Physique conditioning scheme generation method, system and device and storage medium
By constructing controlled reasoning prompts and security control devices, and combining user characteristics and environmental constraints, the problem of standardization and large-scale application of traditional Chinese medicine constitution conditioning programs has been solved. This has enabled the generation of personalized and reliable conditioning programs, and improved the system's adaptability and output quality.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-31
- Publication Date
- 2026-04-03
AI Technical Summary
Traditional Chinese medicine constitution conditioning programs rely on physician experience, making standardization and large-scale application difficult. Digital systems have rigid structures that cannot handle complex situations involving multiple constitutions. General-purpose language models lack professional TCM knowledge, leading to a high risk of output errors and failing to meet the safety and reliability requirements of health management.
By constructing reasoning prompts that include user feature information, candidate solution sets, and rule templates, the large model is guided to perform controlled reasoning within the scope of professional knowledge. Combined with the solar terms and geographical location as constraints, a comprehensive score is calculated and dynamically prioritized. Safety control devices are set up for verification and rollback, and an adaptive closed-loop structure is constructed.
It generates highly personalized and reliable conditioning plans, improving decision-making quality and generation reliability in complex physical conditions, ensuring the accuracy, stability and security of the output, and possessing online evolution and adaptive control capabilities.
Smart Images
Figure CN121789951A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of health management and artificial intelligence technology, and in particular to a method, system, device and storage medium for generating a physical conditioning program. Background Technology
[0002] Traditional Chinese medicine's constitution-regulating programs mainly rely on the physician's experience. After diagnosis through "observation, auscultation, inquiry, and palpation," suggestions for diet, exercise, or massage are manually formulated based on constitution type, seasonal, and regional characteristics. This method is highly subjective, depends on professional experience, and is difficult to standardize and apply on a large scale.
[0003] To improve recommendation efficiency, some digital systems employ rule-based automatic matching, outputting results through a knowledge base and preset rules, such as "recommending yam porridge if the constitution is Qi deficiency." However, such systems are rigid in structure, making it difficult to handle complex situations involving multiple constitutions, and they also cannot dynamically adjust recommended content based on seasonal changes, geographical differences, or individual conditions.
[0004] With the development of Large Language Models (LLM), health question-answering systems based on general models have emerged, allowing users to obtain generative health advice through natural language input. However, due to the lack of professional TCM knowledge constraints in general LLMs, the results suffer from "illusory" outputs and logical distortions, potentially leading to content errors, taboo conflicts, or out-of-bounds generation risks. Furthermore, the model's inference process and output results are uncontrollable, making it difficult to meet the security and reliability requirements of health management applications. While Retrieval-Enhanced Generation (RAG) methods attempt to incorporate knowledge graphs to alleviate these problems, they still rely on static path matching, making it difficult to handle dynamic scenarios such as multiple constitutions, seasonal changes, and regional differences. Moreover, their inference logic is complex, and their output stability is insufficient. Summary of the Invention
[0005] To overcome the shortcomings of existing technologies, the technical problem to be solved by this invention is to provide a method, system, device, and storage medium for generating a body constitution conditioning program, employing the following technical solution: This invention provides a method for generating a body conditioning program, comprising: S1: Receive the user's characteristic information, which includes at least body type, solar term, and geographical location; S2: Based on the above physical type, retrieve the candidate solution set from the solution knowledge base; S3: Construct reasoning prompts, which include the aforementioned feature information, the aforementioned candidate solution set, and rule templates for instructing the large model to score and prioritize solutions; wherein, the solar terms and geographical location are configured as constraints in the aforementioned reasoning prompts. S4: Input the above reasoning prompts into the large model, and dynamically prioritize the above candidate solution set according to the above rule template to obtain the conditioning solution; S5: Output the above treatment plan and the explanation information corresponding to the above feature information.
[0006] As a further improvement, in step S3, the aforementioned reasoning prompt information also includes an output format template, used to limit the output content to include at least one of the following: a list of solutions, a recommended strength level, execution frequency, and precautions; and The aforementioned reasoning prompts also include taboo rules, which restrict the output of solutions that are incompatible with the user's constitution and prohibit the output of solutions not included in the aforementioned candidate solution set.
[0007] As a further improvement, in step S4, the dynamic priority ranking of candidate solutions includes calculating a comprehensive score. The comprehensive score is determined based on multiple matching parameters, which include at least physical fitness matching degree, environmental matching degree, and contraindication penalty item. The comprehensive score is positively correlated with physical fitness matching degree and environmental matching degree, and negatively correlated with contraindication penalty item. When the score difference between candidate solutions is less than a set threshold, an interactive query is triggered to obtain user preferences, and the scores are updated and re-ranked based on user preferences.
[0008] As a further improvement, in step S4, for those with mixed constitutions, a conflict resolution strategy is adopted, specifically: the priority of candidate solutions that are incompatible with any of the constitutions is downgraded, and the priority of candidate solutions that can adapt to multiple constitutions at the same time is increased. The environmental matching degree mentioned above is determined by the solar term information and geographical environment information associated with the candidate schemes in the above scheme knowledge base.
[0009] Further improvements include: The above treatment plan will be subject to compliance verification. When the verification result does not meet the above taboo rules or exceeds the above candidate solution set, a solution replacement operation is performed, and the output is replaced with a secondary solution that meets the constraints. The reason for the replacement is recorded in the above explanation information. Generate audit logs to record operation records and parameters during scoring calculation, priority adjustment, and scheme replacement processes for traceability.
[0010] Another aspect of the present invention provides a system for generating a constitution conditioning plan, comprising: A data access device is used to acquire user characteristic information and retrieve candidate solution sets from a solution knowledge base; The prompt generation device is connected to the aforementioned data access device and is used to generate structured reasoning prompt information based on the aforementioned feature information, candidate scheme set, and preset rule template. The reasoning execution device is connected to the above-mentioned prompt generation device and is used to receive the above-mentioned reasoning prompt information and output the sorted treatment plan and explanation information through the large model; A safety control device, connected to the aforementioned inference execution device, is used to verify the output conditioning scheme and execute scheme replacement when the verification condition is triggered. The aforementioned safety control device is also connected to the aforementioned prompt generation device and inference execution device to form a feedback control mechanism. The aforementioned safety control device is configured to send the verification result and audit information as feedback signals to the aforementioned prompt generation device and inference execution device to dynamically adjust the generation strategy and scheme sorting logic of subsequent inference prompt information.
[0011] As a further improvement, the data access device also includes an update interface for receiving new scheme data input from outside during system operation and synchronizing the new scheme data to the scheme knowledge base.
[0012] Further improvements include: The strategy management unit is used to configure and manage the system's operating parameters, including at least query trigger thresholds, verification conditions, and fallback strategies. The aforementioned strategy management unit can adaptively adjust the above operating parameters based on system operating history or load status.
[0013] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the aforementioned method for generating a constitution conditioning scheme.
[0014] Furthermore, an electronic device is also provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the aforementioned method for generating a constitution conditioning program.
[0015] Compared with the prior art, the beneficial effects of the present invention are: Firstly, this invention constructs reasoning prompts that include user characteristic information, a set of candidate solutions, and rule templates, and incorporates solar terms and geographical location as constraints to guide a large-scale model to perform controlled reasoning within a pre-defined scope of professional knowledge. This mechanism enables the system to fully integrate individual user characteristics with dynamic environmental context to generate deeply personalized treatment plans, and fundamentally avoids the illusion and erroneous recommendation problems that general-purpose large-scale models are prone to in professional fields, thereby improving the decision-making quality and generation reliability in complex physical conditions.
[0016] Secondly, this invention calculates a comprehensive score for candidate solutions and dynamically prioritizes them. Combined with conflict resolution strategies for individuals with multiple constitutions, the system can automatically balance the conditioning needs of different constitutions. This solution effectively solves the problem of rigid logic and inability to simultaneously consider multiple factors when traditional rule engines handle cases with multiple constitutions. Users can obtain balanced and complete conditioning suggestions without multiple queries or manual combination of solutions, thus significantly improving the rationality and efficiency of the recommendations.
[0017] Third, this invention incorporates a safety control device to verify and roll back the output scheme, forming a feedback control loop with the prompt generation module and the inference execution module. This allows the system to dynamically adjust subsequent inference logic based on the verification results of historical outputs. This feedback mechanism transforms the system from a unidirectional inference process into an adaptive closed-loop structure, enabling it to continuously learn and self-correct during operation, thereby maintaining the accuracy, stability, and security of the scheme output in long-term operation.
[0018] Fourth, this invention establishes an update interface in the data access device and configures a strategy management unit to centrally manage system operating parameters. During operation, the system can receive newly added external solution data and synchronize it to the solution knowledge base in real time. Simultaneously, it adaptively adjusts the query trigger threshold and rollback strategy based on historical interaction information or system load. This design enables the system to possess online evolution and adaptive control capabilities, continuously absorbing new knowledge and optimizing operating parameters without service interruption, thereby maintaining the timeliness of knowledge and the stability of system operation in the long term. Attached Figure Description
[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.
[0020] Figure 1 This is a flowchart of the steps of the method of the present invention; Figure 2 This is a schematic diagram of the structural framework of the system of the present invention. Detailed Implementation
[0021] To facilitate understanding by those skilled in the art, the structure of the present invention will now be described in further detail with reference to the accompanying drawings: The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0022] This invention discloses a method, apparatus, device, and storage medium for generating constitution conditioning plans, aiming to achieve safe, accurate, and personalized constitution conditioning plan recommendations by combining professional TCM knowledge with large language model capabilities.
[0023] like Figure 1 As shown, the present invention provides a method for generating a constitution conditioning program, which can be executed by a constitution conditioning program generation system using electronic devices, specifically by one or more processors within the generation system or device, to at least achieve the following steps: Step S1: During the process of generating a body constitution conditioning plan, the system first obtains the user's characteristic information, which includes at least body constitution type, solar term, and geographical location.
[0024] This step can be achieved in several ways: users can fill out a constitution questionnaire in the mobile health management interface, or the system can automatically determine their constitution type based on their long-term monitoring data. For example, based on parameters such as sleep quality, tongue appearance, pulse, and mood fluctuations, it can identify common constitution categories such as "Qi deficiency," "Yin deficiency," and "damp heat."
[0025] In one embodiment, the system automatically reads the terminal's date and location information, corresponding to two types of environmental variables: solar terms and geographical location. Solar term information reflects climate change trends over time, such as "Beginning of Spring," "Summer Solstice," and "White Dew"; geographical location information reflects environmental characteristics over space, such as humidity, altitude, or differences in sunshine. Together, they constitute the constraints of an individual's external environment.
[0026] Specifically, when a user is located in Hainan and the current solar term is "Great Heat," the system identifies their environment as "high temperature and high humidity." If their constitution is "Yin deficiency with damp heat," the system will automatically suppress recommendations for warming solutions during subsequent reasoning processes to avoid misleading "heat-enhancing" approaches and fundamentally reduce the risk of the model generating incorrect recommendations (i.e., large model illusion).
[0027] In one specific embodiment, the system maintains an internal mapping table of environment and attributes, transforming abstract solar terms and geographical locations into specific, computable medical characteristic parameters. For example, the solar term "Great Heat" is not only a time label, but is internally mapped to {temperature: high; humidity: high; TCM characteristics: ["heat", "dampness"]}; the geographical location "Guangzhou" may be mapped to {climate zone: subtropical; TCM characteristics: ["dampness", "heat"]}, achieving adaptation to the time and place.
[0028] In the above embodiments, the introduction of solar terms and geographical location not only makes the conditioning suggestions more time- and space-specific, but also limits the reasoning boundaries of the language model at the algorithm level, preventing it from generating content that deviates from medical principles.
[0029] Step S2: After identifying the user's body type, the system retrieves a set of candidate solutions from its internal solution knowledge base based on that body type.
[0030] Specifically, the scheme includes multi-source data entries formed by expert annotation and model training. Each scheme entry records attributes such as suitable constitution, suitable solar term, contraindicated constitution, efficacy description and precautions.
[0031] In practical applications, when the system identifies a user as having a "Qi deficiency constitution," it will prioritize searching for conditioning programs that match this constitution, such as "Astragalus and Goji Berry Stewed Chicken Soup" or "Eight Pieces of Brocade Exercises," while filtering out entries with contraindications, such as spicy or excessively warming foods. If the user has a mixed constitution, such as "Qi deficiency with dampness," the system will automatically expand the search criteria to cover composite programs that can accommodate multiple constitutions.
[0032] The data for this solution primarily comes from authoritative TCM classics, textbooks, publicly published academic papers, and experience-based solutions reviewed and confirmed by senior physicians, ensuring the professionalism and reliability of the data. As an extension, the solution can also incorporate model-generated samples that have undergone rigorous manual review to continuously optimize its coverage. These solution entries have undergone structured processing and risk grading to ensure their interpretability and safety. The candidate solution set output by the system at this stage forms the input basis for subsequent large-scale model inference and priority calculation.
[0033] Step S3: After obtaining the candidate solution set, the system enters the inference preparation stage. This stage constructs a structured inference prompt message, which serves as input for the large model's inference calculations.
[0034] Specifically, the inference prompts contain three core elements: user characteristic information, a set of candidate solutions, and a rule template. Among these, the solar term and geographical location information are configured as constraints to limit the generation boundaries during model inference and prevent output from deviating from medical common sense. In implementation, the system structures and encodes each element in an instruction-based manner. For example, constitution type is formatted as `"Constitution = Qi Deficiency"`, solar term information is formatted as `"Solar Term = White Dew"`, and geographical location is abstracted as `"Environment = High Humidity Coastal Area"`. The set of candidate solutions is input in list form, with each item including name, suitable constitution, efficacy description, and contraindication constitution identifier.
[0035] Furthermore, the rule template is used to constrain the model's thought process, and its content includes scoring calculation logic, priority ranking rules, and response format limitations. In one embodiment, the rule template instructs the large model to follow the formula "physical fit × weight 1 + environmental fit × weight 2". The comprehensive score is calculated using the "taboo penalty item" method, rather than freely generating conclusions. The purpose is to transform the generative thinking of large models into structured reasoning behavior, effectively reducing the "illusion" problem, that is, preventing the model from generating solutions that do not exist in the candidate set or violate medical logic.
[0036] To further control output quality, the inference prompts also include output format templates to specify the output content and format of the large model, such as a list of solutions, recommendation strength levels, execution frequency, and precautions. In this way, the system ensures that each output has a consistent structure and interpretability, rather than unstructured content in free text format.
[0037] Furthermore, to prevent output from deviating from medical safety boundaries, the system also embeds taboo rules in the inference prompts. Taboo rules are a set of constraint statements jointly defined by traditional Chinese medicine constitution theory and user health records, used to eliminate content incompatible with the user's constitution. For example, when the user's constitution is "Yin deficiency," the rule will prohibit the output of solutions containing characteristics such as "warming and nourishing," "spicy," and "high-calorie diet."
[0038] During the inference execution phase, these taboo rules are directly read by the model as hard constraints at the input level. Compared with the traditional method of filtering through post-processing, this method can control the inference path before generation, reducing the occurrence of illusions and false recommendations from the source.
[0039] Step S4: After the inference prompts are input into the large model, the system enters the scheme ranking stage. The core task of this stage is to dynamically prioritize candidate schemes based on a comprehensive score calculated from multiple parameters. The comprehensive score consists of three parts: physical fit, environmental fit, and taboo penalty, which respectively reflect the suitability, safety, and rationality of the candidate schemes.
[0040] The body constitution matching score is calculated based on the similarity between the target body constitution label recorded in the candidate solution and the user's body constitution label. If they are completely identical, the score is 1; if they belong to similar body constitutions, such as "Qi deficiency" and "blood deficiency", the score is reduced according to the similarity coefficient.
[0041] Environmental matching score is determined by combining solar term information with geographical environment information. For example, during the "Start of Winter" solar term, in the dry and cold environment of northern regions, solutions with "warming and nourishing" and "moistening and drying" characteristics will receive higher scores, while solutions with "clearing heat and promoting diuresis" characteristics will receive relatively lower scores. This matching score calculation mechanism enables the model inference to consider spatiotemporal conditions, achieving personalized and dynamic adaptation.
[0042] The contraindication penalty is used to reflect medical risks. When a candidate solution contains elements that overlap with the user's physical constitution contraindications, such as "fried food" for an individual with a "damp-heat constitution," the system will introduce a negative penalty in the calculation, thereby reducing the overall score of the solution.
[0043] The final comprehensive score is calculated using a weighted function and updated in real time as the model iterates. The weighting coefficients can be adjusted based on the specific compatibility between the contraindications of physical constitution and the individual's physical constitution, and are not limited here.
[0044] Specifically, in one embodiment, when the score difference between two candidate solutions is less than a set threshold, the system triggers an interactive inquiry mechanism, guiding the user to express their preference through brief questions. For example, when the scores of "Astragalus Chicken Soup" and "Codonopsis and Goji Berry Porridge" are similar, the system can prompt the user to choose a preferred "soup" or "porridge" cooking method. The model then dynamically adjusts the scores and reorders them based on the user's answers, ensuring that the results are both scientific and personalized.
[0045] For cases involving combined constitutions, such as "Qi deficiency with dampness" or "Yin deficiency with cold," the system employs a conflict resolution strategy. This includes two aspects: First, any plan that is contraindicated for any body type will be downgraded in priority, and will not be recommended even if the plan has a high degree of matching in a certain body type dimension. Secondly, for compound solutions that can be adapted to multiple body types, the system automatically gives them higher priority. For example, "Yam and Job's Tears Porridge" has both spleen-strengthening and qi-boosting effects as well as dampness-removing effects, making it suitable for both "qi deficiency" and "dampness excess" body types, thus its priority will be significantly increased.
[0046] This conflict resolution mechanism ensures that the choice of solutions has medical logical coherence, avoiding problems such as "solution conflicts" and "self-contradictory suggestions" that occur in traditional models in scenarios with mixed constitutions, thereby improving the credibility of the generated results.
[0047] Step S5: In the model output stage, the treatment plan generated by the system not only includes a list of plans but also explanatory information corresponding to the user's characteristics. The explanatory information includes the reasons for the plan recommendation, matching scores, and explanations of contraindications and risks, helping users understand the basis for the recommendations and thus enhancing the transparency and verifiability of the results.
[0048] The output results undergo compliance verification through a security control process. The verification process automatically compares the generated content with the prohibition rules and candidate solution set to ensure that the output does not exceed the authorized scope. If an anomaly is detected, such as the output containing a solution not in the candidate set or content violating prohibition constraints, the system will automatically perform a solution replacement operation. The replacement logic is based on the principle of prioritizing the second-best option, that is, selecting the second-highest ranked solution that meets the safety conditions for replacement. Simultaneously, the system automatically records the reason for the replacement in the explanation information, such as "The original solution contained hot ingredients, which is incompatible with a Yin-deficient constitution; it has been replaced with a similar warming and nourishing solution." This mechanism ensures that every recommendation output is traceable, enhancing both security and regulatory compliance.
[0049] Preferably, to ensure process traceability and accountability, the system also generates audit logs. These logs record data such as the comprehensive score calculation process, priority adjustment history, protocol replacement events, and parameter version numbers, for subsequent quality assessments or medical reviews.
[0050] This log mechanism can verify the repeatability and compliance of model inference, which is especially important for large models participating in medical-related decision-making.
[0051] In the above embodiments, audit data is also used by the system for self-correction. When it is found that a certain type of solution is frequently replaced or questioned, the system will input the feedback information into the model for fine-tuning to correct the subsequent reasoning logic.
[0052] like Figure 2 As shown, another aspect of the present invention provides a system for generating body constitution conditioning plans. The system includes a data access device, a prompt generation device, a reasoning execution device, and a security control device. The data access device is responsible for collecting user characteristic information and retrieving candidate plans. It integrates a data parsing interface, which can obtain real-time data from terminal applications, health management platforms, or wearable devices. The system provides an update interface for receiving new plan entries during operation, such as new formulas entered by traditional Chinese medicine experts or new dietary therapy plans verified based on user feedback, and automatically synchronizing them to the plan knowledge base for dynamic expansion.
[0053] The prompt generation device is used to generate inference prompts based on the input data. Its core is a structured template engine, which can assemble physical information, seasonal constraints, geographical parameters, and candidate solutions into an input sequence in a fixed format, ensuring that the large model has a standardized input structure when receiving information.
[0054] The inference execution unit connects to the large model inference engine, responsible for reading templated input and outputting sorting results and explanatory text. This unit can run on a local inference engine or a cloud API interface, supporting model version management and secure isolation to prevent unauthorized models from accessing user health data.
[0055] The safety control device plays a closed-loop control role in the system architecture. It embeds a rule engine and validator to perform logical consistency verification on the model output. When an output anomaly is detected, the device directly invokes the replacement logic and updates the audit log, while simultaneously sending the feedback result to the prompt generation device and the inference execution device, forming a data feedback path. This feedback closed-loop mechanism enables the system to self-adjust, dynamically optimizing the prompt template weights and scoring algorithms based on historical operating states and feedback information, achieving self-adaptation.
[0056] Furthermore, it includes a policy management unit, which is used to uniformly manage system parameters, including query trigger thresholds, verification conditions, and fallback strategies. When the system load is too high or the distribution of user data changes, this unit can automatically adjust operating parameters, such as loosening or tightening the trigger threshold, thereby achieving a balance between performance and accuracy.
[0057] Furthermore, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the aforementioned method for generating a constitution conditioning program. Exemplarily, the computer program can be divided into one or more modules, one or more of which are stored in the memory and executed by the processor to complete the present invention. The one or more modules can be a series of computer program instruction segments capable of performing specific functions. These instruction segments describe the execution process of the computer program in the constitution conditioning program generation device, and their functions correspond to the data access device, prompt generation device, inference execution device, security control device, and strategy management unit in the system embodiments of the present invention.
[0058] The aforementioned processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the constitution conditioning solution generation device, connecting the entire device through various interfaces and lines, and executing computer programs to realize the various functions of the method.
[0059] The memory can be used to store computer programs and / or modules. The processor implements various functions of the constitution conditioning program generation method by running or executing the computer programs and / or modules stored in the memory, and by calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area. The program storage area can store the operating system, at least one application program required for a function, such as data access, prompt generation, and model reasoning. The data storage area can store data created based on device usage, such as user characteristic information, program knowledge base, reasoning prompt information, and audit logs. In addition, the memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disk, RAM, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.
[0060] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the aforementioned method for generating a constitution conditioning scheme.
[0061] Wherein, if the aforementioned devices and modules are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.
[0062] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.
Claims
1. A method for generating a body constitution conditioning program, characterized in that, include: S1: Receive user feature information, which includes at least body type, solar term, and geographical location; S2: Based on the described body type, retrieve a set of candidate solutions from the solution knowledge base; S3: Construct reasoning prompt information, which includes: the feature information, the candidate solution set, and a rule template for instructing the large model to perform solution scoring and priority ranking; wherein, the solar term and geographical location are configured as constraints in the reasoning prompt information; S4: Input the reasoning prompt information into the large model, and dynamically prioritize the candidate solution set according to the rule template to obtain the conditioning solution; S5: Output the conditioning plan and the explanation information corresponding to the feature information.
2. The method for generating a body conditioning plan as described in claim 1, characterized in that, In step S3, the inference prompt information further includes an output format template, which limits the output content to include at least one of the following: a list of solutions, a recommended strength level, execution frequency, and precautions; and The reasoning prompts also include taboo rules, which restrict the output of solutions that are incompatible with the user's constitution and prohibit the output of solutions not included in the candidate solution set.
3. The method for generating a body conditioning plan as described in claim 2, characterized in that, In step S4, the dynamic priority ranking of candidate solutions includes calculating a comprehensive score, which is determined based on multiple matching parameters, including at least physical fitness matching degree, environmental matching degree, and contraindication penalty item; the comprehensive score is positively correlated with physical fitness matching degree and environmental matching degree, and negatively correlated with contraindication penalty item; When the score difference between candidate solutions is less than a set threshold, an interactive query is triggered to obtain user preferences, and the scores are updated and re-ranked based on user preferences.
4. The method for generating a body conditioning plan as described in claim 3, characterized in that, In step S4, for those with mixed constitutions, a conflict resolution strategy is adopted, specifically: the priority of candidate solutions that are incompatible with any of the constitutions is downgraded, and the priority of candidate solutions that can adapt to multiple constitutions at the same time is increased. The environmental matching degree is determined by the solar term information and geographical environment information associated with the candidate schemes in the scheme knowledge base.
5. The method for generating a body conditioning plan as described in claim 4, characterized in that, Also includes: The treatment plan shall be subject to compliance verification; When the verification result does not meet the taboo rule or exceeds the candidate solution set, a solution replacement operation is performed, the output is replaced with a secondary solution that meets the constraints, and the reason for replacement is recorded in the explanation information. Generate audit logs to record operation records and parameters during scoring calculation, priority adjustment, and scheme replacement processes for traceability.
6. A system for generating a constitution conditioning plan, characterized in that, include: A data access device is used to acquire user characteristic information and retrieve candidate solution sets from a solution knowledge base; A prompt generation device, connected to the data access device, is used to generate structured reasoning prompt information based on the feature information, the candidate scheme set, and a preset rule template; The reasoning execution device is connected to the prompt generation device and is used to receive the reasoning prompt information and output sorted conditioning schemes and explanation information through the large model; A security control device, connected to the inference execution device, is used to verify the output conditioning scheme and perform scheme replacement when the verification condition is triggered. The security control device is also connected to the prompt generation device and the inference execution device to form a feedback control mechanism. The security control device is configured to send the verification result and audit information as feedback signals to the prompt generation device and the inference execution device to dynamically adjust the generation strategy and scheme sorting logic of subsequent inference prompt information.
7. The body conditioning program generation system as described in claim 6, characterized in that, The data access device also includes an update interface for receiving newly added scheme data from external input during system operation and synchronizing the newly added scheme data to the scheme knowledge base.
8. The body constitution conditioning program generation system as described in claim 7, characterized in that, Also includes: The strategy management unit is used to configure and manage the system's operating parameters, including at least query trigger thresholds, verification conditions, and fallback strategies. The strategy management unit can adaptively adjust the operating parameters based on the system's operating history or load status.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method for generating a body conditioning program as described in any one of claims 1-5.
10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method for generating a body conditioning program as described in any one of claims 1-5.