Automatic comprehensive scoring method and device for smart sports product training

Abstract

The present invention discloses an automatic comprehensive scoring method and device for smart sports product training, which belongs to the technical field of sports products, and includes the following steps: S1: Real-time collection of the user's motion data, physiological data and posture data when the user is training with the smart sports product, and determination of the real-time data of user training based on the smart sports product; S2: Retrieval, sorting and feature extraction of the real-time data of user training based on the smart sports product. The present invention solves the existing problems of not being able to grasp the user's training situation in time, not being able to automatically comprehensively score according to the user's training situation, not being able to give the user personalized guidance, resulting in poor user experience and poor user training effect. The present invention can understand the user's smart sports product training situation in real time, can grasp the user's training situation in time, and can automatically comprehensively score according to the user's training situation, can give the user personalized guidance, and can improve the user experience and user training effect.

Description

Technical Field

[0001] The present invention relates to the technical field of sports products, and in particular to a method and device for automatic comprehensive scoring of smart sports product training. Background Art

[0002] With the rapid development of computer technology, the importance of science and technology in promoting the development of sports has become increasingly prominent. Digital, networked and intelligent sports product training is becoming a hot topic. Smart sports is the application of new technologies such as the Internet of Things technology, mobile Internet technology, 5G communication technology, etc., with the help of various ubiquitous sensors to achieve user monitoring and perception during smart sports product training. It can also use big data algorithms such as cloud computing for comprehensive integration, analysis and processing, and provide intelligent feedback and decision support for various needs.

[0003] The Chinese patent with publication number CN116663954A discloses a sports stadium recommendation method, device, medium and equipment based on the level of intelligence. The method includes: in response to a venue intelligence level test request, sending pre-configured simulation scene data to each smart system configured in the target sports stadium, wherein the simulation scene data of different smart systems are different; obtaining test feedback data of each smart system; inputting the test feedback data of each smart system into a pre-trained smart level evaluation model to determine the smart score of each smart system; determining the smart result of the target sports stadium according to the smart score of each smart system and the preset weight corresponding to each smart system; sending the venue information of the target sports stadium to the target customer according to the smart result; the smart result of the target sports stadium can be obtained reliably and accurately; however, the patent has the following defects:

[0004] The existing smart sports product cannot understand the user's training status in real time, cannot grasp the user's training status in time, cannot automatically make comprehensive scores based on the user's training status, and cannot provide users with personalized guidance, resulting in a poor user experience and poor user training results. Summary of the invention

[0005] The purpose of the present invention is to provide a method and device for automatic comprehensive scoring of smart sports product training, which can understand the user's smart sports product training situation in real time, can timely grasp the user's training situation, and can automatically comprehensively score according to the user's training situation, can give users personalized guidance, can improve user experience and user training effect, and solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] The automatic comprehensive scoring method for smart sports product training includes the following steps:

[0008] S1: collecting the user's sports data, physiological data and posture data in real time when training with the smart sports product, and determining the user's real-time training data based on the smart sports product;

[0009] S2: Retrieve, sort and extract features of real-time user training data based on smart sports products to determine user training feature data based on smart sports products;

[0010] S3: According to the requirements of automatic comprehensive scoring for smart sports product training, select a suitable model architecture, train the model architecture based on the training set, and determine the automatic comprehensive scoring model based on smart sports products;

[0011] S4: Conduct performance testing and optimization adjustments on the automatic comprehensive scoring model based on smart sports products, and determine the best automatic comprehensive scoring model after optimization and adjustment;

[0012] S5: Predict and score the user training feature data based on the automatic comprehensive scoring model to determine the automatic comprehensive scoring result of the user training based on the smart sports product;

[0013] S6: Develop personalized training guidance plans for users based on smart sports products, provide personalized training guidance for users according to the personalized training guidance plans, and provide users with real-time training feedback.

[0014] According to another aspect of the present invention, there is provided an automatic comprehensive scoring device for smart sports product training, which is used to implement the automatic comprehensive scoring method for smart sports product training as described above, comprising:

[0015] The data collection module is used to collect user training motion data, user training physiological data and user training posture data in real time, and determine the real-time user training data based on the smart sports product;

[0016] A data processing module is used to retrieve, sort and extract features of real-time user training data based on smart sports products, and determine user training feature data based on smart sports products;

[0017] The model training module is used to train the model architecture based on the training set and determine the automatic comprehensive scoring model based on the smart sports product;

[0018] The test optimization module is used to test and adjust the automatic comprehensive scoring model based on smart sports products, and determine the best automatic comprehensive scoring model after optimization and adjustment;

[0019] A comprehensive scoring module is used to predict and score the user training feature data based on the automatic comprehensive scoring model, and determine the automatic comprehensive scoring result of the user training based on the smart sports product;

[0020] The intelligent management and control module is used to provide personalized training guidance for users according to the personalized training guidance plan for users based on smart sports products, and provide users with real-time training feedback.

[0021] Preferably, the data acquisition module includes:

[0022] The motion collection unit is used to collect the user's motion situation in real time when training with the smart sports product, and determine the user's training motion data;

[0023] The physiological collection unit is used to collect the physiological conditions of the user in real time when training with the smart sports product, and determine the user's training physiological data;

[0024] The posture collection unit is used to collect the posture of the user in real time when training with the smart sports product, and determine the user's training posture data;

[0025] Among them, based on the real-time collected user training motion data, user training physiological data and user training posture data, the real-time user training data based on the smart sports product is determined.

[0026] Preferably, the motion acquisition unit comprises:

[0027] A type determination unit, used to obtain the movement frequency data and position data of each body part of the user when the user is training with the smart sports product, and determine the user's movement type based on the movement frequency data and position data;

[0028] A mode determination unit, used for matching the sport type with the sports data collection mode of the smart sports product to obtain a target sports data collection mode;

[0029] A data acquisition unit, used to determine the acquisition sensor that needs to work according to the target motion data acquisition method, and determine the acquisition parameters of the acquisition sensor, and collect data from the user according to the acquisition parameters to obtain sensor data;

[0030] A data processing unit is used to match and divide the sensor data with the movement parts to obtain multiple sensor data groups, determine the movement conversion mode of the corresponding sensor data group based on the movement parts, convert the corresponding sensor data group according to the movement conversion mode to obtain the part movement data, and perform abnormal number deletion and correction processing on the part movement data to obtain the target part movement data;

[0031] A sequence determination unit, for horizontally sorting the target part motion data based on all motion parts, obtaining full motion data with position order, and obtaining a full motion data sequence corresponding to the full motion data according to a time acquisition order;

[0032] The model analysis unit is used to input the full motion data sequence into the motion analysis model to obtain the user's motion characteristics, and use the motion characteristics as the user's final user training motion data.

[0033] Preferably, the data processing module includes:

[0034] A data retrieval unit, used to retrieve real-time training data of users based on smart sports products;

[0035] Obtain real-time user training data based on smart sports products;

[0036] Based on the sequential retrieval method, the real-time data of user training based on smart sports products is retrieved;

[0037] Check the consistency of real-time training data of users based on smart sports products;

[0038] Remove data that is beyond the normal range, logically unreasonable or contradictory in the real-time training data of users based on smart sports products;

[0039] Determine the user training real-time data with data consistency;

[0040] Process invalid and missing values ​​for real-time user training data with data consistency;

[0041] Remove invalid data and missing data that are not valuable for the automatic comprehensive scoring of smart sports product training contained in the real-time data of user training based on smart sports products;

[0042] Identify the real-time user training data that is valuable for the automatic comprehensive scoring of smart sports product training;

[0043] A data sorting unit, used to sort the retrieved user training real-time data;

[0044] Obtain the retrieved real-time user training data that is valuable for the automatic comprehensive scoring of smart sports product training;

[0045] Based on the internal sorting method, sort the retrieved user training real-time data that is valuable for the automatic comprehensive scoring of smart sports product training;

[0046] Determine the user training real-time data with arrangement order;

[0047] A feature extraction unit, used for extracting features from the sorted user training real-time data;

[0048] Obtaining sorted and ordered user training real-time data;

[0049] Perform feature extraction on the sorted user training real-time data with arrangement order;

[0050] Determine user training feature data based on smart sports products.

[0051] Preferably, the model training module includes:

[0052] A data storage unit, used to store user training history data based on smart sports products;

[0053] According to the demand for automatic comprehensive scoring of smart sports product training, historical training data of users based on the smart sports product is obtained, and the obtained historical training data of users based on the smart sports product is stored for future use;

[0054] An extraction and division unit, used for extracting and dividing user training history data based on smart sports products;

[0055] According to the requirements of automatic comprehensive scoring of smart sports product training, the stored user training history data based on smart sports products is extracted;

[0056] Divide the user training history data extracted based on smart sports products;

[0057] Determine the training set and test set;

[0058] Architecture training unit, used to train the model architecture;

[0059] Select the appropriate model architecture based on the requirements for automatic comprehensive scoring of smart sports product training;

[0060] Obtain the training set and train the model architecture based on the training set;

[0061] Determine an automatic comprehensive scoring model based on smart sports products.

[0062] Preferably, the test optimization module includes:

[0063] Performance testing unit, used to test the automatic comprehensive scoring model based on smart sports products;

[0064] Get the test set;

[0065] Obtain an automatic comprehensive scoring model based on smart sports products;

[0066] Perform performance testing on the automatic comprehensive scoring model based on smart sports products based on the test set;

[0067] Determine the performance test results based on the automatic comprehensive scoring model;

[0068] An optimization and adjustment unit, used to adjust the automatic comprehensive scoring model based on smart sports products;

[0069] Obtain performance test results based on the automatic comprehensive scoring model;

[0070] Conduct mining and analysis on performance test results based on automatic comprehensive scoring model;

[0071] Determine the optimization adjustment plan based on the automatic comprehensive scoring model;

[0072] Optimize and adjust the automatic comprehensive scoring model based on smart sports products according to the optimization and adjustment plan based on the automatic comprehensive scoring model;

[0073] Determine the best automatic comprehensive scoring model after optimization and adjustment.

[0074] Preferably, the optimization and adjustment unit includes:

[0075] A mining determination unit is used to obtain performance indicators and their corresponding indicator values ​​from the performance test results of the automatic comprehensive scoring model, and based on the performance indicators and their corresponding indicator values, calculate the mining index K of the automatic comprehensive scoring model according to the following formula;

[0076]

[0077] Where n represents the number of performance indicators, ε i represents the indicator weight of the i-th performance indicator, H i represents the index value of the i-th performance index, M ij represents the correlation value between the i-th performance indicator and the j-th other performance indicator;

[0078] The optimization determination unit is used to calculate the optimization adjustment index G based on the mining index of the automatic comprehensive scoring model and the difference between the index value and the standard index value according to the following formula;

[0079]

[0080] Among them, H 0i represents the standard index value of the i-th performance index;

[0081] The optimization and adjustment unit is used to select a corresponding optimization and adjustment scheme from an optimization and adjustment scheme library based on the optimization and adjustment index to optimize and adjust the automatic comprehensive scoring model based on the smart sports product.

[0082] Preferably, the comprehensive scoring module includes:

[0083] A data input unit, used to input user training feature data based on smart sports products;

[0084] Obtain user training feature data based on smart sports products;

[0085] Input the user training feature data based on smart sports products into the optimal automatic comprehensive scoring model after optimization and adjustment;

[0086] A prediction and scoring unit, used to score user training feature data based on smart sports products;

[0087] Based on the best automatic comprehensive scoring model after optimization and adjustment, the user training feature data based on smart sports products is predicted and scored;

[0088] Determine the automatic comprehensive scoring results of user training based on smart sports products.

[0089] Preferably, the intelligent management and control module includes:

[0090] The training formulation unit is used to formulate personalized training guidance plans for users based on smart sports products;

[0091] Obtain automatic comprehensive scoring results of user training based on smart sports products;

[0092] Mining and analyzing the automatic comprehensive scoring results of user training based on smart sports products;

[0093] Determine personalized training guidance plans for users based on smart sports products;

[0094] An interactive guidance unit for providing personalized guidance for user training;

[0095] Obtain personalized training guidance plans for users based on smart sports products;

[0096] Provide personalized training guidance to users based on the personalized training guidance plan of smart sports products, and provide users with real-time training feedback.

[0097] Compared with the prior art, the present invention has the following beneficial effects:

[0098] The present invention determines the real-time data of user training based on the smart sports product by collecting the motion data, physiological data and posture data of the user when training with the smart sports product in real time, retrieves, sorts and extracts features of the real-time data of user training based on the smart sports product, determines the characteristic data of user training based on the smart sports product, predicts and scores the characteristic data of user training based on the automatic comprehensive scoring model, determines the automatic comprehensive scoring result of user training based on the smart sports product, formulates a personalized guidance plan for user training based on the smart sports product, provides personalized guidance for user training according to the personalized guidance plan for user training, provides real-time training feedback for users, can understand the user's smart sports product training situation in real time, can grasp the user's training situation in time, can perform automatic comprehensive scoring according to the user's training situation, can give users personalized guidance, and can improve user experience and user training effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] Figure 1 This is a flow chart of the automatic comprehensive scoring method for smart sports product training of the present invention;

[0100] Figure 2 This is a module structure diagram of the automatic comprehensive scoring device for smart sports product training of the present invention. DETAILED DESCRIPTION

[0101] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0102] In order to solve the existing problems of not being able to understand the user's smart sports product training situation in real time, not being able to grasp the user's training situation in time, not being able to automatically comprehensively score the user's training situation, and not being able to give the user personalized guidance, resulting in a poor user experience and poor user training effect, please refer to Figure 1-Figure 2 , this embodiment provides the following technical solutions:

[0103] An automatic comprehensive scoring device for smart sports product training, including a data acquisition module, a data processing module, a model training module, a test optimization module, a comprehensive scoring module and an intelligent management and control module;

[0104] It should be noted that through the interactive communication between the data acquisition module, data processing module, model training module, test optimization module, comprehensive scoring module and intelligent management and control module, the user's smart sports product training situation can be understood in real time, the user's training situation can be grasped in time, and automatic comprehensive scoring can be performed based on the user's training situation. Personalized guidance can be given to users, which can improve user experience and user training effects.

[0105] Among them, the data collection module is used to collect user training motion data, user training physiological data and user training posture data in real time, and determine the real-time training data of users based on smart sports products;

[0106] In this embodiment, the data acquisition module includes:

[0107] The motion collection unit is used to collect the user's motion situation in real time when training with the smart sports product, and determine the user's training motion data;

[0108] The physiological collection unit is used to collect the physiological conditions of the user in real time when training with the smart sports product, and determine the user's training physiological data;

[0109] The posture collection unit is used to collect the posture of the user in real time when training with the smart sports product, and determine the user's training posture data;

[0110] Among them, based on the real-time collected user training motion data, user training physiological data and user training posture data, the real-time user training data based on the smart sports product is determined.

[0111] Among them, the data processing module is used to retrieve, sort and extract features of real-time user training data based on smart sports products, and determine user training feature data based on smart sports products;

[0112] In one embodiment, the motion acquisition unit comprises:

[0113] A type determination unit, used to obtain the movement frequency data and position data of each body part of the user when the user is training with the smart sports product, and determine the user's movement type based on the movement frequency data and position data;

[0114] A mode determination unit, used for matching the sport type with the sports data collection mode of the smart sports product to obtain a target sports data collection mode;

[0115] A data acquisition unit, used to determine the acquisition sensor that needs to work according to the target motion data acquisition method, and determine the acquisition parameters of the acquisition sensor, and collect data from the user according to the acquisition parameters to obtain sensor data;

[0116] A data processing unit is used to match and divide the sensor data with the movement parts to obtain multiple sensor data groups, determine the movement conversion mode of the corresponding sensor data group based on the movement parts, convert the corresponding sensor data group according to the movement conversion mode to obtain the part movement data, and perform abnormal number deletion and correction processing on the part movement data to obtain the target part movement data;

[0117] A sequence determination unit, for horizontally sorting the target part motion data based on all motion parts, obtaining full motion data with position order, and obtaining a full motion data sequence corresponding to the full motion data according to a time acquisition order;

[0118] The model analysis unit is used to input the full motion data sequence into the motion analysis model to obtain the user's motion characteristics, and use the motion characteristics as the user's final user training motion data.

[0119] In this embodiment, various body parts include arms, wrists, legs, ankles, waist and other positions.

[0120] In this embodiment, the exercise types include running, cycling, walking, etc.

[0121] In this embodiment, the motion conversion method is, for example, determining the motion characteristics according to frequency, swing, etc.

[0122] In this embodiment, the motion analysis model is used to holistically analyze the characteristics of the user's whole body motion, and is obtained by a preset design.

[0123] The beneficial effects of the above design scheme are: by obtaining the movement frequency data and position data of each body part of the user when the user is training with the smart sports product, determining the user's movement type based on the movement frequency data and position data, matching the movement type with the movement data collection method of the smart sports product to obtain a target movement data collection method, using different collection methods for different movement types, reducing the waste of resources for collecting irrelevant data, and providing targeted collection, determining the collection sensors that need to work according to the target movement data collection method, and determining the collection parameters of the collection sensors, collecting data from the user according to the collection parameters to obtain sensor data, matching and dividing the sensor data with the movement parts to obtain multiple groups of sensor data groups, and determining the corresponding sensor data based on the movement parts. A motion conversion method for a data group, converting the corresponding sensor data group according to the motion conversion method to obtain part motion data, and performing abnormal number deletion and correction processing on the part exercise data to obtain target part motion data, thereby ensuring the expression accuracy and data accuracy of the target part motion data; finally, based on all motion parts, horizontally sorting the target part motion data to obtain full motion data with position order, and obtaining a full motion data sequence corresponding to the full motion data in a time acquisition order, inputting the full motion data sequence into a motion analysis model to obtain the user's motion characteristics, and using the motion characteristics as the user's final user training motion data to ensure the comprehensiveness and accuracy of the user training motion data, thereby providing an accurate data basis for further user scoring and personalized recommendations.

[0124] In this embodiment, the data processing module includes:

[0125] A data retrieval unit, used to retrieve real-time training data of users based on smart sports products;

[0126] Obtain real-time user training data based on smart sports products;

[0127] Based on the sequential retrieval method, the real-time data of user training based on smart sports products is retrieved;

[0128] Check the consistency of real-time training data of users based on smart sports products;

[0129] Remove data that is beyond the normal range, logically unreasonable or contradictory in the real-time training data of users based on smart sports products;

[0130] Determine the user training real-time data with data consistency;

[0131] Process invalid and missing values ​​for real-time user training data with data consistency;

[0132] Remove invalid data and missing data that are not valuable for the automatic comprehensive scoring of smart sports product training contained in the real-time data of user training based on smart sports products;

[0133] Identify the real-time user training data that is valuable for the automatic comprehensive scoring of smart sports product training;

[0134] A data sorting unit, used to sort the retrieved user training real-time data;

[0135] Obtain the retrieved real-time user training data that is valuable for the automatic comprehensive scoring of smart sports product training;

[0136] Based on the internal sorting method, sort the retrieved user training real-time data that is valuable for the automatic comprehensive scoring of smart sports product training;

[0137] Determine the user training real-time data with arrangement order;

[0138] A feature extraction unit, used for extracting features from the sorted user training real-time data;

[0139] Obtaining sorted and ordered user training real-time data;

[0140] Perform feature extraction on the sorted user training real-time data with arrangement order;

[0141] Determine user training feature data based on smart sports products.

[0142] Among them, the model training module is used to train the model architecture based on the training set and determine the automatic comprehensive scoring model based on the smart sports product;

[0143] In this embodiment, the model training module includes:

[0144] A data storage unit, used to store user training history data based on smart sports products;

[0145] According to the demand for automatic comprehensive scoring of smart sports product training, historical training data of users based on the smart sports product is obtained, and the obtained historical training data of users based on the smart sports product is stored for future use;

[0146] An extraction and division unit, used for extracting and dividing user training history data based on smart sports products;

[0147] According to the requirements of automatic comprehensive scoring of smart sports product training, the stored user training history data based on smart sports products is extracted;

[0148] Divide the user training history data extracted based on smart sports products;

[0149] Determine the training set and test set;

[0150] Architecture training unit, used to train the model architecture;

[0151] Select the appropriate model architecture based on the requirements for automatic comprehensive scoring of smart sports product training;

[0152] Obtain the training set and train the model architecture based on the training set;

[0153] Determine an automatic comprehensive scoring model based on smart sports products.

[0154] Among them, the test optimization module is used to test and adjust the automatic comprehensive scoring model based on smart sports products, and determine the best automatic comprehensive scoring model after optimization and adjustment;

[0155] In this embodiment, the test optimization module includes:

[0156] Performance testing unit, used to test the automatic comprehensive scoring model based on smart sports products;

[0157] Get the test set;

[0158] Obtain an automatic comprehensive scoring model based on smart sports products;

[0159] Perform performance testing on the automatic comprehensive scoring model based on smart sports products based on the test set;

[0160] Determine the performance test results based on the automatic comprehensive scoring model;

[0161] An optimization and adjustment unit, used to adjust the automatic comprehensive scoring model based on smart sports products;

[0162] Obtain performance test results based on the automatic comprehensive scoring model;

[0163] Conduct mining and analysis on performance test results based on automatic comprehensive scoring model;

[0164] Determine the optimization adjustment plan based on the automatic comprehensive scoring model;

[0165] Optimize and adjust the automatic comprehensive scoring model based on smart sports products according to the optimization and adjustment plan based on the automatic comprehensive scoring model;

[0166] Determine the best automatic comprehensive scoring model after optimization and adjustment.

[0167] Among them, the comprehensive scoring module is used to predict and score the user training feature data based on the automatic comprehensive scoring model, and determine the automatic comprehensive scoring result of the user training based on the smart sports product;

[0168] 1. In one embodiment, the optimization and adjustment unit comprises:

[0169] A mining determination unit is used to obtain performance indicators and their corresponding indicator values ​​from the performance test results of the automatic comprehensive scoring model, and based on the performance indicators and their corresponding indicator values, calculate the mining index K of the automatic comprehensive scoring model according to the following formula;

[0170]

[0171] Where n represents the number of performance indicators, ε i represents the indicator weight of the i-th performance indicator, H i represents the index value of the i-th performance index, M ij represents the correlation value between the i-th performance indicator and the j-th other performance indicator;

[0172] The optimization determination unit is used to calculate the optimization adjustment index G based on the mining index of the automatic comprehensive scoring model and the difference between the index value and the standard index value according to the following formula;

[0173]

[0174] Among them, H 0i represents the standard index value of the i-th performance index;

[0175] The optimization and adjustment unit is used to select a corresponding optimization and adjustment scheme from an optimization and adjustment scheme library based on the optimization and adjustment index to optimize and adjust the automatic comprehensive scoring model based on the smart sports product.

[0176] In this embodiment, the performance indicators include accuracy, recall rate, false positive rate, etc.

[0177] In this embodiment, the range of index values ​​is standardized and the values ​​are (0, 10).

[0178] In this embodiment, the sum of the indicator weights of the performance indicators is 1.

[0179] In this embodiment, the correlation value between the performance indicators is (0, 1).

[0180] In this embodiment, the higher the mining index of the automatic comprehensive scoring model is, the larger the corresponding optimization adjustment range is.

[0181] In this embodiment, the optimization adjustment solution library is pre-designed, and the corresponding relationship between the optimization adjustment index and the solution is pre-established.

[0182] The beneficial effect of the above design scheme is: by obtaining performance indicators and their corresponding indicator values ​​from the performance test results of the automatic comprehensive scoring model, and based on the performance indicators and their corresponding indicator values, the mining index of the automatic comprehensive scoring model is calculated to provide a basis for the optimization and adjustment of the model. Then, based on the mining index of the automatic comprehensive scoring model and the difference between the indicator value and the standard indicator value, the optimization adjustment index is obtained. Based on the optimization adjustment index, the corresponding optimization adjustment scheme is selected from the optimization adjustment scheme library to optimize and adjust the automatic comprehensive scoring model based on the smart sports product, so as to ensure the optimality of the model performance of the obtained automatic comprehensive scoring model.

[0183] In this embodiment, the comprehensive scoring module includes:

[0184] A data input unit, used to input user training feature data based on smart sports products;

[0185] Obtain user training feature data based on smart sports products;

[0186] Input the user training feature data based on smart sports products into the optimal automatic comprehensive scoring model after optimization and adjustment;

[0187] A prediction and scoring unit, used to score user training feature data based on smart sports products;

[0188] Based on the best automatic comprehensive scoring model after optimization and adjustment, the user training feature data based on smart sports products is predicted and scored;

[0189] Determine the automatic comprehensive scoring results of user training based on smart sports products.

[0190] Among them, the intelligent management and control module is used to provide personalized guidance for user training according to the personalized guidance plan for user training based on smart sports products, and provide users with real-time training feedback.

[0191] In this embodiment, the intelligent management and control module includes:

[0192] The training formulation unit is used to formulate personalized training guidance plans for users based on smart sports products;

[0193] Obtain automatic comprehensive scoring results of user training based on smart sports products;

[0194] Mining and analyzing the automatic comprehensive scoring results of user training based on smart sports products;

[0195] Determine personalized training guidance plans for users based on smart sports products;

[0196] An interactive guidance unit for providing personalized guidance for user training;

[0197] Obtain personalized training guidance plans for users based on smart sports products;

[0198] Provide personalized training guidance to users based on the personalized training guidance plan of smart sports products, and provide users with real-time training feedback.

[0199] In order to better demonstrate the automatic comprehensive scoring process of smart sports product training, this embodiment now provides an automatic comprehensive scoring method for smart sports product training, including the following steps:

[0200] S1: collecting the user's sports data, physiological data and posture data in real time when training with the smart sports product, and determining the user's real-time training data based on the smart sports product;

[0201] S2: Retrieve, sort and extract features of real-time user training data based on smart sports products to determine user training feature data based on smart sports products;

[0202] S3: According to the requirements of automatic comprehensive scoring for smart sports product training, select a suitable model architecture, train the model architecture based on the training set, and determine the automatic comprehensive scoring model based on smart sports products;

[0203] S4: Conduct performance testing and optimization adjustments on the automatic comprehensive scoring model based on smart sports products, and determine the best automatic comprehensive scoring model after optimization and adjustment;

[0204] S5: Predict and score the user training feature data based on the automatic comprehensive scoring model to determine the automatic comprehensive scoring result of the user training based on the smart sports product;

[0205] S6: Develop personalized training guidance plans for users based on smart sports products, provide personalized training guidance for users according to the personalized training guidance plans, and provide users with real-time training feedback.

[0206] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0207] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for automatic comprehensive scoring of smart sports product training, based on a device for automatic comprehensive scoring of smart sports product training, is characterized in that: The steps include: S1: collecting the user's sports data, physiological data and posture data in real time when training with the smart sports product, and determining the user's real-time training data based on the smart sports product; S2: Retrieve, sort and extract features of real-time user training data based on smart sports products to determine user training feature data based on smart sports products; S3: According to the requirements of automatic comprehensive scoring for smart sports product training, select a suitable model architecture, train the model architecture based on the training set, and determine the automatic comprehensive scoring model based on smart sports products; S4: Conduct performance testing and optimization adjustments on the automatic comprehensive scoring model based on smart sports products, and determine the best automatic comprehensive scoring model after optimization and adjustment; S5: Predict and score the user training feature data based on the automatic comprehensive scoring model to determine the automatic comprehensive scoring result of the user training based on the smart sports product; S6: Develop personalized training guidance plans for users based on smart sports products, provide personalized training guidance for users according to the personalized training guidance plans, and provide users with real-time training feedback; The automatic comprehensive scoring device for smart sports product training includes: The data collection module is used to collect user training motion data, user training physiological data and user training posture data in real time, and determine the real-time user training data based on the smart sports product; A data processing module is used to retrieve, sort and extract features of real-time user training data based on smart sports products, and determine user training feature data based on smart sports products; The model training module is used to train the model architecture based on the training set and determine the automatic comprehensive scoring model based on the smart sports product; The test optimization module is used to test and adjust the automatic comprehensive scoring model based on smart sports products, and determine the best automatic comprehensive scoring model after optimization and adjustment; A comprehensive scoring module is used to predict and score the user training feature data based on the automatic comprehensive scoring model, and determine the automatic comprehensive scoring result of the user training based on the smart sports product; Intelligent management and control module, used to provide personalized training guidance for users based on personalized training guidance programs of smart sports products, and provide users with real-time training feedback; The test optimization module includes: Performance testing unit, used to test the automatic comprehensive scoring model based on smart sports products; Get the test set; Obtain an automatic comprehensive scoring model based on smart sports products; Perform performance testing on the automatic comprehensive scoring model based on smart sports products based on the test set; Determine the performance test results based on the automatic comprehensive scoring model; An optimization and adjustment unit, used to adjust the automatic comprehensive scoring model based on smart sports products; Obtain performance test results based on the automatic comprehensive scoring model; Conduct mining and analysis on performance test results based on automatic comprehensive scoring model; Determine the optimization adjustment plan based on the automatic comprehensive scoring model; Optimize and adjust the automatic comprehensive scoring model based on smart sports products according to the optimization and adjustment plan based on the automatic comprehensive scoring model; Determine the best automatic comprehensive scoring model after optimization and adjustment; The optimization and adjustment unit comprises: The mining determination unit is used to obtain the performance index and its corresponding index value from the performance test results of the automatic comprehensive scoring model, and based on the performance index and its corresponding index value, calculate the mining index of the automatic comprehensive scoring model according to the following formula: ; Where n represents the number of performance indicators. represents the indicator weight of the i-th performance indicator, represents the index value of the i-th performance index, represents the correlation value between the i-th performance indicator and the j-th other performance indicator; The optimization determination unit is used to calculate the optimization adjustment index G based on the mining index of the automatic comprehensive scoring model and the difference between the index value and the standard index value according to the following formula; in, represents the standard index value of the i-th performance index; The optimization and adjustment unit is used to select a corresponding optimization and adjustment scheme from an optimization and adjustment scheme library based on the optimization and adjustment index to optimize and adjust the automatic comprehensive scoring model based on the smart sports product.

2. The automatic comprehensive scoring method for smart sports product training according to claim 1 is characterized in that: The data acquisition module comprises: The motion collection unit is used to collect the user's motion situation in real time when training with the smart sports product, and determine the user's training motion data; The physiological collection unit is used to collect the physiological conditions of the user in real time when training with the smart sports product, and determine the user's training physiological data; The posture collection unit is used to collect the posture of the user in real time when training with the smart sports product, and determine the user's training posture data; Among them, based on the real-time collected user training motion data, user training physiological data and user training posture data, the real-time user training data based on the smart sports product is determined.

3. The automatic comprehensive scoring method for smart sports product training according to claim 2 is characterized in that: The motion acquisition unit comprises: A type determination unit, used to obtain the movement frequency data and position data of each body part of the user when the user is training with the smart sports product, and determine the user's movement type based on the movement frequency data and position data; A mode determination unit, used for matching the sport type with the sports data collection mode of the smart sports product to obtain a target sports data collection mode; A data acquisition unit, used to determine the acquisition sensor that needs to work according to the target motion data acquisition method, and determine the acquisition parameters of the acquisition sensor, and collect data from the user according to the acquisition parameters to obtain sensor data; A data processing unit is used to match and divide the sensor data with the movement parts to obtain multiple sensor data groups, determine the movement conversion mode of the corresponding sensor data group based on the movement parts, convert the corresponding sensor data group according to the movement conversion mode to obtain the part movement data, and perform abnormal number deletion and correction processing on the part movement data to obtain the target part movement data; A sequence determination unit, for horizontally sorting the target part motion data based on all motion parts to obtain full motion data with position order, and obtaining a full motion data sequence corresponding to the full motion data in a time acquisition order; The model analysis unit is used to input the full motion data sequence into the motion analysis model to obtain the user's motion characteristics, and use the motion characteristics as the user's final user training motion data.

4. The automatic comprehensive scoring method for smart sports product training according to claim 3 is characterized in that: The data processing module comprises: A data retrieval unit, used to retrieve real-time training data of users based on smart sports products; Obtain real-time user training data based on smart sports products; Based on the sequential retrieval method, the real-time data of user training based on smart sports products is retrieved; Check the consistency of real-time training data of users based on smart sports products; Remove data that is beyond the normal range, logically unreasonable or contradictory in the real-time training data of users based on smart sports products; Determine the user training real-time data with data consistency; Process invalid and missing values ​​for real-time user training data with data consistency; Remove invalid data and missing data that are not valuable for the automatic comprehensive scoring of smart sports product training contained in the real-time data of user training based on smart sports products; Identify the real-time user training data that is valuable for the automatic comprehensive scoring of smart sports product training; A data sorting unit, used to sort the retrieved user training real-time data; Obtain the retrieved real-time user training data that is valuable for the automatic comprehensive scoring of smart sports product training; Based on the internal sorting method, sort the retrieved user training real-time data that is valuable for the automatic comprehensive scoring of smart sports product training; Determine the user training real-time data with arrangement order; A feature extraction unit, used for extracting features from the sorted user training real-time data; Obtaining sorted and ordered user training real-time data; Perform feature extraction on the sorted user training real-time data with arrangement order; Determine user training feature data based on smart sports products.

5. The automatic comprehensive scoring method for smart sports product training according to claim 4 is characterized in that: The model training module includes: A data storage unit, used to store user training history data based on smart sports products; According to the demand for automatic comprehensive scoring of smart sports product training, historical training data of users based on the smart sports product is obtained, and the obtained historical training data of users based on the smart sports product is stored for future use; An extraction and division unit, used for extracting and dividing user training history data based on smart sports products; According to the requirements of automatic comprehensive scoring of smart sports product training, the stored user training history data based on smart sports products is extracted; Divide the user training history data extracted based on smart sports products; Determine the training set and test set; Architecture training unit, used to train the model architecture; Select the appropriate model architecture based on the requirements for automatic comprehensive scoring of smart sports product training; Obtain the training set and train the model architecture based on the training set; Determine an automatic comprehensive scoring model based on smart sports products.

6. The automatic comprehensive scoring method for smart sports product training according to claim 5 is characterized in that: The comprehensive scoring module includes: A data input unit, used to input user training feature data based on smart sports products; Obtain user training feature data based on smart sports products; Input the user training feature data based on smart sports products into the optimal automatic comprehensive scoring model after optimization and adjustment; A prediction and scoring unit, used to score user training feature data based on smart sports products; Based on the best automatic comprehensive scoring model after optimization and adjustment, the user training feature data based on smart sports products is predicted and scored; Determine the automatic comprehensive scoring results of user training based on smart sports products.

7. The automatic comprehensive scoring method for smart sports product training according to claim 6 is characterized in that: The intelligent management and control module includes: The training formulation unit is used to formulate personalized training guidance plans for users based on smart sports products; Obtain automatic comprehensive scoring results of user training based on smart sports products; Mining and analyzing the automatic comprehensive scoring results of user training based on smart sports products; Determine personalized training guidance plans for users based on smart sports products; An interactive guidance unit for providing personalized guidance for user training; Obtain personalized training guidance plans for users based on smart sports products; Provide personalized training guidance to users based on the personalized training guidance plan of smart sports products, and provide users with real-time training feedback.

Images