Three-degree-of-freedom snakelike robot control method of variable telescopic structure and robot

Abstract

The invention discloses a three-degree-of-freedom snakelike robot control method of a variable telescopic structure and a robot, the snakelike robot structure is modularly designed, each module comprises a pitching joint, a yawing joint and a joint, and telescopic joints associated with rotating joints are adopted in the structural design of the joints; the whole rotary joint moves by adopting a serpenoid curve, and different gaits of the telescopic joint are matched with the rotary joint; a telescopic joint is combined with a winding curve, the route of the snakelike robot is optimized through a genetic algorithm, and route deviation is solved; a proximity sense distance measuring sensor at the head of the snake-shaped robot detects obstacles, gaits are automatically switched, the height of the obstacles and the lifting angle of the snake-shaped robot are calculated, and a set of multi-mode control system is designed; the parameters can be adjusted according to the topographic features, the gait strategy is generated, the motion performance of the snake-shaped robot is changed, the gait switching effect in the complex terrain is achieved, and the motion flexibility and safety of the snake-shaped robot are improved.

Description

Technical Field

[0001] The present invention relates to the field of robot technology, and in particular to a control method for a three-degree-of-freedom snake-like robot with a variable telescopic structure and the robot. Background Art

[0002] With the rapid development of automation and intelligent technology, snake-like robots have attracted more and more attention due to their flexible movement ability and wide application potential. As a multi-degree-of-freedom bionic robot, snake-like robots play an important role in human public life and military needs. Mechanical structure classification of snake-like robots: The body can be divided into rigid, flexible, and a combination of flexible and rigid. The connection methods include parallel connection, orthogonal connection, universal joint connection, and PR connection. Snake-like robots with various structures are staggered in pairs.

[0003] However, there are many technical problems in the motion control of existing snake-like robots. First, traditional snake-like robots can only move in a single way, such as creeping, winding or lateral movement, and cannot coordinate multiple motion modes at the same time. This limitation leads to the robot's insufficient adaptability and flexibility in complex environments, and it cannot effectively cope with changing terrain and obstacles, reducing its efficiency and effectiveness in practical applications. Second, the existing technology performs poorly in the obstacle-crossing ability of snake-like robots. Most snake-like robots rely on additional devices such as active wheels to help climb over step-type obstacles, which increases mechanical complexity and energy consumption. At the same time, these traditional control methods fail to fully utilize the structural advantages of the robot itself, resulting in low robot movement efficiency when dealing with obstacles of different heights, and even safety hazards such as falling. With the continuous development of information technologies related to robots and artificial intelligence, the trend of snake-like robots is to be able to move quickly and reach positions accurately. Summary of the invention

[0004] The embodiment of the present application provides a control method and robot for a three-degree-of-freedom snake-like robot with a variable telescopic structure, and a control method based on a genetic algorithm. By optimizing the design of the snake-like robot with a variable telescopic structure, the technical difficulties of the snake-like robot in the prior art in terms of a single motion mode and insufficient obstacle-crossing ability are solved. The present application sets a fitness function to optimize the motion trajectory of the robot in different modes of the telescopic joint in real time, thereby achieving multi-modal motion control and enhancing the flexibility and adaptability of the robot. At the same time, combined with virtual robot technology, online programming records the running trajectory in different modes, so that the robot can automatically adjust the motion strategy according to the terrain characteristics, improve the overall operating efficiency, and achieve accurate, safe, and efficient control of the snake-like robot in complex environments.

[0005] The present application provides a control method for a three-degree-of-freedom snake-like robot with a variable telescopic structure, comprising:

[0006] Step S1, a modular design is adopted for the structure of the snake robot. Each module includes pitch, yaw, and telescopic joints. In the joint structure design, a telescopic joint associated with a rotary joint is used.

[0007] Step S2, according to the road surface conditions, the road surface is divided into three types: flat ground, obstacle road surface, and step-type obstacle. Analyze the movement mode, distance transformation, and rhythm transformation of the telescopic joint, and then divide the telescopic joint gait into six modes: single standing wave gait, multiple standing wave gait,脱困 gait, obstacle crossing gait, acceleration and deceleration gait, and stop gait.

[0008] Step S3, the overall rotary joint uses a meandering curve for motion control, and the telescopic joint gait is different under different motion modes.

[0009] Step S4, combine the telescopic joint with the meandering curve, and use the genetic algorithm to optimize the overall trajectory of the snake robot to solve the route deviation caused by the vertex problem of joint angle change.

[0010] Step S5, the proximity ranging sensor at the head of the snake robot detects obstacles, automatically switches the gait, calculates the height of the obstacle, calculates the lifting angle of the snake robot, and designs a multi-modal control system.

[0011] Preferably, the telescopic joint associated with the rotary joint in step S1 is specifically: when the angle of the rotary joint is within the range of ±0.5 degrees, the control system will trigger the corresponding telescopic joint to expand and contract, forming a telescopic control mode based on the joint angle.

[0012] Preferably, the number of standing waves of the telescopic joint gait in step S2 is:

[0013] m ∈ [(n + 2) / (h + 2), (n + h) / (h + 2)]

[0014] Wherein, the number of standing waves m and the number of joint units n satisfy m < n / 2; a gap h is set between multiple standing waves, and h < n is satisfied.

[0015] Preferably, step S4 is specifically:

[0016] Step S41, initialize individuals and populations, and generate a certain number of random trajectories.

[0017] Step S42, set the fitness function; the fitness function is used to evaluate the quality of each motion trajectory, and the specific calculation method can be based on the deviation degree between the current trajectory and the ideal trajectory. Yc is the y coordinate of the current trajectory, Yi is the y coordinate of the ideal trajectory, and n is the number of points of the trajectory.

[0018] Step S43, selection operation: using tournament selection, randomly selecting 3 individuals from the population for comparison each time, and selecting the individual with the highest fitness as the parent;

[0019] Step S44, crossover operation: single-point crossover is adopted, a crossover point is randomly selected, and the crossover probability is set to 80%;

[0020] Step S45, mutation operation; the mutation rate is set to 3%, and 3% of individuals are randomly selected in each generation for small adjustments to introduce new genes;

[0021] Step S46, setting the number of iterations; set it to 200 generations;

[0022] Step S46, setting the termination condition; if the fitness does not significantly improve within 20 consecutive generations, the algorithm is stopped.

[0023] Preferably, the obstacle height calculation formula in step S5 is:

[0024] H=h1+h2=(l+d)tanθ+h2

[0025] Among them, h1 is the height of a single module of the snake-like robot; h2 is the height between a single module of the snake-like robot and an obstacle; l is the length of a single module of the snake-like robot; d is the distance between a single module of the snake-like robot and an obstacle; θ is the lifting angle of a single module of the snake-like robot.

[0026] Preferably, the calculation formula of the snake-like robot lifting angle in step S5 is:

[0027] (1) When the obstacle height is ≤ the length of a single module of the snake-like robot, the robot can lift two joints to rest on the obstacle, and the lifting angle satisfies θ = arcsin(H / l);

[0028] (2) When the obstacle height is ≥ the length of a single module of the snake-like robot, the robot needs to lift more than or equal to three joints to place it on the obstacle, and the lifting angle meets the requirements. n is the number of joints to be lifted.

[0029] The present application also provides a three-degree-of-freedom snake-like robot with a variable telescopic structure, including several robot single modules, a telescopic structure, and a unidirectional passive wheel. The connection modes between the robot single modules include parallel connection, orthogonal connection, universal joint connection, and PR connection.

[0030] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0031] 1. The gait motion is generated by using the serpentine curve proposed by the Hirose team. After an in-depth analysis of the serpentine motion characteristics of biological snakes, the Hirose team simplified its motion curve and obtained a sine wave function for controlling the serpentine motion of the snake-like robot. Unlike the comparative literature, the present invention fits the serpentine curve into a sine wave control function and directly uses it to control the robot motion, which greatly reduces the complexity of the mathematical model, makes the control process simpler, and has better controllability and adjustment flexibility in the direction of motion.

[0032] 2. The present invention uses a genetic algorithm to optimize the initial angle α of the serpentine curve and the wave number kn in the serpentine motion. These two parameters directly determine the serpentine motion trajectory of the snake robot. During the optimization process, the system records the distance of the robot's offset from the x-axis in the simulation environment, and feeds the offset as a fitness standard to the genetic algorithm, continuously adjusting α and kn to reduce the offset. This method avoids a complex feedback control system and directly uses the offset distance as the optimization target, making the motion trajectory more consistent with the ideal path. At the same time, since the optimization parameters are concentrated in the serpentine curve control, the control model is simplified and the optimization efficiency is higher. .

[0033] 3. The present invention realizes the multi-gait mode switching based on the scene, and the switching process relies on the real-time detection feedback of the camera and the sensor. Specifically, the present invention designs an obstacle crossing behavior. When the ranging sensor detects an obstacle, the gait is automatically switched to adapt to the shape and distance of the obstacle. By monitoring the environmental information, during the robot movement, the sensor feedback triggers the automatic switching, which enhances the adaptability to the multi-gait environment.

[0034] 4. The present invention adopts a telescopic joint associated with a rotary joint in the joint structure design. When the angle of the rotary joint is within the range of ±0.5 degrees, the control system will trigger the corresponding telescopic joint to telescope, forming a telescopic control mode based on the joint angle. This structural design refers to the prototype of the servo electric cylinder, and has the functions of adjusting the telescopic force and amount. Unlike the comparative literature, the telescopic joint in the present invention can respond to the angle change of the rotary joint to form a matching relationship, thereby improving the flexible control effect of the robot during movement, especially when dealing with obstacles and climbing. It has good adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 This is a flow chart of the snake-like robot control method in Example 1 of the present application;

[0036] Figure 2 This is a schematic diagram of multi-modal control in Example 1 of the present application;

[0037] Figure 3This is a schematic diagram of a single standing wave straight line gait in Example 1 of the present application;

[0038] Figure 4 This is a schematic diagram of a multi-standing wave straight line gait in Example 1 of the present application;

[0039] Figure 5 This is a schematic diagram of the combination of the telescopic joint and the meandering curve in the first embodiment of the present application;

[0040] Figure 6 This is a schematic diagram of the flattening process of the joint in the first embodiment of the present application;

[0041] Figure 7 This is a schematic diagram of the ideal curve iteration in Example 1 of the present application;

[0042] Figure 8 This is a schematic diagram of route deviation and correction in Example 1 of the present application;

[0043] Fig. 9 This is a schematic diagram of obstacle detection of a snake-like robot in Example 1 of the present application;

[0044] Fig.10 This is a schematic diagram of the head-raising action of the snake-like robot during obstacle detection in Example 1 of the present application;

[0045] Fig.11 This is a schematic diagram of obstacle height calculation for the snake-like robot in Example 1 of the present application;

[0046] Fig.12 This is a schematic diagram of the structure of the snake-like robot in Example 2 of the present application. DETAILED DESCRIPTION

[0047] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0048] A control method for a three-degree-of-freedom snake-like robot with a variable telescopic structure, such as Figure 1 As shown, it includes step S1, adopts modular design of snake-like robot structure, each module includes pitch, yaw and telescopic joints, and the joint structure design adopts telescopic joints associated with rotary joints. When the angle of the rotary joint is within the range of ±0.5 degrees, the control system will trigger the corresponding telescopic joint to extend and retract, forming a telescopic control mode based on the joint angle. With reference to the physiological structure of biological snakes and earthworms, the snake-like robot body is designed to ensure that the robot's structural body has good walking ability, load-bearing capacity, and three-dimensional space movement ability.

[0049] Step S2, such as Figure 2 As shown in Figure 1, the road surface is divided into three types according to the road surface conditions: flat ground, obstacle road surface, and step obstacle. Figure 3 , Figure 4 As shown in the figure, analyze the movement mode, distance transformation, and rhythm transformation of the telescopic joint, and then classify the telescopic joint gait into six modes: single standing wave gait, multiple standing wave gait,脱困 gait, obstacle crossing gait, acceleration and deceleration gait, and stop gait. Among them, the relationship between the number of standing waves of the telescopic joint gait and the body speed is as follows:

[0050] m ∈ [(n + 2) / (h + 2), (n + h) / (h + 2)]

[0051] Among them, the number of standing waves m and the number of joint units n satisfy m < n / 2; a gap h is set between multiple standing waves, and h < n; when the robot enters the motion cycle, the average forward speed of the multiple standing wave motion is n / (h + 2) times that of the single standing wave motion.

[0052] Step S3, as Figure 5 shown, in step S3, the overall rotation joint uses a serpentine curve for motion control, and the telescopic joint gait is different under different motion modes.

[0053] The Serpenoid serpentine curve is a mathematical model used to simulate the movement of snake-like organisms or snake-shaped robots. The shape of this curve is similar to a sine wave, but its characteristic lies in simulating the twisting motion of the snake's body when moving forward on a plane. The shape of the Serpenoid curve can be described by parametric equations that involve sine and cosine functions to simulate the lateral and longitudinal movements of the snake's body. Each part of the snake will fluctuate with different phases and frequencies, and the overall effect is a continuously advancing curve that swings periodically in the left and right directions. The specific formula is as follows:

[0054]

[0055] δσ = αcosβσ + cσ (2)

[0056] Among them, α is the swing amplitude of the curve, β is the number of waveform periods within a unit length; c is the offset angle of the waveform. By adjusting the Serpenoid curve parameters, the propagation amplitude, frequency, and offset angle of the serpentine curve can be flexibly changed, thereby adjusting the propagation direction of the curve. To more intuitively see how the Serpenoid curve controls the snake-shaped robot, the above curve is deformed and sorted, and the simplified curve formula is as follows:

[0057]

[0058] Among them, βi(S) is the rotation radian of the rotation joint i; α is the initial angle of the rotation joint, that is, the swing amplitude; Kn is the number of S waves during the serpentine process, π is the circumference ratio, s is the total arc length, L is the total length of the snake-shaped robot, and N is the number of modules.

[0059] A serpentine curve is used to generate gait motion, and the serpentine curve is fitted into a sinusoidal wave control function, which is directly used for controlling the robot motion. This greatly reduces the complexity of the mathematical model, makes the control process simpler, and has better controllability and adjustment flexibility in the direction of motion.

[0060] The system records the distance that the robot deviates from the x-axis in the simulation environment, and feeds the offset back to the genetic algorithm as a fitness criterion, continuously adjusting α and kn to reduce the offset.

[0061] The target trajectory is a preset straight path, and the offset can be measured by calculating the Euclidean distance between the robot's current position and the target trajectory: The smaller the offset, the higher the fitness, so the fitness function can be defined as: fitness = 1 / (1 + total offset). The inverse form is used here to ensure that the smaller the offset, the closer the fitness is to 1. The optimization goal is to minimize the offset. The initial population is randomly generated, and each individual contains a set of initial values ​​​​of α and Kn. The population size P can be set to dozens to hundreds of individuals. Individuals with higher fitness are selected to enter the next generation. Common methods include roulette selection and tournament selection. Select parent individuals P1 = [α1, Kn1] and P2 = [α2, Kn2], and obtain offspring individuals after crossover: C = [α1, Kn2] or C = [α2, Kn1]. Add a small random perturbation to a control parameter (such as α or Kn) to avoid falling into a local optimum. α new =α old +N(0,δ) where N(0,σ) is Gaussian noise with mean 0 and standard deviation σ. In each generation, the simulation is run using the control parameters of the current individual (i.e., the current values ​​of α and Kn): the robot's offset from the predetermined path is recorded.

[0062] The smaller the offset, the higher the fitness, and the genetic algorithm will iterate in the direction of reducing the offset. The offset of each individual is fed back to the genetic algorithm as the fitness to ensure that individuals with smaller offsets are selected. When the fitness reaches a certain preset threshold, or the number of iterations reaches the maximum value, the algorithm stops. A set of optimized α and Kn parameters can be obtained.

[0063] Step S4, step S4, combining the telescopic joint with the meandering curve, using a genetic algorithm to optimize the overall trajectory of the snake-like robot, and solving the route deviation caused by the vertex problem of joint angle change.

[0064] like Figure 6 As shown in the figure, the joint angles are flattened by genetic algorithm, as shown in the figure. Figure 7As shown, the steps can be divided into selection-crossover-mutation-fusion, which is one generation; set a suitable fitness value function, select a curve in a suitable range, select the one that meets the conditions in one generation and fuse it with the ideal copy again, and repeat this cycle.

[0065] The specific operations are as follows:

[0066] Step S41, initializing individuals and populations, and generating a certain number of random trajectories.

[0067] Each individual represents a motion trajectory, which consists of a series of x and y coordinates corresponding to a series of time points. When initializing the population, a certain number of random trajectories can be generated. These trajectories should cover the potential motion space to ensure diversity. The number is set to 50 to ensure diversity with limited computing resources.

[0068] Step S42, setting the fitness function.

[0069] The fitness function is used to evaluate the quality of each motion trajectory. The specific calculation method can be based on the deviation between the current trajectory and the ideal trajectory. Yc is the y coordinate of the current trajectory, Yi is the y coordinate of the ideal trajectory, and n is the number of points in the trajectory. Using negative values ​​makes the fitness higher (the smaller the deviation), the better.

[0070] Step S43, select an operation.

[0071] Using tournament selection, three individuals are randomly selected from the population for comparison each time, and the individual with the highest fitness is selected as the parent.

[0072] Step S44, crossover operation.

[0073] Using single-point crossover, a crossover point is randomly selected and the crossover probability is set to 80%, which means that 80% of the fathers will undergo the crossover operation.

[0074] Step S45, mutation operation.

[0075] The mutation operation is used to introduce new genes to increase the diversity of the population. The mutation rate is set to 3%, and 3% of individuals are randomly selected in each generation for small adjustments to introduce new genes.

[0076] Step S46, setting the number of iterations.

[0077] The number of generations is set to 200, which is sufficient to optimize the path without over-consuming computing resources.

[0078] Step S46, setting termination conditions

[0079] If the fitness does not improve significantly within 20 consecutive generations, the algorithm is stopped.

[0080] like Figure 8 As shown, the red line is the route deviation caused by the vertex phenomenon, the blue line is the result after correction by the genetic algorithm, and the red straight line is the center line of the ideal curve.

[0081] Step S5, the proximity ranging sensor on the head of the snake-like robot detects an obstacle and automatically switches gait;

[0082] like Fig. 9 , Fig.10 As shown in Figure 1, if there is an obstacle and it reaches the specified range, the robot switches to a straight gait, the robot head is raised, the proximity ranging sensor continues to work, and when the distance is detected to change suddenly, the monitoring stops and the pitch joint angle is recorded, as shown in Figure 1. Fig.11 As shown, the obstacle height can be calculated at this time.

[0083] Obstacle height: H = h1 + h2 = (l + d) tanθ + h2

[0084] Among them, h1 is the height of a single module of the snake-like robot; h2 is the height between a single module of the snake-like robot and an obstacle; l is the length of a single module of the snake-like robot; d is the distance between a single module of the snake-like robot and an obstacle; θ is the lifting angle of a single module of the snake-like robot.

[0085] Case 1: When the obstacle height is ≤ the length of a single module of the snake robot, the robot only needs to lift two joints to place it on the obstacle, and the lifting angle satisfies θ = arcsin (H / l).

[0086] Case 2: When the obstacle height is ≥ the length of the single module of the snake robot, the robot needs to lift more than or equal to three joints to place on the obstacle, and the lifting angle meets n is the number of joints to be lifted.

[0087] Based on the above two situations, we can conclude that the lifting angle of the robot to climb over obstacles is closely related to the length of its own module. The longer the module, the smaller the lifting angle required, and the smaller the energy consumption. Therefore, when lifting the same angle to climb over obstacles, the snake-like robot with variable telescopic length joints can climb over higher obstacles, and the energy supply during climbing is only satisfied by the other telescopic joints that are not climbing.

[0088] Embodiment 2

[0089] A three-degree-of-freedom snake-like robot with a variable telescopic structure, such as Fig.12 As shown, it includes several robot single modules, a retractable structure, and a unidirectional passive wheel. The connection modes between the robot single modules include parallel connection, orthogonal connection, universal joint connection, and PR connection.

[0090] The embodiments of this specific implementation mode are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, all equivalent changes made according to the structure, shape, and principle of the present invention should be included in the protection scope of the present invention. Although the preferred embodiments of the present invention have been described, once the technical personnel in this field know the basic creative concept, they can make other changes and modifications to these embodiments. Therefore, the attached claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. Obviously, the technical personnel in this field can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. In this way, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and variations.

Claims

1. A control method for a three-degree-of-freedom snake-like robot with a variable telescopic structure, characterized in that: Including: Step S1: Design the structure of the snake robot using a modular approach. Each module includes pitch, yaw, and telescopic joints. In the joint structure design, a telescopic joint associated with a rotary joint is adopted. Step S2: Classify the road surface into three types according to the road surface state: flat ground, obstacle road surface, and step-type obstacle. Analyze the movement mode, distance transformation, and rhythm transformation of the telescopic joint, and then divide the gait of the telescopic joint into six modes: single standing wave gait, multiple standing wave gait,脱困 gait, obstacle crossing gait, acceleration and deceleration gait, and stop gait. Step S3: The rotary joint as a whole uses a serpentine curve for motion control, and the gait of the telescopic joint is different under different motion modes. Step S4: Combine the telescopic joint with the serpentine curve, and use the genetic algorithm to optimize the overall trajectory of the snake robot to solve the route deviation caused by the vertex problem of joint angle change. Step S5: The proximity ranging sensor at the head of the snake robot detects obstacles, automatically switches the gait, calculates the height of the obstacle, calculates the lifting angle of the snake robot, and designs a multi-modal control system.

2. The snake-like robot control method according to claim 1, characterized in that: The telescopic joint associated with the rotary joint in step S1 is specifically: when the angle of the rotary joint is within the range of ±0.5 degrees, the control system will trigger the corresponding telescopic joint to expand and contract, forming a telescopic control mode based on the joint angle.

3. The snake-like robot control method according to claim 1, characterized in that: The number of standing waves in the gait of the telescopic joint in step S2 is: m ∈ [(n + 2) / (h + 2), (n + h) / (h + 2)] where the number of standing waves m and the number of joint units n satisfy m < n / 2; a gap h is set between multiple standing waves, and h < n.

4. The snake-like robot control method according to claim 1, characterized in that: Step S4 is specifically: Step S41: Initialize individuals and populations, and generate a certain number of random trajectories. Step S42: Set the fitness function; the fitness function is used to evaluate the quality of each motion trajectory, and the specific calculation method can be based on the deviation degree between the current trajectory and the ideal trajectory. Yc is the y coordinate of the current trajectory, Yi is the y coordinate of the ideal trajectory, and n is the number of points on the trajectory. Step S43: Selection operation; use tournament selection, randomly select 3 individuals from the population for comparison each time, and select the individual with the highest fitness as the parent. Step S44: Crossover operation; adopt single-point crossover, randomly select a crossover point, and set the crossover probability to 80%. Step S45: Mutation operation; set the mutation rate to 3%, and randomly select 3% of the individuals in each generation for small adjustments to introduce new genes. Step S46: Set the number of iterations; set it to 200 generations. Step S46: Set the termination condition. If the fitness does not increase significantly within 20 consecutive generations, stop the algorithm.

5. The snake-like robot control method according to claim 1, characterized in that: The calculation formula for the height of the obstacle in step S5 is: H = h1 + h2 = (l + d)tanθ + h2 where h1 is the height of a single module of the snake robot; h2 is the height between a single module of the snake robot and the obstacle; l is the length of a single module of the snake robot; d is the distance between a single module of the snake robot and the obstacle; θ is the lifting angle of a single module of the snake robot.

6. The snake-like robot control method according to claim 1, characterized in that: The calculation formula for the lifting angle of the snake robot in step S5 is: (1) When the obstacle height is ≤ the length of a single module of the snake-like robot, the robot can lift two joints to rest on the obstacle, and the lifting angle satisfies θ = arcsin(H / l); (2) When the obstacle height is ≥ the length of a single module of the snake-like robot, the robot needs to lift more than or equal to three joints to place it on the obstacle, and the lifting angle meets the requirements. n is the number of joints to be lifted.

7. A snake-like robot using the snake-like robot control method according to any one of claims 1 to 6, characterized in that: It includes several robot single modules, a retractable structure, and a unidirectional passive wheel. The connection modes between the robot single modules include parallel connection, orthogonal connection, universal joint connection, and PR connection.

Images