A gait control adjustment method for multi-legged robots based on fuzzy control

Abstract

The invention discloses a gait control and adjustment method of a multi-legged robot based on fuzzy control. First, IMU data is collected to obtain quaternions; then the roll angle and pitch angle of the multi-legged robot are calculated by formula as input variables; and the input variables are fuzzified by determining the fuzzy control rule base. After that, the fuselage height coefficient Kh, the stride distance coefficient KI and the leg lift height coefficient Kf are obtained by fuzzy decision logic reasoning and defuzzification. The fuselage height, stride distance and leg lift height of the multi-legged robot in the motion state are changed in real time by the three coefficients, so that the multi-legged robot can perform adaptive gait adjustment according to environmental information and control the multi-legged robot to move. The method of the invention sets artificial constraints on the key motion indicators of the hexapod robot without the need for additional sensing equipment other than the IMU. Therefore, the cost of the robot and the complexity of its hardware components are low, thereby reducing the cost and hardware requirements.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a gait control adjustment method for a multi-legged robot based on fuzzy control, aiming to improve the motion stability and obstacle surmounting capability of the multi-legged robot in complex terrain. Background Art

[0002] At present, there are many research results on the gait control of multi-legged robots, including methods based on classical control theory, optimization algorithms and deep learning. These results provide a variety of solutions for the gait control of multi-legged robots and provide technical support for their application in different fields. However, traditional multi-legged robots usually walk in a fixed gait. When encountering a relatively rough road surface, the defects of the fixed gait are obviously exposed. If the height of the body is too high, the center of gravity will be raised, affecting stability; if the height of the body is too low, it will be difficult to cross obstacles. Similarly, a large stride distance can easily lead to poor stability, and a small stride distance will reduce the movement speed. Lifting the leg too high will increase unnecessary energy consumption, while lifting the leg too low will affect the ability to cross obstacles.

[0003] Therefore, multi-legged robots need to rely on a variety of sensors to obtain environmental information and implement complex gait planning. Usually, they need to include visual sensors, tactile sensors, inertial measurement units, etc. to perceive the surrounding environment, identify the terrain, and adjust the robot's walking path in time. The integration of these sensors will undoubtedly increase the total weight of the robot, making the body bulky.

[0004] The gait control method based on fuzzy control sets artificial constraints on the key motion indicators (i.e., speed and angular velocity) of the multi-legged robot without the need for additional sensing equipment other than the IMU. Therefore, the cost of the robot and the complexity of its hardware components are low. Fuzzy control can achieve intelligent regulation and control of the robot's gait through fuzzy rules and fuzzy logical operations to adapt to different environments and task requirements. The fuzzy control method has good robustness and adaptability, and can effectively deal with the motion control problems of multi-legged robots in complex environments. Therefore, the gait control method based on fuzzy control is of great significance in improving the motion performance of multi-legged robots and enhancing their adaptability and stability. Summary of the invention

[0005] At present, in order to ensure the stability and autonomy of multi-legged robots, multiple sensors and complex control modes are mainly used, which impose great limitations on these robots. The gait control method based on fuzzy control relies only on the inertial measurement unit, thereby reducing the cost and hardware requirements. The gait adaptive control method based on environmental information provides an innovative solution for the adaptive movement of multi-legged robots in complex environments. The purpose of the present invention is to provide a gait control adjustment method for a multi-legged robot based on fuzzy control, so that the multi-legged robot can autonomously adapt to environmental changes according to environmental information during movement, adjust its gait, and improve movement stability and obstacle crossing ability.

[0006] To achieve the above object, the present invention provides a gait control and adjustment method for a multi-legged robot based on fuzzy control, comprising the following steps:

[0007] S1: IMU data acquisition: Use the inertial measurement unit (IMU) to read the three-axis attitude angle of the multi-legged robot in real time during movement, obtain environmental information, and obtain quaternion.

[0008] S2: Input variable determination: The quaternion is calculated by formula to obtain the roll angle β and pitch angle γ of the multi-legged robot as the input variables of the control system.

[0009] S3: Fuzzy control processing: Fuzzy processing is performed on the input variables by determining the fuzzy control rule base.

[0010] S4: Output variable generation: After fuzzification processing, fuzzy decision logic reasoning and defuzzification are used to obtain the body height coefficient Kh, stride distance coefficient KI and leg lift height coefficient Kf required to control the gait of the multi-legged robot.

[0011] S5: Multi-legged robot motion adjustment: The body height, stride distance and leg lift height of the multi-legged robot in motion are changed in real time through the above three coefficients, so that the multi-legged robot can adaptively adjust its gait according to environmental information and control the multi-legged robot to move.

[0012] The beneficial effects of the present invention are as follows:

[0013] The method of the present invention enhances the stability of a multi-legged robot when moving in an unknown and complex environment. The method of the present invention sets artificial constraints on the key motion indicators (i.e., speed and angular velocity) of the hexapod robot without requiring additional sensing equipment other than the IMU. Therefore, the cost of the robot and the complexity of its hardware components are low, thereby reducing the cost and hardware requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a gait control flow chart of an embodiment of the present invention.

[0015] Figure 2 A membership function graph is input for the embodiment of the present invention.

[0016] Figure 3 The membership function graph is output for the embodiment of the present invention.

[0017] Figure 4 This is a corresponding relationship diagram between fuzzy control input and output variables in an embodiment of the present invention.

[0018] Figure 5 This is a diagram of the adaptive adjustment changes of the hexapod robot according to an embodiment of the present invention.

[0019] Figure 6 1 is a comparison of the roll and pitch changes of the hexapod robot according to the embodiment of the present invention.

[0020] Figure 7 The figure is a comparison of the rolling and pitching variation ranges per unit time of the hexapod robot according to the embodiment of the present invention. DETAILED DESCRIPTION

[0021] The technical solution of the present invention is further described in detail below in conjunction with specific embodiments.

[0022] Example 1

[0023] like Figure 1 The method for adjusting the gait control of a multi-legged robot based on fuzzy control comprises the following steps:

[0024] Step (1): IMU data collection.

[0025] First, during the movement of the multi-legged robot, the inertial measurement unit (IMU) installed on the body monitors its three-axis attitude angle in real time, and obtains the quaternion q=[w,v,y,z] through the IMU, where w is the real part and v, y, and z are the imaginary parts.

[0026] Step (2): Determine the input variables.

[0027] The posture information including the roll angle β and pitch angle γ is calculated by the formula. These posture information reflect the posture changes of the robot on the current terrain.

[0028] The pitch angle calculation formula is as follows:

[0029] γ=atan2(2(wy+vz),1-2(y 2 +z 2 ));

[0030] The roll angle calculation formula is as follows:

[0031] β=atan2(2(wz+vy),1-2(v 2 +y2 );

[0032] Among them, atan2 is the inverse trigonometric function.

[0033] The posture information obtained by calculating the quaternion is used as the input variable, where the roll angle β reflects the degree of the multi-legged robot's left and right tilt, and the pitch angle γ reflects the degree of the front and back tilt.

[0034] Step (3): Determine the fuzzy control rule base.

[0035] The fuzzy control rule base is the basis of fuzzy logic reasoning, is built on the basis of language variables, is also the core of fuzzy control, and directly affects the performance of fuzzy control.

[0036] In the fuzzy control method of the present invention, the control rule conditions use two variables, namely the roll angle β and the pitch angle γ, thereby obtaining three multi-legged robot gait control parameters, namely the body height coefficient Kh, the stride distance coefficient KI and the leg lift height coefficient Kf.

[0037] The fuzzy rule design of the present invention is as follows:

[0038] When both β and γ increase, it means that the ground is rugged and the fuselage height needs to be increased to lower the center of gravity. At the same time, the stride distance should be reduced to slow down the movement speed and the leg lift height should be increased to improve the obstacle crossing ability. That is, Kh decreases, KI decreases, and Kf increases.

[0039] When only one of β or γ increases or decreases, it means that the ruggedness of the ground is stable, so the gait control parameters of the three multi-legged robots are also stable.

[0040] When both β and γ decrease, it means that the ground is flat at this time, so the fuselage height can be raised, the stride distance can be increased, and the leg lift height can be reduced, that is, Kh can be increased, KI can be increased, and Kf can be reduced.

[0041] The specific fuzzy control rules are shown in Table 1:

[0042] Table 1 Fuzzy control rule base

[0043]

[0044] Among them, NB means negative big, NM means negative medium, NS means negative small, ZO means zero, PS means positive small, PM means positive medium, and PB means positive big; XS means extremely small, S means small, MS means medium small, M means medium, MB means medium large, B means large, and XB means extremely large.

[0045] Figure 4 1 is a corresponding relationship diagram between the fuzzy control inputs β and γ and the output variables Kh, KI, and Kf according to an embodiment of the present invention.

[0046] Since the Gaussian membership function has good continuity and smoothness, and can be adjusted by adjusting the center point and standard deviation, it has strong adjustability and adaptability. Therefore, the present invention selects the Gaussian membership function as the membership function of the input variable and the output variable fuzzy value.

[0047] Gaussian membership function:

[0048]

[0049] in, is the value of the variable; =0.15 (fixed value), is the parameter that controls the smoothness of the Gaussian curve; c is the center point of the Gaussian function.

[0050] In the input membership function, the value range of c is c=[-1,-0.6667,-0.3333,0,0.3333,0.6667,1]. Take c=-1 and substitute it into the Gaussian membership function to get the fuzzy subset NB (negative large); similarly, the remaining six subsets NM (negative medium), NS (negative small), ZO (zero), PS (positive small), PM (positive medium), and PB (positive large) can be obtained.

[0051] In the output Kh membership function, the value range of c is c=[0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1]; take c=0.4 and substitute it into the Gaussian membership function to calculate and obtain the fuzzy subset XS (extremely small); similarly, the remaining six subsets S (small), MS (medium small), M (medium), MB (medium large), B (large), and XB (extremely large) can be obtained.

[0052] In the output KI and Kf membership functions, the value range of c is c=[0.2, 0.3333, 0.4667, 0.6, 0.7333, 0.8667, 1], and the seven subset definitions are the same as the output Kh membership function.

[0053] Step (4): Fuzzy control processing.

[0054] The input variables are fuzzified to obtain the fuzzy values ​​of the input variables. Specifically, for the input variables, let the basic domain of the roll angle β be [-K1, K1], and take the quantization factor as 1 / K1, so that its fuzzy domain is [-1, 1]; the basic domain of the pitch angle γ is [-K2, K2], and take the quantization factor as 1 / K2, so that its fuzzy domain is [-1, 1] (where K1 and K2 are determined by the attitude angle range of the multi-legged robot in motion when there is no adaptive control adjustment). Seven fuzzy subsets are used to calculate the fuzzy domain of the input variables, and the seven fuzzy subsets are NB (negative large), NM (negative medium), NS (negative small), ZO (zero), PS (positive small), PM (positive medium), and PB (positive large). Figure 2 Shown are the input β and γ membership functions.

[0055] Step (5): Output variable generation.

[0056] For the output variables, fuzzy decision logic reasoning is performed according to the determined fuzzy control rules to obtain the fuzzy values ​​of the fuselage height coefficient Kh, the stride distance coefficient KI and the leg lift height coefficient Kf. The actual robot's workspace is analyzed, and the minimum leg lift height and the minimum fuselage height are set according to the ruggedness of the terrain. The minimum stride distance is set to maintain a continuous motion state, and its safety and rationality are considered. Specifically, the minimum values ​​of the three coefficients can be adjusted according to the range allowed by the actual robot, thereby determining its domain.

[0057] Seven fuzzy subsets are used to calculate the fuzzy domain of the output variable. The seven fuzzy subsets are XS (extremely small), S (small), MS (medium small), M (medium), MB (medium large), B (large), and XB (extremely large). Figure 3 Is the output Kh, KI, Kf membership function.

[0058] The fuzzy value of the output variable obtained by fuzzy control library reasoning is precisely calculated to obtain the exact control value. The present invention adopts the centroid method for precise calculation, which contains all the information in the fuzzy subset and has a smoother output reasoning control. The principle is to take the centroid of the area enclosed by the fuzzy membership function curve and the horizontal coordinate as the final output value of the fuzzy reasoning, that is:

[0059] (1)

[0060] in, The final result of defuzzification is the output coefficient value. m is the value of When , the number of intersections with the membership function; is a point in the fuzzy domain of the output membership function; is the membership function value of the point; m The values ​​together constitute the fuzzy domain of the output variable.

[0061] The fuzzy values ​​of the output variables are defuzzified to obtain clear fuselage height coefficient Kh, stride distance coefficient KI and leg lift height coefficient Kf.

[0062] Step (6): Multi-legged robot motion adjustment.

[0063] The support phase of a multi-legged robot refers to the stage in the gait cycle when the robot's legs touch the ground. Its characteristic is that it needs to maintain a stable posture to support the robot's mass. The swing phase is the stage in the gait cycle when the robot's legs are away from the ground. Its task is to lift the legs and move the center of gravity to prepare for the next support. These two phases together constitute the gait of the multi-legged robot, ensuring the stability and efficiency of the robot during movement.

[0064] According to the obtained body height coefficient Kh, stride distance coefficient KI and leg lift height coefficient Kf, the gait parameters of the multi-legged robot are adjusted in real time. Specifically:

[0065] The body height Z2 of the multi-legged robot is adjusted by the body height coefficient Kh so that it can maintain an appropriate center of gravity height in rough terrain.

[0066] The stride distance coefficient KI is used to adjust the stride distance Y2 of the multi-legged robot so that it can maintain an appropriate step length in complex terrain and improve movement stability.

[0067] The leg-lifting height Z1 of the multi-legged robot is adjusted by the leg-lifting height coefficient Kf, so that it can flexibly cope with obstacles of different heights during the obstacle crossing process.

[0068] According to the above requirements, the constraints required for the trajectory planning of the multi-legged robot are shown in Table 2.

[0069] Table 2 Multi-legged robot trajectory planning constraint parameters

[0070]

[0071] According to the seven constraints in the table above, a polynomial is used to express the relationship between the distance and time of a single foot of the robot moving in three directions:

[0072] Let P = (X, Y, Z),

[0073] ;

[0074]

[0075] ;

[0076] in, ~ is the coefficient, t is the time, and P is the coordinate position.

[0077] Substituting the seven constraints into formula (2) we obtain:

[0078] (3)

[0079] In the formula, is the initial time, the default is 0; is the time taken from the initial position to the final position, which is 0.5 movement cycle in the triangular gait; is the middle moment, that is, ( + ) / 2. The remaining constraints , , is a configurable parameter. is the initial position, which is obtained by reading the motor angle and performing forward kinematics calculation (forward kinematics is a general method for robot kinematics analysis, which will not be described in detail in the present invention).

[0080] The initial stride distance is fixed at L. is the target location, specifically , is the steering angle.

[0081] The initial fixed value of the leg-lifting height is H. is the middle position, specifically .make , , , All are 0 (i.e., the values ​​in the three directions of X, Y, and Z are all 0). Formula (3) can be simplified to obtain the coefficients ~ The values ​​are:

[0082] (4)

[0083] Substitute the polynomial coefficients obtained by formula (4) into formula (2) to obtain the polynomials about time t in the x, y, and z directions, and establish the parametric equation of t. The parametric equations are combined to obtain the motion trajectory of a single foot in three-dimensional space. The motion trajectories of the corresponding feet in three-dimensional space are calculated for all the feet of the multi-legged robot respectively, and the gait control adjustment of the multi-legged robot is realized according to the obtained motion trajectory.

[0084] Through the above steps, the multi-legged robot can dynamically adjust its gait parameters according to the environmental information obtained in real time, and achieve stable and efficient movement in complex terrain.

[0085] Example 2

[0086] Taking a hexapod robot as an example, an experimental scenario was designed. Since the sand ground is soft, undulating and unpredictable, in order to verify the effectiveness of the present invention, a hexapod robot sand ground experimental platform was built to conduct experiments. The gait control method based on fuzzy control proposed in the present invention was used and the changes in the roll angle and pitch angle of the hexapod robot in the experiment were recorded.

[0087] In this experiment, the vertical distance between the body coordinate system {OB} and the world coordinate system {OW} is recorded as the fuselage height, and the initial fuselage height is 362.43mm. In this experiment, the time T is 1s, and the overall motion speed V is 0.1m / s, so the initial motion stride, that is, the initial stride distance, is 0.1m. Affected by the mechanical design limit of the hexapod robot, the initial leg lift height is set to 150mm. Therefore, the minimum fuselage height coefficient used by the hexapod robot is 0.4, the minimum stride distance coefficient is 0.2, and the minimum leg lift height coefficient is 0.2.

[0088] Experimental results show that the gait control method based on fuzzy control significantly optimizes the movement performance of the hexapod robot on different terrains. Figure 5 The following are the adaptive adjustment change diagrams of the hexapod robot according to the embodiment of the present invention, wherein (a) is the gait adaptive adjustment parameter change curve diagram, (b) is the multi-legged robot body height change diagram, (c) is the multi-legged robot stride distance change diagram, and (d) is the multi-legged robot leg lift height change diagram. Specifically, it is as follows:

[0089] Compare Figure 6 (Where (a) is the roll change of the multi-legged robot, and (b) is the pitch change of the multi-legged robot). Before any control method is adopted, the roll angle change range of the six-legged robot during movement is (-1.7682°, 1.2725°), and the pitch angle change range is (-3.506°, 4.231°). After using the adaptive control method of the present invention, the roll angle change range is (-1.3864°, 1.1081°), and the pitch angle change range is (-2.6476°, 2.5148°). The unstable vibration range of the six-legged robot has been significantly improved, reducing the risk of overturning. The roll angle span is stably maintained within 2.5°, and the pitch angle span can also be stably controlled within 5°. Since it is moving on unknown non-flat sand and is an evaluation adjustment made to the experienced environmental conditions, there is a small amplitude vibration. Comparison Figure 7 (where (a) is the roll change amplitude, and (b) is the pitch change amplitude). After adopting the adaptive control method of the present invention, the change amount of the hexapod robot's roll angle per unit time is significantly reduced and more stable, and no large changes will occur. The degree of change can be stabilized within ±0.5°, and the degree of pitch angle change can also be roughly maintained within ±1°.

[0090] In summary, the fuzzy control-based gait control and adjustment method for a multi-legged robot provided by the present invention can effectively improve the motion stability and obstacle surmounting ability of the multi-legged robot in complex terrains, and has broad application prospects.

[0091] The above contents are further detailed descriptions of the present invention in combination with specific / preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, they can also make several substitutions or modifications to these described embodiments without departing from the concept of the present invention, and these substitutions or modifications should be regarded as belonging to the protection scope of the present invention.

[0092] Parts of the present invention that are not described in detail belong to the well-known technologies of those skilled in the art.

Claims

1. A gait control and adjustment method of a multi-legged robot based on fuzzy control, characterized in that: The following steps are involved: S1: IMU data acquisition: Use the inertial measurement unit (IMU) to read the three-axis attitude angle of the multi-legged robot in real time during movement, obtain environmental information, and obtain quaternion; S2: Input variable determination: The quaternion is calculated by the formula to obtain the roll angle β and pitch angle γ of the multi-legged robot as the input variables of the control system; S3: Fuzzy control processing: fuzzification of input variables is performed by determining the fuzzy control rule base; S4: Output variable generation: After fuzzification processing, the body height coefficient Kh, stride distance coefficient KI and leg lift height coefficient Kf required to control the gait of the multi-legged robot are obtained by using fuzzy decision logic reasoning and defuzzification; S5: Multi-legged robot motion adjustment: The multi-legged robot body height, stride distance and leg lift height are changed in real time through the above three coefficients in the motion state, so as to realize the multi-legged robot to make adaptive gait adjustment according to environmental information and control the multi-legged robot to move; The specific method of step S1 is as follows: First, during the movement of the multi-legged robot, the inertial measurement unit (IMU) installed on the robot monitors its three-axis attitude angle in real time, and obtains the quaternion q = [w, v, y, z] through the IMU, where w is the real part and v, y, z are the imaginary parts; The specific method of step S2 is as follows: The posture information including the roll angle β and the pitch angle γ are calculated by the formula; these posture information reflect the posture changes of the robot on the current terrain; The pitch angle calculation formula is as follows: γ=atan2(2(wy+vz),1-2(y 2 +with 2 )) The roll angle calculation formula is as follows: β=atan2(2(wz+vy),1-2(v 2 +y 2 ) Among them, atan2 is the inverse trigonometric function; The posture information obtained by quaternion calculation is used as the input variable, where the roll angle β reflects the degree of the multi-legged robot's left and right tilt, and the pitch angle γ reflects the degree of the front and back tilt; The specific method for determining the fuzzy control rule base is as follows: The fuzzy rules are designed as follows: When both β and γ increase, it means that the ground is rugged at this time, and the fuselage height needs to be increased to lower the center of gravity, while the stride distance needs to be reduced to slow down the movement speed, and the leg lift height needs to be increased to improve the obstacle crossing ability, that is, Kh decreases, KI decreases, and Kf increases; When only one of β or γ increases or decreases, it means that the ruggedness of the ground is stable, so the gait control parameters of the three multi-legged robots are also stable; When both β and γ decrease, it means the ground is flat, so the fuselage height is raised, the step distance is increased, and the leg height is reduced, that is, Kh is increased, KI is increased, and Kf is reduced; The specific fuzzy control rules are as follows: When β belongs to the fuzzy subset NB, NM, NS, ZO, PS, PM, PB, and γ belongs to the fuzzy subset NB: Kh, KI, and Kf are XS / XS / XB, S / S / B, MS / MS / MB, M / M / M, MS / MS / MB, S / S / B, and XS / XS / SB respectively; When β belongs to the fuzzy subset NB, NM, NS, ZO, PS, PM, PB, and γ belongs to the fuzzy subset NM: Kh, KI, and Kf are S / S / B, MS / MS / MB, M / M / M, MB / MB / MS, M / M / M, MS / MS / MB, and S / S / B respectively; When β belongs to the fuzzy subset NB, NM, NS, ZO, PS, PM, PB, and γ belongs to the fuzzy subset NS: Kh, KI, and Kf are MS / MS / MB, M / M / M, MB / MB / MS, B / B / S, MB / MB / MS, M / M / M, and MS / MS / MB respectively; When β belongs to the fuzzy subset NB, NM, NS, ZO, PS, PM, PB, and γ belongs to the fuzzy subset ZO: Kh, KI, and Kf are M / M / M, MB / MB / MS, B / B / S, XB / XB / XS, B / B / S, MB / MB / MS, and M / M / M respectively; When β belongs to the fuzzy subset NB, NM, NS, ZO, PS, PM, PB, and γ belongs to the fuzzy subset PS: Kh, KI, and Kf are MS / MS / MB, M / M / M, MB / MB / MS, B / B / S, MB / MB / MS, M / M / M, and MS / MS / MB respectively; When β belongs to the fuzzy subset NB, NM, NS, ZO, PS, PM, PB, and γ belongs to the fuzzy subset PM: Kh, KI, and Kf are S / S / B, MS / MS / MB, M / M / M, MB / MB / MS, M / M / M, MS / MS / MB, and S / S / B respectively; When β belongs to the fuzzy subset NB, NM, NS, ZO, PS, PM, PB, and γ belongs to the fuzzy subset PB: Kh, KI, and Kf are XS / XS / XB, S / S / B, MS / MS / MB, M / M / M, MS / MS / MB, S / S / B, and XS / XS / SB respectively; Among them, NB means negative big, NM means negative medium, NS means negative small, ZO means zero, PS means positive small, PM means positive medium, PB means positive big; XS means very small, S means small, MS means medium small, M means medium, MB means medium large, B means large, XB means extremely large; Select Gaussian membership function as the membership function of the fuzzy value of the input variable and the output variable; Gaussian membership function: Where x is the value of the variable; σ = 0.15, which is the parameter for controlling the smoothness of the Gaussian curve; c is the center point of the Gaussian function; In the input membership function, the value range of c is c = [-1, -0.6667, -0.3333, 0, 0.3333, 0.6667, 1]; take c = -1, substitute it into the Gaussian membership function to calculate, and get the fuzzy subset NB; similarly, get the remaining six subsets NM, NS, ZO, PS, PM, PB; In the output Kh membership function, the value range of c is c = [0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1]; take c = 0.4, substitute it into the Gaussian membership function to calculate, and get the fuzzy subset XS; similarly, get the remaining six subsets S, MS, M, MB, B, XB; In the output KI and Kf membership functions, the value range of c is c = [0.2, 0.3333, 0.4667, 0.6, 0.7333, 0.8667, 1], and the seven subsets are defined in the same way as the output Kh membership function; The specific method of fuzzy processing of input variables is as follows: The input variables are fuzzified to obtain the fuzzy values ​​of the input variables; specifically, for the input variables, the basic domain of the roll angle β is [-K1, K1], and the quantization factor is 1 / K1, so that its fuzzy domain is [-1, 1]; the basic domain of the pitch angle γ is [-K2, K2], and the quantization factor is 1 / K2, so that its fuzzy domain is [-1, 1]; wherein K1 and K2 are determined by the attitude angle range of the multi-legged robot in motion when there is no adaptive control adjustment; seven fuzzy subsets are used to calculate the fuzzy domain of the input variables, and the seven fuzzy subsets are NB, NM, NS, ZO, PS, PM, and PB respectively; The specific method of step S4 is as follows: For the output variables, fuzzy decision logic reasoning is performed according to the determined fuzzy control rules to obtain the fuzzy values ​​of the fuselage height coefficient Kh, the stride distance coefficient KI and the leg lift height coefficient Kf; the actual robot's workspace is analyzed, and the minimum leg lift height and the minimum fuselage height are set according to the ruggedness of the terrain. The minimum stride distance is set to maintain a continuous motion state. Specifically, the minimum values ​​of the three coefficients are adjusted according to the range allowed by the actual robot, thereby determining its domain; Seven fuzzy subsets are used to calculate the fuzzy domain of the output variable. The seven fuzzy subsets are XS, S, MS, M, MB, B, and XB. The fuzzy value of the output variable obtained by fuzzy control library reasoning is precisely calculated to obtain the exact control value; the centroid method is used for precise calculation, and its principle is to take the centroid of the area enclosed by the fuzzy membership function curve and the horizontal coordinate as the final output value of fuzzy reasoning, that is: Among them, v0 is the final result of defuzzification, that is, the output coefficient value; m is the value of the coefficient when x=v k When , the number of intersections with the membership function; v k is a point in the fuzzy domain of the output membership function; μ v (v k ) is the membership function value of the point; m v k The values ​​together constitute the fuzzy domain of the output variable; Defuzzify the fuzzy value of the output variable to obtain clear fuselage height coefficient Kh, stride distance coefficient KI and leg lift height coefficient Kf; The specific method of step S5 is as follows: The support phase of a multi-legged robot refers to the stage in the gait cycle when the robot's legs touch the ground. Its characteristic is that it needs to maintain a stable posture to support the robot's mass. The swing phase is the stage in the gait cycle when the robot's legs are away from the ground. Its task is to lift the legs and move the center of gravity to prepare for the next support. These two phases together constitute the gait of the multi-legged robot, ensuring the stability and efficiency of the robot during movement. According to the obtained body height coefficient Kh, stride distance coefficient KI and leg lift height coefficient Kf, the gait parameters of the multi-legged robot are adjusted in real time; specifically: Adjust the body height Z2 of the multi-legged robot by the body height coefficient Kh; Adjust the stride distance Y2 of the multi-legged robot by the stride distance coefficient KI; Adjust the leg-lifting height Z1 of the multi-legged robot by using the leg-lifting height coefficient Kf; According to the above requirements, the constraints required for the trajectory planning of the multi-legged robot are as follows: (1) The constraint parameter is the initial position; the symbol is P0 = (X0, Y0, Z0); (2) The constraint parameter is the middle position; the symbol is P1 = (X1, Y1, Z1); (3) The constraint parameter is the final position; the symbol is P2 = (X2, Y2, Z2); (4) The constraint parameter is the initial velocity; the symbol is V0; (5) The constraint parameter is the final velocity; the symbol is V2; (6) The constraint parameter is the initial acceleration; the symbol is a0; (7) The constraint parameter is the final acceleration; the symbol is a2; According to the seven constraints in the table above, a polynomial is used to express the relationship between the distance and time of a single foot of the robot moving in three directions: Let P = (X, Y, Z), Among them, c0~c6 are coefficients, t is time, and P is the coordinate position; Substituting the seven constraints into formula (2) we get: Where t0 is the initial moment, which defaults to 0; t2 is the time taken from the initial position to the final position, which is 0.5 motion cycles in the triangular gait; t1 is the intermediate moment, which is (t0+t2) / 2; the remaining constraints P0, P1, and P2 are configurable parameters, P0 is the initial position, which is obtained by reading the motor angle and performing positive kinematic calculations; The initial step distance is fixed at L, P2 is the target position, specifically (X0+KI*Lcosθ, Y0+KI*Lsinθ, Z0*Kh), and θ is the steering angle; The initial fixed value of the leg-lifting height is H, and P1 is the middle position, specifically (X0+0.5KI*Lcosθ, Y0+0.5KI*Lsinθ, Z0*Kh-Kf*H); let V0, V2, a0, and a2 all be 0, and simplify formula (3) to obtain the values ​​of coefficients c0 to c6: Substitute the polynomial coefficients obtained by formula (4) into formula (2) to obtain the polynomials about time t in the x, y, and z directions, and establish the parametric equation of t. The parametric equations are combined to obtain the motion trajectory of a single foot in three-dimensional space. The motion trajectory of the corresponding foot in three-dimensional space is calculated for all the feet of the multi-legged robot, and the gait control adjustment of the multi-legged robot is realized according to the obtained motion trajectory.

Images