A treadmill parameter control method based on scenario recommendation

Abstract

The present application relates to the technical field of program control, and particularly relates to a treadmill parameter control method based on a scenario recommendation. The method comprises the following steps: acquiring the user's step frequency, impact rate and knee joint alignment and other biomechanical parameters in real time, inputting the parameters into a machine learning model trained by a labeled data set, and outputting an injury risk index in real time. The system then performs cascading screening with the injury risk index as the highest priority, and recommends the most suitable training scenario from a structured scenario library. Finally, the system constructs a safety correction factor based on the injury risk index, dynamically corrects the ideal parameter curve of the recommended scenario in real time, and generates the final control parameters. When the deviation exceeds the threshold, the system also issues a biological feedback instruction to the user to guide the user to actively adjust the running posture. The present application ensures that the treadmill parameters can be safely and smoothly self-adaptively intervened in complex scenario training, and effectively prevents overuse injuries.

Description

Technical Field

[0001] This invention relates to the field of program control technology, and in particular to a method for controlling treadmill parameters based on scenario-based recommendations. Background Technology

[0002] Existing treadmill control systems and virtual running software, such as Zwift and Peloton, while significantly enhancing user immersion and training motivation through virtual scene simulation, social competition, and power tracking, suffer from significant flaws in their core control logic. These systems primarily adjust treadmill parameters based on preset training plans or macroscopic physiological indicators, focusing on achieving training goals. However, these software programs fail to acquire and analyze crucial biomechanical parameters during the user's run, making it impossible for existing systems to assess the risk of injury in real time or to provide preventative intervention before such risks materialize, leaving users exposed to potential sports injuries.

[0003] Due to a lack of biomechanical data, current technologies cannot achieve truly personalized and safe adaptive control. For users with abnormal running posture or pre-existing injuries, the system cannot provide early warnings or targeted feedback instructions, such as guiding users to increase stride frequency or avoid knee valgus. This forces users to passively stop training after pain occurs, unable to fundamentally correct incorrect posture, thus hindering long-term, healthy exercise. Summary of the Invention

[0004] To overcome the above shortcomings, this invention provides a treadmill parameter control method based on scenario-based recommendations, which aims to achieve real-time assessment of individual injury risk and adaptive intervention of training scenarios and treadmill parameters with safety as the highest priority.

[0005] In a first aspect, the present invention provides the following technical solution: a treadmill parameter control method based on scenario-based recommendation, comprising:

[0006] Receive and store the user's basic information and exercise risk preferences, and generate default exercise scenarios and corresponding parameter curves;

[0007] The system acquires the user's physiological and biomechanical parameters in real time during operation, and assesses the risk index of damage to specific joints caused by the user's running posture based on the biomechanical parameters.

[0008] Based on the user's exercise risk preferences, physiological parameters, and injury risk index, the system matches and recommends the most suitable training scenario from a pre-set scenario library.

[0009] The damage risk index is used to construct a safety correction factor, which is then used to dynamically correct the parameter curve of the training scenario in real time to generate the final control parameters.

[0010] The treadmill is adjusted based on the final control parameters, a virtual running screen is displayed to the user terminal according to the training scenario, and feedback instructions are provided based on biomechanical parameters.

[0011] Preferably, the biomechanical parameters include at least: cadence, stride length, vertical ground reaction force, ground contact time, and knee alignment data.

[0012] Preferably, the real-time assessment step of the damage risk index includes:

[0013] The real-time acquired biomechanical parameters are used as input to construct a multidimensional feature vector;

[0014] Based on a pre-collected labeled running posture dataset, a machine learning model is trained to establish a correlation model between injury risk and biomechanical characteristics.

[0015] The multidimensional feature vector is input into the machine learning model, which outputs the probability value of the injury risk type of the current running posture in real time, and the probability value is converted into an injury risk index.

[0016] Preferably, the step of matching and recommending the most suitable training scene from a preset scene library includes:

[0017] Each training scenario in the scenario library is structurally labeled, and the structural labels include core risk indicators, physiological load levels, and motion target types;

[0018] The importance of users' exercise risk preferences, physiological parameters, and real-time injury risk index is ranked, and cascaded screening is performed based on the ranking results and the structured tags. The screening results are then used as the most suitable training scenario.

[0019] Preferably, if there are multiple filtering results, a unique recommendation score is calculated based on the historical usage frequency of the scene in the filtering results and the user's completion rate of the scene, and the single scene with the highest score is selected as the final recommendation result; if there are no filtering results, the filtering conditions are relaxed in order of importance from low to high until the scene with the lowest safe load preset in the scene library is matched.

[0020] Preferably, the real-time dynamic correction of the parameter curves of the training scenario includes:

[0021] Obtain the ideal parameter curve preset for the most suitable training scenario, the curve including the target speed and the target slope;

[0022] Based on the real-time damage risk index, a safety correction factor is constructed;

[0023] The ideal parameter curve is adjusted in real time based on the safety correction factor to generate the final control parameters.

[0024] Preferably, real-time adjustment of the ideal parameter curve includes:

[0025] The damage risk index is normalized and mapped to a preset risk range;

[0026] Establish an inverse mapping function between the normalized damage risk index and the safety correction factor;

[0027] The Sigmoid function is used to calculate the safety correction factor, and the ideal parameter curve is adjusted in real time using the safety correction factor to avoid sudden changes in the treadmill parameters.

[0028] Preferably, the step of providing feedback instructions based on biomechanical parameters includes:

[0029] Real-time acquisition and calculation of the deviation between the user's biomechanical parameters and the preset ideal correction target values;

[0030] When the deviation exceeds a preset correction threshold, a biofeedback command is sent to the user terminal in real time, which is used to guide the user to actively adjust their running posture.

[0031] Secondly, the present invention provides the following technical solution: a treadmill parameter control system based on scenario-based recommendation, comprising:

[0032] The default configuration generation module is used to receive and store the user's basic information and exercise risk preferences, and generate default exercise scenarios and corresponding parameter curves.

[0033] The injury risk quantification module is used to acquire the user's physiological and biomechanical parameters in real time during operation, and to assess the injury risk index of the user's running posture to specific joints in real time based on the biomechanical parameters.

[0034] The training scenario matching module is used to match and recommend the most suitable training scenario from a preset scenario library based on the user's exercise risk preferences, physiological parameters and injury risk index.

[0035] The recommended scenario correction module is used to construct a safety correction factor from the damage risk index and dynamically correct the parameter curve of the training scenario in real time to generate the final control parameters.

[0036] The training feedback module is used to adjust the treadmill based on the final control parameters, display a virtual running screen to the user terminal according to the training scenario, and provide feedback instructions based on biomechanical parameters.

[0037] The present invention has the following beneficial effects:

[0038] 1. This invention utilizes a machine learning model trained on a labeled running posture dataset to acquire multi-dimensional biomechanical parameters such as cadence, impact rate, and knee alignment in real time, achieving accurate assessment of running posture risk. By converting the probability values ​​output by the model into quantifiable injury risk indices, this method overcomes the shortcomings of existing technologies that rely solely on macroscopic physiological indicators, providing a method to quantify abstract injury risks.

[0039] 2. This invention establishes a safety priority cascading screening mechanism. This mechanism sets the real-time injury risk index as the highest priority, ensuring that safety factors always take precedence over training objectives and user preferences when recommending scenarios. This risk-based recommendation method fundamentally guarantees the safety and relevance of the scenarios selected by users during exercise, preventing users from aggravating potential injuries due to inappropriate scenario selection.

[0040] 3. This invention constructs a safety correction factor based on an injury risk index for real-time dynamic correction of treadmill speed and incline. This factor is preferably calculated using a sigmoid smoothing function, ensuring the gradual and continuous adjustment of parameters and effectively avoiding secondary injuries caused by sudden parameter changes. Attached Figure Description

[0041] Figure 1 This is a flowchart of a treadmill parameter control method based on scenario-based recommendation proposed in this invention;

[0042] Figure 2 This is a structural diagram of a treadmill parameter control system based on scenario-based recommendation proposed in this invention. Detailed Implementation

[0043] The technical solutions in 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.

[0044] Example 1

[0045] In a first embodiment of the present invention, the present invention provides a treadmill parameter control method based on scenario-based recommendations, such as... Figure 1 As shown, it includes the following steps:

[0046] S1. Receive and store the user's basic information and exercise risk preferences, and generate default exercise scenarios and corresponding parameter curves;

[0047] Specifically, the system receives and stores key user data, including basic information and exercise risk preferences, through the treadmill's human-computer interface or a linked mobile application. Basic information includes the user's age, height, weight, and gender, used to calculate fundamental physiological indicators such as maximum heart rate, basal metabolic rate, and body mass index. Exercise risk preferences include whether the user has a history of injuries to specific joints such as the knee, ankle, or lower back, as well as the user's preferences for training intensity and environment. Based on the user's input, the system selects and loads a default scenario and its parameter curve from a pre-set scenario library. The parameter curve includes the initial target speed and initial target incline.

[0048] S2. Real-time acquisition of the user's physiological and biomechanical parameters during operation, and based on the biomechanical parameters, real-time assessment of the risk index of damage to specific joints caused by the user's running posture.

[0049] Preferably, the biomechanical parameters include at least: cadence, stride length, vertical ground reaction force, ground contact time, and knee alignment data.

[0050] Specifically, biomechanical parameters are closely related to common knee joint diseases. Cadence and stride length directly determine the impact dispersion efficiency of each step. Too low a cadence often leads to an excessively long stride, with the foot landing too far from the body's center of gravity, increasing ground impact and knee extension torque. This is a significant cause of iliotibial band syndrome and patellofemoral pain syndrome. Vertical ground reaction force and ground contact time are the most direct indicators for assessing impact rate and load magnitude. An excessively high impact rate is a major cause of patellofemoral pain syndrome and patellar tendinitis. Knee valgus is the primary biomechanical trigger for iliotibial band syndrome and patellofemoral pain syndrome. Knee valgus means that when the foot lands, the knee collapses inward, causing excessive friction of the iliotibial band on the lateral epicondyle of the femur, or leading to abnormal patellar trajectory and anterior knee pain. Various biomechanical sensors on a treadmill can acquire parameters such as cadence, stride length, vertical ground reaction force, and ground contact time, while motion capture cameras can acquire knee alignment data.

[0051] Through the above steps, real-time, quantitative, and high-precision biomechanical parameters can be obtained, providing a reliable data foundation for subsequent damage risk assessment.

[0052] Preferably, the real-time assessment step of the damage risk index includes:

[0053] The real-time acquired biomechanical parameters are used as input to construct a multidimensional feature vector;

[0054] Based on a pre-collected labeled running posture dataset, a machine learning model is trained to establish a correlation model between injury risk and biomechanical characteristics.

[0055] The multidimensional feature vector is input into the machine learning model, which outputs the probability value of the injury risk type of the current running posture in real time, and the probability value is converted into an injury risk index.

[0056] The system uses the biomechanical parameters collected in real time in step S2 as input to construct a multidimensional feature vector. .

[0057] Running data was collected from a large number of runners, including healthy runners and those known to have specific injury risks. This data was labeled by experts or therapists for each set of running data, with the following categories: Category 0, low-risk / healthy running form; Category 1, high-risk iliotibial band syndrome; Category 2, high-risk patellofemoral pain syndrome; and Category 3, high-risk patellar tendinitis. This labeling ensured that the machine learning model could accurately translate the feature vectors... It is associated with a specific category of injury risk.

[0058] It is preferable to use classification algorithms such as random forest, gradient boosting tree, or support vector machine to train the model to learn the input feature vector. With damage label The nonlinear mapping relationship between them. During the user's running process, the system continuously updates the new multidimensional feature vectors. The input is fed into the trained machine learning model. The model outputs one or more probability values. This indicates that the current running posture belongs to a specific injury risk category. The possibility of [something]. To facilitate subsequent parameter control and scene matching, the system needs to [something] probability values. This is transformed into a unified, quantifiable damage risk index. Usually Zoom to The range. For example: if The risk index is obtained by multiplying it by a scaling factor of 100. ,in This represents the probability value corresponding to iliotibial band syndrome. This indicates the risk index corresponding to iliotibial band syndrome.

[0059] Through the above steps, the machine learning model achieves a real-time nonlinear correlation between biomechanical characteristics and damage risk, ensuring that damage risk assessment can be quantified.

[0060] S3. Based on the user's exercise risk preferences, physiological parameters, and injury risk index, match and recommend the most suitable training scenario from the preset scenario library.

[0061] Preferably, the step of matching and recommending the most suitable training scene from a preset scene library includes:

[0062] Each training scenario in the scenario library is structurally labeled, and the structural labels include core risk indicators, physiological load levels, and motion target types;

[0063] The importance of users' exercise risk preferences, physiological parameters, and real-time injury risk index is ranked, and cascaded screening is performed based on the ranking results and the structured tags. The screening results are then used as the most suitable training scenario.

[0064] Preferably, if there are multiple filtering results, a unique recommendation score is calculated based on the historical usage frequency of the scene in the filtering results and the user's completion rate of the scene, and the single scene with the highest score is selected as the final recommendation result; if there are no filtering results, the filtering conditions are relaxed in order of importance from low to high until the scene with the lowest safe load preset in the scene library is matched.

[0065] Specifically, during system initialization, each training scenario All data is assigned structured attributes and stored in the scenario database. These structured attributes include core risk indicators, physiological load levels, and exercise goal types. Core risk indicators describe the types of injuries that need to be corrected or avoided, such as ITBS correction and high-impact avoidance. Physiological load levels describe the stress the scenario places on heart rate and physical fitness, such as low load or moderate to high intensity. Exercise goal types describe the training objectives of the scenario, such as fat loss, endurance improvement, and basal recovery.

[0066] The system employs a rule engine to implement cascading filtering. Input data is sorted and processed hierarchically according to a preset importance ranking until a unique result is obtained. The system starts from the scenario library and performs filtering in the order of P1, P2, and P3. In one implementation, the sorting is performed according to core risk indicators, physiological load levels, and movement target types. P1: Safety filtering, used to determine the maximum injury risk index for each category in real time. Has the safety threshold been exceeded? This forces the exclusion of all scenarios that would exacerbate the current highest risk, such as excluding steep downhill scenarios when the risk is high, thus obtaining a pool of safe scenarios. P2: Target filtering, used for user-based exercise goals and real-time physiological parameters such as heart rate and fatigue assessment. The physiological load level of the scenarios is determined, and scenarios that do not match the current goal or physiological state are excluded. For example, if the goal is fat loss but the real-time heart rate is low, all low-load scenarios are excluded. This results in a pool of scenarios that are suitable for the goal. P3: Preference filtering and uniqueness, used for user-based sports risk preferences. All scenarios are scored, and the single scenario with the highest score is selected as the final recommendation result. .

[0067] like or If the result is empty, the system automatically discards the P3 and P2 filtering conditions until the scenario with the lowest safe load is matched as the final recommendation, ensuring continuous system operation and user safety. Multiple filtering results: In the P3 filtering, historical usage frequency and historical completion rate are introduced as scoring weights to ensure the uniqueness of the final recommendation.

[0068] By employing a three-stage cascaded screening process based on risk priority, the system ensures that, under complex motion objectives, it can recommend the safest and most suitable training scenario in real time and uniquely.

[0069] S4. Construct a safety correction factor from the damage risk index to dynamically correct the parameter curve of the training scenario in real time, so as to generate the final control parameters.

[0070] Preferably, the real-time dynamic correction of the parameter curves of the training scenario includes:

[0071] Obtain the ideal parameter curve preset for the most suitable training scenario, the curve including the target speed and the target slope;

[0072] Based on the real-time damage risk index, a safety correction factor is constructed;

[0073] The ideal parameter curve is adjusted in real time based on the safety correction factor to generate the final control parameters.

[0074] Preferably, real-time adjustment of the ideal parameter curve includes:

[0075] The damage risk index is normalized and mapped to a preset risk range;

[0076] Establish an inverse mapping function between the normalized damage risk index and the safety correction factor;

[0077] The Sigmoid function is used to calculate the safety correction factor, and the ideal parameter curve is adjusted in real time using the safety correction factor to avoid sudden changes in the treadmill parameters.

[0078] Specifically, the ideal parameter curve for the final recommended scene is obtained from the scene matching module. This curve defines the target speed under risk-free conditions. and target slope Simultaneously, the system receives the damage risk index output in real time from step S3. .

[0079] The system first evaluates the real-time damage risk index. Normalize it and map it to a unified risk interval for function calculation , and obtain . The system establishes an inverse mapping function to ensure that the safety correction factor and are negatively correlated. To ensure the smoothness of the correction process and avoid sudden changes in the treadmill speed and slope, it is preferable to use the Sigmoid function for inverse mapping calculation. This step can be expressed by the following formula:

[0080] ;

[0081] where represents the lowest safety correction limit value, represents the risk critical point, that is, the risk value at which the correction intensity starts to increase rapidly, represents the steepness coefficient, which controls the response speed of the system to risk changes.

[0082] The system uses to adjust the ideal parameter curve in real time to generate the final control parameters sent to the treadmill drive system . Among them, the final speed is the product of the ideal speed and the correction factor, which ensures that when the risk increases, the speed is forced to decrease. This part can be described as:

[0083] ;

[0084] When a specific injury risk index exceeds the preset high-risk threshold, the system will trigger a mandatory slope adjustment. The system will be corrected to a low-impact uphill interval to force the runner to use the hip muscles and reduce the knee joint load and impact. The corrected slope can be described as:

[0085] ;

[0086] where represents the corrective slope, and this value is a static, biomechanics-based safety value, usually set to .

[0087] Through the safety correction factor based on the Sigmoid function, a smooth inverse mapping between risk and parameters is achieved, effectively avoiding sudden changes in treadmill parameters, and at the same time ensuring an adaptive safety control with higher risk and stronger correction.

[0088] S5. Adjust the treadmill according to the final control parameters, display a virtual running screen on the user terminal according to the training scenario, and provide feedback instructions according to the biomechanical parameters.

[0089] Preferably, the step of providing feedback instructions based on biomechanical parameters includes:

[0090] Real-time acquisition and calculation of the deviation between the user's biomechanical parameters and the preset ideal correction target values;

[0091] When the deviation exceeds a preset correction threshold, a biofeedback command is sent to the user terminal in real time, which is used to guide the user to actively adjust their running posture.

[0092] Specifically, the system first determines the ideal correction target value that the user needs to achieve based on the current training scenario and risk type, such as a specific cadence range or a safe knee alignment angle. Then, the system continuously calculates the deviation between the user's current performance and this target value using real-time collected biomechanical parameters. A correction threshold is preset in the system. The feedback instruction model is only activated when the calculated deviation exceeds this threshold.

[0093] Once a deviation triggers a threshold, the system sends real-time biofeedback commands to the user's terminal, such as the treadmill screen, headphones, or wearable devices. The strength of the feedback is determined by the magnitude of the deviation; the larger the deviation, the stronger the command and the more urgent the prompt. The command type is selected based on the biomechanical parameters of the deviation, choosing the most targeted feedback type. Commands for cadence and rhythm are typically issued via auditory signals, guiding the user to speed up or slow down their cadence; commands for posture, such as knee valgus or trunk instability, are primarily issued via visual signals. The screen displays a comparison image or trajectory line between the user's real-time posture and their ideal posture.

[0094] By calculating biomechanical deviations in real time and providing targeted biofeedback, the system can effectively guide users to actively correct their running posture, thereby reducing the risk of sports injuries at the source.

[0095] Example 2

[0096] A running enthusiast developed iliotibial band syndrome due to improper running form, including dynamic knee valgus and low cadence. Current treadmill control software primarily focuses on entertainment, social interaction, and macroscopic physiological indicators, lacking the ability to monitor the biomechanics of the user's running posture. Therefore, when the runner's knee valgus and low cadence triggered the risk of iliotibial band syndrome, the software was unable to adjust the treadmill parameters to correct the incline or reduce the speed, forcing the runner to endure pain or involuntarily stop training.

[0097] To address the aforementioned problems, this invention provides a treadmill parameter control system based on scenario-based recommendations, the structure of which is as follows: Figure 2As shown, the system includes a default configuration generation module, a damage risk quantification module, a training scenario matching module, a recommended scenario correction module, and a training feedback module. The specific implementation process of this system is as follows:

[0098] The default configuration generation module is responsible for system initialization and safety baseline setting. It receives and stores the user's basic information and key exercise risk preferences. Subsequently, this module adjusts the safety intervention threshold for iliotibial band syndrome based on the user's risk preferences, making it more sensitive and thus establishing a personalized safety baseline. Finally, this module generates a default initial exercise scenario and its corresponding parameter curves for initial startup.

[0099] The injury risk quantification module acquires real-time physiological and crucial biomechanical parameters of the user during running, such as knee alignment, cadence, and impact force data. These parameters are then fed into a trained machine learning model to assess and output a real-time injury risk index for specific joints caused by the user's current running posture.

[0100] The training scenario matching module receives the user's exercise risk preferences, physiological parameters, and injury risk index output by the quantification module. The system then performs a cascading filtering process with the injury risk index as the highest priority. Under the premise of ensuring safety, the system comprehensively considers the user's training goals and physical condition, matching and recommending the safest and most suitable unique training scenario from a pre-set scenario library.

[0101] The recommended scenario correction module receives an injury risk index and constructs a safety correction factor based on it. To avoid abrupt changes in treadmill speed and incline, the system uses a smoothing function mechanism to calculate this correction factor. This module uses the correction factor to dynamically adjust the preset ideal parameter curves of the training scenario in real time, such as forcibly reducing speed and adjusting incline, ultimately generating control parameters that are sent to the treadmill drive system.

[0102] The training feedback module receives and executes the final control parameters generated by the correction module, thereby adjusting the treadmill's actual operating state. Simultaneously, it displays a virtual running screen to the user terminal, enhancing immersion. Furthermore, based on the deviation between real-time biomechanical parameters and ideal values, this module sends biofeedback commands to the user, guiding them to actively adjust their running posture and thus fundamentally eliminate injury-causing factors.

[0103] The system empowers the system with decision-making capabilities through a damage risk quantification module, then determines the training objectives through a scenario matching module, and finally provides dual protection for the training process through a scenario recommendation correction module and a training feedback module.

[0104] It is worth noting that although this system / device only discloses the default configuration generation module, the damage risk quantification module, the training scenario matching module, the recommended scenario correction module, and the training feedback module, it does not mean that this device is limited to the above-mentioned basic functional modules. On the contrary, what this invention intends to express is that, based on the above-mentioned basic functional modules, those skilled in the art can arbitrarily add one or more functional modules in combination with existing technology to form an infinite number of embodiments or technical solutions. That is to say, this system / device is open rather than closed. The fact that this embodiment only discloses a few basic functional modules should not be taken as a reason to believe that the scope of protection of the claims of this invention is limited to the above-disclosed basic functional modules.

[0105] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the meaning consistent with their meaning in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless specifically defined.

[0106] For the sake of simplicity, the method embodiments are described as a series of actions. However, those skilled in the art should understand that the embodiments of the present invention are not limited to the described order of actions, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to the embodiments of the present invention.

[0107] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for controlling treadmill parameters based on scenario-based recommendations, characterized in that, include: Receive and store the user's basic information and exercise risk preferences, and generate default exercise scenarios and corresponding parameter curves; The system acquires the user's physiological and biomechanical parameters in real time during operation, and assesses the risk index of damage to specific joints caused by the user's running posture based on the biomechanical parameters. Based on the user's exercise risk preferences, physiological parameters, and injury risk index, the system matches and recommends the most suitable training scenario from a pre-set scenario library. The damage risk index is used to construct a safety correction factor, which is then used to dynamically correct the parameter curve of the training scenario in real time to generate the final control parameters. The treadmill is adjusted based on the final control parameters, a virtual running screen is displayed to the user terminal according to the training scenario, and feedback instructions are provided based on biomechanical parameters.

2. The treadmill parameter control method based on scenario-based recommendation according to claim 1, characterized in that, The biomechanical parameters include at least: cadence, stride length, vertical ground reaction force, ground contact time, and knee alignment data.

3. The treadmill parameter control method based on scenario-based recommendation according to claim 1, characterized in that, The real-time assessment steps for the damage risk index include: The real-time acquired biomechanical parameters are used as input to construct a multidimensional feature vector; Based on a pre-collected labeled running posture dataset, a machine learning model is trained to establish a correlation model between injury risk and biomechanical characteristics. The multidimensional feature vector is input into the machine learning model, which outputs the probability value of the injury risk type of the current running posture in real time, and the probability value is converted into an injury risk index.

4. The treadmill parameter control method based on scenario-based recommendation according to claim 1, characterized in that, The step of matching and recommending the most suitable training scene from a preset scene library includes: Each training scenario in the scenario library is structurally labeled, and the structural labels include core risk indicators, physiological load levels, and motion target types; The importance of users' exercise risk preferences, physiological parameters, and real-time injury risk index is ranked, and cascaded screening is performed based on the ranking results and the structured tags. The screening results are then used as the most suitable training scenario.

5. The treadmill parameter control method based on scenario-based recommendation according to claim 4, characterized in that, If there are multiple filtering results, a unique recommendation score is calculated based on the historical usage frequency of the scene in the filtering results and the user's completion rate of the scene, and the single scene with the highest score is selected as the final recommendation result. If no results are found, the filtering conditions are relaxed sequentially from low to high importance until a scenario with the lowest preset safety load in the scenario library is matched.

6. The treadmill parameter control method based on scenario-based recommendation according to claim 1, characterized in that, The real-time dynamic correction of the parameter curves in the training scenario includes: Obtain the ideal parameter curve preset for the most suitable training scenario, the curve including the target speed and the target slope; Based on the real-time damage risk index, a safety correction factor is constructed; The ideal parameter curve is adjusted in real time based on the safety correction factor to generate the final control parameters.

7. The treadmill parameter control method based on scenario-based recommendation according to claim 6, characterized in that, Real-time adjustment of the ideal parameter curve includes: The damage risk index is normalized and mapped to a preset risk range; Establish an inverse mapping function between the normalized damage risk index and the safety correction factor; The Sigmoid function is used to calculate the safety correction factor, and the ideal parameter curve is adjusted in real time using the safety correction factor to avoid sudden changes in the treadmill parameters.

8. The treadmill parameter control method based on scenario-based recommendation according to claim 1, characterized in that, The provision of feedback instructions based on biomechanical parameters includes: Real-time acquisition and calculation of the deviation between the user's biomechanical parameters and the preset ideal correction target values; When the deviation exceeds a preset correction threshold, a biofeedback command is sent to the user terminal in real time, which is used to guide the user to actively adjust their running posture.

9. A treadmill parameter control system based on scenario-based recommendation, characterized in that, include: The default configuration generation module is used to receive and store the user's basic information and exercise risk preferences, and generate default exercise scenarios and corresponding parameter curves. The injury risk quantification module is used to acquire the user's physiological and biomechanical parameters in real time during operation, and to assess the injury risk index of the user's running posture to specific joints in real time based on the biomechanical parameters. The training scenario matching module is used to match and recommend the most suitable training scenario from a preset scenario library based on the user's exercise risk preferences, physiological parameters and injury risk index. The recommended scenario correction module is used to construct a safety correction factor from the damage risk index and dynamically correct the parameter curve of the training scenario in real time to generate the final control parameters. The training feedback module is used to adjust the treadmill based on the final control parameters, display a virtual running screen to the user terminal according to the training scenario, and provide feedback instructions based on biomechanical parameters.

Images