A dynamic bicycle with a motion feedback intervention system

Abstract

The present invention discloses a spinning bike with a motion feedback intervention system, which belongs to the field of bio-information feedback intervention for virtual reality depression disorders. The spinning bike collects the user's heart rate variability and electroencephalogram through a heart rate belt detection module and an electroencephalogram detection module, and the photoelectric speed measurement module monitors the rotation of the driving wheel in real time to obtain a speed pulse signal. The single-chip computer module obtains the target speed by combining the real-time heart rate variability, the real-time electroencephalogram with the target heart rate variability and the target electroencephalogram through a PID neural network decoupling control algorithm based on sparrow search, and reduces the system error through a nonlinear fault-tolerant control law based on an adaptive neural network to ensure the stable operation of the system, and finally transmits the real-time speed value and the target speed value to the VR glasses worn by the user in real time, and the user can adjust the speed of the spinning bike in time according to the target speed value and the real-time speed value, control the exercise intensity, enhance the user's emotional regulation ability, and improve the emotional regulation effect.

Description

Technical Field

[0001] The invention relates to the field of virtual reality depression disorder bioinformation feedback intervention, and in particular to a spinning bike with a motion feedback intervention system. Background Art

[0002] With the development of the times, depressive disorder has gradually become a major disease that endangers people's health. It is a classic of the new medical model. Whether from the perspective of etiology, clinic or therapy, depressive disorder is a model of biological, psychological and social models. Therefore, when intervening in depressive disorder, it is necessary to emphasize comprehensive and holistic considerations.

[0003] At present, the clinical diagnosis of mental disorders such as depression is mainly based on medical history, interviews, observations, scale filling, and questionnaire evaluation. The multi-channel physiological instrument is used to collect the patient's physiological responses for auxiliary diagnosis. The above-mentioned detection methods rely on the doctor's subjective experience and are interfered by the surrounding scene environment. In addition, there are shortcomings such as slow inquiry speed, low efficiency, expensive instruments and complex operation. In addition, there are drug interventions and non-drug interventions for people with depressive disorders. Drug treatment has large side effects, is prone to recurrence, and the intervention effect is difficult to evaluate. In addition, there is often overlap between multiple abnormal states, resulting in poor evaluation and intervention effects. Therefore, non-drug intervention technology is very important. Summary of the invention

[0004] The purpose of the present invention is to provide a spinning bike with a motion feedback intervention system, which can timely adjust the exercise intensity of the spinning bike user, enhance the user's emotion regulation ability, and improve the emotion regulation effect.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A dynamic bicycle with a motion feedback intervention system, comprising: a photoelectric speed measurement module, a heart rate belt detection module, an electroencephalogram detection module, a single chip computer module and VR glasses;

[0007] The signal input end of the single-chip microcomputer module is respectively connected to the signal output end of the photoelectric speed measurement module, the signal output end of the heart rate belt detection module and the signal output end of the electroencephalogram detection module, and the signal output end of the single-chip microcomputer module is connected to the signal input end of the VR glasses;

[0008] The photoelectric speed measurement module is used to detect the speed of the spinning bike in real time and form a speed pulse signal;

[0009] The heart rate belt detection module is used to detect the periodic changes of the cardiac current of the spinning bike user in real time to obtain the real-time heart rate variability;

[0010] The electroencephalogram detection module is used to detect changes in the user's electroencephalogram signals in real time to obtain a real-time electroencephalogram;

[0011] The single-chip computer module is used to calculate the real-time speed value through the speed pulse signal, and at the same time calculate the target speed value through the real-time heart rate variability, the real-time electroencephalogram, the target heart rate variability and the target electroencephalogram through the PID neural network decoupling control algorithm based on sparrow search, and ensure the stable operation of the system through the nonlinear fault-tolerant control law based on the adaptive neural network, and then transmit the real-time speed value and the target speed value to the VR glasses worn by the user in real time;

[0012] The user adjusts the speed of the spinning bike in time according to the target speed value and the real-time speed value to control the exercise intensity.

[0013] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0014] The present invention discloses a spinning bike with a motion feedback intervention system. A heart rate belt detection module and an electroencephalogram detection module are used to collect the heart rate variability and electroencephalogram of a user. A photoelectric speed measurement module is used to monitor the rotation of a driving wheel in real time to obtain a speed pulse signal. A single-chip computer module uses a sparrow search-based PID neural network decoupling control algorithm to combine the real-time heart rate variability, the real-time electroencephalogram with the target heart rate variability and the target electroencephalogram to obtain a target speed. A nonlinear fault-tolerant control law based on an adaptive neural network is used to reduce system errors to ensure stable operation of the system. Finally, the real-time speed value and the target speed value are transmitted to VR glasses worn by the user in real time. The user can adjust the speed of the spinning bike in time according to the target speed value and the real-time speed value, control the exercise intensity, enhance the user's emotion regulation ability, and improve the emotion regulation effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0016] Figure 1 A schematic diagram of the principle of a spinning bike of a motion feedback intervention system provided by an embodiment of the present invention;

[0017] Figure 2 A circuit diagram of a single chip microcomputer provided in an embodiment of the present invention;

[0018] Figure 3 A connection circuit diagram of a photoelectric speed measurement module and a single-chip microcomputer module provided in an embodiment of the present invention;

[0019] Figure 4A connection circuit diagram of a Bluetooth module and a single-chip microcomputer module provided in an embodiment of the present invention;

[0020] Figure 5 A flow chart of a PID neural network decoupling control algorithm based on sparrow search provided by an embodiment of the present invention;

[0021] Figure 6 A flow chart of a nonlinear fault-tolerant control law algorithm based on an adaptive neural network provided in an embodiment of the present invention.

[0022] Explanation of symbols: photoelectric speed measurement module-1, single chip computer module-2, power module-3, Bluetooth module-4, heart rate belt detection module-5, electroencephalogram detection module-6, photoelectric sensor-7, amplifier circuit-8, VR glasses-9. DETAILED DESCRIPTION

[0023] 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.

[0024] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] like Figure 1 As shown, an embodiment of the present invention provides a spinning bike with a motion feedback intervention system, comprising: a photoelectric speed measurement module 1, a heart rate belt detection module 5, an electroencephalogram detection module 6, a single chip computer module 2 and VR glasses 9.

[0026] The signal input end of the single-chip microcomputer module 2 is respectively connected to the signal output end of the photoelectric speed measurement module 1, the signal output end of the heart rate belt detection module 5 and the signal output end of the electroencephalogram detection module 6, and the signal output end of the single-chip microcomputer module 2 is connected to the signal input end of the VR glasses 9.

[0027] The photoelectric speed measuring module 1 is used to detect the speed of the spinning bike in real time and form a speed pulse signal. The heart rate belt detection module 5 is used to detect the periodic changes of the cardiac current of the spinning bike user in real time and obtain the real-time heart rate variability. The electroencephalogram detection module 6 is used to detect the changes of the user's brain electrical signal in real time and obtain the real-time electroencephalogram.

[0028] The single-chip computer module 2 is used to calculate the real-time speed value through the speed pulse signal, and at the same time calculate the target speed value from the real-time heart rate variability, real-time electroencephalogram, target heart rate variability and target electroencephalogram through the PID neural network decoupling control algorithm based on sparrow search, and ensure the stable operation of the system through the nonlinear fault-tolerant control law based on the adaptive neural network, and then transmit the real-time speed value and the target speed value to the VR glasses 9 worn by the user in real time.

[0029] The user adjusts the speed of the spinning bike in time according to the target speed value and the real-time speed value to control the exercise intensity.

[0030] The real-time speed value and the target speed value are fed back into the virtual reality scene. The user can adjust the real-time speed value in time according to the target speed value, thereby enhancing the user's emotion regulation ability, improving the emotion regulation effect, alleviating depression, and effectively treating patients with early depressive disorders.

[0031] The spinning bike of the motion feedback intervention system further includes a Bluetooth module 4. The Bluetooth module 4 is respectively connected to the heart rate belt detection module 5, the electroencephalogram detection module 6, the single chip computer module 2 and the VR glasses 9. The Bluetooth module 4 is mainly responsible for data communication.

[0032] The photoelectric speed measurement module 1 is installed on the driving wheel of the spinning bike, the heart rate belt detection module 5 installs the heart rate belt on the user's chest, the electroencephalogram detection module 6 installs the electroencephalogram sensor on the user's head, the single-chip computer module 2 is respectively connected to the power module 3, the Bluetooth module 4, and the photoelectric speed measurement module 1, and finally the Bluetooth module 4 is wirelessly connected to the heart rate belt detection module 5, the electroencephalogram detection module 6 and the cycling virtual scene. The photoelectric speed measurement module 1 monitors the rotation of the driving wheel of the spinning bike in real time, obtains the speed pulse signal, and transmits it to the single-chip microcomputer module 2. The heart rate belt detection module 5 detects the periodic changes of the cardiac current in the skin in real time to obtain the heart rate variability (HRV). At the same time, the electroencephalogram detection module 6 detects the changes of the electroencephalogram signal in real time to obtain the electroencephalogram (EEG) index, and feeds the result back to the single-chip microcomputer module 2 through the Bluetooth module 4. The single-chip microcomputer module 2 calculates the real-time speed value through the speed pulse signal, and at the same time, the real-time HRV, EEG and the target HRV, EEG obtained by the experiment are decoupled by the PID neural network decoupling control algorithm based on sparrow search (SSA) to obtain the target speed value, and the nonlinear fault-tolerant control law based on the adaptive neural network is used to ensure the stable operation of the system. The Bluetooth module 4 transmits the real-time HRV, real-time EEG, real-time speed value and target speed value to the cycling virtual scene in real time.

[0033] Furthermore, the dynamic bicycle of the motion feedback intervention system further includes: a power module 3. The power module 3 is respectively connected to the photoelectric speed measurement module 1 and the single chip computer module 2. The power module 3 is used to supply power to the entire device.

[0034] like Figure 3 As shown, the photoelectric speed measurement module 1 includes: a photoelectric sensor 7, a first resistor, a second resistor, a third resistor and an amplifier circuit 8. The photoelectric sensor 7, the first resistor, the second resistor, the third resistor and the amplifier circuit 8 are all installed at the driving wheel of the spinning bike. The positive power line of the power module 3 is connected to the positive power line of the photoelectric sensor 1 through the first resistor (resistor R1), and the negative power line of the power module 3 is connected to the negative power line of the photoelectric sensor 7 through the second resistor (resistor R2). The signal output end of the photoelectric sensor 7 is connected to the signal input end of the amplifier circuit 8 through the third resistor (resistor R3), and the signal output end of the amplifier circuit 8 is connected to the signal input end of the single-chip module 2. The photoelectric sensor 7 includes a light transmitter and a receiver. The light transmitter emits light, and the photoelectric sensor 7 works to output a pulse signal, thereby completing a control action. The grating disk is installed at the center of the driving wheel hub of the spinning bike, and the speed pulse signal of the spinning bike can be measured in real time by the photoelectric speed measurement method. Among them, the amplifier circuit 8 includes a transistor P and a resistor R4.

[0035] Exemplarily, the model of the photoelectric sensor 7 is a H42B6 slot-type photoelectric sensor, with a light emitter and a receiver installed on both sides of the slot. The light emitter emits infrared light or visible light. When there is no obstruction between the light emitter and the receiver, the receiver can receive light. When there is an obstruction between the light emitter and the receiver, the receiver cannot receive light. The positive and negative power lines of the power module 3 are connected to the positive and negative power lines of the photoelectric sensor 7, and the positive and negative power lines and data lines of the photoelectric sensor 7 are connected to the positive and negative power lines and data lines of the amplifier circuit 8. The positive and negative power lines of the amplifier circuit 8 are respectively connected to the VCC pin and the GND pin of the single-chip module 2, and the signal output line of the amplifier circuit 8 is connected to the P3.2 pin of the single-chip module 2. The grating disk is installed at the center of the driving wheel hub of the spinning bike, with a diameter of 64 mm and 50 grating grids.

[0036] This example uses the Polar H10 heart rate belt to measure the periodic changes of the cardiac current in the skin through electrodes on both sides. The time series signal of the RR interval is transformed into the frequency domain using the Fourier transform method to form a spectrum curve. The shape of the spectrum curve is analyzed to obtain HRV.

[0037] This example uses a portable three-lead EEG depression disorder identification system, measures the EEG signals on the scalp surface through EEG sensor electrodes, and innovatively proposes a continuous acquisition mode for the three-lead EEG signals of the forehead. It uses the asynchronous brain-like neuromorphic chip developed by the team for psychophysiological computing, and performs three-lead EEG acquisition on the prefrontal cortex (the command and control center of the brain), achieving accurate quantification of the negative emotional cognitive bias of depressed patients, and calculating the next target speed by extracting EEG features.

[0038] The photoelectric speed measuring module 1 monitors the rotation of the driving wheel of the spinning bike in real time, obtains the speed pulse signal, and transmits it to the single-chip computer module 2. Based on the speed pulse signal, the single-chip computer module 2 calculates the speed of the spinning bike. The calculation method is as follows:

[0039] The speed of the spinning bike is:

[0040] In the formula, V is the speed of the spinning bike, in revolutions per second, T is the sampling time of the speed pulse signal, in milliseconds, m is the count value of the microcontroller pulse counter within the T time, in pieces, and n is the number of gratings on the grating disk, in pieces. Afterwards, the real-time speed value of the spinning bike will be transmitted in real time to the virtual reality scene displayed by the VR glasses 9 through the Bluetooth module 4.

[0041] This embodiment uses the STC89C52 model single-chip microcomputer to realize the collection and processing of dynamic bicycle motion data. The circuit diagram is shown in Figure 2 .like Figure 4 As shown, the model of Bluetooth module 4 is HC-05, the VCC pin of Bluetooth module 4 is connected to the VCC pin of single-chip module 2, the GND pin of Bluetooth module 4 is connected to the GND pin of single-chip module 2, the RXD pin of Bluetooth module 4 is connected to the P3.1 pin of single-chip module 2, and the TXD pin of Bluetooth module 4 is connected to the P3.0 pin of single-chip module 2. When the Bluetooth serial port module leaves the factory, it is in slave mode by default. In order to complete the pairing of two Bluetooth modules 4, the Bluetooth module 4 connected to the single-chip module 2 is set as a slave, and the Bluetooth module 4 connected to one end of the PC is set as a host. The Bluetooth module 4 connected to one end of the PC is connected to the PC because the PC cannot be directly connected to the serial port, so a USB-to-TTL conversion module is used to achieve connection with the PC. After that, the baud rate is set to 9600bps, 8 data bits are set, and there is no parity check bit. After the pairing is completed, select the appropriate COM port, and open the corresponding COM port when communicating with the PC to complete the serial port communication and realize wireless signal transmission with the PC.

[0042] Based on the 51 MCU development board, the MCU P3.2 pin receives the pulse signal from the photoelectric sensor 7, counts and calculates the pulse signal, and transmits it to the SBUF register, and then transmits it to the PC wirelessly using the Bluetooth module 4. Because the STC89C52 MCU has 8k bytes of Flash, 512 bytes of RAM, and powerful processing capabilities, it also has 40 I / O port pins, which meets the number of I / O ports required by the data acquisition device, and also has 3 16-bit timers and counters, 4 external interrupts, which meet the timing and counting conditions required for data acquisition.

[0043] PID neural network decoupling control algorithm based on sparrow search:

[0044] The PID neural network can be divided into three layers: input layer, hidden layer and output layer. The PID neural network of n control variables contains n parallel identical sub-networks. Each sub-network is independent of each other and connected to each other through network connection weights. The input layer of each sub-network has two neurons, which receive the target value and current value of the control variable respectively. The hidden layer of each sub-network is composed of proportional element, integral element and differential element, which correspond to proportional control, integral control and differential control in the PID controller respectively. The output layer is the control law.

[0045] In the process of control, the PID neural network corrects the weight according to the gradient correction method based on the control quantity error, so that the control quantity is constantly close to the control target value. The control object of the PID neural network is a complex coupled system with multiple inputs and outputs, and the control quantities of the system are coupled with each other.

[0046] The learning algorithm used by the PID neural network is the gradient learning method. The initial weights are obtained randomly, and the weights may fall into the local optimal value during the learning process. Therefore, the sparrow search algorithm (SSA) is used to optimize the initial weights of the neural network.

[0047] The Sparrow Search Algorithm (SSA) is mainly inspired by the foraging behavior and anti-predation behavior of sparrows. The algorithm is relatively novel and has the advantages of strong optimization ability and fast convergence speed.

[0048] The mathematical model of the sparrow search algorithm is established, and the main rules are as follows:

[0049] (1) The discoverer usually has a higher energy reserve and is responsible for searching for areas with abundant food in the entire population, providing foraging areas and directions for all participants. In the model establishment, the level of energy reserve depends on the fitness value (Fitness Value) corresponding to the sparrow individual. In this example, the fitness function is the control error and.

[0050] (2) Once a sparrow discovers a predator, it will begin to chirp as an alarm signal. When the alarm value is greater than the safety value, the discoverer will take the joiner to another safe area to forage.

[0051] (3) The identities of discoverers and joiners change dynamically. As long as they can find a better source of food, every sparrow can become a discoverer, but the proportion of discoverers and joiners in the entire population remains unchanged. In other words, if one sparrow becomes a discoverer, another sparrow must become a joiner.

[0052] (4) The lower the energy of the joiners, the worse their foraging position in the entire population. Some hungry joiners are more likely to fly to other places to forage for more energy.

[0053] (5) During the foraging process, the joiners can always search for the finder that provides the best food, and then obtain food from the best food or forage around the finder. At the same time, some joiners may constantly monitor the finder and compete for food resources in order to increase their own predation rate.

[0054] (6) When aware of danger, sparrows at the edge of the group will quickly move to a safe area to gain a better position, while sparrows in the middle of the group will move randomly to get closer to other sparrows.

[0055] The real-time heart rate variability, real-time electroencephalogram, target heart rate variability and target electroencephalogram are calculated by the PID neural network decoupling control algorithm based on sparrow search to obtain the target speed value, including:

[0056] 1-1: Establish a PID neural network, use the sparrow search algorithm to optimize the initial weights of the PID neural network, and use the gradient learning method to learn the weights of the PID neural network.

[0057] 1-2: The seven features extracted from heart rate variability and electroencephalogram are used as control quantities; the control quantities include the maximum and central values ​​of power spectral density (PSD), sample entropy (SampEn), frontal alpha wave asymmetry (FAA), alpha wave power variation value (APV) and Lempel-Ziv complexity (LZC). The PID neural network of the seven control quantities contains seven parallel identical sub-networks, each of which is independent of each other and connected to each other through network connection weights.

[0058] 1-3: Input the target value and current value of the control quantity into the PID neural network, and correct the weight according to the gradient correction method based on the control quantity error during the control process, so that the control quantity continuously approaches the target value of the control quantity, and finally obtains the target speed value. Among them, the target value and current value of the control quantity are input, and then each sub-network passes through the proportional element, integral element and differential element, and the output is the control law.

[0059] The process of the PID neural network decoupling control algorithm based on sparrow search is as follows Figure 5 shown.

[0060] Nonlinear fault-tolerant control law based on adaptive neural network:

[0061] Aiming at the feedback problem of the exercise intervention system, a nonlinear fault-tolerant control law based on adaptive neural network is designed, considering the influence of interference, acquisition equipment failure and model uncertainty in the process of collecting the user's heart rate variability (HRV) and electroencephalogram (EEG) indicators through the heart rate belt detection module 5 and the electroencephalogram detection module 6. The control algorithm combines continuous terminal sliding mode control, radial basis neural network and adaptive control method. First, based on the sliding mode control theory, a fast terminal sliding mode surface is designed to ensure that the system tracking error can converge to zero in a finite time. Then, based on the terminal sliding mode surface, a terminal sliding mode control law based on adaptive radial basis neural network estimation is proposed. The neural network with adaptive parameters is used to approximate the system parameters and improve the anti-interference performance, and the smooth continuous control strategy is used to eliminate the tremor phenomenon in the terminal sliding mode. The convergence and global stability of the closed-loop system are proved by Lyapunov's analysis method.

[0062] The present invention designs a terminal sliding mode control law for a motion feedback intervention system. Based on the terminal sliding mode surface, a radial basis function neural network with adaptive parameters is used to approximate system parameters, and adaptive control items are designed to compensate for disturbances and approximate process errors, so as to enhance the closed-loop stability of the system and improve transient performance.

[0063] When a certain execution component fails during the operation of the system, the system model will change greatly during the whole process, and the control accuracy of the terminal sliding mode control law may be poor. Therefore, the radial basis function neural network can be used to approximate the system parameters. When the system parameters change, the radial basis function neural network is used to approximate and add adaptive control items to compensate for disturbances and errors in the approximate process, enhance the closed-loop stability of the system and improve transient performance.

[0064] The motion feedback intervention system can be modeled as a three-channel system, so the terminal sliding surface is designed as follows:

[0065] In order to ensure that the system error e converges to zero in a finite time, let x 1 =e, Design the following non-singular terminal sliding surface:

[0066] s=x 2 +C 1 x 1 +C 2 sig l (x 1 )

[0067] Where: s = [s 1 ,s 2 ,s 3 ] T ∈R 3 , 0 <l<1。C 1and C 2 Positive definite, and satisfies C i =diag(c i1 ,c i2 ,c i3 ), i = 1, 2. 1 、s 2 、s 3 Represent the sliding surfaces of the three channels, R 3 represents the real number domain; x 1 、x 2 Represent the first and second state variables respectively, C 1 and C 2 represents the coefficient of the first and second variables, c i1 、c i2 、c i3 They respectively represent the three-channel data that make up the i-th variable coefficient.

[0068] sig l (x 1 ) is defined as: sig l (x 1 )=[sig l (x 11 ),sig l (x 12 ),sig l (x 13 )] T , and define sig l (x)=|x| l sgn(x), where the sign function sgn(x) is defined as:

[0069]

[0070] Among them, sig l (x 1 ) represents the activation function, l represents the exponent, sig l (x 11 ), sig l (x 12 ), sig l (x 13 ) represent the activation functions of the three channels respectively.

[0071] Adaptive radial basis neural network control is introduced as follows:

[0072] The designed control law makes the system error converge to zero in a finite time:

[0073] u=G -1 (x)(u tsm +u RBFNN +u ad )

[0074] Where: G(x) is the system function; u tsm is the terminal sliding mode control term; u RBFNN is the nonlinear function F(x) approximated by radial basis neural network; uad It is an adaptive control term used to compensate for disturbances and approximate process errors, enhance system closed-loop stability and improve transient performance. The adaptive control term is defined as:

[0075]

[0076] In the formula, and are the estimated values ​​of the unknown actual interference ξ and the adaptive gain τ, respectively, μ is an arbitrary positive value, ||s|| 1 Represents the 1-norm of the sliding surface.

[0077] The process of the nonlinear fault-tolerant control law algorithm based on adaptive neural network is as follows Figure 6 shown.

[0078] The present invention designs a terminal sliding mode adaptive radial basis neural network control law for motion intervention systems with stuck actuators and abnormal feedback. The designed controller can ensure that the system converges to zero in a short time, and the continuously changing control law completely eliminates the chattering phenomenon in the sliding mode control. In addition, the controller only uses the information of the control input matrix, uses the radial basis neural network to approximate the system parameters, and uses adaptive control items to compensate for external interference and errors in the approximation process, so that it has good robustness.

[0079] Combining real-time monitoring equipment with exercise equipment, the appropriate exercise intensity is calculated to help patients with early-stage depression to relieve depression and improve their mental state. According to experimental data, the effectiveness of this method in regulating early-stage depression symptoms is over 85%.

[0080] The advantages of the present invention compared with the prior art are:

[0081] (1) The present invention proposes a spinning bike with a motion feedback intervention system, proposes a personalized motion feedback index based on the fusion of heart rate variability (HRV) and electroencephalogram (EEG), and constructs a motion feedback intervention model, which has an auxiliary therapeutic effect on patients with depressive disorders and has basically no side effects compared with drug treatment;

[0082] (2) Aiming at the motion intervention system with stuck execution components and abnormal feedback, the present invention proposes a PID neural network decoupling control algorithm based on sparrow search algorithm (SSA) and a nonlinear fault-tolerant control law based on adaptive neural network, which have good anti-interference performance and good robustness.

[0083] (3) The present invention can adjust the speed of the spinning bike in real time according to the user's condition to achieve the best treatment effect, ensuring the scientificity and safety of the treatment. At the same time, the user rides in a virtual scene, which solves the problem of exercisers being bored during the fitness process and increases the entertainment value of the spinning bike.

[0084] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0085] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A spinning bike with a motion feedback intervention system, characterized in that: include: Photoelectric speed measurement module, heart rate belt detection module, EEG detection module, single chip microcomputer module and VR glasses; The signal input end of the single-chip microcomputer module is respectively connected to the signal output end of the photoelectric speed measurement module, the signal output end of the heart rate belt detection module and the signal output end of the electroencephalogram detection module, and the signal output end of the single-chip microcomputer module is connected to the signal input end of the VR glasses; The photoelectric speed measurement module is used to detect the speed of the spinning bike in real time and form a speed pulse signal; The heart rate belt detection module is used to detect the periodic changes of the cardiac current of the spinning bike user in real time to obtain the real-time heart rate variability; The electroencephalogram detection module is used to detect changes in the user's electroencephalogram signals in real time to obtain a real-time electroencephalogram; The single-chip computer module is used to calculate the real-time speed value through the speed pulse signal, and at the same time calculate the target speed value through the real-time heart rate variability, the real-time electroencephalogram, the target heart rate variability and the target electroencephalogram through the PID neural network decoupling control algorithm based on sparrow search, and ensure the stable operation of the system through the nonlinear fault-tolerant control law based on the adaptive neural network, and then transmit the real-time speed value and the target speed value to the VR glasses worn by the user in real time; The user timely adjusts the speed of the spinning bike according to the target speed value and the real-time speed value to control the exercise intensity; The photoelectric speed measurement module includes: a photoelectric sensor, a first resistor, a second resistor, a third resistor and an amplifier circuit; The photoelectric sensor, the first resistor, the second resistor, the third resistor and the amplifier circuit are all installed at the driving wheel of the spinning bike; The positive power line of the power module is connected to the positive power line of the photoelectric sensor through a first resistor, and the negative power line of the power module is connected to the negative power line of the photoelectric sensor through a second resistor; The signal output end of the photoelectric sensor is connected to the signal input end of the amplifier circuit via the third resistor, and the signal output end of the amplifier circuit is connected to the signal input end of the single-chip computer module; The EEG detection module is based on an asynchronous brain-like neuromorphic chip, and adopts a continuous acquisition mode of EEG signals for the three leads of the forehead to perform three-lead EEG acquisition on the prefrontal cortex to obtain the real-time EEG of the user.

2. The spinning bike according to the motion feedback intervention system of claim 1, characterized in that: The photoelectric speed measurement module is installed on the driving wheel of the spinning bike, the heart rate belt of the heart rate detection module is installed on the chest of the user, and the electroencephalogram sensor of the electroencephalogram detection module is installed on the head of the user.

3. The spinning bike according to the motion feedback intervention system of claim 1, characterized in that: Also includes: Bluetooth module; The Bluetooth module is connected to the heart rate belt detection module, the electroencephalogram detection module, the single-chip computer module and the VR glasses respectively.

4. The spinning bike according to the motion feedback intervention system of claim 1, characterized in that: Also includes: Power module; The power supply module is connected with the photoelectric speed measurement module and the single chip computer module respectively.

5. The spinning bike of the motion feedback intervention system according to claim 1, characterized in that: The real-time heart rate variability, real-time electroencephalogram, target heart rate variability and target electroencephalogram are calculated by the PID neural network decoupling control algorithm based on sparrow search to obtain the target speed value, including: A PID neural network is established, an initial weight of the PID neural network is optimized by using a sparrow search algorithm, and a gradient learning method is used to learn the weight of the PID neural network; the PID neural network includes 7 parallel identical sub-networks; The seven features extracted from heart rate variability and electroencephalogram were used as control quantities; The target value and current value of the control quantity are input into the PID neural network, and the weight is corrected according to the gradient correction method based on the control quantity error during the control process, so that the control quantity continuously approaches the target value of the control quantity, and finally the target speed value is obtained.

6. The spinning bike of the motion feedback intervention system according to claim 1, characterized in that: The determination process of the nonlinear fault-tolerant control law based on the adaptive neural network is as follows: Establish a non-singular terminal sliding surface s = x2 + C1x1 + C2sig l (x1); where s represents the sliding surface, s = [s1, s2, s3] T ∈R 3 , s1, s2, s3 represent the sliding surfaces of the three channels, R 3 represents the real number domain; x1 and x2 represent the first and second state variables respectively, x1=e, e is the system error; C1 and C2 represent the first and second variable coefficients, C1 and C2 are positive definite, and satisfy C i =diag(c i1 , c i2 , c i3 ), i=1,2,c i1 、c i2 、c i3 Respectively represent the three-channel data that make up the i-th variable coefficient; sig l (x1) represents the activation function, l represents the exponent, 0<l<1, sig l (x1)=[sig l (x 11 ), sig l (x 12 ), sig l (x 13 )] T , sig l (x 11 ), sig l (x 12 ), sig l (x 13 ) represent the activation functions of the three channels, sig l (x)=|x| l sgn(x), sgn(x) represents the sign function, Variable x = x 11 、x 12 or x 13 ; The adaptive control term used to compensate for disturbances and process errors is defined as In the formula, u ad is the adaptive control term, and are the estimated values ​​of the unknown actual disturbance ξ and the adaptive gain τ, respectively, μ is an arbitrary positive value, and ||ε|||1 represents the 1-norm of the sliding surface; On the basis of the non-singular terminal sliding surface, the radial basis function neural network with adaptive parameters is used to approximate the system parameters, and an adaptive control term is added to obtain the nonlinear fault-tolerant control law based on adaptive neural network that makes the system error converge to zero in a finite time: u=G -1 (x)(u tsm +u RBFNN +u ad ), where u represents the control law and G(x) is the system function; u tsm is the terminal sliding mode control term; u RBFNN is the nonlinear function F(x) approximated by radial basis neural network.

7. The spinning bike of the motion feedback intervention system according to claim 1, characterized in that: The calculation formula of real-time speed value is: Where V is the speed of the spinning bike, T is the sampling time of the speed pulse signal, m is the count value of the microcontroller pulse counter within T time, and n is the number of gratings on the grating disk.

8. The spinning bike of the motion feedback intervention system according to claim 1, characterized in that: The VR glasses are used to display a virtual cycling scene.

Images