Search

CN-121983245-A - Modularized food culture system based on intelligent sensing and preparation method thereof

CN121983245ACN 121983245 ACN121983245 ACN 121983245ACN-121983245-A

Abstract

The application relates to the technical field of large health foods, in particular to an intelligent sensing-based modularized food system and a preparation method thereof, wherein the system comprises an intelligent sensing module, a constitution identification and formula decision module, a food product preparation module and a timing management module, wherein physiological and environmental data of a user are collected through multi-source sensor fusion, a three-part medicine and food homologous formula system and a four-section variable temperature gradient precise extraction process are provided, a dynamic constitution identification model based on a federal learning framework is utilized for precise analysis, a reference formula can be subjected to personalized adaptation through a generation and fine adjustment technology, the extraction process is subjected to real-time simulation and parameter optimization through combining a digital twin and process optimization model, and a taking scheme is planned through a dynamic timing management module with cognitive feedback capability. The application realizes the technical closed loop from intelligent perception, accurate decision making, self-adaptive production to closed loop optimization, and greatly improves the accuracy, effectiveness and user compliance of diet intervention.

Inventors

  • Han Kunzhao

Assignees

  • 惠州市润元科技产业有限公司

Dates

Publication Date
20260505
Application Date
20260123

Claims (9)

  1. 1. The modularized food and beverage system based on intelligent sensing is characterized by comprising an intelligent sensing module, a constitution identification and formula decision module, a food and beverage product preparation module and a time sequence management module, wherein, The intelligent sensing module is used for collecting physiological state data and environment data of a user through a multi-source sensor fusion network, and cleaning and marking quality of the original data by the built-in data credibility evaluation unit; The physique identification and formula decision module is used for carrying out physique analysis on the physiological state data through a dynamic physique identification model constructed based on a federal learning framework, matching a reference formula from a predefined medicinal and edible homologous raw material formula system, and outputting a personalized formula adjustment scheme through a generated formula fine adjustment engine to obtain a target formula; the dietetic product preparation module is used for preparing raw materials in the target formula by adopting a four-section variable-temperature gradient precise extraction process to obtain a preparation product, and integrating a process digital twin model and a process parameter optimization algorithm to realize simulation, prediction and parameter dynamic optimization of the production process; the time sequence management module is used for planning the taking time sequence for the prepared product and dynamically adjusting the time sequence rule according to the feedback data and the physiological response of the user.
  2. 2. The intelligent sensing-based modular food farming system of claim 1, wherein the predefined food and drug homologous raw material formulation system is comprised of a functionally complementary and independently prepared day formulation, person formulation, and ground formulation.
  3. 3. The intelligent sensing-based modular food farming system of claim 1, wherein the analyzing the physiological state data for user physique through a dynamic physique recognition model constructed based on a federal learning framework, matching a benchmark recipe from a predefined pharmaceutical and food homologous raw material recipe system, and outputting a personalized recipe adjustment scheme through a generated recipe fine tuning engine comprises: building a universal constitution identification reference model, training the universal constitution identification reference model on the basis of a federal learning framework in a central server, and performing model personalized fine adjustment under privacy protection by using local data at user terminal equipment to form a dynamically updated user-specific constitution identification model; And inputting the physiological state data into the exclusive physique identification model of the user, inputting the characteristic vector output by the exclusive physique identification model into a condition generation type countermeasure network, and generating a raw material proportion fine adjustment scheme aiming at the current user state.
  4. 4. The intelligent sensing-based modular food farming system of claim 1, wherein the integrated process digital twin model and process parameter optimization algorithm enables simulation, prediction and parameter dynamic optimization of a production process comprising: establishing a digital twin model of key equipment of a four-section variable-temperature gradient accurate extraction process; Inputting the real-time production data and the preset optimization targets into a process optimization algorithm, performing simulation calculation through the digital twin model, and dynamically recommending and adjusting process parameters of each stage.
  5. 5. The intelligent sensing-based modular food curing system of claim 2, wherein the raw material composition and parts by weight ratio of the day formula, the person formula and the ground formula comprise: The heaven part module comprises 8-12 parts of rose, 4-6 parts of chrysanthemum, 2-4 parts of mint, 4-6 parts of lotus leaf, 8-12 parts of poria cocos, 3-5 parts of dried orange peel, 5-7 parts of almond, 2-4 parts of bamboo leaf, 1-3 parts of seville orange flower and 1-2 parts of tea; The human part module comprises 8-12 parts of wild jujube seed, 6-10 parts of longan pulp, 8-12 parts of lily, 8-12 parts of poria cocos, 8-12 parts of lotus seed, 6-8 parts of walnut kernel, 1-2 parts of red ginseng/American ginseng slice, 2-4 parts of dark plum, 1-3 parts of sweet osmanthus and 1-2 parts of tea, wherein the weight ratio of the walnut kernel to the red ginseng/American ginseng slice is (3.5-4.5): 1; the land module comprises 10-14 parts of black sesame, 8-12 parts of rhizoma polygonati, 8-12 parts of medlar, 8-12 parts of mulberry, 13-17 parts of Ficus simplicissima lour, 10-14 parts of Chinese yam, 5-7 parts of black date and 4-6 parts of black tea.
  6. 6. The modularized dietetic preparation method based on intelligent sensing is characterized by adopting a four-section variable-temperature gradient precise extraction process, and classifying and sequentially feeding raw materials of the formulas of the heaven, the human and the earth according to physical textures, comprising the following steps of: Putting and extracting rhizome and seed raw materials in the formula at a first high temperature stage; The fruit and peel raw materials in the formula are put into and extracted in a second medium-temperature stage; Putting and extracting the flower and leaf raw materials in the formula at a third low-temperature stage; And (5) carrying out flavor integration and blending in the final integration stage.
  7. 7. The intelligent sensing-based modularized dietetic preparation method according to claim 6, wherein the four-stage variable-temperature gradient accurate extraction process is characterized in that the extraction temperature in the first high-temperature stage is 98+/-2 ℃, the extraction time is 35-45 minutes, the extraction temperature in the second medium-temperature stage is 88+/-2 ℃, the extraction time is 15-25 minutes, the extraction temperature in the third low-temperature stage is 78+/-2 ℃, the extraction time is 5-10 minutes, the integration temperature in the final integration stage is 70+/-2 ℃, and the integration time is 2-5 minutes.
  8. 8. The method for preparing the modularized dietetic therapy based on intelligent sensing according to claim 6, wherein poria cocos, lotus seeds and red ginseng/American ginseng slices are added in a first high-temperature stage, longan pulp, lily, walnut kernels and dark plums are added in a second medium-temperature stage, and wild jujube kernels and sweet osmanthus are added in a third low-temperature stage when preparing the human formula.
  9. 9. The intelligent sensing-based modularized dietetic preparation method according to claim 6, wherein one or more of kudzuvine root powder, cold extract moringa oleifera leaf powder, acerola cherry powder, konjak powder, honey, mogroside and natural maple syrup are added as natural flavoring agents or form regulators in the final stage of integration or in the subsequent blending stage so as to meet the requirements of different flavors and product forms.

Description

Modularized food culture system based on intelligent sensing and preparation method thereof Technical Field The application relates to the technical field of large health foods, in particular to a modularized diet system based on intelligent sensing and a preparation method thereof. Background At present, dietetic products based on medicinal and edible substances face a series of systematic technical bottlenecks, namely, firstly, single function of the product, mismatching with the composite sub-health state of human bodies, secondly, the lack of quantitative paths for butt joint with modern production technology in traditional health preserving theory, rough compatibility and process, and more prominent contradiction is represented in a preparation link, namely, a one-pot stewing process for generally adopting all raw materials to be boiled for a long time in the industry. The linear thinking mode completely ignores the essential difference of physical textures and chemical component thermal stability of different raw materials, and causes at least three defects of 1) heat-sensitive components (such as volatile oil and partial glycosides) in flower and leaf raw materials are destroyed or dissipated at high temperature, 2) active components in compact rhizome and seed raw materials are extracted incompletely due to insufficient temperature or time, 3) process fluctuation is large, and quality among product batches is unstable. In addition, "good medicine bitter taste" seriously affects user experience and long-term compliance, and product application is disjointed with the circadian rhythm of human body, further weakening the intervention effect. More importantly, the degree of "intellectualization" of the existing scheme is shallow, stays in the data acquisition and simple rule matching stage, and lacks continuous learning capability based on data, accurate modeling and control capability of the production process, and closed loop capability of performing system self-optimization according to effect feedback. This results in a static stiffness of the intervention scheme, and a dynamic accurate adaptation of "one person to one side" cannot be truly achieved. Although intelligent technologies such as federal learning and digital twinning are developed in the respective fields, in the field of large health dietetic industry, there is a lack of systematic solutions capable of deeply coupling the front-edge intelligent technology with a fixed modular formula system designed based on a specific theory (such as three-talent theory) and a special physical production process (such as four-stage variable-temperature gradient extraction) thereof, and forming a complete technical closed loop with continuous optimization capability. The core contradiction that the prior art fails to solve is how a static recipe process system responds to dynamic personal physiological states and how complex personalized decisions precisely drive and optimize the entity production process. Therefore, a solution that can carry out all-round innovation from theory to practice and from design to application and can deeply merge with the leading edge intelligent technology is needed to systematically solve the above-mentioned problems. Disclosure of Invention The application provides a modularized dietetic system based on intelligent sensing, which aims to solve the problems of single function, extensive production process, low user compliance, insufficient intelligent degree of the system, lack of self-adaptive optimization capability and the like of the traditional dietetic product. The embodiment of the first aspect of the application provides an intelligent sensing-based modularized food culture system, which comprises an intelligent sensing module, a constitution identification and formula decision module, a food culture product preparation module and a time sequence management module, wherein the intelligent sensing module is used for collecting physiological state data and environment data of a user through a multi-source sensor fusion network and internally arranging a data credibility evaluation unit to clean and mark quality of original data, the constitution identification and formula decision module is used for carrying out constitution analysis on the physiological state data of the user through a dynamic constitution identification model constructed based on a federal learning framework, matching a reference formula from a predefined medicine and food homologous raw material formula system and outputting an individualized formula adjustment scheme through a generated formula fine adjustment engine to obtain a target formula, the food culture product preparation module is used for preparing raw materials in the target formula through a four-section variable temperature gradient precise extraction process to obtain a prepared product, integrating a process digital twin model and a process parameter optimization algorithm to real