Active compliance control method for valve-controlled hydraulic mechanical leg based on active disturbance rejection control

Abstract

The invention discloses an active-disturbance-rejection-control-based active compliance control method for a valve-controlled hydraulic mechanical leg, which combines an active-disturbance-rejection controller of an inner ring and an admittance controller of an outer ring, and designs an active compliance controller considering both position control precision and compliance. For nonlinearity and uncertainty of the valve-controlled hydraulic mechanical leg, an expansion state observer is designed, a dynamic coupling item of the valve-controlled hydraulic mechanical leg is estimated, then compensation is performed in a feedforward mode in active-disturbance-rejection control, and better tracking performance is obtained. The external force is calculated according to the mathematical model of the valve-controlled hydraulic mechanical leg and the collected signal of the force sensor, and the expected trajectory of the active disturbance rejection control is compensated through the admittance controller, so that the active compliant control of the valve-controlled hydraulic mechanical leg is realized.

Description

Technical Field

[0001] The invention belongs to the technical field of servo control of hydraulic mechanical legs, and in particular relates to an active compliance control method of valve-controlled hydraulic mechanical legs based on anti-disturbance control. Background Art

[0002] Compared with wheeled robots and tracked robots, legged robots have stronger adaptability to unknown environments. Compared with bipedal robots, quadrupedal robots have better stability and simpler structures than hexapod robots. The commonly used driving methods for the leg joints of quadrupedal robots are motor drive, hydraulic drive and pneumatic drive. Hydraulic drive is widely used in quadrupedal robots due to its strong load-bearing capacity and explosive power. Nowadays, hydraulic drive units are widely used in hydraulic quadrupedal robots. The hydraulic drive unit integrates the displacement sensor, force sensor and servo valve on the hydraulic cylinder, which simplifies the structure and improves the control effect.

[0003] The control strategy of the robot is the most critical factor affecting the robot's adaptability. Robots replace humans to perform dangerous tasks such as disaster rescue and military operations. Different loads and road conditions bring different impact forces to the foot end, causing the robot's movement to become unstable and the structure to be damaged. Therefore, a simple position control strategy cannot meet the robot's control needs, while active compliance control can make the robot's movement compliant, thereby reducing external impact. Force-position hybrid control, impedance control, and admittance control are commonly used active compliance control methods. Admittance control takes into account the dynamic relationship between the force and position of the controlled system, and involves force control and position control at the same time. It can ensure the accuracy of the robot's movement while avoiding excessive contact force. In view of the advantages of admittance control, it has been applied in many fields.

[0004] Position-based admittance control has higher control accuracy, and the most critical factor affecting its control accuracy is the position control strategy. Since hydraulic quadruped robots have strong nonlinearity and uncertainty, traditional PID control cannot meet the high dynamic control requirements of mechanical legs. The self-disturbance rejection control proposed by Han Jingqing uses an extended state observer to observe the unmodeled disturbances of the system. It has good dynamic tracking performance for systems with high uncertainty and high nonlinearity, and has the advantages of simple structure, convenient parameter adjustment, and no need for precise modeling. The bandwidth method proposed by Gao Zhiqiang can be used to determine the gain parameters of the extended state observer, which enhances the practicality of self-disturbance rejection control. Summary of the invention

[0005] The purpose of the present invention is to provide a method for active compliance control of a valve-controlled hydraulic mechanical leg based on active anti-disturbance control, which can not only ensure the control accuracy of the mechanical leg, but also ensure that the mechanical leg has a compliant response to external forces.

[0006] The technical solution to achieve the purpose of the present invention is: a valve-controlled hydraulic mechanical leg active compliance control method based on active disturbance rejection control, comprising the following steps:

[0007] Step 1: Establish a mathematical model of the valve-controlled hydraulic mechanical leg system and proceed to step 2.

[0008] Step 2: Based on the mathematical model of the valve-controlled hydraulic mechanical leg system, design an active compliance controller for the valve-controlled hydraulic mechanical leg based on anti-disturbance control, and then proceed to step 3.

[0009] Step 3: Use Lyapunov stability theory to prove the stability of the active compliant controller of the valve-controlled hydraulic mechanical leg based on anti-disturbance control, and obtain the result that both the system observation error and tracking error are bounded.

[0010] Compared with the prior art, the present invention has the following significant advantages: (1) it uses position control as the inner loop and admittance control as the outer loop, which ensures the control accuracy of the robotic legs while enabling the robotic legs to respond flexibly when subjected to external forces; (2) the inner loop of the position control adopts anti-disturbance control, which has a simple structure, does not require precise modeling, is easy to implement, and has good control performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 This is a model diagram of the valve-controlled hydraulic mechanical leg of the present invention.

[0012] Figure 2 This is a schematic diagram of the structure of the valve-controlled hydraulic mechanical leg of the present invention. The shaded part in the right figure represents the same component.

[0013] Figure 3 It is an interaction diagram between the hydraulic mechanical leg of the present invention and the external force when in motion.

[0014] Figure 4 It is the expected trajectory of the three joints when the mechanical leg and foot are subjected to external force under the controller of the present invention.

[0015] Figure 5 It's the mechanical legs. Figure 4 Comparison of tracking errors of three joints under the trajectory of ADRC and PI control.

[0016] Figure 6 It is the control input of three joints under the action of the valve-controlled hydraulic mechanical leg active compliance controller based on self-disturbance rejection control designed by the present invention. DETAILED DESCRIPTION

[0017] Combination Figure 1 and Figure 2 The present invention provides a valve-controlled hydraulic mechanical leg active compliance control method based on active disturbance rejection control, comprising the following steps:

[0018] Step 1: Establish a mathematical model of the hydraulic mechanical leg system, which is formed by coupling the mechanical system and the hydraulic system, as follows:

[0019] According to the Lagrangian method, the mechanical dynamics model of the valve-controlled hydraulic mechanical leg joint space is described as:

[0020]

[0021] Among them, θ is the joint angle of the mechanical leg, is the angular velocity of the mechanical leg joint, is the angular acceleration of the mechanical leg joint, τ is the control torque of the mechanical leg joint, M, C, G are the inertia matrix, centrifugal force and Coriolis force matrix and gravity moment vector in the mechanical leg joint space respectively.

[0022] Define the actuator extension displacement as x p , then the angular velocity of the joint of the mechanical leg and actuator extension speed There are the following relationships:

[0023]

[0024] Where J is the Jacobian matrix.

[0025] Then, from equations (1) and (2), we can get the mechanical dynamics model expression of the mechanical leg actuator space:

[0026]

[0027] Where F is the output of the actuator, M p , C p , G p are the inertia matrix, centrifugal force and Coriolis force matrix and gravity moment vector in the mechanical leg actuator space, respectively. is the actuator extension acceleration.

[0028] The output of the actuator has the following relationship with the pressure in the two chambers of the actuator and the effective area:

[0029]

[0030] Among them, P1 is the oil inlet chamber pressure, P2 is the oil discharge chamber pressure, A1 is the oil inlet chamber effective area, and A2 is the oil discharge chamber effective area.

[0031] The pressure dynamic equation of the actuator is:

[0032]

[0033] Among them, β e is the elastic modulus of hydraulic oil, V 01is the initial volume of the oil inlet chamber, V 02 is the initial volume of the oil discharge chamber, C t is the internal leakage coefficient, Q1 is the oil inlet flow rate, Q2 is the oil discharge flow rate, is the first-order derivative of P1 with respect to time, is the first-order derivative of P2 with respect to time.

[0034] Since the bandwidth of the servo valve is much higher than the system bandwidth, the spool displacement of the servo valve is considered to be proportional to the control input, and the hydraulic oil flow of the two chambers is expressed as:

[0035]

[0036] Among them, k t is the control input gain, u is the control input, P s is the system oil supply pressure, P r is the system return oil pressure, sign is the sign function, expressed as:

[0037]

[0038] Define the system's state variables x = [x1, x2, x3, x4] T , where the state variable x1 = [x p1 ,x p2 ,x p3 ] T , state variables State variable x3 = [P 11 ,P 21 ,P 31 ] T , state variable x4=[P 12 ,P 22 ,P 32 ] T , x p1 、x p2 、x p3 are the extension displacements of the three actuators, are the extension speeds of the three actuators, P 11 , P 21 , P 31 , are the oil inlet chamber pressures of the three actuators, P 12 , P 22 , P 32 , are the oil discharge chamber pressures of the three actuators respectively, and T represents transposition.

[0039] The nonlinear model of the system is written in the following state space form:

[0040]

[0041] in, represents the first-order derivative of x1 with respect to time, represents the first-order derivative of x2 with respect to time, represents the first-order derivative of x3 with respect to time, represents the first-order derivative of x4 with respect to time, Indicates M P The inverse of , y represents the intermediate variable.

[0042] Step 2: Based on the mathematical model of the valve-controlled hydraulic mechanical leg system, an active compliance controller of the valve-controlled hydraulic mechanical leg based on active disturbance rejection control is designed, as follows:

[0043] Step 2-1: Design a valve-controlled hydraulic mechanical leg anti-disturbance controller to achieve joint position control of the mechanical leg. The specific steps are as follows:

[0044] Let variable z1 = x1, variable z2 = x2, variable Then formula (8) can be rewritten as:

[0045]

[0046] in, is the first-order derivative of z1 with respect to time, is the first-order derivative of z2 with respect to time, is the first-order derivative of z3 with respect to time, f represents the intermediate variable, and E represents the intermediate variable.

[0047]

[0048] for The first derivative with respect to time, C p The first derivative with respect to time, G p The first derivative with respect to time.

[0049] Expand f to a new system state z4, and the first-order derivative of z4 with respect to time If w(t) is a time-dependent function, the system is expanded into a new linear control system:

[0050]

[0051] The corresponding extended state observer is designed as:

[0052]

[0053] Among them, β 01 , β 02 , β 03 , β 04Both are observer gains, and the tracking error ε1=x1-x 1d , x 1d is the expected trajectory, η1 is the estimation error, is the estimate of z1, is the estimate of z2, is the estimate of z3, is the estimate of z4, for The first derivative with respect to time, for The first derivative with respect to time, for The first derivative with respect to time, for For the first-order derivative with respect to time, the observer gain is chosen so that its characteristic equation s 4 +β 01 s 3 +β 02 s 2 +β 03 s+β 04 Satisfies the Hurwitz condition, s represents a complex variable; using the bandwidth method, the observer gain is β 01 =4ω0, Adjusting the observer gain ω0 realizes the selection of observer parameters.

[0054] The tracking error is defined as:

[0055]

[0056] Among them, ε1, ε2, ε3 are tracking errors, y d is the desired trajectory, the virtual control law Virtual Control Law Yes d The first derivative with respect to time, is the first-order derivative of α1 with respect to time, K1,K2∈R 3×3 , is the controller gain, R 3×3 Represents a 3×3 real matrix.

[0057] According to the estimated value of the state of the robotic leg system by the extended state observer, high-precision control of the robotic leg system is achieved. Therefore, after replacing the system state with its corresponding estimated value, the tracking error of the system is defined as:

[0058]

[0059] in, are the tracking errors under the estimated value of the system state, and the estimation of the virtual control law Estimation of virtual control law yes Estimates.

[0060] The design control input u is:

[0061]

[0062] Among them, E -1 is the inverse of E, K3∈R 3×3 , is the controller gain.

[0063] Step 2-2: Design an admittance controller to compensate for the desired trajectory of the ADRC and realize active compliance control of the robotic leg. The specific steps are as follows:

[0064] Admittance control takes the desired trajectory as input and calculates the output F of the hydraulic actuator based on the mechanical dynamics model. When the mechanical leg is subjected to external force, the actuator output F and the force feedback F of the force sensor are calculated. b The difference between the two is the external force F on the mechanical leg. L :

[0065] F L =FF b (17)

[0066] The desired trajectory of the three joints of the robotic leg is compensated by admittance control, so that it exhibits the characteristics of a mass-spring damping system; then the robotic leg is subjected to an external force F L Under the action of z Described as:

[0067]

[0068] in, Represents y z The first derivative with respect to time, Represents y z The second derivative with respect to time;

[0069] Since the range of motion of the joints of the robotic leg is limited, a saturation function is introduced to make the final desired trajectory within the range of motion space of the robotic leg. The final desired trajectory is:

[0070]

[0071] Among them, the constant γ=(1-β)ω c , constant β satisfies 0<β<1, constant ω c is the limit rotation angle of the joint, the expected trajectory after compensation

[0072] Finally, through the anti-disturbance controller, the robot leg follows the final desired trajectory y with high accuracy. dz Movement,The controller finally designed is divided into two parts, the inner and outer parts. The inner loop's anti-disturbance control makes the robotic leg move along the desired trajectory, and the outer loop's admittance control is used to compensate for the desired trajectory of the robotic leg, so that the robotic leg shows compliance when subjected to external forces.

[0073] Step 3: Use Lyapunov stability theory to prove the stability of the active compliant controller of the valve-controlled hydraulic mechanical leg based on the anti-disturbance control, and obtain the results that the system observation error and tracking error are bounded, as follows:

[0074] The Lyapunov function is defined as follows:

[0075]

[0076] Among them, V1, V2, and V3 are all Lyapunov functions. The derivatives of V1, V2, and V3 are obtained respectively:

[0077]

[0078] in, is the first-order derivative of V1 with respect to time, is the first-order derivative of V2 with respect to time, is the first-order derivative of V3 with respect to time, and the controller (16) is substituted into get:

[0079]

[0080] also,

[0081]

[0082] Where I is a unit vector, σ1, σ2, σ3, and σ4 are all constants. Substituting equation (22) into equation (21), we get:

[0083]

[0084] Where H is a constant:

[0085]

[0086] Among them, max is the maximum value function, according to Young's inequality we can get:

[0087]

[0088] Among them, the intermediate variable ψ=2min{λ min {K1},λ min {K2},λ min{K3-I / 2}}, λ min {K1} is the smallest eigenvalue of K1, λ min {K2} is the smallest eigenvalue of K2, λ min {K3-I / 2} is the minimum eigenvalue of K3-I / 2, min is the minimum function, ζ=1 / 2H 2 is a positive definite constant, so V3 is bounded:

[0089] V3≤V3(0)e -ψt +ζ / ψ (26)

[0090] Where e is a natural constant, t is time, and the tracking error ε = [ε1ε2ε3] T The convergence region of is defined as:

[0091]

[0092] Where Λ represents the convergence region of the tracking error.

[0093] Example

[0094] In order to verify the performance of the designed controller, the physical parameters of the valve-controlled hydraulic mechanical leg in the simulation are shown in Table 1.

[0095] Table 1 System physical parameters

[0096] Physical parameters Numeric Physical parameters Numeric <![CDATA[k t1 (m 3 / s / V / Pa 1 / 2 )]]> <![CDATA[1.5×10 -8 ]]> <![CDATA[P s (Well)]]> <![CDATA[21×10 6 ]]> <![CDATA[k t2 (m 3 / s / V / Pa 1 / 2 )]]> <![CDATA[1.5×10 -8 ]]> <![CDATA[P r (Well)]]> 0 <![CDATA[k t3 (m 3 / s / V / Pa 1 / 2 )]]> <![CDATA[1.4×10 -8 ]]> <![CDATA[C t (m 5 / (N·s))]]> <![CDATA[2.38×10 -13 ]]> <![CDATA[β e (Well)]]> <![CDATA[8×10 8 ]]>

[0097] Where, the control input gain k t =[k t1 ,k t2 ,k t3 ],k t1 ,k t2 ,k t3 are the control input gains of each joint of the robotic leg. To verify the effect of compliant control, external force is applied to the foot of the robotic leg. The lateral hip joint mainly resists lateral impact, while the longitudinal hip joint and knee joint mainly resist vertical impact. Therefore, vertical and horizontal forces are applied to the foot, such as Figure 3 As shown. The law of change of force with time t is designed as:

[0098]

[0099] Among them, F1 represents the vertical force, F2 represents the horizontal force, t represents time, and N represents the unit of force, Newton.

[0100] Depend on Figure 4 It can be seen that under admittance control, when the foot of the robotic leg is subjected to external force, the joint angle of the robotic leg shows a compliant response. The inner loop's anti-disturbance control controls the robotic leg to track the desired trajectory.

[0101] In order to verify the effect of the active disturbance rejection control in the present invention, the position controller of the inner loop compares the active disturbance rejection control with the traditional PI control. The parameters of the PI control are set as K P =[2000,3,4],K I =[5000,50,40]. In the ADRC, the observer gain of the three joints is ω0 = [180,200,220], and the controller gain of the ADRC is K1 = [800,900,800], K2 = [600,800,700], K3 = [300,600,600].

[0102] Depend on Figure 5 It is found that under the same expected trajectory, the maximum error and steady-state error of the three joints' ADRC are smaller than those of the traditional PI control, and it has higher control accuracy. Figure 6 It is the control input of the controller of the present invention. As can be seen from the figure, the control input is a continuous signal and can be operated in an actual controller.

Claims

1. A method for active compliance control of valve-controlled hydraulic mechanical legs based on active disturbance rejection control, characterized in that: The following steps are involved: Step 1: Establish a mathematical model of the valve-controlled hydraulic mechanical leg system, and then proceed to step 2; Step 2: Based on the mathematical model of the valve-controlled hydraulic mechanical leg system, design an active compliance controller of the valve-controlled hydraulic mechanical leg based on active disturbance rejection control, and then proceed to step 3; Step 3: Use Lyapunov stability theory to prove the stability of the active compliant controller of the valve-controlled hydraulic mechanical leg based on anti-disturbance control, and obtain the result that both the system observation error and tracking error are bounded.

2. The active compliance control method of valve-controlled hydraulic mechanical legs based on active disturbance rejection control according to claim 1 is characterized in that: In step 1, a mathematical model of the valve-controlled hydraulic mechanical leg system is established as follows: According to the Lagrangian method, the mechanical dynamics model of the valve-controlled hydraulic mechanical leg joint space is described as: Among them, θ is the joint angle of the mechanical leg, is the angular velocity of the mechanical leg joint, is the angular acceleration of the mechanical leg joint, τ is the control torque of the mechanical leg joint, M, C, and G are the inertia matrix, centrifugal force and Coriolis force matrix, and gravity moment vector in the mechanical leg joint space, respectively; Define the actuator extension displacement as x p , then the angular velocity of the joint of the mechanical leg and actuator extension speed There are the following relationships: Where J is the Jacobian matrix; Then, from equations (1) and (2), we can get the mechanical dynamics model expression of the mechanical leg actuator space: Where F is the output of the actuator, M p , C p , G p are the inertia matrix, centrifugal force and Coriolis force matrix and gravity moment vector in the mechanical leg actuator space, respectively. is the actuator extension acceleration; The output of the actuator has the following relationship with the pressure in the two chambers of the actuator and the effective area: Among them, P1 is the oil inlet chamber pressure, P2 is the oil discharge chamber pressure, A1 is the oil inlet chamber effective area, and A2 is the oil discharge chamber effective area; The pressure dynamic equation of the actuator is: Among them, β e is the elastic modulus of hydraulic oil, V 01 is the initial volume of the oil inlet chamber, V 02 is the initial volume of the oil discharge chamber, C t is the internal leakage coefficient, Q1 is the oil inlet flow rate, Q2 is the oil discharge flow rate, is the first-order derivative of P1 with respect to time, is the first-order derivative of P2 with respect to time; Since the bandwidth of the servo valve is much higher than the system bandwidth, the spool displacement of the servo valve is considered to be proportional to the control input, and the hydraulic oil flow of the two chambers is expressed as: Among them, k t is the control input gain, u is the control input, P s is the system oil supply pressure, P r is the system return oil pressure, sign is the sign function, expressed as: Define the system's state variables x = [x1, x2, x3, x4] T , where the state variable x1 = [x p1 ,x p2 ,x p3 ] T , state variables State variable x3 = [P 11 ,P 21 ,P 31 ] T , state variable x4=[P 12 ,P 22 ,P 32 ] T , x p1 、x p2 、x p3 are the extension displacements of the three actuators, are the extension speeds of the three actuators, P 11 , P 21 , P 31 , are the oil inlet chamber pressures of the three actuators, P 12 , P 22 , P 32 , are the oil discharge chamber pressures of the three actuators respectively, and T represents transposition; The nonlinear model of the system is written in the following state space form: in, represents the first-order derivative of x1 with respect to time, represents the first-order derivative of x2 with respect to time, represents the first-order derivative of x3 with respect to time, represents the first-order derivative of x4 with respect to time, Indicates M P The inverse of , y represents the intermediate variable.

3. The active compliance control method of valve-controlled hydraulic mechanical legs based on active disturbance rejection control according to claim 2 is characterized in that: In step 2, based on the mathematical model of the valve-controlled hydraulic mechanical leg system, an active compliance controller of the valve-controlled hydraulic mechanical leg based on active disturbance rejection control is designed. The specific steps are as follows: Step 2-1, design a valve-controlled hydraulic mechanical leg anti-disturbance controller to achieve joint position control of the mechanical leg; Step 2-2: Design an admittance controller to compensate for the desired trajectory of the ADRC and realize active compliant control of the robotic leg.

4. The active compliance control method of valve-controlled hydraulic mechanical legs based on active disturbance rejection control according to claim 3 is characterized in that: In step 2-1, a valve-controlled hydraulic mechanical leg anti-disturbance controller is designed to realize the joint position control of the mechanical leg. The specific steps are as follows: Let variable z1 = x1, variable z2 = x2, variable Then formula (8) can be rewritten as: in, is the first-order derivative of z1 with respect to time, is the first-order derivative of z2 with respect to time, is the first-order derivative of z3 with respect to time, f represents the intermediate variable, and E represents the intermediate variable; in, for The first derivative with respect to time, C p The first derivative with respect to time, G p The first derivative with respect to time; k t represents the control input gain; Expand f to a new system state z4, and the first-order derivative of z4 with respect to time If w(t) is a time-dependent function, the system is expanded into a new linear control system: The corresponding extended state observer is designed as: Among them, β 01 , β 02 , β 03 , β 04 Both are observer gains, and the tracking error ε1=x1-x 1d , x 1d is the expected trajectory, η1 is the estimation error, is the estimate of z1, is the estimate of z2, is the estimate of z3, is the estimate of z4, for The first derivative with respect to time, for The first derivative with respect to time, for The first derivative with respect to time, for First derivative with respect to time; The observer gain is chosen so that its characteristic equation s 4 +β 01 s 3 +β 02 s 2 +β 03 s+β 04 Satisfies the Hurwitz condition, s represents a complex variable; using the bandwidth method, the observer gain is β 01 =4ω0, Adjust the observer gain ω0 to select the observer parameters; The tracking error is defined as: Among them, ε1, ε2, ε3 are tracking errors, y d is the desired trajectory, the virtual control law Virtual Control Law Yes d The first derivative with respect to time, is the first-order derivative of α1 with respect to time, and the controller gains K1, K2∈R 3×3 , R 3×3 represents a 3×3 real matrix; The control of the robotic leg system is realized according to the estimated value of the state of the robotic leg system by the extended state observer. Therefore, after replacing the system state with its corresponding estimated value, the tracking error of the system is defined as: in, are the tracking errors under the estimated value of the system state, and the estimation of the virtual control law Estimation of virtual control law yes estimates; The design control input u is: Among them, E -1 is the inverse of E, the controller gain K3∈R 3×3 .

5. The active compliance control method of valve-controlled hydraulic mechanical legs based on active disturbance rejection control according to claim 4 is characterized in that: In step 2-2, the admittance controller is designed to compensate for the desired trajectory of the ADRC and realize the active compliance control of the robotic leg. The specific steps are as follows: The admittance controller takes the desired trajectory as input and calculates the output F of the hydraulic actuator according to the mechanical dynamics model. When the mechanical leg is subjected to external force, the actuator output F and the force feedback F of the force sensor are b The difference between the two is the external force F on the mechanical leg. L : F L =F-F b (17) The desired trajectory of the three joints of the robotic leg is compensated by admittance control, so that it exhibits the characteristics of a mass-spring damping system; then the robotic leg is subjected to an external force F L Under the action of z Described as: in, Represents y z The first derivative with respect to time, Represents y z The second derivative with respect to time; Since the range of motion of the joints of the robotic leg is limited, a saturation function is introduced to make the final desired trajectory within the range of motion space of the robotic leg. The final desired trajectory is: Among them, the constant γ=(1-β)ω c , constant β satisfies 0<β<1, constant ω c is the limit rotation angle of the joint, the expected trajectory after compensation Finally, through the anti-disturbance control, the robot leg follows the final desired trajectory y dz Movement,The controller finally designed is divided into two parts, the inner and outer parts. The inner loop's anti-disturbance control makes the robotic leg move along the desired trajectory, and the outer loop's admittance control is used to compensate for the desired trajectory of the robotic leg, so that the robotic leg shows compliance when subjected to external forces.

6. The active compliance control method for valve-controlled hydraulic mechanical legs based on active disturbance rejection control according to claim 5 is characterized in that: In step 3, the stability of the active compliant controller of the valve-controlled hydraulic mechanical leg based on the anti-disturbance control is proved by using the Lyapunov stability theory, and the results that the observation error and tracking error of the system are bounded are obtained, as follows: The Lyapunov function is defined as follows: Among them, V1, V2, and V3 are all Lyapunov functions; The stability is proved by using Lyapunov stability theory, and the results show that the system observation error and tracking error are bounded.

Images