TWI442956B - Intelligent control method and system for treadmill - Google Patents

Description

Intelligent treadmill control method and system

The invention relates to a control technology of a treadmill, in particular to a smart treadmill control method and system using an artificial neural network.

The use of treadmills for physical training and energy consumption is a fairly common form of exercise. In general, the treadmill's control panel provides users with an understanding of the running process data, including running time, running distance, running speed, etc., as well as some design and planning of the running period. Basically, the basic design on the treadmill's control panel provides users with more detailed running participation information, which helps the sports participants to understand and record the individual's participation in the sports, and allows the user to adjust the treadmill accordingly. Track speed.

In addition to manually adjusting the track speed of the treadmill by the information provided by the control panel, as disclosed in the Chinese Patent No. 200815064, the body temperature, heartbeat physiological signal, and position detection signal can be used to automatically adjust the treadmill. Speed and slope to change the intensity of exercise.

The present invention applies a neural network, and outputs relevant control parameters according to various physiological signals of the user, sports environment conditions, and feedback of the current treadmill crawler speed, so as to automatically control the track speed of the treadmill.

Therefore, the object of the present invention is to provide a smart treadmill control Method of production.

Therefore, the smart treadmill control method of the present invention comprises the following steps: (a) performing a learning calculation according to a plurality of expert training samples to obtain a neural network model including a plurality of weighting values and a plurality of partial weight values; (b) providing a human-machine interface for a user to enter a gender, an age, a body mass index, and a cardio-pulmonary endurance index; (c) sensing one of the user's current heartbeat signals, a current body temperature signal, a current breathing signal, and a current indoor temperature signal; (d) obtaining a corresponding heart rate, an integrated temperature value, a breathing rate, and an indoor temperature value; (e) receiving a current treadmill track speed value, and a Current treadmill gradient value; and (f) the gender, the age, the body mass index, the cardiorespiratory endurance index, the heart rate, the body temperature value, the respiration rate, the room temperature value, the current treadmill track speed The value, and the current treadmill gradient value, is entered into the neural network model to determine a treadmill adjustment parameter set to adjust a treadmill track speed and a treadmill slope.

Another object of the present invention is to provide a smart treadmill control system.

Therefore, the smart treadmill control system of the present invention comprises a human machine interface unit, a physiological signal measuring unit, an environment sensing unit, and a microcomputer unit. The human interface unit is for a user to input a gender, an age, a body mass index, and a cardiopulmonary endurance index. The physiological signal measuring unit includes a heartbeat sensor for sensing one of the current heartbeat signals of the user, a body temperature sensor for sensing a current body temperature signal of the user, and measuring One of the users is currently breathing one of the breathing signals The sensor, the physiological signal measuring unit is further configured to generate a heart rate, an integrated temperature value, and a breathing rate according to the current heartbeat signal, the current body temperature signal, and the current breathing signal. The environment sensing unit is configured to sense a current indoor temperature signal and generate a current indoor temperature value. The microcomputer unit is configured to perform a learning calculation according to a plurality of expert training samples to obtain a neural network model including a plurality of weighting values and a plurality of partial weight values, and used the gender, the age, and the body mass index The cardiorespiratory endurance index, the heart rate, the body temperature value, the respiration rate, the indoor temperature value, a current treadmill track speed, and a current treadmill gradient are input into the neural network model to obtain one of the correspondences. The treadmill adjusts the parameter set for adjusting the treadmill track speed and a treadmill slope.

The invention applies a neural network, and according to the various physiological signals of the user, the sports environment conditions, and the current treadmill track speed and slope of the feedback, the relevant treadmill adjustment parameter group is obtained, and the treadmill track speed can be automatically controlled. With the slope, it is indeed possible to achieve the object of the present invention.

The above and other technical contents, features and advantages of the present invention will be apparent from the following detailed description of the preferred embodiments.

Referring to FIG. 1 and FIG. 2, a preferred embodiment of the smart treadmill control system 1 of the present invention comprises a human interface unit 11, a physiological signal measuring unit 12, an environmental sensing unit 13, and a microcomputer unit 14.

The human interface unit 11 is for a user 2 to input a gender, an age, a body mass index (B.M.I.), and A cardiopulmonary endurance index. The physiological signal measuring unit 12 includes a heartbeat sensor 121 for sensing one of the current heartbeat signals of the user 2, and a body temperature sensor 122 for sensing one of the current body temperature signals of the user 2, And a respiratory sensor 123 for sensing one of the current respiratory signals of the user 2; and the physiological signal measuring unit 12 is further configured to respond to the current heartbeat signal, the current body temperature signal, and the current respiratory signal Produces a heart rate, an integrated temperature, and a breathing rate. The environment sensing unit 13 is configured to sense a current indoor temperature and generate an indoor temperature value.

The microcomputer unit 14 includes a central processing unit (CPU) 141, a read only memory (ROM) 142, and a random access memory (RAM) 143. The read-only memory 142 is configured to store a plurality of program instructions; when the program instructions are loaded into the central processing unit 141, the following operations are performed: learning calculus is performed according to the expert training samples to obtain a complex number a neural network model such as a weighted value and a plurality of partial weights; and the gender, the age, the body mass index, the cardiorespiratory endurance index, the current heart rate, the current body temperature, the current respiratory rate, the current indoor Temperature, the current treadmill crawler speed, and the current treadmill gradient, inputting the neural network model, and obtaining a treadmill adjustment parameter set for adjusting the treadmill track speed of one of the treadmills 3, and A treadmill slope; the random access memory 143 is configured to store the weighted values and the offset values.

Referring to Figures 1 and 3, a preferred embodiment of the smart treadmill control method of the present invention comprises the following steps.

First, as shown in steps 41-47, the microcomputer unit 14 must first perform learning calculations to establish such a neural network model, the steps of which are as follows.

In step 41, the number of layers of one of the neural network models is determined in advance, and the number of neurons corresponding to each layer.

In step 42, the weighted values of the neural network model and the bias values are initialized and uniformly set by a random random number.

In steps 43-45, each expert training sample including an input vector and a target output vector is read; then the weighted values and the correction calculations of the partial weights are performed, and the weighted values and the offsets are updated. Weight.

In steps 46-47, it is determined whether the neural network of the type converges; if so, the learning algorithm of the neural network is terminated, and the weighted values and the partial weights are stored; otherwise, steps 43-45 are repeated.

Next, referring to FIG. 1, FIG. 4 and FIG. 5, as shown in steps 51-56, the treadmill adjustment parameter set is further determined by using the neural network model, wherein the treadmill adjustment parameter group includes a speed increasing parameter. , a deceleration parameter, an increase in slope parameters, and a reduction in slope parameters.

In step 51, the gender, the age, the body mass index, and the cardiorespiratory endurance index are input through the human interface unit 11.

In step 52, the physiological signal measuring unit 12 senses the current heartbeat signal, the body temperature signal, and the breathing signal, and correspondingly generates the heart rate, the body temperature value, and the breathing rate. The environment sensing unit 13 senses the current indoor temperature and generates the indoor temperature value.

In step 53, the microcomputer unit 14 obtains the heart rate, the body temperature value, the breathing rate, and the indoor temperature value.

In steps 54-56, the microcomputer unit 14 compares the gender (represented by x ₁ ), the age (expressed as x ₂ ), the body mass index (expressed as x ₃ ), and the cardiorespiratory endurance index (expressed as x ₄ ) ), the heart rate (expressed as x ₅ ), the body temperature value (expressed as x ₆ ), the respiration rate (expressed as x ₇ ), the room temperature value (expressed as x ₈ ), and the current treadmill returned The track speed value (expressed as x ₉ ), and the current treadmill slope value (expressed as x ₁₀ ), and enters (x ₁ ~ x ₁₀ ) into the neural network model to determine the treadmill adjustment The parameter group (y ₁ , y ₂ , y ₃ , y ₄ ), wherein y _{1 is} the speed increasing parameter, y _{2 is} the deceleration parameter, y _{3 is} the increasing slope parameter, and y _{4 is} the decreasing slope parameter. Then, the microcomputer unit 14 outputs the treadmill adjustment parameter set (y ₁ , y ₂ , y ₃ , y ₄ ), and returns the current treadmill crawler speed, and the current treadmill gradient.

Finally, as shown in step 57, the treadmill speed control unit 32 controls the controller 321 to drive a servo motor 322 according to the treadmill adjustment parameter set to control a treadmill track mechanism 323 to adjust the treadmill track speed. If the speed increase parameter (y ₁ ) is 1, the treadmill track speed is increased, and if the deceleration parameter (y ₂ ) is 1, the treadmill track speed is reduced. The treadmill gradient control unit 33 controls the controller 331 to drive a servo motor 332 according to the treadmill adjustment parameter set to control a treadmill slope mechanism 333 to adjust the treadmill gradient. If the increased gradient parameter (y ₃ ) is 1, the treadmill gradient is increased, and if the reduced gradient parameter (y ₄ ) is 1, the treadmill gradient is reduced.

The above steps 52-57 are repeated until the user 2 presses a treadmill stop button (not shown) to stop the operation of the treadmill 3, and the treadmill 3 is operated through a display unit 31. The treadmill crawler speed and the treadmill slope are displayed to the user 2.

In summary, the present invention applies a neural network, and according to the various physiological signals of the user 2, the sports environment conditions, and the current treadmill track speed value and the slope value, the relevant treadmill adjustment parameter group can be obtained. The treadmill track speed and slope are automatically controlled; indeed, the object of the invention can be achieved.

The above is only the preferred embodiment of the present invention, and the scope of the invention is not limited thereto, that is, the simple equivalent changes and modifications made by the scope of the invention and the description of the invention are All remain within the scope of the invention patent.

1‧‧‧Smart Treadmill Control System

11‧‧‧Human Machine Interface Unit

12‧‧‧physiological signal measuring unit

121‧‧‧heartbeat sensor

122‧‧‧body temperature sensor

123‧‧‧Respiratory sensor

13‧‧‧Environmental Sensing Unit

14‧‧‧Microcomputer unit

141‧‧‧CPU

142‧‧‧ROM

143‧‧‧RAM

18‧‧‧Speed adjustment unit

2‧‧‧Users

3‧‧‧Treadmill

31‧‧‧Display unit

32‧‧‧Treadmill Speed Control Unit

321‧‧‧Servo Controller

322‧‧‧Servo motor

323‧‧‧Treadmill crawler mechanism

33‧‧‧Treadmill Slope Control Unit

331‧‧‧Servo Controller

332‧‧‧Servo motor

333‧‧‧Treadmill slope mechanism

41~47‧‧‧Steps

51~57‧‧‧Steps

1 is a block diagram showing a preferred embodiment of the smart treadmill control system of the present invention; FIG. 2 is a schematic view showing the user and the treadmill; FIG. 3 is a flow chart illustrating the nerves used in the present invention. FIG. 4 is a schematic diagram illustrating a neural network model used in the present invention; and FIG. 5 is a flow chart illustrating a preferred embodiment of the smart treadmill control method of the present invention.

51~57‧‧‧Steps

Claims (12)

A smart treadmill control method comprises the following steps: (a) performing a learning calculation according to a plurality of expert training samples to obtain a neural network model including a plurality of weighting values and a plurality of partial weight values; (b) Providing a human-machine interface for a user to input a gender, an age, a body mass index, and a cardio-pulmonary endurance index; (c) sensing one of the user's current heartbeat signals, a current body temperature signal, and a current breath a signal, and a current indoor temperature signal; (d) obtaining a corresponding heart rate, an integrated temperature value, a breathing rate, and an indoor temperature value; (e) receiving a current treadmill track speed value, and a current treadmill a slope value; and (f) the gender, the age, the body mass index, the cardiorespiratory index, the heart rate, the body temperature value, the breathing rate, the room temperature value, the current treadmill track speed value, and The current treadmill slope value is entered into the neural network model to obtain a treadmill adjustment parameter set to adjust a treadmill track speed and a treadmill slope. The smart treadmill control method according to claim 1, wherein the step (a) comprises the following sub-steps: (a-1) initializing the weighted values and the partial weights by uniformly random random numbers (a-2) reading each expert training sample; (a-3) performing the weighted values and the modified calculation of the partial weights; (a-4) updating the weighted values and the partial weights; and (a-5) repeating steps (a-2) through (a-4) until the neural network of the class converges. The smart treadmill control method according to claim 1, wherein the treadmill adjustment parameter set includes a speed increasing parameter and a deceleration parameter. The smart treadmill control method according to claim 3, wherein if the speed increase parameter is 1, the treadmill track speed is increased, and if the deceleration parameter is 1, the treadmill track speed is reduced. The smart treadmill control method according to claim 3, wherein the treadmill adjustment parameter set further comprises an increase gradient parameter and a decrease gradient parameter. The smart treadmill control method according to claim 5, wherein if the increased gradient parameter is 1, the treadmill gradient is increased, and if the reduced gradient parameter is 1, the treadmill gradient is reduced. A smart treadmill control system comprising: a human-machine interface unit for a user to input a gender, an age, a body mass index, and a cardiorespiratory endurance index; a physiological signal measuring unit, comprising: for sensing One of the user's current heartbeat sensors, a heartbeat sensor, a body temperature sensor for sensing one of the user's current body temperature signals, and a sensory sense of one of the current respiratory signals of the user The physiological signal measuring unit is further configured to generate a heart rate, an integrated temperature value, and a breathing rate according to the current heartbeat signal, the current body temperature signal, and the current breathing signal; An environment sensing unit for sensing a current indoor temperature signal and generating a current indoor temperature value; and a microcomputer unit for performing a learning calculation according to the plurality of expert training samples to obtain a plurality of weighting values And a plurality of neural network models such as partial weights, and used to determine the gender, the age, the body mass index, the cardiorespiratory endurance index, the heart rate, the body temperature value, the breathing rate, the indoor temperature value, and At present, the treadmill track speed value, and a current treadmill gradient value, are input into the neural network model to obtain a treadmill adjustment parameter for adjusting the treadmill track speed and a treadmill gradient. Smart treadmill control system according to item 7 of the patent application scope The micro-computer unit first initializes the weighted values and the partial weights by uniformly random random numbers, and then reads each expert training sample, and corresponding target parameter groups of each expert training sample, and then performs the weighting values. And the correction calculation of the partial weight values, and updating the weighted values and the partial weight values, and calculating until the neural network of the class converges to obtain the neural network model. The smart treadmill control system according to claim 7, wherein the treadmill adjustment parameter set includes a speed increasing parameter and a deceleration parameter. The smart treadmill control system according to claim 9, wherein if the speed increasing parameter is 1, the speed adjusting unit increases the treadmill track speed, and if the deceleration parameter is 1, the speed adjusting unit Reduce the treadmill crawler speed. The smart treadmill control system according to claim 9, wherein the treadmill adjustment parameter set further comprises an increasing slope parameter and a decreasing slope parameter. The smart treadmill control system according to claim 11, wherein if the increased gradient parameter is 1, the treadmill gradient is increased, and if the reduced gradient parameter is 1, the treadmill gradient is reduced.