions; glycemic index, used to assess whether it meets diabetes restrictions; purine content, used to assess whether it meets gout restrictions; and data on the content of nutrients such as calories, protein, fat, and carbohydrates. This application does not limit these aspects.

[0052] Next, the nutritional characteristics of the candidate dishes are compared and verified against the aforementioned health constraints one by one. If a dish's ingredient list does not contain any dietary restrictions or allergens, the content of each nutrient does not exceed the corresponding chronic disease restriction threshold, and there are no dietary restrictions between its ingredients and the currently taken medication, then the nutritional characteristics of the dish are determined to meet all health constraints. In this case, the dish passes the health filter, and its corresponding historical cooking information is retained and identified as the target historical cooking dish information. The system summarizes all target historical cooking dish information that passes the health filter to construct a target cooking event set.

[0053] If a candidate dish's ingredient list matches any item on the subject's dietary restrictions or allergen list, or if its sodium content, glycemic index, or purine content exceeds the corresponding chronic disease restriction threshold, or if its ingredients have known dietary contraindications with the subject's current medications, then the dish's nutritional characteristics are deemed to fail the health constraints, meaning the dish does not pass the health filter. In this case, the system directly removes the corresponding historical cooking information from the candidate set, preventing it from proceeding to the subsequent confidence assessment stage, effectively ensuring dietary health and safety.

[0054] After completing the health filtering, the system performs a confidence assessment on each target historical cooking dish in the target cooking event set.

[0055] In one possible implementation, cooking frequency data of historical cooking dishes from a multi-dimensional historical cooking memory bank is obtained; scene similarity data of each target cooking event is determined based on the target cooking event set; cooking preference data is determined based on the cooking flavor information, cooking ingredient information, and cooking method information of the target historical cooking dishes; health gain data is determined based on the nutritional component characteristic data of the target historical cooking dishes; and the scene similarity data, cooking frequency data, cooking preference data, and health gain data are weighted and summed to obtain the confidence information of each target historical cooking dish.

[0056] Specifically, cooking frequency data refers to the number of times a target historical dish has been cooked in a multidimensional historical cooking memory database. The database extracts the historical cooking frequency of each target historical dish within a preset time period. This data reflects the user's actual selection frequency of that dish in their daily life. Higher cooking frequency indicates a higher level of acceptance and liking for the dish. Cooking frequency data can be directly obtained from historical records in the multidimensional historical cooking memory database.

[0057] Then, based on the matching results between each historical cooking event in the target cooking event set and the current scene information, scene similarity data corresponding to each event is obtained. Scene similarity data is used to characterize the degree of matching between the historical scene information corresponding to each target historical cooking event and the current multi-dimensional scene information. The higher the scene similarity, the closer the current scene is to the occurrence context of the historical event, and the greater the reference value of the historical event for predicting the current cooking intention.

[0058] Next, the above three types of information are matched with the user's preset or historically learned taste preference vector, ingredient preference vector, and cooking method preference vector, respectively, to obtain taste matching sub-data, ingredient matching sub-data, and cooking method matching sub-data. These three sub-data are then weighted and fused to obtain the cooking preference data. Here, cooking taste information can refer to information describing the flavor characteristics of a dish, such as ratings for dimensions like spiciness, saltiness, sweetness, acidity, and umami; cooking ingredient information can refer to the main types of ingredients used in the dish; and cooking method information can refer to the cooking techniques used, such as stir-frying, steaming, roasting, stewing, boiling, and frying.

[0059] In one possible implementation, the cooking flavor information is matched with the flavor preference vector to obtain flavor matching sub-data; the cooking method information is matched with the cooking method preference vector to obtain method preference sub-data; the cooking ingredient information is matched with the ingredient preference vector to obtain ingredient preference sub-data; and the flavor matching sub-data, method preference sub-data, and ingredient preference sub-data are weighted and fused to obtain cooking preference data.

[0060] Specifically, the cooking flavor information is obtained from the historical cooking information of the target dish. The cooking flavor information is quantitative data describing the flavor characteristics of the dish. It is represented as a flavor label vector, which includes five dimensions of rating values: spiciness, saltiness, sweetness, sourness, and umami. The rating range of each dimension is, for example, a floating-point number from 0 to 5.

[0061] The taste preference vector is a multidimensional vector used to represent the taste habits of an object, which is either pre-stored in a multidimensional historical cooking memory bank or implicitly learned through historical cooking parameter adjustment records.

[0062] During the matching process, the system calculates the similarity between the flavor annotation vector and the object's flavor preference vector. For example, a cosine similarity algorithm can be used to calculate the cosine of the angle between the two vectors as the flavor matching sub-data. Alternatively, Euclidean distance or Manhattan distance can be used to calculate the degree of difference between the two vectors, and the normalized difference is then inverted to obtain the matching degree. The value of the flavor matching sub-data ranges from 0 to 1. The higher the value, the more the flavor characteristics of the dish match the object's personal taste habits, and the greater the likelihood that the user will choose the dish.

[0063] Secondly, the cooking method information is obtained from the target historical cooking recipe information. Cooking method information refers to the main cooking methods used in the recipe, including but not limited to one or more combinations of the six common cooking methods: stir-frying, steaming, roasting, stewing, boiling, and frying. The cooking method preference vector is a pre-stored six-dimensional weighted vector representing the object's preference for different cooking methods. Each dimension corresponds to a cooking method, and the sum of all dimensions of this vector is normalized to 1.

[0064] During the matching process, the system determines the primary cooking method type for the target dish based on its historical cooking method information. Then, it queries the object's cooking method preference vector for the weight value corresponding to that cooking method type and uses this weight value as the method preference sub-data. If the dish uses a combination of multiple cooking methods, the average or maximum weight value of each method can be used as the method preference sub-data. The value of the method preference sub-data ranges from 0 to 1; a higher value indicates that the cooking method of the dish is more in line with the user's operating habits, and the fewer obstacles the user will encounter when actually preparing it.

[0065] Next, the cooking ingredient information is retrieved from the target historical cooking dish information. Cooking ingredient information refers to the list of main ingredients used in the dish, which may include ingredient names and quantities. The ingredient preference vector is a pre-stored data structure representing the object's degree of preference for different ingredients. It may include a list of positively preferred ingredients (i.e., ingredients the user prefers and their corresponding ratings) and a list of negatively preferred ingredients (i.e., ingredients the user dislikes and their corresponding penalty values).

[0066] During the matching process, the system compares the ingredient list of the target's historical cooked dishes with the object's ingredient preference vector: it calculates the number of ingredients in the ingredient list belonging to the positive preference ingredient list and the weighted sum of their scores to obtain a positive matching score; it also calculates the number of ingredients in the ingredient list belonging to the negative preference ingredient list and the weighted sum of their penalty values ​​to obtain a negative penalty score. The positive matching score and the negative penalty score are combined to obtain the ingredient preference sub-data. For example, the negative penalty score can be subtracted from the positive matching score and normalized. The value of the ingredient preference sub-data ranges from 0 to 1; the higher the value, the more the ingredients used in the dish match the object's ingredient preferences.

[0067] Finally, obtain the preset weight coefficients corresponding to each matching sub-data. For example, the first weight coefficient corresponds to the taste matching sub-data, the second weight coefficient corresponds to the method preference sub-data, and the third weight coefficient corresponds to the ingredient preference sub-data. Each weight coefficient can be preset according to the actual application scenario, and the sum of all weights is 1. For example, it can be set to a weight of 0.5 for the taste matching sub-data, 0.2 for the method preference sub-data, and 0.3 for the ingredient preference sub-data. The specific formula is as follows. CookingPref=ω1×TasteMatch+ω2×MethodMatch+ω3×IngredientMatch; In this dataset, CookingPref represents the cooking preference data, TasteMatch represents the taste matching sub-data, MethodMatch represents the method preference sub-data, and IngredientMatch represents the ingredient preference sub-data. ω1, ω2, and ω3 are the corresponding preset weight coefficients. Through matching and weighted fusion of these three dimensions, the system can integrate the compatibility of dishes with user preferences in terms of taste, cooking method, and ingredients into a comprehensive cooking preference dataset.

[0068] Then, the system acquires the user's real-time health data (such as daily nutrient intake, recent exercise records, weight change trends, and real-time vital signs data), and combines this with the nutritional characteristics data of the target historical cooked dishes (such as calories, protein, fat, dietary fiber, and sodium content) to calculate the degree to which the dishes supplement nutrient deficiencies, their suitability for post-exercise recovery, their alignment with weight management goals, and their suitability for real-time vital signs (such as blood sugar and blood pressure). By combining these sub-scores, health gain data is obtained.

[0069] Finally, the preset weight coefficients corresponding to the data of each of the above evaluation dimensions are obtained. Scene similarity data corresponds to the first weight, i.e., scene similarity weight α; cooking frequency data corresponds to the second weight, i.e., cooking frequency weight β; cooking preference data corresponds to the third weight, i.e., cooking preference weight γ; and health gain data corresponds to the fourth weight, i.e., health gain weight δ. Each weight coefficient can be preset according to the actual application scenario. For example, it can be set to scene similarity weight 0.25, cooking frequency weight 0.15, cooking preference weight 0.30, and health gain weight 0.30, with the sum of all weights being 1. The system calculates the confidence information of each target's historical cooking dish information according to the following formula: Confidence score is calculated as follows: α × Scene similarity data + β × Cooking frequency data + γ × Cooking preference data + δ × Health gain data. Through this weighted summation, the system integrates the evaluation results from multiple dimensions into a comprehensive confidence score. This confidence score is a numerical rating between 0 and 1, comprehensively and objectively reflecting the overall performance of each candidate dish across multiple dimensions, including scene matching, historical popularity, user preference fit, and health gain. A higher confidence score indicates that the system is more confident that the dish is the user's most likely cooking choice at this time.

[0070] In another possible implementation, time decay penalty and conflict penalty can also be obtained, and then the confidence information = α × scene similarity data + β × cooking frequency data + γ × cooking preference data + δ × health gain data is subtracted from the time decay penalty and conflict penalty.

[0071] Specifically, based on the calculated scene similarity data, cooking frequency data, cooking preference data, and health gain data, the system also obtains the time decay penalty term and conflict penalty term corresponding to the historical cooking information of each target. Both the time decay penalty term and the conflict penalty term are negative adjustment factors used to correct the above weighted summation result.

[0072] Confidence information = α × scene similarity data + β × cooking frequency data + γ × cooking preference data + δ × health gain data - ε × time decay penalty - ζ × conflict penalty. ε and ζ are the preset weight coefficients for the time decay penalty and conflict penalty, respectively.

[0073] The time decay penalty term characterizes the aging effect of historical memory records over time. Specifically, the further back in time a historical cooking event occurred, the lower its reference value for predicting current cooking intentions. The system calculates the time decay penalty value based on the time difference between the timestamp of the historical cooking event and the current time, combined with the corresponding decay type of the event (e.g., exponential decay for daily dishes, annual cosine composite decay for seasonal dishes, and impulse activation decay for holiday dishes). The further back in time the historical record is from the current time, the larger its decay penalty value, and the stronger its reduction effect on the confidence score.

[0074] The conflict penalty term characterizes whether there is a resource conflict between the candidate cooking dish and the current scenario. Specifically, the system detects whether the ingredients or equipment resources required for the candidate cooking dish are available in the current scenario: if the required main ingredients have been exhausted by other cooking plans for the day, or the required kitchen appliances are currently occupied by other tasks, a scenario conflict is determined to exist, and a corresponding conflict penalty value is applied. The more severe the conflict, the larger the penalty value, and the stronger the reduction effect on the confidence score.

[0075] By using time decay penalty terms and conflict penalty terms, we can avoid the undue influence of overly distant historical records on current predictions and avoid recommending dishes that are not practically feasible in the current scenario, thereby further improving the accuracy of intent prediction and the rationality of recommendation results.

[0076] In step S107, if the confidence information of the target historical cooking dish information is greater than or equal to the first preset confidence threshold, the kitchen appliances in the preset space are controlled to perform active cooking preparation based on the target historical cooking dish information.

[0077] In the embodiments of this specification, the first preset confidence threshold may refer to a pre-set confidence threshold used to determine whether to perform active cooking preparation.

[0078] Proactive cooking preparation can refer to cooking-related preparatory operations that are actively performed before the target object issues an explicit instruction. Kitchen appliances can refer to kitchen appliances with intelligent control functions that can be used for cooking operations, including but not limited to smart ovens, smart steam ovens, smart stoves, smart microwave ovens, smart air fryers, smart refrigerators, etc., and this application does not limit this.

[0079] In one possible implementation, after obtaining the confidence information corresponding to each candidate cooking recipe, the system compares each confidence information with a first preset confidence threshold. The first preset confidence threshold is a pre-set threshold value used to determine whether to trigger active cooking preparation. This threshold can be flexibly set according to the actual application scenario, for example, it can be set to a value of 0.7 or 0.8, or it can be dynamically adjusted according to the type of kitchen appliance and the degree of impact of operation. This application does not limit this.

[0080] If the confidence level of a target historical cooking dish is greater than or equal to a first preset confidence threshold, it indicates that the system has a high confidence level in predicting that dish, meaning the system is confident that the dish is the most likely cooking option for the user. In this case, the system does not need to wait for the user's active command; it automatically generates the corresponding control command based on the target historical cooking dish information and sends the control command to the target kitchen appliance in the preset space to control the kitchen appliance to perform active cooking preparation.

[0081] The specific details of this proactive cooking preparation are determined based on information recorded in the target historical cooking recipe information, and include, but are not limited to, at least one of the following operations: For equipment preheating, if the target historical cooking dish requires preheating equipment such as an oven, steamer, or air fryer, the system calculates the target temperature for that equipment. For example, if the system determines the preheating temperature based on the recipe, it controls the corresponding kitchen appliance to start its preheating program in advance. For instance, if the system predicts that the user is preparing "roasted chicken wings," and that the dish requires an oven, it automatically preheats the oven to the temperature specified in the recipe (e.g., 200℃). Since ovens and similar equipment typically require 10 to 20 minutes of preheating time, preheating ensures the equipment is ready when the user actually begins cooking, effectively eliminating waiting time and significantly improving the user's cooking experience.

[0082] Preset cooking programs are loaded from local storage or the cloud based on the target historical cooked recipes. These preset programs include, but are not limited to, cooking temperature profiles, cooking time parameters, and device operating modes such as hot air circulation and steam modes. By automatically loading preset programs, users no longer need to repeatedly set up the device and adjust parameters before each cooking session, reducing their operational burden.

[0083] If the target historical cooking dish requires the use of frozen ingredients, the system can determine whether it needs to be thawed in advance based on the comparison between the ingredients required for the recipe and the refrigerator inventory, and send a thaw reminder notification to the user terminal, or control a device with thaw function to automatically start the thaw program. This application does not limit the above content.

[0084] In addition, while controlling kitchen appliances to perform active cooking preparation, the system can also push notifications to the user terminal, informing the user of the currently executed operation and its reason, such as "The oven has been preheated to 200℃, and cooking is expected to begin in 15 minutes." This notification can also include a cancel option; if the user does not agree with the system's prediction, they can cancel the operation by clicking the cancel button. Upon receiving the cancel command, the system immediately stops the device's active operation to avoid wasting resources.

[0085] In one possible implementation, when the confidence level of the target historical cooking dish information is less than a first preset confidence threshold but greater than or equal to a second preset confidence threshold, the interaction module controlling the kitchen appliance sends a recommendation prompt about the historical cooking dish information to family members and listens for feedback instructions from family members regarding the recommendation prompt; the first preset confidence threshold is greater than the second preset confidence threshold. If a feedback instruction confirms the cooking instruction, the kitchen appliance is controlled to load the corresponding preset cooking program; or, if a feedback instruction rejects the cooking instruction, the kitchen appliance is controlled to remain in standby mode.

[0086] Specifically, if the confidence level of the target historical cooking recipe information is less than a first preset confidence threshold but greater than or equal to a second preset confidence threshold, it indicates that the system's prediction of the recipe has a certain degree of reliability, but has not yet reached a level of certainty sufficient for direct automatic execution. In this case, the system does not directly execute proactive cooking preparation, but instead controls the interaction module of the kitchen appliance to send recommended prompts about the historical cooking recipe information to family members and listens for feedback from family members regarding the recommended prompts. The first preset confidence threshold is greater than the second preset confidence threshold. The interaction module refers to the human-computer interface on the kitchen appliance, such as a touch screen, button panel, or a user terminal device connected to the kitchen appliance, such as a mobile app or smart speaker. The recommended prompts include the name of the dish predicted by the system, a summary of the recommendation reason (e.g., "Pork ribs were detected in the refrigerator, and it is currently a weekend dinner time; braised pork ribs are recommended for you"), and confirmation and rejection options for the user.

[0087] After sending the recommendation message, the system enters a waiting state, listening for feedback commands input by family members through the interaction module. If a confirmation cooking command is received within a preset time, such as 30 seconds, indicating that the user approves of the system's recommendation, the system controls the kitchen appliance to load the corresponding preset cooking program and perform the corresponding cooking preparation work. If a rejection cooking command is received, or if no feedback command is received within the preset time, the system controls the kitchen appliance to remain in standby mode, without performing any cooking preparation operations, and records this prediction and user feedback in the log for subsequent model optimization iterations.

[0088] In one possible implementation, if the confidence level of the target historical cooking recipe information is less than a second preset confidence threshold, the kitchen appliance is controlled to remain in standby mode or to display cooking recommendation information in the interactive module.

[0089] Specifically, after obtaining the confidence level of the target historical cooked dish information, if the confidence level is less than a second preset confidence threshold, it indicates that the system's prediction confidence for that dish is low, meaning the matching degree between the current scenario information and historical cooking events is insufficient to support reliable intent prediction. In this case, the system does not trigger any active intervention, keeping the kitchen appliance in standby mode to avoid malfunctions caused by mispredictions under low confidence levels, resulting in energy waste or unnecessary disturbance to the user.

[0090] In one possible implementation, even if the confidence level is less than a second preset confidence threshold, the system does not perform proactive cooking preparation, but it can still display cooking recommendations in the interactive module. Specifically, the system displays recommendations for one or more candidate dishes in a non-intrusive manner, such as in a corner area of ​​the device display screen or in the recommendation section of the app, for the user to actively view and select. If the user is interested in the recommendations, they can actively click to view details and manually trigger the cooking program; if the user is not interested, they can ignore the recommendations, and the device remains in standby mode without causing any interference to the user. Through the organic combination of the above three-level tiered intervention strategies, the system can perform differentiated intervention actions at different confidence levels, effectively balancing the relationship between proactive service and user control, ensuring that opportunities to provide proactive intelligent services are not missed, and that the user experience is not reduced due to frequent mispredictions or low-quality recommendations.

[0091] Figure 4 This is a block diagram illustrating an intelligent intervention device for a kitchen appliance according to an exemplary embodiment. (Refer to...) Figure 4 The device may include: The multi-dimensional scene information acquisition module 401 is used to acquire multi-dimensional scene information of a preset space within a preset time period; the preset space refers to the cooking-related area where the kitchen appliance is installed. The candidate cooking event set matching module 403 is used to match the multi-dimensional scene information with multiple historical cooking events in the multi-dimensional historical cooking memory bank to obtain a candidate cooking event set; the candidate cooking event set includes the historical cooking events matched by the multi-dimensional scene information. In one possible implementation, the candidate cooking event set matching module 403 includes: The scene feature vector determination unit is used to input the multi-dimensional scene information into a preset vector encoding model for vector encoding processing to obtain multiple scene feature vectors within the preset time period. The initial similarity information determination unit is used to perform similarity calculation processing between the multiple scene feature vectors and each historical cooking event in the preset multi-dimensional historical cooking memory bank to obtain multiple initial similarity information. The target similarity information determination unit is used to perform weighted summation processing on the multiple initial similarity information and multiple preset scene weights to obtain the target similarity information that matches each of the historical cooking events with the multidimensional scene information; The candidate cooking event set determination unit is used to combine the historical cooking events whose target similarity information is greater than a preset scene similarity threshold to obtain the candidate cooking event set.

[0092] The confidence assessment module 405 is used to perform health filtering and confidence assessment on the historical cooking dish information in the matched historical cooking events to obtain the confidence information of the target historical cooking dish information after health filtering. In one possible implementation, the confidence assessment module 405 includes: The target cooking event set determination unit is used to perform health filtering processing on the historical cooking dish information in each of the matched historical cooking events based on the health data of the object information in the preset space, so as to obtain the target cooking event set; the target cooking dataset includes the target historical cooking dish information after health filtering. The confidence assessment unit is used to perform confidence assessment processing on the target historical cooking dish information after health filtering based on confidence assessment conditions, so as to obtain the confidence information of each target historical cooking dish information after health filtering.

[0093] In one possible implementation, the confidence assessment module 405 includes: A health constraint determination unit is used to determine the health constraint corresponding to the health data based on the health data of the object information. The feature data extraction unit is used to extract the nutritional component feature data of the historical cooked dishes information; The historical dish information retention unit is used to retain the historical dish information when the nutritional component characteristic data of the historical dish information meets the health constraints, and to determine the target historical dish information after health filtering based on the historical dish information that meets the health constraints. A target cooking event set construction unit is used to construct the target cooking event set based on the target historical cooking dish information; Alternatively, a dish removal unit is used to remove the historical cooking dish information if the nutritional component characteristic data of the historical cooking dish information does not meet the health constraints.

[0094] In one possible implementation, the confidence assessment module 405 includes: The flavor matching sub-data determination unit is used to match the cooking flavor information with the flavor preference vector to obtain flavor matching sub-data; The cooking method preference sub-data determination unit is used to match the cooking method information with the cooking method preference vector to obtain cooking method preference sub-data; The ingredient preference sub-data determination unit is used to match the cooking ingredient information with the ingredient preference vector to obtain ingredient preference sub-data; The cooking preference sub-data determination unit is used to perform weighted fusion processing on the flavor matching sub-data, the method preference sub-data, and the ingredient preference sub-data to obtain the cooking preference data.

[0095] The active cooking module 407 is used to control the kitchen appliances in the preset space to perform active cooking preparation based on the target historical cooking information when the confidence information of the target historical cooking information is greater than or equal to a first preset confidence threshold.

[0096] In one possible implementation, the active cooking module 407 also includes: The recommendation prompt unit is used to control the interaction module of the kitchen appliance to send recommendation prompt information of the historical cooked dish information to family members when the confidence information of the target historical cooked dish information is less than the first preset confidence threshold and greater than or equal to the second preset confidence threshold, and to listen for the feedback instructions of the family members on the recommendation prompt information; the first preset confidence threshold is greater than the second preset confidence threshold; A preset cooking program control unit is used to control the kitchen appliance to load the corresponding preset cooking program when the feedback instruction is a confirmation cooking instruction; Alternatively, a standby state holding unit is used to control the kitchen appliance to remain in standby state when the feedback instruction is a rejection cooking instruction.

[0097] In one possible implementation, the active cooking module 407 also includes: The display recommendation unit is used to control the kitchen appliance to maintain the standby state or display cooking recommendation information in the interaction module when the confidence information of the target historical cooking dish information is less than the second preset confidence threshold.

[0098] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0099] Figure 5 This is a block diagram illustrating an electronic device for intelligent intervention in kitchen appliances according to an exemplary embodiment. The electronic device may be a terminal, and its internal structure diagram may be as follows: Figure 5 As shown, the electronic device includes a processor, memory, network interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for intelligent intervention of kitchen appliances. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the device's casing, or an external keyboard, touchpad, or mouse. Those skilled in the art will understand that Figure 5 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the electronic device to which the present application is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0100] Figure 6 This is a block diagram illustrating an electronic device for intelligent intervention of kitchen appliances according to an exemplary embodiment. The electronic device may be a server, and its internal structure diagram may be as follows: Figure 6As shown, the electronic device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements a smart intervention method for kitchen appliances. Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the electronic device to which the present application is applied. The specific electronic device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements. In an exemplary embodiment, an electronic device is also provided, including: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the intelligent intervention method for kitchen appliances as described in the embodiments of this application.

[0101] In an exemplary embodiment, a computer-readable storage medium is also provided, which, when executed by a processor of an electronic device, enables the electronic device to perform the intelligent intervention method for kitchen appliances according to the embodiments of this application. The computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, or optical data storage device, etc.

[0102] In an exemplary embodiment, a computer program product containing instructions is also provided, which, when run on a computer, causes the computer to execute the intelligent intervention method for kitchen appliances in the embodiments of this application.

[0103] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0104] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0105] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for intelligent intervention of kitchen appliances, characterized in that, include: Obtain multi-dimensional scene information of a preset space within a preset time period; The preset space refers to the cooking-related area where the kitchen appliances are installed; The multidimensional scene information is matched with multiple historical cooking events in the multidimensional historical cooking memory database to obtain a candidate cooking event set; the candidate cooking event set includes the historical cooking events matched with the multidimensional scene information. The historical cooking dish information in the matched historical cooking events is subjected to health filtering and confidence evaluation to obtain the confidence information of the target historical cooking dish information after health filtering. If the confidence level of the target historical cooking dish information is greater than or equal to a first preset confidence threshold, the kitchen appliances in the preset space are controlled to perform active cooking preparation based on the target historical cooking dish information.

2. The intelligent intervention method for kitchen appliances according to claim 1, characterized in that, The multi-dimensional scene information includes object information within the preset space; the process of performing health filtering and confidence assessment on the historical cooking dish information in the matched historical cooking events to obtain the confidence information of the target historical cooking dish information after health filtering includes: Based on the health data of the object information within the preset space, the historical cooking dish information in each of the matched historical cooking events is subjected to health filtering processing to obtain a target cooking event set; the target cooking dataset includes the target historical cooking dish information after health filtering. Based on the confidence assessment conditions, the target historical cooking dish information after health filtering is subjected to confidence assessment processing to obtain the confidence information of each target historical cooking dish information after health filtering.

3. The intelligent intervention method for kitchen appliances according to claim 2, characterized in that, The step involves performing health filtering on the historical cooking dish information in each matched historical cooking event based on the health data of the object information, to obtain a target cooking event set, including: Based on the health data of the object information, determine the health constraints corresponding to the health data; Extract the nutritional component feature data of the historical cooked dishes; If the nutritional component feature data of the historical cooking dish information meets the health constraints, the historical cooking dish information is retained, and the target historical cooking dish information after health filtering is determined based on the historical cooking dish information that meets the health constraints. Based on the target historical cooking dish information, construct the target cooking event set; Alternatively, if the nutritional component characteristics data of the historical cooked dishes do not meet the health constraints, the historical cooked dishes information may be removed.

4. The intelligent intervention method for kitchen appliances according to claim 3, characterized in that, The confidence assessment process, based on confidence evaluation conditions, evaluates the target historical cooking information after health filtering to obtain the confidence information of each target historical cooking information after health filtering, including: Obtain the cooking frequency data of the historical cooking dishes in the multidimensional historical cooking memory database; Based on the target cooking event set, determine the scene similarity data for each target cooking event; Based on the cooking flavor information, cooking ingredient information, and cooking method information of the target historical cooking dishes, cooking preference data is determined; Based on the nutritional component characteristic data of the target historical cooked dishes, determine the health gain data; The scene similarity data, cooking frequency data, cooking preference data, and health gain data are weighted and summed to obtain the confidence information of each target historical cooking dish information.

5. The intelligent intervention method for kitchen appliances according to claim 4, characterized in that, The object information within the preset space includes the object's taste preference vector, cooking method preference vector, and ingredient preference vector. The step of determining cooking preference data based on the cooking flavor information, cooking ingredient information, and cooking method information of the target historical cooking dish information includes: The cooking flavor information is matched with the flavor preference vector to obtain flavor matching sub-data. The cooking method information is matched with the cooking method preference vector to obtain method preference sub-data; The cooking ingredient information is matched with the ingredient preference vector to obtain ingredient preference sub-data; The flavor matching sub-data, the method preference sub-data, and the ingredient preference sub-data are weighted and fused to obtain the cooking preference data.

6. The intelligent intervention method for kitchen appliances according to claim 1, characterized in that, The method further includes: If the confidence level of the target historical cooking dish information is less than the first preset confidence threshold and greater than or equal to the second preset confidence threshold, the interaction module of the kitchen appliance is controlled to send a recommendation prompt message of the historical cooking dish information to family members, and listen for the feedback instructions of the family members regarding the recommendation prompt message; the first preset confidence threshold is greater than the second preset confidence threshold; Upon receiving a feedback instruction confirming a cooking instruction, the kitchen appliance is controlled to load the corresponding preset cooking program. Alternatively, if the feedback instruction is a rejection of cooking, the kitchen appliance can be kept in standby mode.

7. The intelligent intervention method for kitchen appliances according to claim 6, characterized in that, The method further includes: If the confidence level of the target historical cooking recipe information is less than the second preset confidence threshold, the kitchen appliance is controlled to maintain the standby state or to display cooking recommendation information in the interaction module.

8. The intelligent intervention method for kitchen appliances according to claim 1, characterized in that, The step of matching the multi-dimensional scene information with multiple historical cooking events in a multi-dimensional historical cooking memory database to obtain a candidate cooking event set includes: The multi-dimensional scene information is input into a preset vector encoding model for vector encoding processing to obtain multiple scene feature vectors within the preset time period. The similarity calculation is performed between the multiple scene feature vectors and each historical cooking event in the preset multi-dimensional historical cooking memory database to obtain multiple initial similarity information. The multiple initial similarity information and multiple preset scene weights are weighted and summed to obtain the target similarity information that matches each of the historical cooking events with the multidimensional scene information; The historical cooking events whose target similarity information is greater than a preset scene similarity threshold are combined to obtain the candidate cooking event set.

9. An intelligent intervention device for kitchen appliances, characterized in that, include: The multi-dimensional scene information acquisition module is used to acquire multi-dimensional scene information of a preset space within a preset time period; The preset space refers to the cooking-related area where the kitchen appliances are installed; The candidate cooking event set matching module is used to match the multi-dimensional scene information with multiple historical cooking events in the multi-dimensional historical cooking memory bank to obtain a candidate cooking event set; the candidate cooking event set includes the historical cooking events matched by the multi-dimensional scene information. The confidence assessment module is used to perform health filtering and confidence assessment on the historical cooking dish information in the matched historical cooking events to obtain the confidence information of the target historical cooking dish information after health filtering. An active cooking module is used to control kitchen appliances in a preset space to perform active cooking preparation based on the target historical cooking information when the confidence information of the target historical cooking information is greater than or equal to a first preset confidence threshold.

10. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the intelligent intervention method for kitchen appliances as described in any one of claims 1 to 8.

Images