CN117226852A - Soft exoskeletons control method and device - Google Patents

Abstract

The embodiment of the invention provides a soft exoskeleton control method and a device, which belong to the technical field of robots, and can calculate the gas flow estimation of an electromagnetic proportional valve by utilizing a gas flow model and a driving voltage of the electromagnetic proportional valve, generate a flow control command by utilizing a dynamics model and a gas flow estimation of an actuator, extract a target flow from the flow control command, and control the electromagnetic proportional valve to work according to the gas flow estimation and the target flow, so that the actuator of the soft exoskeleton can work by obtaining the torque and the bending angle of the soft exoskeleton based on the target flow and the dynamics model of the actuator, and the dynamics model is obtained by taking the product of the torque and the angular speed of the soft exoskeleton as output power and constructing based on the output power, so that the dynamics model of the soft exoskeleton actuator is not limited by inertia quality, resistance and traction, thereby improving the stability and the robustness of the soft exoskeleton control.

Description

Software exoskeleton control method and device

Technical field

The present invention relates to the field of robotic technology, and specifically to a soft exoskeleton control method and device.

Background technique

A mechanical exoskeleton, or powered exoskeleton, is a mechanical device made of a steel frame that can be worn by people. This equipment can provide extra energy for the movement of the limbs. Mechanical exoskeletons include rigid exoskeletons and soft exoskeletons. Soft exoskeletons are a type of flexible exoskeleton made of special soft materials. The deformation of the software itself transmits power and enables movement. The emerging wearable exoskeleton robot is a soft exoskeleton that assists human limb movement by driving the rotation of human bone joints. It can enhance the strength of human limbs and replace doctors in providing limb movement training during the rehabilitation process.

Since the soft exoskeleton is restricted by the uncertain inertial mass of the human limb and autonomous force, the general control method based on Euler-Lagrangian dynamic modeling is not suitable for wearable pneumatic soft exoskeleton actuators. Therefore, there is an urgent need for a high-stability and robust control method suitable for soft exoskeletons.

Contents of the invention

In view of this, the purpose of the present invention is to provide a soft exoskeleton control method, device and electronic equipment, which can improve the problems of low stability and robustness caused by the current exoskeleton control method being applied to soft exoskeletons.

In order to achieve the above objects, the technical solutions adopted in the embodiments of the present invention are as follows:

In a first aspect, an embodiment of the present invention provides a method for controlling a soft exoskeleton, which is applied to the control system of the soft exoskeleton. The soft exoskeleton further includes an electromagnetic proportional valve and an actuator. The method includes:

Based on the driving voltage of the electromagnetic proportional valve at the current moment, use the gas flow model of the electromagnetic proportional valve to calculate the gas flow estimate of the electromagnetic proportional valve;

Using the dynamic model of the actuator and the gas flow estimate, a motion estimate is calculated; wherein the motion estimate includes at least one of a torque estimate and a bending angle estimate;

Generate a flow control command according to the control command at the current moment, the motion estimate and the gas flow estimate;

Extract the target flow rate from the flow control command, and calculate the target drive voltage of the electromagnetic proportional valve based on the gas flow estimate and the target flow rate;

Control the electromagnetic proportional valve to operate according to the target driving voltage, so that the actuator can obtain the torque or bending angle of the soft exoskeleton based on the target flow rate and the dynamic model of the actuator;

Wherein, the dynamic model of the actuator is constructed based on the product of the torque and the angular velocity of the soft exoskeleton as the output power.

Further, the actuator further includes an actuation controller, and the method further includes the step of constructing the actuation controller, which step includes:

The product of the torque and the angular velocity of the soft exoskeleton is used as the output power of the soft exoskeleton;

Construct a dynamic model of the actuator based on the output power and the gas flow output by the gas flow model;

The dynamic model is processed using a linearization algorithm to obtain an actuation controller.

Further, the step of constructing a dynamic model of the actuator based on the output power and the gas flow output by the gas flow model includes:

Use the ideal gas equation of state to express the gas volume in the actuator chamber;

Use the product of the gas pressure and the gas flow rate input by the solenoid proportional valve to express the input gas power;

Express the accumulated potential energy of the actuator based on the product of the gas volume and the gas pressure;

The input energy is expressed according to the input gas power, the output energy is expressed according to the output power, and the bending angle of the soft exoskeleton is expressed according to the angular velocity;

The gas flow output by the gas flow model is used as system input, the input gas energy, the accumulated potential energy, the output energy and the bending angle are used as system state variables, and the output power is used to calculate the dynamic assist torque, The static torque is calculated based on the principle of virtual work, and the nonlinear state space model of the actuator is constructed.

Further, the dynamic model includes:

in, ,/> ,/> Characterizing gas flow,/> Characterizing gas pressure,/> Characterizing the molar gas constant,/> Characterizing temperature,/> Characterizes the molar mass of air,/> Characterize gas quality,/> Characterizes dynamic response time delay,/> Characterizes the height of the exoskeleton cavity,/> Characterizes the cross-sectional area of the exoskeleton lumen,/> Characterizes conversion efficiency,/> Characterizes the input gas energy,/> Represents accumulated potential energy,/> Characterizes the output energy,/> Characterizes the bending angle of the soft exoskeleton.

Further, the step of calculating the motion estimate using the dynamic model of the actuator and the gas flow estimate includes:

When the control mode is the dynamic assist mode, the torque estimate of the actuator is obtained by using the dynamic model of the actuator based on the gas flow estimate and the gas pressure and temperature of the inner cavity of the soft exoskeleton. quantity;

When the control mode is the limb passive mode, based on the gas flow estimate and the gas pressure and temperature of the soft exoskeleton cavity, the bending angle of the actuator is obtained by using the dynamic model of the actuator estimator.

Further, the method also includes the step of constructing a gas flow model of the electromagnetic proportional valve, which step includes:

Based on the experimental input voltage and experimental gas flow of the electromagnetic proportional valve, the voltage flow equation of the electromagnetic proportional valve is fitted; wherein the voltage flow equation represents the input voltage, flow coefficient and effective opening of the electromagnetic proportional valve area relationship;

Based on the Sanville flow formula and the voltage flow equation, a gas flow model of the electromagnetic proportional valve is obtained.

Further, the step of fitting the voltage flow equation of the electromagnetic proportional valve based on the experimental input voltage and experimental gas flow of the electromagnetic proportional valve includes:

Using the least squares method, curve fitting was performed on the experimental input voltage and experimental gas flow rate of the electromagnetic proportional valve to obtain the voltage flow equation.

Further, the gas flow model includes:

in, Characterizing gas flow,/> Characterizing the voltage-flow equation,/> Characterizing the molar gas constant,/> Characterizing temperature,/> Indicates the gas source pressure,/> Characterizes the gas pressure in the chamber,/> Characterizes the driving voltage.

Further, the linearization algorithm includes any one of Jacobian matrix linearization, Taylor expansion and nonlinear dynamic inverse control.

In a second aspect, embodiments of the present invention provide a soft exoskeleton control device, which is applied to the control system of the soft exoskeleton. The soft exoskeleton further includes an electromagnetic proportional valve and an actuator. The soft exoskeleton control device Including flow estimation module, motion estimation module, estimation calculation module, feedback calculation module and control module;

The flow estimation module is used to calculate the gas flow estimate of the electromagnetic proportional valve based on the driving voltage of the electromagnetic proportional valve at the current moment and using the gas flow model of the electromagnetic proportional valve;

The motion estimation module is used to calculate a motion estimate using the dynamic model of the actuator and the gas flow estimate; wherein the motion estimate includes a torque estimate and a bending angle estimate. at least one;

The estimation calculation module is used to generate a flow control command according to the control command at the current moment, the motion estimate and the gas flow estimate;

The feedback calculation module is used to extract the target flow rate from the received flow control command, and calculate the target drive voltage of the electromagnetic proportional valve according to the gas flow estimate and the target flow rate;

The control module is used to control the electromagnetic proportional valve to operate according to the target driving voltage, so that the actuator obtains the soft exoskeleton based on the target flow rate and the dynamic model of the actuator. torque or bending angle;

Wherein, the dynamic model of the actuator is constructed based on the product of the torque and the angular velocity of the soft exoskeleton as the output power.

In a third aspect, embodiments of the present invention provide an electronic device, including a processor and a memory. The memory stores machine-executable instructions that can be executed by the processor. The processor can execute the machine-executable instructions. To implement the software exoskeleton control method described in the first aspect.

In a fourth aspect, embodiments of the present invention provide a storage medium on which a computer program is stored. When the computer program is executed by a processor, the software exoskeleton control method as described in the first aspect is implemented.

According to the software exoskeleton control method and device provided by the embodiment of the present invention, the proportional valve controller of the software exoskeleton can calculate the gas flow rate of the electromagnetic proportional valve based on the driving voltage of the electromagnetic proportional valve at the current moment and using the gas flow model of the electromagnetic proportional valve. Flow estimate, and use the actuator's dynamic model and gas flow estimate to calculate the motion estimate, thereby generating a flow control command based on the control command, motion estimate, and gas flow estimate at the current moment, from the flow control command The target flow rate is extracted from the gas flow estimate and the target flow rate, and the target driving voltage of the electromagnetic proportional valve is calculated, thereby controlling the electromagnetic proportional valve to work according to the target driving voltage, so that the actuator of the soft exoskeleton is based on the target flow rate and the actuator The dynamic model of the soft exoskeleton is used to obtain the torque and bending angle of the soft exoskeleton to perform work, and the dynamic model of the actuator of the soft exoskeleton is based on the output power of the product of the torque and angular velocity of the soft exoskeleton. It is constructed so that the dynamic model of the soft exoskeleton actuator is not restricted by the inertial mass of the human limb and the exoskeleton, thereby greatly improving the stability and robustness of the soft exoskeleton control.

In order to make the above objects, features and advantages of the present invention more obvious and easy to understand, the preferred embodiments are described in detail below along with the accompanying drawings.

Description of drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore do not It should be regarded as a limitation of the scope. For those of ordinary skill in the art, other relevant drawings can be obtained based on these drawings without exerting creative efforts.

Figure 1 shows a schematic structural diagram of a software exoskeleton control system provided by an embodiment of the present invention.

Figure 2 shows a control principle diagram of the software exoskeleton control method provided by the embodiment of the present invention.

FIG. 3 shows one of the schematic flow charts of the software exoskeleton control method provided by the embodiment of the present invention.

Figure 4 shows the second schematic flowchart of the software exoskeleton control method provided by the embodiment of the present invention.

Figure 5 shows the third schematic flowchart of the software exoskeleton control method provided by the embodiment of the present invention.

Figure 6 shows a block diagram of a software exoskeleton control device provided by an embodiment of the present invention.

FIG. 7 shows a block diagram of an electronic device provided by an embodiment of the present invention.

Explanation of reference signs: 100-soft exoskeleton control system; 10-soft exoskeleton; 20-control box; 21-electromagnetic proportional valve; 22-pressure sensor; 23-pressure solenoid valve; 24-exhaust solenoid valve; 25- Vacuum solenoid valve; 26-controller; 27-pneumatic pump; 28-temperature sensor; 30-angle sensor; 40-force sensor; 50-inertial measurement unit; 60-soft exoskeleton control device; 601-flow estimation module; 602 -Motion estimation module; 603-estimation calculation module; 604-feedback calculation module; 605-control module; 70-electronic equipment.

Detailed ways

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, not all of them. The components of the embodiments of the invention generally described and illustrated in the figures herein may be arranged and designed in a variety of different configurations.

Accordingly, the following detailed description of embodiments of the invention provided in the appended drawings is not intended to limit the scope of the claimed invention, but rather to represent selected embodiments of the invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without any creative work fall within the scope of protection of the present invention.

It should be noted that relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that these entities or operations are mutually exclusive. any such actual relationship or sequence exists between them. Furthermore, the terms "comprises," "comprises," or any other variations thereof are intended to cover a non-exclusive inclusion such that a process, method, article, or apparatus that includes a list of elements includes not only those elements, but also those not expressly listed other elements, or elements inherent to the process, method, article or equipment. Without further limitation, an element defined by the statement "comprises a..." does not exclude the presence of additional identical elements in a process, method, article, or apparatus that includes the stated element.

The emerging wearable exoskeleton robot assists human limb movement by driving the rotation of human bone joints. It can enhance the strength of human limbs and replace doctors in providing limb movement training during the rehabilitation process. Wearable soft exoskeleton is a type of flexible exoskeleton made of special soft materials. The deformation of the software itself transmits power and realizes movement. Rigid exoskeleton actuators are capable of generating the torque and precise motion required for assisted locomotion; however, biomechanics in terms of mechanical complexity, weight and size, exoskeleton and human joint alignment issues, compliance and safety of human joint motion Machine-to-machine coupling issues hinder its true potential for widespread use. In contrast, the soft exoskeleton method is light in weight and highly flexible, and can adapt to human movement and prevent damage to the human body during tasks. In particular, wearable soft exoskeletons made of folded and folded high-elastic fabrics have complete Wearable and lightweight potential.

After compressed air is injected into the actuator of the soft exoskeleton, the gas pressure gives the actuator a bending stiffness and an expanded arc length, allowing it to drive the exoskeleton robot. The basic physical laws of a soft exoskeleton are more complex than those of a rigid exoskeleton. Its inherent complexity is reflected in the scalability, elastic deformation, and infinite freedom of movement of a highly flexible structure of soft materials, which requires the design of special constraints and structural restrictions. Useless freedom of movement, efficiently transmits pneumatic energy to the optimal part to help limb bending, achieving the best biomechanical human-machine coupling. Structural restraint methods include fiber restraint, rib structure and independent air chamber structure bellows structure. Mathematical modeling methods based on structure and material properties include: piecewise constant curvature method, Euler-Bernoulli beam equation, Cosserat rod theory, using superelastic material stress and strain theory and geometric analysis methods to establish the bending angle of the software actuator. Mathematical models, finite element modeling methods.

Current soft exoskeleton control systems usually follow the Euler-Lagrangian dynamics model of traditional robots. The force of the Euler-Lagrangian dynamics model is equal to the product of the overall mass and acceleration. However, when the soft exoskeleton is worn until the human limb moves as one, the actuator model is restricted by the uncertain human limb load: 1) The inertial mass in the model is the sum of the two; the limb inertial mass varies from person to person. , especially holding and placing objects will lead to changes in inertial mass, requiring additional sensors for calibration; 2) The traction or resistance of human limbs on the exoskeleton due to arbitrary and independent force, leading to uncertainty in the resultant force in the model.

Different from traditional robots, wearable soft exoskeleton robots are designed to provide power assistance and provide dynamic torque or static torque in the same direction as the bending direction of human limbs. Obviously, the force assist of the soft exoskeleton is different from the force of the Euler-Lagrangian dynamic model, which leads to instability when the control method based on the Euler-Lagrangian dynamic model is applied to the control of the soft exoskeleton. The problem of low performance and robustness.

Based on the above considerations, embodiments of the present invention provide a soft exoskeleton control method, which can improve the problems of low stability and robustness existing in the current soft exoskeleton control.

The software exoskeleton control method provided by the embodiment of the present invention can be applied to the software exoskeleton control system 100 as shown in Figure 1. The software exoskeleton control system 100 can include a software exoskeleton 10, a control box 20, an angle sensor 30, Force sensor 40 and inertial measurement unit 50. The control box 20 is equipped with a controller 26 (i.e. microprocessor), an air source pressure sensor, an electromagnetic proportional valve 21, an inner cavity air pressure sensor 22, a temperature sensor 28, and an electric pneumatic pump. 27. Pressure solenoid valve 23, vacuum solenoid valve 25, exhaust solenoid valve 24 and PWM voltage drive, etc.

The soft exoskeleton 10 may be a hand soft exoskeleton 10 robot, and may include power-assisting parts such as the thumb, index finger, middle finger, ring finger, little finger, wrist, and elbow, and any one of the power-assisting parts of the thumb, index finger, middle finger, ring finger, and little finger. The force sensor 40, the flexible angle sensor 30, the temperature sensor 28, the inner cavity air pressure sensor 22, the electromagnetic proportional valve 21 and the air source pressure sensor are connected in sequence. The force sensor 40 and the inertial measurement unit 50 are connected in sequence to the power-assisted parts such as the wrist and elbow. , temperature sensor 28, inner cavity air pressure sensor 22, electromagnetic proportional valve 21 and air source pressure sensor.

The controller 26 includes controllers of the electromagnetic proportional valve 21, the pneumatic pump 27, the pressure solenoid valve 23, the vacuum solenoid valve 25, the exhaust solenoid valve 24 and the electromagnetic proportional valve 21, as well as the torque controller, the angle controller and the flow controller.

The flexible angle sensor 30 is used to measure the bending angle of the connected finger.

The inertial measurement unit 50 is used to measure the bending angle of the wrist and elbow.

The pressure solenoid valve 23 is used to inflate the soft exoskeleton 10 .

The vacuum solenoid valve 25 is used to pump air to the soft exoskeleton 10 .

The exhaust solenoid valve 24 is used to deflate the soft exoskeleton 10 .

The electromagnetic proportional valve 21 is used to control the gas flow of the booster part where it is located.

The controller 26 is used to control the driving voltage of each electromagnetic proportional valve 21.

In a possible implementation, the embodiment of the present invention provides a software exoskeleton control method. The control principle of the software exoskeleton control method is shown in Figure 2, including a controller, an electromagnetic proportional valve and a software exoskeleton. The mathematical model of the actuator and electromagnetic proportional valve is the gas flow model H1, and the mathematical model of the actuator is the dynamic model H2. The gas flow estimate can be obtained based on the gas flow model of the electromagnetic proportional valve and the power of the actuator is used. The chemical model and gas flow estimate are used to calculate the motion estimate. The gas flow estimate and motion estimate are fed back to the controller to adjust the intake air flow of the electromagnetic proportional valve, and then control the torque of the software exoskeleton based on the intake air flow. or bending angle to achieve closed-loop control and form a stable gas flow control system.

It should be noted that in Figure 2 Is the torque controller,/> Angle controller,/> Is the flow controller,/> Represents gas mass flow,/> Represents the gas flow estimate,/> Represents the torque estimate,/> represents the bending angle estimate, u represents the system input. Torque controller, angle controller and flow controller can adopt PID control method or sliding mode control method.

The control system in Figure 2 can have two control strategies: torque control and bending angle control. (1) Torque control is used in the dynamic assist mode (under the dynamic assist module, the power of the soft exoskeleton is added to the human limbs to increase the force generated by the human muscles). At this time, the switch closed, switch/> Open. In dynamic assist mode, torque command/> It can be a constant value (constant assist), or it can be an output signal from a myoelectric sensor placed on the corresponding skin surface of the upper limb, as a torque command for controlling the soft exoskeleton robot/> , the force changes with the strength of the electromyographic signal. (2) The bending angle control is used in the passive mode of the limb (for example, in the early stage of rehabilitation, the limb cannot bend on its own and is pulled by the soft exoskeleton). At this time, the switch/> closed, switch/> Open. The control command at this time can be an angle command issued by the caregiver/> .

Referring to Figure 3, the software exoskeleton control method may include the following steps. In this embodiment, the software exoskeleton control method is applied to the software exoskeleton control system 100 in FIG. 1 as an example.

S11. Based on the driving voltage of the electromagnetic proportional valve at the current moment, use the gas flow model of the electromagnetic proportional valve to calculate the gas flow estimate of the electromagnetic proportional valve.

S13, use the dynamic model of the actuator and the gas flow estimate to calculate the motion estimate.

It should be noted that in the dynamic assist mode, the motion estimate may include a torque estimate, and in the limb passive mode, the motion estimate may include a bending angle estimate.

S15: Generate a flow control command based on the control command, motion estimate and gas flow estimate at the current moment.

S17, extract the target flow rate from the flow control command, and calculate the target drive voltage of the electromagnetic proportional valve based on the estimated gas flow rate and the standard flow rate.

S19, control the electromagnetic proportional valve to work according to the target driving voltage, so that the actuator can obtain the torque or bending angle of the soft exoskeleton based on the target flow rate and the dynamic model of the actuator.

In this embodiment, the dynamic model of the actuator is constructed based on the product of the torque and the angular velocity of the soft exoskeleton 10 as the output power.

Taking the electromagnetic proportional valve 21 as an example of the electromagnetic proportional valve 21 connected to the wrist assist part, when receiving the flow control command, the controller 26 obtains the driving voltage of the electromagnetic proportional valve 21 connected to the wrist assist part at the current time, and uses the electromagnetic The gas flow model of the proportional valve 21 is used to calculate the gas flow estimate of the electromagnetic proportional valve 21 .

According to the control mode selected by the user (which can be a dynamic assist mode or a limb passive mode), the controller 26 uses the dynamic model of the actuator and the gas flow estimate to calculate the motion estimate. In the dynamic assist mode, the motion estimator is the torque estimator, and in the limb passive mode, the motion estimator is the bending angle estimator.

Furthermore, the controller 26 calculates the target flow rate based on the motion estimate, the control command at the current time, and the gas flow rate estimate, and generates a flow control command. The controller 26 can calculate the difference between the control quantity in the control command and the bending angle or torque according to the bending angle or torque in the control command, and then search or calculate the corresponding gas flow difference according to the difference, and then according to the difference The gas flow difference and gas flow estimate are used to calculate the target flow and generate a flow control command.

The target flow rate of the electromagnetic proportional valve 21 connected to the wrist assist unit is extracted from the flow control command. Furthermore, the target drive voltage of the electromagnetic proportional valve 21 is calculated based on the estimated gas flow rate and the target flow rate.

For example, based on the difference between the gas flow estimate and the target flow, the difference correspondence can be found in the preset correspondence table between the flow difference and the voltage adjustment value of the electromagnetic proportional valve 21 connected to the wrist assist part. voltage adjustment value to obtain the target driving voltage. You can also directly use the target driving voltage calculation formula to calculate the target driving voltage.

After obtaining the target driving voltage, the controller 26 controls the electromagnetic proportional valve 21 connected to the wrist assist part to operate according to the target driving voltage, so that the gas flow rate of the electromagnetic proportional valve 21 reaches the target flow rate, thereby actuating the soft exoskeleton 10 The device can obtain the torque and bending angle of the wrist assist part based on the target flow rate of gas and the dynamic model of the brake, so as to realize the movement of the wrist assist part.

In the above-mentioned soft exoskeleton control method, the dynamic model of the soft exoskeleton actuator is based on the product of the soft exoskeleton's torque and angular velocity as the output power, and is constructed based on the output power, so that the power of the soft exoskeleton actuator The learning model is not restricted by the inertial mass, resistance and traction force of the human limbs and exoskeleton, and at the same time achieves closed-loop torque-free and angle-sensorless control, which can greatly improve the stability and robustness of the soft exoskeleton control.

For step S13, the calculated motion estimation amount is different in different control modes. When the control mode is the dynamic assist mode, based on the gas flow estimate and the gas pressure and temperature of the inner cavity of the soft exoskeleton, the dynamic model of the actuator is used to obtain the torque estimate of the actuator. When the control mode is the limb passive mode, the bending angle estimate of the actuator is obtained by using the dynamic model of the actuator based on the gas flow estimate and the gas pressure and temperature in the soft exoskeleton cavity.

Furthermore, the actuator of the soft exoskeleton can include a brake controller. In order to realize that the control of the soft exoskeleton actuator is not restricted by the inertial mass, resistance and traction force of the human limb and the exoskeleton, the ideal gas state equation and The energy conversion principle is used to construct the dynamic model of the soft exoskeleton actuator. In a possible implementation, referring to Figure 4, the step of building a brake controller can be implemented in the following manner.

S21, the product of the torque and the angular velocity of the soft exoskeleton is used as the output power of the soft exoskeleton.

S23: Construct a dynamic model of the actuator based on the output power and the gas flow output by the gas flow model.

S25, use the linearization algorithm to process the dynamics model to obtain the actuation controller.

Further, since the gas kinetic energy input to the actuator, the potential energy accumulated by the actuator, the external rotational kinetic energy and energy loss comply with the law of energy conservation, step S23 can be implemented in the following manner:

23-1, use the ideal gas equation of state to express the gas volume in the actuator chamber.

23-2. Use the product of the gas pressure and the gas flow rate input by the solenoid proportional valve to express the input gas power.

23-3, based on the product of gas volume and gas pressure, represents the accumulated potential energy of the actuator.

23-4. The input gas energy is expressed according to the input gas power, the output energy is expressed according to the output power, and the bending angle of the soft exoskeleton is expressed according to the angular velocity.

It should be understood that the derivative of the input gas energy is the input gas power, the derivative of the output energy is the output power, and the derivative of the bending angle is the angular velocity. Therefore, the input gas power, output power and angular velocity can respectively express the input gas energy, output energy and bending angle.

23-5, use the gas flow output by the gas flow model as the system input, use the input gas energy, accumulated potential energy, output energy and bending angle as the system state variables, use the output power to calculate the dynamic assist torque, and calculate the static torque based on the principle of virtual work, Construct a nonlinear state space model of the actuator.

The gas output by the solenoid proportional valve enters the actuator cavity of the soft exoskeleton through the air conduit. Therefore, the gas mass M flowing from the solenoid proportional valve to the actuator is the integral of the gas mass flow rate G, which can be expressed as: , where G represents the gas mass flow rate,/> represents the gas density.

The gas parameters in the actuator chamber can be expressed as: , where p is the gas pressure in the chamber (Pa), V is the gas volume (m³), T is the temperature (K), R is the molar gas constant =8.314J/(mol/> K), M is the mass of the gas (g), the molar mass of air μ =29g/mol, and T is the temperature (K).

Therefore, the input gas power of the solenoid proportional valve is the product of gas flow and gas pressure, which can be expressed as: , P _in represents the input gas power.

right Find the derivative, and/> , the change rate of the gas volume in the actuator chamber can be obtained, which can be expressed as:/> .

The potential energy accumulated in the actuator cavity (i.e. inner cavity) is equal to the product of gas pressure and gas volume, that is . Rate of change of accumulated potential energy in actuator chamber/> It can be expressed as: , where,/> Represents the rate of change of accumulated potential energy in the actuator chamber,/> ,/> .

further, ,/> .

The output function of the actuator is equal to the input gas power minus the rate of change of the accumulated potential energy, which can be expressed as: , P _out represents the output power, and eta represents the conversion efficiency.

The bending angle of each assisting part of the soft exoskeleton is expressed by the geometric formula of the axial expansion of the actuator, which is: ,/> represents the volume of the actuator cavity when not inflated, S represents the cross-sectional area of the inner cavity, /> is the height of the actuator cavity,/> Represents the bending angle of the actuator (i.e., soft exoskeleton).

right By derivation, we get:/> , at this time/> Substitute and you will get/> .

The output power of the actuator (ie, soft exoskeleton) can be expressed as the product of torque and angular velocity. Therefore, the torque can be expressed as: ,/> Represents torque.

Based on the above principle, let the system state variable be: ;/> ;/> ;/> . Among them,/> Represents the input gas energy,/> Represents the potential energy stored by the actuator, that is, accumulated potential energy,/> Represents the output energy. Let the system input variable be:/> . The system output variable is:/> ;/> .

and ;/> ;/> ;/> , therefore, the dynamic equation can be obtained: .

Further, the static torque of the actuator is calculated according to the principle of virtual work. When the angular velocity When equal to zero, use the principle of virtual work to calculate torque/> , the static torque is used as the dynamic torque/> The initial value of , and considering that the gas temperature of the soft exoskeleton does not change much, the gas can be regarded as an isothermal expansion process, that is, the conversion efficiency is 100%, to simplify the actuator model.

Will ,/> ,/> and Substituting into the above dynamic equation, we can get/> ,Right now .

further, It can be expressed as:/> , ,/> is a nonlinear function, p is the measured value of the gas pressure sensor,/> is the system input.

Furthermore, considering that the gas pressure in the inner cavity of the actuator determines the bending stiffness of the software, and the process of establishing gas pressure, the gas flow rate in the conduit, and the elastic strain process lead to a dynamic response time lag in the system, when introducing dynamic response into the system output lag, at this time, the output of the nonlinear system is: ,/> , , that is,/> ,/> Represents dynamic response delay.

Therefore, the dynamic model can be expressed as: , where,/> .

For step S25, in order to simplify the design and state estimation of the human-machine coupling controller when using the subsequent dynamic model, any one of Jacobian matrix linearization, Taylor expansion and nonlinear dynamic inverse control is included.

Jacobian matrix linearization, that is, the derivative function Jacobian matrix linearization method, uses the state transition Jacobian matrix derived from the nonlinear state space model of the actuator system (i.e., the dynamic model of the actuator), the input Jacobian matrix , output the Jacobian matrix, which is used to design linear controllers, and design Kalman speed and torque estimators, you can get a soft exoskeleton controller for speed-free and torque-free.

A linear controller can also be designed by studying the Taylor expansion linearized state space model of the actuator system's nonlinear state space model (i.e., the dynamic model of the actuator).

By studying the nonlinear dynamic inverse control method of the dynamic model of the soft exoskeleton actuator and the nonlinear system feedback linearization control algorithm, the nonlinear equation of the soft exoskeleton actuator is converted into an easy-to-control linear form, and the linear form can also be obtained Activate the controller.

Furthermore, in order to improve the effectiveness of the dynamic model and the accuracy of the linearization algorithm, the prepared exoskeleton can be installed on the test bench of the artificial bionic upper limb model with biomechanical properties to conduct gas-driven exoskeleton experiments. Use an air mass flow meter to record the gas mass flow output by the proportional valve, a pressure sensor to record the gas source pressure of the proportional valve and the pressure in the actuator cavity, a force/torque meter to record the output torque of the actuator, and an angle sensor to record the bending angle. After obtaining various data, conduct experimental verification, analyze the error of the dynamic model, and optimize the structure and parameters of the dynamic model (that is, the controller model obtained after linearization) based on the error.

Experimental verification can be divided into three methods: direct drive experiment, human-machine coupling control experiment, and angle-free and torque-less sensor control experiment.

When using direct drive experiments, multiple excitation voltage signals are designed for the dynamic model to drive the movement of the exoskeleton, and the experimental data are recorded. The experimental data are compared with the computer simulation results of the dynamic model, and the errors of the dynamic model are obtained and analyzed. To optimize the kinetic model structure and parameters until an exact matching (i.e. optimized) kinetic model is obtained.

When using human-machine coupling control experiments, an artificial bionic upper limb model is used to measure and analyze the errors of the linearization control algorithm, which is used to optimize the linearization algorithm of the model.

Using angle- and torque-free sensor control experiments, the accuracy of the linearized estimator of the dynamic model can be verified and optimized.

In order to further verify the versatility and effectiveness of the dynamic model and optimize the dynamic model, driving experiments can be conducted using silicone soft exoskeleton actuators with different constraint structures and soft exoskeleton actuators made by folding high-elastic fabrics.

In the above steps S21 to S25 and its further implementation, by assuming the system state variable as x=[E _in ,E _p ,E _out , ], system input u = G , system output/> , a nonlinear state space model H2, the dynamic model, of a soft exoskeleton actuator was established. In this dynamic model, the gas pressure in the actuator cavity will determine the stiffness required for bending of the soft actuator, and the dynamic response time of the system caused by the process of gas pressure, gas flow rate in the conduit, and elastic strain process will be introduced. Hysteresis, using power to calculate the dynamic torque of the assist, and the principle of virtual work to calculate the static torque, so that it is not restricted by the inertial mass, resistance or traction force of the human body. At the same time, the linearization algorithm is used to process the dynamic model of the actuator, and the errors and inaccuracies caused by linearization can be studied and analyzed to provide accurate pneumatic soft exoskeleton actuators for controller and state estimator design. The linearization algorithm of the model, that is, the linearized brake controller, improves the controllability and observability of the dynamic model and ensures that the feedback information required for control can be measured or observed, thereby helping to improve software exoskeleton control. of robustness.

Furthermore, the soft exoskeleton control method provided by the embodiment of the present invention may also include the step of constructing a gas flow model of the electromagnetic proportional valve, considering that the gas flow of the electromagnetic proportional valve is approximately a one-dimensional isentropic flow of ideal gas through the shrinking nozzle. , that is, the gas mass flow rate flowing through the electromagnetic proportional valve is determined by factors such as the effective opening area of the valve and the pressure ratio of the inlet and outlet ports. Therefore, the Sanville gas mass flow formula and the ideal gas state equation can be introduced to establish the nonlinearity of the electromagnetic proportional valve. Model. In a possible implementation, referring to Figure 5, it can be implemented through the following steps.

S31. Based on the experimental input voltage and experimental gas flow of the electromagnetic proportional valve, fit the voltage flow equation of the electromagnetic proportional valve.

In this embodiment, the voltage-flow equation represents the relationship between the input voltage, the flow coefficient, and the effective opening area of the electromagnetic proportional valve 21 .

S32, based on the Sanville flow formula and voltage flow equation, obtain the gas flow model of the electromagnetic proportional valve.

Multiple sets of experimental input voltages and experimental gas flow rates of the electromagnetic proportional valve can be obtained through flow meters and sensors, and then the least squares method, stepwise regression method, polynomial fitting, logarithmic fitting, and gamma adjustment can be used to analyze the electromagnetic proportional valve. Curve fitting is performed on multiple sets of experimental input voltages and experimental gas flows to obtain the voltage flow equation. In this embodiment, the voltage flow equation can be expressed as .

The gas flow rate of the solenoid proportional valve follows the Sanville flow formula. Therefore, the gas flow rate of the solenoid proportional valve can be expressed using the Sanville flow formula.

The Sanville flow formula is: , where,/> Characterizes the gas flow, C represents the flow coefficient of the proportional valve (measuring the flow capacity of the valve), A represents the effective opening area of the proportional valve,/> Characterizing the molar gas constant,/> Characterizing temperature,/> Indicates the gas source pressure,/> Characterizes the gas pressure in the chamber,/> Characterizing the driving voltage,/> as a calculation method.

By substituting the voltage flow equation into the Sanville flow formula, the gas flow model of the electromagnetic proportional valve can be obtained. The gas flow model can be expressed as: , at this time,/> Represents the driving voltage of the electromagnetic proportional valve. In this model, the input variable is the proportional valve driving voltage u, the output variable is the gas flow rate G , and the state variables are the gas source pressure and the inner cavity gas pressure.

Through the above steps S31 and S32, a highly accurate gas flow model of the electromagnetic proportional valve can be obtained.

In the above-mentioned soft exoskeleton control method, it is proposed to use the physical principles of the ideal gas state equation, gas energy conversion, and bending work of the soft actuator to establish a dynamic model of the actuator of the soft exoskeleton. This dynamic model is not subject to human influence. The constraints of limb inertial mass, resistance or traction are applicable to pneumatic soft exoskeleton actuators of different structures. Based on this dynamic model and combined with the gas flow model of the electromagnetic proportional valve, the driving voltage of the electromagnetic proportional valve is feedback adjusted. This can greatly improve the stability and robustness of software exoskeleton control.

Based on the same concept as the above-mentioned software exoskeleton control method, in a possible implementation, a software exoskeleton control device 60 is also provided, which can be applied to the software exoskeleton control system 100 in FIG. 1 . Referring to FIG. 6 , the software exoskeleton control device 60 may include a flow estimation module 601 , a motion estimation module 602 , an estimation calculation module 603 , a feedback calculation module 604 and a control module 605 .

The flow estimation module 601 is used to calculate the gas flow estimate of the electromagnetic proportional valve based on the driving voltage of the electromagnetic proportional valve at the current moment and using the gas flow model of the electromagnetic proportional valve.

The motion estimation module 602 is used to calculate the motion estimate using the dynamic model of the actuator and the gas flow estimate. The motion estimate includes at least one of a torque estimate and a bending angle estimate.

The estimation calculation module 603 is configured to generate a flow control command based on the control command at the current moment, the motion estimate, and the gas flow estimate.

The feedback calculation module 604 is used to extract the target flow rate from the received flow control command, and calculate the target drive voltage of the electromagnetic proportional valve according to the gas flow estimate and the target flow rate.

The control module 605 is used to control the electromagnetic proportional valve to operate according to the target driving voltage, so that the actuator can obtain the torque and bending angle of the soft exoskeleton based on the target flow rate and the dynamic model of the actuator.

Among them, the dynamic model of the actuator is based on the product of the torque and angular velocity of the soft exoskeleton as the output power, and is constructed based on the output power.

Furthermore, a model building module may also be included, and the model building model is used to implement the above steps S21-S25 and steps S31-S32.

In the above-mentioned software exoskeleton control device 60, through the synergy of the flow estimation module 601, the motion estimation module 602, the estimation calculation module 603, the feedback calculation module 604 and the control module 605, based on the driving voltage of the electromagnetic proportional valve at the current moment, the electromagnetic The gas flow model of the proportional valve calculates the gas flow estimate of the electromagnetic proportional valve, and uses the dynamic model of the actuator and the gas flow estimate to calculate the motion estimate, so as to calculate the motion estimate based on the control command and motion estimate at the current moment. and the gas flow estimate to generate a flow control command, extract the target flow rate from the flow control command, calculate the target drive voltage of the electromagnetic proportional valve based on the gas flow estimate and the target flow rate, thereby controlling the electromagnetic proportional valve to work according to the target drive voltage, so that Based on the target flow rate and the dynamic model of the actuator, the actuator of the soft exoskeleton obtains the torque and bending angle of the soft exoskeleton to work, and the dynamic model of the actuator of the soft exoskeleton is based on the soft exoskeleton. The product of torque and angular velocity is used as the output power and is constructed based on the output power, so that the dynamic model of the soft exoskeleton actuator is not restricted by the inertial mass of the human limb and the exoskeleton, thereby greatly improving the control of the soft exoskeleton stability and robustness.

For specific limitations on the software exoskeleton control device 60, please refer to the above limitations on the control method of the software exoskeleton 10, which will not be described again here. Each module in the above-mentioned software exoskeleton control device 60 can be implemented in whole or in part by software, hardware and combinations thereof. Each of the above modules may be embedded in or independent of the processor in the electronic device 70 in the form of hardware, or may be stored in the memory of the electronic device 70 in the form of software, so that the processor can call and execute operations corresponding to the above modules.

In one embodiment, an electronic device 70 is provided. The electronic device 70 may be a terminal, and its internal structure diagram may be as shown in FIG. 7 . The electronic device 70 includes a processor, memory, communication interface and input device connected through a system bus. Wherein, the processor of the electronic device 70 is used to provide computing and control capabilities. The memory of the electronic device 70 includes non-volatile storage media and internal memory. The non-volatile storage medium stores operating systems and computer programs. This internal memory provides an environment for the execution of operating systems and computer programs in non-volatile storage media. The communication interface of the electronic device 70 is used for wired or wireless communication with external terminals. The wireless mode can be implemented through WIFI, operator network, near field communication (NFC) or other technologies. When the computer program is executed by the processor, the control method of the software exoskeleton 10 provided in the above embodiments is implemented.

The structure shown in Figure 7 is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the electronic device 70 to which the solution of the present invention is applied. The specific electronic device 70 may include other components than those in Figure 7 More or fewer parts are shown, or certain parts are combined, or have different parts arrangements.

In one embodiment, the software exoskeleton control device 60 provided by the present invention can be implemented in the form of a computer program, and the computer program can run on the electronic device 70 as shown in FIG. 7 . The memory of the electronic device 70 can store various program modules that make up the software exoskeleton control device 60, such as the flow estimation module 601, the motion estimation module 602, the estimation calculation module 603, the feedback calculation module 604 and the control module shown in Figure 6 605. The computer program composed of each program module causes the processor to execute the steps in the method for controlling the software exoskeleton 10 described in this specification.

For example, the electronic device 70 shown in FIG. 7 may perform step S11 through the flow estimation module 601 in the software exoskeleton control device 60 shown in FIG. 6 . The electronic device 70 may perform step S13 through the motion estimation module 602. The electronic device 70 may perform step S15 through the estimation calculation module 603 . The electronic device 70 may perform step S17 through the feedback calculation module 604. The electronic device 70 may perform step S19 through the control module 605.

In one embodiment, an electronic device 70 is provided, including a memory and a processor. The memory stores machine-executable instructions. When the processor executes the machine-executable instructions, it implements the following steps: based on the electromagnetic proportional valve at the current moment The driving voltage of the electromagnetic proportional valve is used to calculate the gas flow estimate of the electromagnetic proportional valve; the motion estimate is calculated using the dynamic model of the actuator and the gas flow estimate; according to the control command at the current moment , motion estimate and gas flow estimate, generate a flow control command; extract the target flow rate from the flow control command, and calculate the target drive voltage of the electromagnetic proportional valve according to the gas flow estimate and standard flow rate; control the electromagnetic proportional valve according to the target The driving voltage is operated so that the actuator obtains the torque and bending angle of the soft exoskeleton based on the target flow rate and the dynamic model of the actuator.

In one embodiment, a storage medium is provided with a computer program stored thereon. When the computer program is executed by a processor, the following steps are implemented: based on the driving voltage of the electromagnetic proportional valve at the current moment, utilizing the gas flow rate of the electromagnetic proportional valve. model to calculate the gas flow estimate of the electromagnetic proportional valve; use the actuator's dynamic model and gas flow estimate to calculate the motion estimate; based on the control command, motion estimate and gas flow estimate at the current moment, generate Flow control command; extract the target flow rate from the received flow control command, calculate the target drive voltage of the electromagnetic proportional valve based on the gas flow estimate and the standard flow rate; control the electromagnetic proportional valve to operate according to the target drive voltage to make the actuator Based on the target flow rate and the dynamic model of the actuator, the torque and bending angle of the soft exoskeleton are obtained.

In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device implementations described above are only illustrative. For example, the flowcharts and block diagrams in the accompanying drawings show the possible implementation architecture, functions and functions of the devices, methods and computer program products according to various embodiments of the present invention. operate. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code that contains one or more components for implementing the specified logical function(s). Executable instructions. It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two consecutive blocks may actually execute substantially in parallel, or they may sometimes execute in the reverse order, depending on the functionality involved. It will also be noted that each block of the block diagram and/or flowchart illustration, and combinations of blocks in the block diagram and/or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts. , or can be implemented using a combination of specialized hardware and computer instructions.

In addition, each functional module in each embodiment of the present invention can be integrated together to form an independent part, each module can exist alone, or two or more modules can be integrated to form an independent part.

If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention essentially or the part that contributes to the existing technology or the part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in various embodiments of the present invention. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .

The above descriptions are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims (10)

1. A soft exoskeleton control method, characterized in that, applied to the control system of the soft exoskeleton, the soft exoskeleton further includes an electromagnetic proportional valve and an actuator, and the method includes: Based on the driving voltage of the electromagnetic proportional valve at the current moment, use the gas flow model of the electromagnetic proportional valve to calculate the gas flow estimate of the electromagnetic proportional valve; Using the dynamic model of the actuator and the gas flow estimate, a motion estimate is calculated; wherein the motion estimate includes at least one of a torque estimate and a bending angle estimate; Generate a flow control command according to the control command at the current moment, the motion estimate and the gas flow estimate; Extract the target flow rate from the flow control command, calculate the target drive voltage of the electromagnetic proportional valve according to the gas flow estimate and the target flow rate; control the electromagnetic proportional valve to operate according to the target drive voltage , so that the actuator can obtain the torque or bending angle of the soft exoskeleton based on the target flow rate and the dynamic model of the actuator; Wherein, the dynamic model of the actuator is constructed based on the product of the torque and the angular velocity of the soft exoskeleton as the output power. 2. The soft exoskeleton control method according to claim 1, wherein the actuator further includes an actuation controller, and the method further includes the step of constructing the actuation controller, which step includes: The product of the torque and the angular velocity of the soft exoskeleton is used as the output power of the soft exoskeleton; Construct a dynamic model of the actuator based on the output power and the gas flow output by the gas flow model; The dynamic model is processed using a linearization algorithm to obtain an actuation controller. 3. The software exoskeleton control method according to claim 2, characterized in that the step of constructing a dynamic model of the actuator based on the output power and the gas flow output by the gas flow model, include: Use the ideal gas equation of state to express the gas volume in the actuator chamber; Use the product of the gas pressure and the gas flow rate input by the solenoid proportional valve to express the input gas power; Express the accumulated potential energy of the actuator based on the product of the gas volume and the gas pressure; The input gas energy is expressed according to the input gas power, the output energy is expressed according to the output power, and the bending angle of the soft exoskeleton is expressed according to the angular velocity; The gas flow output by the gas flow model is used as system input, the input gas energy, the accumulated potential energy, the output energy and the bending angle are used as system state variables, and the output power is used to calculate the dynamic assist torque, The static torque is calculated based on the principle of virtual work, and the nonlinear state space model of the actuator is constructed. 4. The software exoskeleton control method according to any one of claims 1 to 3, characterized in that the dynamic model includes: in, ,/> ,/> Characterizing gas flow, Characterizing gas pressure,/> Characterizing the molar gas constant,/> Characterizing temperature,/> Characterizes the molar mass of air,/> Characterize gas quality,/> Characterizes dynamic response time delay,/> Characterizes the height of the exoskeleton cavity,/> Characterizes the cross-sectional area of the exoskeleton lumen,/> Characterizes conversion efficiency,/> Characterizes the input energy,/> Represents accumulated potential energy,/> Characterizes the output energy,/> Characterizes the bending angle of the soft exoskeleton. 5. The software exoskeleton control method according to any one of claims 1 to 3, characterized in that the motion estimate is calculated using the dynamic model of the actuator and the gas flow estimate. The steps include: When the control mode is the dynamic assist mode, the torque estimate of the actuator is obtained by using the dynamic model of the actuator based on the gas flow estimate and the gas pressure and temperature of the inner cavity of the soft exoskeleton. quantity; When the control mode is the limb passive mode, based on the gas flow estimate and the gas pressure and temperature of the soft exoskeleton cavity, the bending angle of the actuator is obtained by using the dynamic model of the actuator estimator. 6. The software exoskeleton control method according to any one of claims 1 to 3, characterized in that the method further includes the step of constructing a gas flow model of the electromagnetic proportional valve, which step includes: Based on the experimental input voltage and experimental gas flow of the electromagnetic proportional valve, the voltage flow equation of the electromagnetic proportional valve is fitted; wherein the voltage flow equation represents the input voltage, flow coefficient and effective opening of the electromagnetic proportional valve area relationship; Based on the Sanville flow formula and the voltage flow equation, a gas flow model of the electromagnetic proportional valve is obtained. 7. The software exoskeleton control method according to claim 6, characterized in that the step of fitting the voltage flow equation of the electromagnetic proportional valve based on the experimental input voltage and experimental gas flow of the electromagnetic proportional valve ,include: Using the least squares method, curve fitting was performed on the experimental input voltage and experimental gas flow rate of the electromagnetic proportional valve to obtain the voltage flow equation. 8. The software exoskeleton control method according to claim 6, wherein the gas flow model includes: in, Characterizing gas flow,/> Characterizing the voltage-flow equation,/> Characterizing the molar gas constant,/> Characterizing temperature,/> Indicates the gas source pressure,/> Characterizes the gas pressure in the chamber,/> Characterizes the driving voltage. 9. The software exoskeleton control method according to claim 2, wherein the linearization algorithm includes any one of Jacobian matrix linearization, Taylor expansion and nonlinear dynamic inverse control. 10. A soft exoskeleton control device, characterized in that, applied to the control system of the soft exoskeleton, the soft exoskeleton further includes an electromagnetic proportional valve and an actuator, and the soft exoskeleton control device includes a flow estimation module, motion estimation module, estimation calculation module, feedback calculation module and control module; The flow estimation module is used to calculate the gas flow estimate of the electromagnetic proportional valve based on the driving voltage of the electromagnetic proportional valve at the current moment and using the gas flow model of the electromagnetic proportional valve; The motion estimation module is used to calculate a motion estimate using the dynamic model of the actuator and the gas flow estimate; wherein the motion estimate includes a torque estimate and a bending angle estimate. at least one; The estimation calculation module is used to generate a flow control command according to the control command at the current moment, the motion estimate and the gas flow estimate; The feedback calculation module is used to extract the target flow rate from the flow control command, and calculate the target drive voltage of the electromagnetic proportional valve according to the gas flow estimate and the target flow rate; The control module is used to control the electromagnetic proportional valve to operate according to the target driving voltage, so that the actuator obtains the soft exoskeleton based on the target flow rate and the dynamic model of the actuator. torque or bending angle; Wherein, the dynamic model of the actuator is constructed based on the product of the torque and the angular velocity of the soft exoskeleton as the output power.