Artificial intelligence training guide device and method

Abstract

As a preferred embodiment of the present application, there is provided an Artificial Intelligence (AI) training guidance device, characterized by including: a Personal Maximum Weight (PMW) prediction unit that predicts a PMW of a training machine to be used by a user, based on a predicted muscle strength value calculated based on user data 预测 ; and a PMW guidance unit that provides a PMW 预测 related to the training machine by supplementing a user individual objectification index in the PMW 个体 .

Description

Technical Field

[0001] This invention relates to a method for automatically providing a target weight for a training machine, taking into account the individual characteristics of the user. Background Technology

[0002] Weight training machines are available in various forms to suit different body parts or purposes, primarily designed for training the upper and lower body through hand or foot movements. Users can move the selected weight using the machine's training mechanisms.

[0003] However, users may find it difficult to determine whether they are using the training machine correctly according to their physical characteristics, training performance capabilities, or the purpose of the training machine.

[0004] Existing technical documents

[0005] Patent documents

[0006] KR 10-1968621B1 Summary of the Invention

[0007] Technical issues

[0008] As a preferred embodiment of the present invention, the artificial intelligence (AI) training guidance device provides an initial target weight for the training machine suitable for the user using only user data including the user's gender, age, weight, height, body mass index (BMI), and body fat percentage.

[0009] In a preferred embodiment of the present invention, the artificial intelligence training guidance device automatically adjusts the target weight of the training machine according to the user's training objectives.

[0010] As a preferred embodiment of the present invention, a method is provided in which, when the same initial target weight is provided to each training machine of a first user and a second user who have the same gender, age, weight, height, body mass index, and body fat percentage, the target weight suitable for the first user and the second user is reset and provided for each training machine used by the first user and the second user respectively by further reflecting individual objectification indices including the training records of the first user and the second user.

[0011] As a preferred embodiment of the present invention, the artificial intelligence training guidance device learns and analyzes the user's individual objectification index, continuously updates the differentiated training machine target weight for each user, and provides a method for the user to utilize the training machine in an optimal way.

[0012] Technical solution

[0013] As a preferred embodiment of the present invention, an artificial intelligence (AI) training guidance device is provided, characterized in that it includes: a personal maximum weight (PMW) prediction unit, which predicts the PMW of the training machine to be used by the user based on a predicted muscle strength value calculated based on user data. 预测 ; and the PMW guidance unit, which, through the PMW 预测 The system supplements the user individual objectification index to provide PMW related to the training machine. 个体 The user objectification index includes at least a portion of the regularity of the weight, number of repetitions (Reps), number of sets, training trajectory, movement speed, and number of repetitions per set unit (Reps) of the training machine as confirmed by the user when using the training machine within a preset period.

[0014] As a preferred embodiment of the present invention, the artificial intelligence training guidance device is characterized in that it further includes: a training target setting unit, which is based on the PMW 预测 The initial target weight of the training machine that the user wants to use is automatically set.

[0015] In a preferred embodiment of the present invention, the artificial intelligence training guidance device is characterized by further comprising: an artificial intelligence training target setting unit, which is based on the PMW 个体 The target weight of the training machine that the user wishes to use is automatically updated.

[0016] In a preferred embodiment of the present invention, the PMW guidance unit includes a PMW update unit. When the objectification index is greater than or equal to the first benchmark value, the PMW update unit updates the PMW index. 个体 Updated to be better than PMW 预测 A larger value will be PMW when the objectification index is less than or equal to the second benchmark value. 个体 Updated to be better than PMW 预测 A smaller value. The PMW update unit updates the updated PMW based on the objectification index every preset period. 个体 .

[0017] In a preferred embodiment of the present invention, the PMW guidance unit includes a display unit, which displays PMW. 个体 The target weight of the training machine is automatically updated based on this. The display unit further displays the amount of change in the target weight of the training machine during the preset period.

[0018] In a preferred embodiment of the present invention, when the training trajectory is used as the objectification index, the PMW guidance unit uses the start point of ascent, start point of descent, average speed of ascent interval, average speed of descent interval and height confirmed by the training trajectory to determine the integrity of the training trajectory, and then uses the integrity level of the training trajectory after numerical conversion as the objectification index.

[0019] In a preferred embodiment of the present invention, when the number of repetitions (Reps) is used as the objectification index, the PMW guidance unit determines the integrity of the number of repetitions (Reps) based on the regularity between the training trajectories of each of the total number of repetitions (Reps) constituting a group, and uses the integrity level of the number of repetitions (Reps) after numerical conversion as the objectification index.

[0020] As another preferred embodiment of the present invention, an artificial intelligence training guidance method is provided, characterized by comprising the following steps: a PMW prediction unit predicts the PMW of the training machine to be used by the user based on the predicted muscle strength value calculated based on user data. 预测 ; and the PMW guidance unit via the PMW 预测 The system supplements the user individual objectification index to provide PMW related to the training machine. 个体 The user objectification index includes at least a portion of the regularity of the training machine's weight, number of repetitions (Reps), number of sets, training trajectory, movement speed, and number of repetitions per set unit (Reps) as confirmed by the user when using the training machine within a preset period.

[0021] Beneficial effects

[0022] As a preferred embodiment of the present invention, when user data including at least a portion of the user's gender, age, weight, height, body mass index (BMI), and body fat percentage is input, the artificial intelligence (AI) training guidance device can automatically set a target weight suitable for the user for each training machine.

[0023] As a preferred embodiment of the present invention, the artificial intelligence training guidance device continuously updates the target weight of the training machine used by the user by learning and analyzing the user's training records, thereby providing a method for the user to use the training machine in an optimal way.

[0024] In a preferred embodiment of the present invention, the artificial intelligence training guidance device receives new suggestions on how to use the training machine based on the user's training records and changes in the user's physical condition, thereby improving the training effect.

[0025] As a preferred embodiment of the present invention, the artificial intelligence training guidance device can also automatically set a target weight suitable for the user for each training machine according to the user's training purpose. Attached Figure Description

[0026] Figure 1 As a preferred embodiment of the present invention, an intelligent gym system utilizing an artificial intelligence (AI) training guidance device is shown.

[0027] Figure 2 As a preferred embodiment of the present invention, an example of a training machine for implementing an artificial intelligence training method is shown in a smart gym.

[0028] Figure 3 As a preferred embodiment of the present invention, an internal configuration diagram of a training machine and a smart gym server is shown in which the artificial intelligence training guidance method is implemented in a smart gym system.

[0029] Figure 4 An internal configuration diagram of an artificial intelligence training guidance device is shown as a preferred embodiment of the present invention.

[0030] Figure 5 An example of the user interface of an artificial intelligence training guidance device is shown as a preferred embodiment of the present invention.

[0031] Figure 6 As a preferred embodiment of the present invention, an example of a training trajectory detected in a training machine that implements an artificial intelligence training guidance method is shown.

[0032] Figures 7 to 9 As a preferred embodiment of the present invention, an example is shown where the objectification index calculation unit performs analysis based on the number of repetitions.

[0033] Figure 10 As a preferred embodiment of the present invention, an example is shown of a machine learning processing unit classifying large datasets of training trajectories.

[0034] Figure 11 An example of measuring the training trajectory of two users on a chest press machine is shown as a preferred embodiment of the present invention.

[0035] Figure 12 A flowchart of an artificial intelligence training guidance method is shown as a preferred embodiment of the present invention. Detailed Implementation

[0036] The invention will now be described with reference to the accompanying drawings of various embodiments thereof, so that those skilled in the art may readily implement the invention.

[0037] Figure 1As a preferred embodiment of the present invention, an intelligent gym system utilizing an artificial intelligence training guidance device is shown.

[0038] The intelligent gym system 100 includes an intelligent gym server 110, at least one training machine 100a, 100b, 100c, ..., 100n, at least one user terminal 120, and an administrator terminal 130.

[0039] exist Figure 1 In this system, the smart gym server 110 can communicate with the first smart gym 112, the second smart gym 114, and the nth smart gym 116, which are physically located in different locations, and can send and receive data with at least one training machine 100a and 100b arranged in the first smart gym 112, at least one training machine 100c arranged in the second smart gym 114, and at least one training machine 100n arranged in the nth smart gym 116.

[0040] In a preferred embodiment of the present invention, the smart gym refers to a physical space that provides the smart gym server 110 with user training records of using training machines. The smart gym server 110 then learns from and analyzes these training records to provide suitable training prescriptions for the user. The smart gym can be implemented in fitness centers, gyms, or spaces equipped with training machines.

[0041] Users (USER A, USER B, USER C, ..., USER N) who access the smart gym for training can enter after identity verification. For example, users can verify their membership by tagging their user terminal to a self-service machine or drone terminal at the entrance of the smart gym using Near Field Communication (NFC) or Radio Frequency Identification (RFID), or by verifying their membership using biometric information such as facial recognition at the drone terminal.

[0042] Information related to users undergoing membership verification can be transmitted via a network from the smart gym server 110 to at least one of the training machines 100a, 100b, 100c, ... to 100n. For example, the smart gym server 200 can transmit user-related information to the training machine of the user tagging terminal 120. In a preferred embodiment of the invention, user-related information is used in the term "user data," which includes at least some or all of the user's gender, age, weight, height, body mass index (BMI), and body fat percentage.

[0043] The smart gym server 110 can provide guidance to a first user (USER A) and a second user (USER B) on suitable training methods, intensity levels, and sequences. Both users utilize at least one training machine 100a and 100b arranged within the smart gym 112. Furthermore, the smart gym server 110 can provide target weights, recommended speeds, and other values ​​for each of the training machines 100a and 100b. Additionally, the smart gym server 110 can receive training records from both users (USER A and USER B) who have used each of the training machines 100a and 100b. Furthermore, it can receive health information such as heart rate, blood pressure, and pulse, or log information from the user terminal 120.

[0044] The smart gym server 110 can be implemented as a cloud server. The smart gym server 110 can integrate and manage information collected from each training machine within a smart fitness center located in different locations. For example, the smart gym server 110 can integrate and manage detailed records of a first user's use of a training machine in a first smart gym 112 and in a second smart gym 114.

[0045] In a preferred embodiment of the present invention, at least one training machine 100a, 100b, 100c, ..., 100n may be a stretching machine, a weight training machine, or an aerobic training machine. A weight training machine includes free weight equipment and mechanical equipment. At least one training machine 100a, 100b, 100c, ..., 100n provides training guidance suitable for the user through a display installed on the training machine or a display capable of wired or wireless communication with the training machine. For example, taking a stretching machine as an example, it provides training guidance related to the stretching exercises the user will perform through a smart mirror capable of wired or wireless communication with the stretching machine. However, it is not limited to this; training guidance can be provided through various output methods such as speakers and vibration.

[0046] At least one training machine 100a, 100b, 100c, ..., 100n can communicate with the smart gym server 110, user terminal 120 and administrator terminal 130 via wired or wireless means.

[0047] In a preferred embodiment of the present invention, the user terminal 120 can be implemented in the form of a smartphone, smartwatch, handheld device, wearable device, etc. Furthermore, the user terminal 120 can be equipped with application software for using the smart gym system. The user terminal 120 can receive training sequence information from the smart gym server 110. The training sequence refers to a training plan that takes into account the user's physical strength and training ability. The training sequence includes information such as the list of training machines the user will use, the target weight for each training machine, and the number of uses.

[0048] When a user uses at least one training machine 100a, 100b, 100c, ..., 100n within the smart gym system 100, communication can be established via NFC and RFID tags using the terminal 120, or authentication can be performed using the user's physical characteristics. After user authentication is complete, the smart gym server 110 can transmit user data to the training machine tagged by the user.

[0049] Figure 3 As a preferred embodiment of the present invention, an internal configuration diagram of a training machine and a smart gym server is shown in which the artificial intelligence training guidance method is implemented in a smart gym system.

[0050] In a preferred embodiment of the present invention, the training machine 300 in the smart gym can communicate with the smart gym server 380, the user terminal 390, and the external server 388.

[0051] In a preferred embodiment of the present invention, the training machine 300 includes a processor 310, a sensing unit 320, a communication unit 340, a training guidance unit 360, and a display 370. Additionally, it may further include a camera unit 330 and an image processing unit 350. The processor 310 may further include an artificial intelligence processing unit 312 as needed.

[0052] Further reference Figure 2 The shoulder pressure machine 200 illustrates an example of a training machine 300 used in a smart gym employing artificial intelligence-guided training methods. See below for reference. Figure 2 and Figure 3 Describe the training machine 300.

[0053] The sensing unit 220 can be disposed in the frame structure 213 of the training machine body 210. The frame structure 213 includes a basic frame 213a, a guide rail 213b, and a connecting line 213c. The sensing unit 220 measures the distance DS220 from the sensing unit 220 to the pin structure 215 in real time or in units of a preset time t by irradiating a laser beam onto the pin structure 215 and receiving the reflected laser beam. The sensing unit 220 can detect in real time at least one of the position, moving speed, and moving direction of the weight component 211 selected by the pin structure 215. In addition, when the user pushes the handle 212 of the training machine 200 to move the weight plate, the sensing unit 220 can measure the distance DS220 from it to the pin structure 215 into which the weight plate is inserted, and detect the training trajectory based on this distance.

[0054] The communication unit 340 receives user input via the display unit 230 and sends and receives user data from the user database (DB) 382 of the smart gym server 380. The communication unit 340 can also communicate with an external server 388.

[0055] The training guidance unit 360 can provide the user with information such as user data received from the smart gym server 380, the target weight of the training machine, the movement speed guidance of the training machine, breathing guidance when using the training machine, and the training sequence. (Reference) Figure 5 and Figure 10 This describes an example of how the training guidance unit 360 provides guidance to the user on the movement speed of the training machine.

[0056] Figure 10 The graph illustrates an example of classifying training trajectories for a specific training machine, obtained from the population via machine learning processing unit 384, into seven patterns (1010 to 1070). Initially, the population can be based on a preset number of n people, but can be used after continuously accumulating data from users of the smart gym.

[0057] The machine learning processing unit 384 can classify the overall training trajectory patterns and determine the proportion of the population belonging to each pattern. The machine learning processing unit 384 selects the training trajectory with the best training effect from the seven analyzed training trajectories. In this case, the machine learning processing unit 384 selects the second training trajectory 1020 and the sixth training trajectory 1060, which have the smallest variance compared to the preset benchmark trajectory 1012 and the best training effect. The second training trajectory 1020 accounts for 21.7% of the total, and the sixth training trajectory 1060 accounts for 23.1%.

[0058] The training guidance processing unit 386 can confirm which mode the detected user training trajectory belongs to. Furthermore, it can determine changes in the user's training trajectory based on training load, number of training sessions, and time. For example, if the first user's training trajectory when using the chest press machine at 60 kg belongs to the second training trajectory 1020, but the first user's training trajectory when using the chest press machine at 70 kg belongs to the first training trajectory 1010, the training guidance processing unit 386 can determine that the suitable load for the first user is 60 kg.

[0059] Training guidance processing unit 386 provides user-appropriate movement speed guidance VG1 to the training machine via training guidance unit 360. Figure 5 (530). Movement speed guide VG1 530 refers to the movement speed suitable for the user to achieve the desired training effect when using the training machine. For example, when the user pushes the shoulder press machine in accordance with the movement speed guide VG1 530, the speed V1 520 measured when the user pushes the shoulder press machine is displayed together with the baseline 510.

[0060] The Smart Gym Server 380 can calculate predicted muscle strength values ​​for specific muscle groups based on user data. Additionally, it determines the PMW (Power Factor) of the training machine the user will use based on the predicted muscle strength values. 预测 And based on the user's training objectives and PMW 预测 The system automatically sets the initial target weight for the training machine the user will use. In a preferred embodiment of the invention, PMW (Personal Maximum Weight) refers to the muscle strength an individual can exert with maximum effort against the resistance of the weight. Examples of PMW include 1RM, 4RM, etc. 预测 This indicates that the PMW (Power Muscle Strength) of a user is predicted based on the predicted muscle strength value. The Smart Gym Server 380 can confirm the PMW of each training machine. 预测 .

[0061] The intelligent gym server 380 can provide the training guidance unit 360 of the training machine 300 with the initial target weight that the user will use on the training machine. The training guidance unit 360 can display the initial target weight that the user will use on the training machine on the display 370. Figure 5 An example of ,550 is shown in Figure 5 .

[0062] The Smart Gym Server 380 can also predict PMW. 个体 The PMW 个体 It is in PMW 预测 The system supplements the user's maximum muscle strength value using the Individual Objectivity Index. In this case, the Smart Gym Server 380 updates the PMW-based system... 预测The initial target weight set for the training machine can be updated to reflect new target weights that further reflect the individual characteristics of the user. Examples of objectivity indices can be found in training records, including the weight of the training machine, number of repetitions (Reps), number of sets, training trajectory, movement speed, and regularity of repetitions per set unit (Reps) observed while using the training machine.

[0063] The camera unit 330 can be embedded in the training machine 300 or communicate wirelessly or wiredly with the training machine to capture the user's training posture. The image processing unit 350 can analyze the user's training posture captured by the camera unit 330 through the processor 310. In addition, the artificial intelligence processing unit 312 can learn the image processing results of the training posture captured when the user trains according to the target weight of the training machine, and transmit the learning results to the smart gym server 380. The artificial intelligence processing unit 312 can also process or learn the user's objectification index.

[0064] The intelligent gym server 380 includes a user database 382, ​​a machine learning processing unit 384, and a training guidance processing unit 386. The user database 382 stores and manages user data, including the user's gender, age, weight, height, body mass index, and body fat percentage.

[0065] The machine learning processing unit 384 learns and processes objectification indices, which include training records from a user's use of a training machine in a smart gym. An example of the machine learning processing unit 384 learning objectification indices is shown below. Figures 6 to 10 .

[0066] The machine learning processing unit 384 can, based on the objectivity index of the training machines used by users in the smart gym within a certain period, process PMW. 预测 Updated to PMW 个体 The machine learning processing unit 384 can process PMW when the objectification index is greater than or equal to the first baseline value. 个体 Updated to be better than PMW 预测 A larger value will result in PMW when the objectification index is less than or equal to the second benchmark value. 个体 Updated to be better than PMW 预测 Smaller values.

[0067] When the machine learning processing unit 384 determines, based on objectification indices confirmed by the user's training records, that the user's training ability is greater than the PMW predicted from the user's specific muscle strength... 预测 At this time, the user's maximum muscle strength can be reset to a value higher than PMW. 预测 Larger PMW 个体In contrast, when the machine learning processing unit 384 determines, based on an objectification index confirmed by the user's training records, that the user's training ability is less than the PMW predicted from the user's specific muscle strength,... 预测 At this time, the user's maximum muscle strength can be reset to a value higher than PMW. 预测 Smaller PMW 个体 .

[0068] Therefore, the machine learning processing unit 384 can process the PMW of both the first and second users even if they have the same gender, age, weight, height, body mass index, and body fat percentage. 预测 When predicted to be the same, further objectification indices reflecting individual user training records, such as those of the first and second users, can be used to learn and predict the appropriate PMW for each training machine used by the first and second users. 个体 .

[0069] In other words, the machine learning processing unit 384 can learn and predict PMW under the same user data related to user physical conditions such as gender, age, weight, height, body mass index, and body fat percentage. 个体 The PMW 个体 It is the maximum muscle strength value that further reflects the individual objectification index, which varies according to each user's muscle development, range of motion, body proportions, training habits, and training achievement rate.

[0070] The sensing unit 320 can also confirm the training trajectory based on distance information sensed during the user's use of the shoulder press machine. The machine learning processing unit 384 can use at least a portion of the rising start point 611, falling start point 613, rising interval speed V1 S610, falling interval speed V2 S620, rising interval average speed, falling interval average speed, and height H 612 confirmed by the training trajectory detected during the user's use of the shoulder press machine to determine the completeness of the training trajectory, and use the completeness level of the training trajectory as an objectification index after numerical conversion.

[0071] In a preferred embodiment of the present invention, when the number of repetitions (Reps) is used as an objectification index, the machine learning processing unit 384 determines the integrity of the number of repetitions (Reps) based on the regularity among the training trajectories of each of the total number of repetitions (Reps) constituting a group, and then uses the integrity level of the number of repetitions (Reps) after numerical conversion.

[0072] Training guidance processing unit 386 can also be based on PMW received from machine learning processing unit 384 个体 It provides the target weight for the updated training machine.

[0073] Figure 4 An internal configuration diagram of an artificial intelligence training guidance device is shown as a preferred embodiment of the present invention. The artificial intelligence training guidance device 400 includes a smart gym server or terminal.

[0074] The artificial intelligence training guidance device 400 includes a PMW prediction unit 410, a PMW guidance unit 420, a training goal setting unit 430, an artificial intelligence training goal setting unit 440, and a display unit 450. The PMW prediction unit 410 further includes a muscle strength prediction unit 412 and a detection unit 414. The PMW guidance unit 420 further includes an objectification index calculation unit 422 and a PMW update unit 424.

[0075] The muscle strength prediction unit 412 predicts a muscle strength value related to a specific muscle strength of the user based on user data. As shown in Equation 1, the muscle strength prediction unit 412 uses the user's gender, body mass index, body fat percentage, and age information to calculate the predicted muscle strength value related to a specific muscle strength of the user. When the user's gender, body mass index, body fat percentage, and age information are all the same, the same predicted muscle strength value will be calculated. An example of a specific muscle strength could be grip strength.

[0076] [Mathematical Expression 1]

[0077] Predicted muscle strength = A * gender + B * age + C * body fat percentage + D * body mass index + E

[0078] In mathematical formula 1, A, B, C, D, and E represent preset values.

[0079] Detection unit 414 detects from an external server ( Figure 3 (388) Obtain or detect the percentile value to which the predicted muscle strength value belongs from preset adult muscle strength percentile values ​​(as shown in Tables 1 and 2). Table 1 shows the percentile of relative grip strength for adult males, and Table 2 shows the percentile of relative grip strength for adult females.

[0080] [Table 1]

[0081]

[0082] [Table 2]

[0083]

[0084] For example, when the muscle strength prediction unit 412 calculates a predicted muscle strength value of 55.9 for a 32-year-old male, the detection unit 414 detects the percentile value 40 to which the predicted muscle strength value of 55.9 for a 32-year-old male belongs from the preset adult muscle strength percentile values ​​shown in Table 1.

[0085] PMW prediction unit 410 utilizes PMW values ​​matched to the predicted muscle strength values ​​detected from detection unit 414. 预测 Percentile value, predicting the PMW of the training machine the user wants to use. 预测 In a preferred embodiment of the present invention, the PMW prediction unit 410 can pre-store the PMW of the training machine. 预测 Percentile values, or downloaded from the smart gym server, wherein the PMW of the training machine 预测 Percentile values ​​are obtained from the population using the training machine.

[0086] The smart gym server has pre-stored a mapping table, which stores the PMW associated with each training machine. 预测 Percentile, PMW 预测 Percentile values ​​are further reflected in the data of users using each training machine in the smart gym, and are updated in preset periods. Furthermore, the machine learning processing unit 384 of the smart gym server 380 generates PMW data based on the PMW data initially obtained from a preset population. 预测 A percentile mapping table. Then, the machine learning processing unit 384 can continue to collect the user's PMW data, thereby providing PMW data in predetermined time intervals. 预测 The percentile mapping table is updated, where the user is a user of the training machines provided in the smart gym, and the PMW... 预测 Percentile values ​​are generated based on PMW data initially obtained from a predefined population.

[0087] Tables 3 to 5 show examples of the mapping used by the PMW prediction unit 410. For ease of explanation, Tables 3 to 5 only show a portion of the PMW 10th percentile to PMW 90th percentile values. It should be noted that the values ​​shown are for the purpose of understanding one embodiment of the invention.

[0088] As a preferred embodiment of the present invention, an example of the PMW percentile values ​​and mapping table of the training machines obtained by the PMW prediction unit 410 from the total number of chest press machines is shown in Table 3.

[0089] [Table 3]

[0090] PMW percentile Initial total PMW (kg) Updated total PMW (kg) PMW90 102.7 107.8 … PMW50 67.2 67.8 PMW40 65.1 64.7 … PMW10 24.6 23.9

[0091] As a preferred embodiment of the present invention, an example of the PMW percentile values ​​and mapping table of the training machine obtained by the PMW prediction unit 410 from the total number of seated leg extension trainers is shown in Table 4.

[0092] [Table 4]

[0093] <![CDATA[PMW 预测 percentile value <![CDATA[PMW of the initial population 预测 (kg)]]> <![CDATA[PMW of the overall to be updated 预测 (kg)]]> PMW90 103.3 107.8 ... ... ... PMW50 67.3 69.2 PMW40 66.5 66.8 ... ... ... PMW10 24.6 24.9

[0094] As a preferred embodiment of the present invention, a PMW for a training machine is provided. 预测 The percentile mapping is shown in Table 5, for example, where the PMW of the training machine... 预测 Percentile values ​​are obtained from the population for the shoulder press by the PMW prediction unit 410.

[0095] [Table 5]

[0096] PMW percentile Initial total PMW (kg) Updated total PMW (kg) PMW90 64.5 65.5 ... ... ... PMW50 43.3 43.2 PMW40 41.2 40.2 ... ... ... PMW10 16.1 16.3

[0097] When the percentile value 40 of the predicted muscle strength value of 55.9 for a 32-year-old male is detected from the preset adult muscle strength percentile values ​​(e.g., Table 1), the PMW prediction unit 410 predicts the PMW of the training machine corresponding to the percentile value 40.

[0098] Referring to Tables 3 to 5, training objective setting unit 430 uses "initial total PMW (kg)" as a benchmark to set the PMW of the chest press machine. 预测 The predicted weight is 65.1 kg, corresponding to the percentile value of PMW40. This will be the weight of the PMW seated leg extension trainer. 预测 The predicted weight is 66.5 kg, corresponding to the percentile value of PMW40. Additionally, it will use the PMW shoulder press machine. 预测 The predicted weight is 41.2 kg, which corresponds to the percentile value of PMW40.

[0099] In another preferred embodiment of the invention, in addition to the initially preset total PMW data, when large amounts of data on users using the training machines in the smart gym are further collected, the PMW prediction unit 410 can utilize updated PMW data, such as the "Updated Total PMW (kg)" in Tables 3 to 5. In this case, the PMW of the chest press machine can be used... 预测 Updated to 64.7 kg, corresponding to the percentile value of PMW40. This allows the PMW of the seated leg extension trainer to... 预测 Updated to 66.8 kg, corresponding to the percentile value of PMW40. Additionally, the PMW rating of the shoulder press machine can be adjusted. 预测 Updated to 40.2 kg, corresponding to the percentile value of PMW40.

[0100] The training objective setting unit 430 is based on the PMW of each training machine predicted by the PMW prediction unit 410. 预测 The initial target weight for each training machine is set. The training target setting unit 430 can further reflect the user's training objectives received through the user interface, thereby setting the initial target value for each training machine.

[0101] The training goal setting unit 430 can set different initial goal values ​​for each training machine based on whether the user's training purpose is free training 541, standard training 542, muscle building 543, or physical training 544.

[0102] Taking the free training 541 as an example, the initial target value of the training machine can be set to PMW. 预测 *P Free (%). Taking the standard training 542 as an example, the initial target value of the training machine can be set to PMW. 预测 *P 标准 (%). Taking Muscle Gain 543 as an example, the initial target value of the training machine can be set to PMW. 预测 *P 增肌 (%). Additionally, taking physical training 544 as an example, the initial target value of the training machine can be set to PMW. 预测 *P 体能训练 (%).

[0103] As a preferred embodiment of the present invention, P 增肌 The percentage (%) can be set to 65% to 85% and the number of repetitions can be set to 6 to 12. As a preferred embodiment of the invention, when the training objective is muscle gain, the reflected weight P... 增肌 (%) can utilize preset values. Additionally, weight P 增肌 (%) can be readjusted by receiving objective index feedback related to the initial target weight suggested to the user, such as the completeness of the user's training trajectory, the training trajectory for each user training session, etc.

[0104] Assuming the user selects the training objective as Standard 542 through the user interface provided by display 500, when the shoulder pressure machine's PMW... 预测 When the target weight is 40.2 kg, the training target setting unit can reflect P 标准 (%) = 60% and then set an initial target weight of 25kg for the user. 40.2kg * 60% = 24.12kg, and the closest weight supported by the training machine to 25kg can be set as the initial target weight.

[0105] As a preferred embodiment of the present invention, the PMW guidance unit 420 guides the PMW prediction unit 410 based on the PMW predicted by the PMW prediction unit 410. 预测 The system supplements the user individual objectification index to provide PMW related to the training machine. 个体 The PMW guidance unit 420 predicts the same PMW for the same training machine by reflecting each user's training record. 预测 To cater to diverse users, we provide PMW tailored to each user's needs. 个体 The same user uses the same training machine and predicts the same PMW. 预测In this case, the PMW guidance unit 420 also reflects the user's training records over a certain period of time, thereby providing the user with a PMW suitable for the current point in time. 个体 The PMW guidance unit 420 can utilize PMW data that reflects the user's physical condition, fitness level, and training logs. 个体 We will continue to provide training plans that are suitable for our users.

[0106] The PMW guidance unit 420 may display at least one of the following information in the display unit 450: muscles activated when the user uses the selected training machine, PMW 个体 Or based on PMW 个体 The calculated target weight.

[0107] The PMW guidance unit 420 includes an objectification index calculation unit 422 and a PMW update unit 424. The objectification index calculation unit 422 calculates the objectification index based on at least a portion of the regularity of the training machine's weight, repetitions (Reps), number of sets, training trajectory, movement speed, and number of repetitions per set unit (Reps) as confirmed by the user when using the training machine within a preset period.

[0108] When the objectification index calculated by the objectification index calculation unit 422 is greater than or equal to the first benchmark value, the PMW update unit 424 will update the PMW. 个体 Updated to be better than PMW 预测 A larger value, when less than or equal to the second baseline value, will result in PMW. 个体 Updated to be better than PMW 预测 Smaller values. The PMW update unit 424 can update the updated PMW based on an objective index every preset period. 个体 .

[0109] The artificial intelligence training objective setting unit 440 sets the PMW calculated from the PMW guidance unit 420. 个体 The target weight for each training machine is updated based on this benchmark. Furthermore, the AI ​​training goal setting unit 440 can further reflect the user's training objectives received through the user interface, thereby setting the initial target value for each training machine. The target weight for each training machine can be updated differently depending on whether the user's training objective is free training 541, standard training 542, muscle building 543, or strength and conditioning training 544.

[0110] As a preferred embodiment of the present invention, refer to Figure 6The description describes an example of the objectification index calculation unit 422 using the training trajectory as an objectification index. The objectification index calculation unit 422 uses the rising start point 611, the falling start point 613, the rising interval speed V1 S610, the falling interval speed V2 S620, the rising interval average speed, the falling interval average speed, and the height H 612 to determine the completeness of the training trajectory, and converts the completeness of the training trajectory into a numerical value for use as the objectification index.

[0111] For example, the objectification index calculation unit 422 determines the starting point 611 of the rise according to the preset standard between 0 and t. a The time is appropriate, exceeding t a If it is too late, then it is too late. Additionally, based on the preset standards for the appropriate starting point 611, completeness can be assigned scores ranging from high, medium, and low, or from 1 to 10 points. Similarly, the speed of the upward interval V1 S610 can be judged according to preset standards: if it is between Va and Vb, it is appropriate; if it exceeds Vb, it is too fast; if it is below Va, it is too slow. Based on the judgment results of being appropriate, too fast, or too slow, scores ranging from high, medium, and low, or from 1 to 10 points, can be assigned respectively.

[0112] Using the method described above, the objectification index calculation unit 422 determines the completeness of the training trajectory detected by the user when using the training machine by at least a portion of the rising start point 611, the falling start point 613, the rising interval speed V1 S610, the falling interval speed V2 S620, the rising interval average speed, the falling interval average speed, and the height H 612.

[0113] Taking the integrity value of the training trajectory as an objectification index as an example, when the objectification index is greater than or equal to the first benchmark value (e.g., 8 points), the PMW update unit 424 will update the PMW. 个体 Updated to be better than PMW 预测 A larger value. As an example, when a user's objectivity index is 9 points and the first benchmark is 8 points, the PMW can be... 个体 Updated to be better than PMW 预测 The value is greater than n% (where n is a natural number). Additionally, when the objectification index is less than or equal to the second benchmark value (e.g., 6 points), PMW can be... 个体 Updated to be better than PMW 预测 Smaller values. Additionally, when the objectification index is greater than the second benchmark value but less than the first benchmark value, PMW can be... 个体 Keep as PMW 预测 value.

[0114] As another preferred embodiment of the present invention, refer to Figures 7 to 9The description of the objectification index calculation unit 422 is an example of using the number of repetitions (Reps) as an objectification index. The objectification index calculation unit 422 judges the integrity of the number of repetitions (Reps) based on the regularity among the training trajectories of each of the total number of repetitions constituting a group, and uses the integrity level of the number of repetitions (Reps) after numerical conversion as the objectification index.

[0115] Figures 7 to 9 This illustrates an example of the training trajectory detected during 12 repetitions of a seated leg press exercise performed by users 1 through 3 in one group. Figures 7 to 9 In the diagram, the X-axis represents time, and the Y-axis represents the displacement of the training machine.

[0116] The objectification index calculation unit 422 determines whether all repetitions in group 1 have been completed. It determines whether all 12 repetitions were performed, or whether some repetitions were performed after abandoning the task midway. Furthermore, the objectification index calculation unit 422 judges the completeness of the repetition counts (Reps) based on the regularity among the training trajectories of each of the first user's 12 repetition counts (711 to 714, 721 to 724, 731 to 734), the second user's 12 repetition counts (811 to 814, 821 to 724, 831 to 834), and the third user's 12 repetition counts (911 to 914, 921 to 924, 931 to 934). In this case, the judgment can be made after dividing the regularity into initial, intermediate, and late stages.

[0117] For example, in Figure 7 The first user and Figure 9 In the initial patterns of the third user's first to fourth training sessions (711 to 744, 911 to 914), the four training trajectories were almost constant, so the initial patterns can be considered to be relatively high. In this case, the initial patterns can be scored as 4 points. Figure 8 The second user's training trajectory remained constant from the first to the third training sessions (811 to 813), but the fourth training session (814) deviated significantly, making it difficult to conclude that its initial regularity was high. In this case, the initial regularity could be scored as 3. However, this is only one example of assigning a value to determine the initial regularity after judging whether the training trajectories are consistent. Various methods can be used to determine regularity, such as those based on the distance between training trajectories or the length of the time series.

[0118] The objectification index calculation unit 422 can determine that the first user's 12 repetitions have a high regularity, the second user's 4th to 7th repetitions (i.e., mid-term regularity) have a low regularity, and the third user's 0th and 2nd repetitions (i.e., late-term regularity) have a low regularity, and assign scores accordingly.

[0119] The objectification index calculation unit 422 can also determine the standard deviation of the rising starting point, the standard deviation of the rising starting point, the standard deviation of the height, the standard deviation of the speed in the rising interval, and the standard deviation of the speed in the falling interval in each of the 12 repetitions of each user's training trajectory, and give scores accordingly with preset standards.

[0120] In addition, the objectification index calculation unit 422 further calculates the objectification index with reference to the execution time, which is the time for all repetitions that constitute a group.

[0121] Figure 11 As a preferred embodiment of the present invention, a chest press machine with the same PMW is shown. 预测 An example of training trajectory detection related to the number of repetitions of the first and second users. Figure 11 In the diagram, the X-axis represents time and the Y-axis represents displacement.

[0122] Initially, users 1120 and 1130 can be based on the same PMW. 预测 And the same training plan is provided. A training plan for a chest press machine can include the initial target weight, repetitions, movement speed guidance, breathing method guidance, etc., for the user to train on the optimal training trajectory 1110.

[0123] refer to Figure 11 It can be seen that the displacement of the first user 1120 is higher than that of the optimal training trajectory 1110, but the starting point of ascent, the starting point of descent, the average velocity of the ascent interval, and the average velocity of the descent interval are similar. In addition, it can be seen that the variance 1120a between the first training trajectories is relatively small.

[0124] It can be seen that the displacement of the second user (1130) is relatively high, and the average velocity at the starting point and during the descent interval is too fast. Additionally, it can be seen that the variance (1130a) between the first training trajectories is relatively large.

[0125] The PMW guidance unit 420 can enable the first user 1120's PMW 个体 The value increases, making the second user's PMW 1130. 个体 The value decreased.

[0126] Figure 12 A flowchart of an artificial intelligence training guidance method is shown as a preferred embodiment of the present invention.

[0127] The PMW prediction unit predicts the PMW of the training machine the user intends to use, based on predicted muscle strength values ​​calculated from user data. 预测 S1210. The training objective setting unit is based on PMW. 预测The initial target weight S1220 of the training machine is set automatically.

[0128] The PMW guidance unit, through PMW 预测 The system supplements the user individual objectification index to provide PMW related to the training machine. 个体 The user individual objectification index includes training records based on an initial target weight over a certain period, etc. (S1230). The artificial intelligence training target setting unit is based on PMW. 个体 The target weight S1240 of the training machine is automatically set.

[0129] The method according to embodiments of the present invention can be implemented in the form of program instructions executable by various computer means, and thus written to a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., individually or in combination. The program instructions recorded on the medium may be specifically designed and configured for the present invention, or may be known and available to those skilled in the art of computer software.

[0130] The present invention has been described above with reference to limited embodiments and accompanying drawings, but the present invention is not limited thereto, and those skilled in the art can make various modifications and variations based on it.

Claims

1. An artificial intelligence training guidance device, characterized in that, include: The Individual Maximum Weight (PMW) prediction unit uses a PMW value that is matched with the percentile value of the predicted muscle strength calculated using a predefined mathematical formula based on the user's data, including gender, age, body mass index, and body fat percentage. 预测 Percentile values ​​are used to predict the PMW of the training machine that the user wants to use. 预测 Among them, the percentile value to which the predicted muscle strength value belongs is detected from the preset adult muscle strength percentile values; and PMW guidance unit, which, through the PMW 预测 The system supplements the user individual objectification index to provide PMW related to the training machine. 个体 , Wherein, PMW refers to the muscle strength that an individual can exert when resisting the resistance of the weight with maximum effort. The individual objectification index includes the regularity of the training trajectory and the number of repetitions per set unit of the training machine as confirmed by the user when using the training machine within a preset period, and includes one or more of the weight, number of repetitions, number of sets, and movement speed of the training machine. The training trajectory is the curve of the movement displacement of the training machine over time. The regularity of the training trajectory and the number of repetitions per group unit is converted into numerical values ​​and used as an individual objectification index. When the individual objectification index is greater than or equal to a first benchmark value, the PMW... 个体 Updated to be more than PMW 预测 A larger value, when the individual objectification index is less than or equal to the second benchmark value, indicates that the PMW... 个体 Updated to be more than PMW 预测 Smaller values, When the training trajectory is used as the individual objectification index, the PMW guidance unit is configured to use the start point of ascent, start point of descent, average speed of the ascent interval, average speed of the descent interval, and height identified through the training trajectory to determine the completeness of the training trajectory, and then use the completeness level of the training trajectory as the individual objectification index after numerical conversion. When the number of repetitions is used as the individual objectification index, the PMW guidance unit is configured to determine the integrity of the number of repetitions based on the regularity between the training trajectories of each of the total number of repetitions constituting a group and the execution time of all the repetitions constituting a group, and then use the integrity level of the number of repetitions as the individual objectification index after numerical conversion.

2. The artificial intelligence training guidance device according to claim 1, characterized in that, Further includes: Training target setting unit, which is based on the PMW 预测 The initial target weight of the training machine that the user wants to use is automatically set.

3. The artificial intelligence training guidance device according to claim 1, characterized in that, Further includes: The artificial intelligence training target setting unit is based on the PMW. 个体 The target weight of the training machine that the user wishes to use is automatically updated.

4. The artificial intelligence training guidance device according to claim 1, characterized in that, include: The PMW update unit updates the PMW when the individual objectification index is greater than or equal to the first benchmark value. 个体 Updated to be better than PMW 预测 A larger value will be added when the individual objectification index is less than or equal to the second benchmark value. 个体 Updated to be better than PMW 预测 Smaller values.

5. The artificial intelligence training guidance device according to claim 4, characterized in that, The PMW update unit updates the PMW based on the individual objectification index every preset period. 个体 .

6. The artificial intelligence training guidance device according to claim 1, characterized in that, include: The PMW guidance unit includes a display unit that displays PMW. 个体 The target weight of the training machine is automatically updated based on this.

7. An artificial intelligence training guidance method, characterized in that, Includes the following steps: The Individual Maximum Weight (PMW) prediction unit uses a PMW value that is matched with the percentile of the predicted muscle strength value calculated using a predefined mathematical formula based on user data, including gender, age, body mass index, and body fat percentage. 预测 Percentile values ​​are used to predict the PMW of the training machine that the user wants to use. 预测 Among them, the percentile value to which the predicted muscle strength value belongs is detected from the preset adult muscle strength percentile values; and The PMW guidance unit, through the PMW 预测 The system supplements the user individual objectification index to provide PMW related to the training machine. 个体 , Wherein, PMW refers to the muscle strength that an individual can exert when resisting the resistance of the weight with maximum effort. The individual objectification index includes the regularity of the training trajectory and the number of repetitions per set unit of the training machine as confirmed by the user when using the training machine within a preset period, and includes one or more of the weight, number of repetitions, number of sets, and movement speed of the training machine. The training trajectory is the curve of the movement displacement of the training machine over time. The regularity of the training trajectory and the number of repetitions per group unit is converted into numerical values ​​and used as an individual objectification index. When the individual objectification index is greater than or equal to a first benchmark value, the PMW... 个体 Updated to be more than PMW 预测 A larger value, when the individual objectification index is less than or equal to the second benchmark value, indicates that the PMW... 个体 Updated to be more than PMW 预测 Smaller values, When the training trajectory is used as the individual objectification index, the PMW guidance unit is configured to use the start point of ascent, start point of descent, average speed of the ascent interval, average speed of the descent interval, and height identified through the training trajectory to determine the completeness of the training trajectory, and then use the completeness level of the training trajectory as the individual objectification index after numerical conversion. When the number of repetitions is used as the individual objectification index, the PMW guidance unit is configured to determine the integrity of the number of repetitions based on the regularity between the training trajectories of each of the total number of repetitions constituting a group and the execution time of all the repetitions constituting a group, and then use the integrity level of the number of repetitions as the individual objectification index after numerical conversion.

8. The artificial intelligence training guidance method according to claim 7, characterized in that, Further steps include: The training objective setting unit is based on the PMW. 预测 The initial target weight of the training machine is set automatically.

9. The artificial intelligence training guidance method according to claim 7, characterized in that, Further steps include: The artificial intelligence training objective setting unit is based on the PMW 个体 The target weight of the training machine is automatically updated.

10. The artificial intelligence training guidance method according to claim 7, characterized in that, Provide PMW 个体 The step involves updating the updated PMW based on the individual objectification index at preset intervals. 个体 .

11. The artificial intelligence training guidance method according to claim 7, characterized in that, Provide PMW 个体 The steps are displayed via the display unit. 个体 Or it can be shown as the maximum change in weight that can be lifted when using the training machine.

12. A computer-readable recording medium, characterized in that, Its storage enables the computing device to execute the following instructions: The Individual Maximum Weight (PMW) prediction unit uses a PMW value that is matched with the percentile of the predicted muscle strength based on a predefined mathematical formula that utilizes user data, including gender, age, body mass index, and body fat percentage. 预测 Percentile values ​​are used to predict the PMW of the training machine that the user wants to use. 预测 Among them, the percentile value to which the predicted muscle strength value belongs is detected from the preset adult muscle strength percentile values; and The PMW guidance unit, through the PMW 预测 The system supplements the user individual objectification index to provide PMW related to the training machine. 个体 , Wherein, PMW refers to the muscle strength that an individual can exert when resisting the resistance of the weight with maximum effort. The individual objectification index includes the regularity of the training trajectory and the number of repetitions per set unit of the training machine as confirmed by the user when using the training machine within a preset period, and includes one or more of the weight, number of repetitions, number of sets, and movement speed of the training machine. The training trajectory is the curve of the movement displacement of the training machine over time. The regularity of the training trajectory and the number of repetitions per group unit is converted into numerical values ​​and used as an individual objectification index. When the individual objectification index is greater than or equal to a first benchmark value, the PMW... 个体 Updated to be more than PMW 预测 A larger value, when the individual objectification index is less than or equal to the second benchmark value, indicates that the PMW... 个体 Updated to be more than PMW 预测 Smaller values, When the training trajectory is used as the individual objectification index, the PMW guidance unit is configured to use the start point of ascent, start point of descent, average speed of the ascent interval, average speed of the descent interval, and height identified through the training trajectory to determine the completeness of the training trajectory, and then use the completeness level of the training trajectory as the individual objectification index after numerical conversion. When the number of repetitions is used as the individual objectification index, the PMW guidance unit is configured to determine the integrity of the number of repetitions based on the regularity between the training trajectories of each of the total number of repetitions constituting a group and the execution time of all the repetitions constituting a group, and then use the integrity level of the number of repetitions as the individual objectification index after numerical conversion.

Images