Four-legged robot navigation coordinate point position acquisition device and navigation precision evaluation method

Abstract

The present application relates to a kind of four-legged robot navigation coordinate point position acquisition device, and navigation precision evaluation method based on the acquisition device, the image information collected by acquisition device, the head and tail coordinates of actual position of robot are calculated, compared with the two coordinates of preset position, the center distance and deviation angle between actual position and preset position are calculated quickly, for the accuracy of the navigation path before and after the software and hardware change is evaluated quickly, the center distance and deviation angle can be used as correction amount when programming in later period, make up the navigation error generated under different motion modes of robot, improve the precision of navigation.

Description

Technical Field

[0001] The present invention relates to the field of robot automatic control, and in particular to a navigation coordinate point acquisition device and a navigation accuracy evaluation method for a quadruped robot. Background Art

[0002] Carrying out reconnaissance, detection, and search and rescue work at the scene of dangerous disasters such as toxic, oxygen-deficient, or smoke-filled accidents is one of the core tasks of firefighting and emergency rescue. However, rescue workers often face threats to their personal safety, and manual operations often face problems such as insufficient information collection.

[0003] It is of great significance to use mobile robots, especially quadruped robots, to replace or assist rescue workers in entering high-risk scenarios. Currently, quadruped robots have excellent adaptability to complex terrain such as ruins, stone piles, and stairs. They can cross high obstacles or penetrate deep into indoor environments, and are highly flexible in narrow spaces. Quadruped robots can detect toxic smoke and oxygen-deficient environments, track heat sources and distant small targets. Fire reconnaissance solutions integrate application modules such as visible light and infrared dual-spectrum gimbals, gas sensors, omnidirectional cameras, rotating laser scanners, and microphones. Combined with an intelligent information background, they provide visual information for rescue decision-making and can transport rescue supplies, assisting firefighters and emergency rescue personnel in post-disaster search and rescue and reconnaissance and rescue work. At the same time, through the collection of full-dimensional on-site data such as residue sampling, gas sampling, and three-dimensional terrain construction, combined with intelligent algorithms and analysis, they provide important reference for disaster tracing and predictive prevention measures.

[0004] Due to the complex terrain, there will be multiple obstacles on the robot's path. In order to avoid obstacles, the robot needs to monitor the environment in real time through video, radar, etc., continuously plan the path, and reach the navigation end point.

[0005] See also Figure 9 However, due to various factors, such as the robot's movement mode (high-speed movement, low-speed movement, jumping movements), deviations in navigation programming, and errors in hardware coordination, the robot cannot follow the preset path to the preset position when starting from the starting position, resulting in a deviation between the actual path and the actual position. Therefore, to reach the navigation destination, it is necessary to continuously correct the errors during the process to reach the navigation destination coordinates. This error correction process is very time-consuming.

[0006] During the robot's navigation programming, hardware assembly, and hardware selection processes, it is necessary to continuously test and optimize the software and hardware so that the preset position and the actual position are the same.

[0007] Therefore, evaluating the deviation between the actual position and the preset position is an important criterion for measuring the matching degree of software and hardware. Therefore, how to quickly calculate the deviation is one of the technical problems to be solved by this application. Summary of the Invention

[0008] The purpose of the present invention is to provide a navigation coordinate point acquisition device and a navigation accuracy evaluation method for a quadruped robot. The head and tail coordinates of the actual position of the robot are calculated through the image information collected by the acquisition device, and compared with the two coordinates of the preset position. The center distance and deviation angle between the actual position and the preset position are quickly calculated, which is used to quickly evaluate the accuracy of the navigation path before and after software and hardware changes. At the same time, the collected center distance and deviation angle can be used as correction values ​​in later software programming to compensate for the navigation errors generated by the robot in different motion modes and improve the accuracy of navigation.

[0009] In order to achieve the above-mentioned object of the invention, the present invention adopts the following technical solutions:

[0010] A quadruped robot navigation coordinate point acquisition device includes multiple acquisition modules arranged in a matrix. The acquisition module includes a glass plate, a closed shell, a camera and an outer frame. The outer frame is cubical, the camera is fixed to the bottom of the outer frame, the glass plate is located on the upper surface of the outer frame, the camera has a lens, the lens is aimed at the glass plate, the closed shell is used to enclose the area between the lens and the glass plate, and the lens collects light spots above the glass plate to form a picture with light spots.

[0011] Preferably, the surface of the glass plate is provided with a shading film for filtering ambient light, and the light transmittance of the shading film is greater than 50%. The light source in the environment is filtered by the shading film to avoid exposure to ambient light sources such as fluorescent lamps and the sun, which may cause the camera to be unable to accurately identify the light spot.

[0012] Preferably, a weight sensing module is provided under the glass plate, that is, only when the robot walks over and the weight suddenly increases (the increased weight is the total weight of the robot), the current position of the robot can be accurately determined.

[0013] Preferably, the collection module further includes a preset laser emitter, which is located below the glass plate.

[0014] Compared with the existing technology, the quadruped robot navigation coordinate point acquisition device using the above technical solution has the following beneficial effects:

[0015] The navigation coordinate point acquisition device can be placed on a flat floor indoors, with the robot to be tested placed on top of it. By emitting lasers downward (towards the glass plate of the acquisition module) at multiple detection points on the robot's body, a bright light spot is formed on the surface of the glass plate. The image of the light spot is then recorded through a lens and camera, and the center position of the light spot is identified using image recognition software, allowing evaluators to quickly obtain the robot's current position and coordinate values ​​for subsequent evaluation calculations.

[0016] The present invention also discloses another technical solution: a method for evaluating the navigation accuracy of a quadruped robot, which includes a starting position, a preset position, and an actual position, and specifically includes the following steps:

[0017] Step 1) Divide the front and rear ends of the robot into two picked coordinate points, namely the front coordinate and the rear coordinate, and take the midpoint on the line connecting the front and rear ends to obtain the mid coordinate;

[0018] Step 2) When the robot is in the starting position, it plans its navigation path to the preset position. The preset position generates preset coordinates, which correspond to the following three coordinates:

[0019] Preset front coordinate: front 0 (X 01 ,Y 01 )

[0020] Preset rear coordinate: rear 0 (X 02 ,Y 02 )

[0021] Preset center coordinate: Center 0 (X 03 ,Y 03 ) The preset middle coordinate is the midpoint of the line connecting the preset front and rear coordinates;

[0022] The robot is enabled to move along a preset path to a preset position, while the evaluator marks the three coordinates of the preset position on a plane;

[0023] Step 3) After the robot has traveled, it follows the actual path and reaches the actual position. The evaluator marks the actual coordinates of the actual position on the plane. The actual coordinates correspond to the following three coordinates:

[0024] Actual front coordinate: Front 1 (X 11 ,Y 11 )

[0025] Actual rear coordinate: rear 1 (X 12 ,Y 12 )

[0026] Actual mid-coordinate: mid-1 (X 13 ,Y 13 ), where the actual mid-coordinate is the midpoint of the line connecting the actual front and rear coordinates;

[0027] Step 4) Collect center distance data: that is, measure or calculate the distance between center 0 and center 1 to obtain the center distance D between center 0 and center 1;

[0028] Step 5) Collect the offset angle data: that is, measure or calculate the angle between the straight line formed by the first 0 and the last 0 and the straight line formed by the first 1 and the last 1 to obtain the offset angle θ;

[0029] Step 6) Evaluate the accuracy of the navigation programming and hardware coordination in this activity by counting the center distance D and offset angle θ after the robot's navigation action.

[0030] Compared with the existing technology, the quadruped robot navigation accuracy evaluation method using the above technical solution has the following beneficial effects:

[0031] Through the above method, the center distance D and offset angle θ between the actual position and the preset position can be quickly calculated, which is used to evaluate the matching of software and hardware during the implementation process. At the same time, after obtaining the center distance D and offset angle θ, the correction code can be added at the end of the software without modifying the software programming, and the actual measured center distance D and offset angle θ can be input for correction of the software code under different activity modes (high-speed movement, low-speed movement, normal movement, jumping movement, etc.).

[0032] There are currently two methods for measuring the center distance D. The first is to directly use a marker / marker to mark the preset and actual positions on a plane. By recording the front 0, back 0, front 1, back 1 in the preset and actual positions, drawing the center 0 and center 1, and directly measuring the distance between the center 0 and center 1, the center distance D can be obtained. The offset angle θ can be obtained by measuring the angle between the lines of front 0, back 0 and front 1, back 1. In other words, the preset front and back coordinates and the actual front and back coordinates are both recorded by the evaluator drawing coordinate points on the plane, and the positions of each generated coordinate point are recorded.

[0033] The second measurement method involves installing a navigation coordinate point acquisition device within the assessment area. Using the device's camera and lens, combined with the robot's laser spot, image recognition software records the laser spot's position. Once a coordinate system is established, the laser spot's location can be determined. Therefore, the current center distance D and offset angle θ can be directly calculated by a computer using a formula.

[0034] The steps for calculating the center distance D are as follows:

[0035] In step 4, the center distance D is calculated by the coordinate values ​​of center 0 and center 1 using a distance calculation formula. The calculation formula is as follows:

[0036] Center distance D=

[0037] The steps for calculating the offset angle θ are as follows:

[0038] In step 5, the offset angle θ is calculated by the angle calculation formula using the coordinate values ​​of front 0, back 0, front 1, and back 1. The calculation formula is as follows:

[0039] Offset angle θ=

[0040] Preferably, the coordinate values ​​of 0 and 1 are obtained as follows:

[0041] Step 2.1) The assessor collects the coordinates of the starting position and calculates the preset coordinates of the preset position (X 01 ,Y 01 ) and, after the preset coordinates are 0 (X 02 ,Y 02 ), and record the two coordinate points of the preset position in the plane;

[0042] Step 3.1) Start the robot and allow it to move to the actual location according to the navigation programming. The evaluator records the two coordinate points of the actual location;

[0043] Step 3.2) Draw the coordinate system with the line connecting the preset front coordinate and the preset rear coordinate as the Y axis and the point where the preset rear coordinate is located as the origin.

[0044] Calculate the preset center coordinate: Center 0 (X 03 ,Y 03 )

[0045] , , where Y 02 =0;

[0046] Step 3.3) Check the actual front coordinate and the actual back coordinate in the coordinate system drawn in step 3.2) and record the two coordinate points of the actual position: front 1 (X 11 ,Y 11 ) and after 1 (X 12 ,Y 12 );

[0047] Calculate the actual coordinates: Center 1 (X 13 ,Y 13 )

[0048] , ;

[0049] Preferably, the robot is provided with laser emitters at its head and tail, and the robot is placed on the above-mentioned quadruped robot navigation coordinate point acquisition device, and the laser emitters irradiate laser light onto the glass plate of the acquisition module to form light spots 1 in front and 1 in the back;

[0050] The preset front and rear coordinates are obtained by navigation programming calculation, and the actual front and rear coordinates are obtained by shooting or taking pictures, and by collecting the light spot of the laser transmitter.

[0051] Preferably, the robot's moving feet are also provided with four laser emitters, and the laser emitters on the head, tail and moving feet use lasers of three different colors, and form light spots of different colors on the surface of the glass plate: front 1, rear 1 and foot light spots.

[0052] When the lens observes the light spots on the surface of the glass plate and four foot light spots appear, it is determined that the robot is in the current acquisition module. The lens captures the picture and uses the processor to supplement the coordinate system in the picture. By observing the positions of the front and back light spots in the coordinate system, the actual coordinates of the front and back light spots are calculated. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Schematic diagram of the quadruped robot navigation accuracy evaluation method of the present invention.

[0054] Figure 2 Plot the coordinates of the center distance D and the offset angle θ for Example 1 (normal speed).

[0055] Figure 3 Coordinate plot of center distance D and offset angle θ in Example 1 (high speed).

[0056] Figure 4 Schematic diagram of the structure of the quadruped robot navigation coordinate point acquisition device in Example 2.

[0057] Figure 5 This is a schematic diagram of the quadruped robot standing on the collection module in Example 2.

[0058] Figure 6 This is a structural diagram of the acquisition module in Example 2.

[0059] Figure 7 This is a schematic diagram of coordinate generation in Example 2.

[0060] Figure 8 Schematic diagram of generating the light spot coordinates in Example 2.

[0061] Figure 9Schematic diagram of the robot's travel path in the background technology.

[0062] Reference numerals: 1, robot; 10, mobile foot; 11, foot light spot; 2, acquisition module; 20, glass plate; 21, sealing shell; 22, camera; 23, lens; 24, outer frame; DETAILED DESCRIPTION

[0063] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0064] Example 1:

[0065] Example 1 discloses a method for evaluating the navigation accuracy of a quadruped robot, which specifically includes the following steps:

[0066] Step 1) The front end and the rear end of the robot 1 are divided into two picked coordinate points, namely the front coordinate and the rear coordinate, and the midpoint on the line connecting the front end and the rear end is taken to obtain the mid coordinate.

[0067] Step 2) See Figure 1 When robot 1 is at the starting position, it plans its navigation path to the preset position, which generates preset coordinates. The preset coordinates correspond to the following three coordinates:

[0068] Preset front coordinate: front 0 (X 01 ,Y 01 )

[0069] Preset rear coordinate: rear 0 (X 02 ,Y 02 )

[0070] Preset center coordinate: Center 0 (X 03 ,Y 03 ) The preset middle coordinate is the midpoint of the line connecting the preset front and rear coordinates; at the same time, the evaluator marks the three coordinates of the preset position on the plane.

[0071] Step 3) After the robot 1 has traveled, it follows the actual path and arrives at the actual position. The evaluator marks the actual coordinates of the actual position on the plane. The actual coordinates correspond to the following three coordinates:

[0072] Actual front coordinate: front 1 (X 11 ,Y 11 )

[0073] Actual rear coordinate: rear 1 (X 12 ,Y 12 )

[0074] Actual coordinates: Center 1 (X 13,Y 13 ), where the actual mid-coordinate is the midpoint of the line connecting the actual front and rear coordinates.

[0075] Step 4) Collect center distance data: connect center 0 and center 1, measure the distance between center 0 and center 1, and obtain the center distance D between center 0 and center 1, that is, Figure 2 middle Figure 3 The length of the middle D.

[0076] Step 5) Collect the offset angle data: measure or calculate the angle between the straight line formed by the first 0 and the last 0 and the straight line formed by the first 1 and the last 1, and obtain the offset angle θ, that is, Figure 2 middle Figure 3 The angle of θ.

[0077] Step 6) Evaluate the accuracy of the navigation programming and hardware coordination in this activity by counting the center distance D and offset angle θ after robot 1’s navigation action.

[0078] For ease of understanding, the applicant provides the coordinate parameters of the points between the actual position and the preset position during normal movement ( Figure 2 ) and the point coordinate parameters between the actual position and the preset position during high-speed movement ( Figure 3 ).

[0079] Example 2:

[0080] This embodiment is based on Example 1, and the method for measuring the center distance D and the offset angle θ is changed from manual drawing and measurement to using a camera to capture images, and automatically acquiring data through image recognition software to meet the needs of a large number of experimental evaluations.

[0081] See attached Figure 4 In this embodiment, the applicant discloses a navigation coordinate point acquisition device for a quadruped robot 1. The acquisition device can be constructed as follows: Figure 7 The large coordinate system shown in FIG. 1 , the acquisition device includes a plurality of acquisition modules 2 arranged in a matrix, each acquisition module 2 corresponding to a coordinate area (such as Figure 8 shown).

[0082] See also Figure 5 and Figure 6 The acquisition module 2 includes a glass plate 20, a sealing shell 21, a camera 22 and an outer frame 24. The camera 22 is fixed to the bottom of the outer frame 24, and the glass plate 20 is located on the upper surface of the outer frame 24. The camera 22 has a lens 23, and the lens 23 is aimed at the glass plate 20. The sealing shell 21 is used to seal the area between the lens 23 and the glass plate 20. The lens 23 collects the light spot above the glass plate 20 to form a picture with a light spot.

[0083] The glass plate 20 is made of thick explosion-proof glass, mainly used to bear the trampling of the robot 1's moving feet 10 to prevent damage, while the sealing shell 21 protects the lens, etc. The outer frame 24 is a cube, which can allow multiple acquisition modules 2 to be assembled to form a large acquisition matrix.

[0084] A preset laser emitter 25 is provided inside the closed shell 21 , and the preset laser emitter 25 emits multiple laser beams to illuminate the glass plate 20 to display the coordinates of the preset position for reference by the evaluator.

[0085] The surface of the glass plate 20 is provided with a shading film for filtering ambient light. The shading film has a light transmittance greater than 50%, which can filter most daily light, sunlight and reflections from smooth objects, preventing the light spot on the surface of the glass plate 20 hit by the robot 1 from being properly recognized by the lens 23 of the camera 22.

[0086] A weight sensing module is provided under the glass plate 20. The weight sensing module is mainly used to count the weight increase. If the weight suddenly increases by 30-90 kg (the general weight of a robot) in a short period of time, it can be determined that the robot has now moved to the glass plate 20 of the current collection module 2.

[0087] The head end, tail end and four moving legs 10 of the robot 1 are each provided with a laser emitter, and the laser colors emitted by the laser emitters at the head end, tail end and four moving legs 10 are also different.

[0088] In this embodiment, the method for evaluating the navigation accuracy of a quadruped robot includes the following specific steps:

[0089] Step 1) Place the robot 1 on any acquisition module 2 of the navigation coordinate point acquisition device and define this position as the starting position.

[0090] Step 2) When the robot 1 is in the starting position, it plans its navigation path to the preset position, which generates preset coordinates. The preset coordinates correspond to the following three coordinates: the preset middle coordinate is the midpoint of the line connecting the preset front and rear coordinates; and simultaneously, the preset laser emitter 25 irradiates the surface of the glass plate 20 with a laser, forming a light spot to mark the three coordinates of the preset position:

[0091] Preset front coordinate: front 0 (X 01 ,Y 01 )

[0092] Preset rear coordinate: rear 0 (X 02 ,Y 02 )

[0093] Preset center coordinate: Center 0 (X 03 ,Y 03 ).

[0094] Step 3) As robot 1 moves, it follows the actual path and reaches the actual position. Simultaneously, the robot's laser emitter continuously forms multiple light spots on the surface of glass plate 20 . The laser emitters on the head, tail, and locomotion foot 10 use three different colors of laser light, forming light spots of different colors on the surface of glass plate 20: front 1, rear 1, and foot spot 11.

[0095] Step 3.1) Robot 1 reaches the current collection module 2. The weight sensor of collection module 2 detects the weight of robot 1 and preliminarily determines that it is located within the current collection module 2.

[0096] Step 3.2) The camera 22 is started and the lens 23 is used to collect light spots in the current field of view. When four light spots of the same color are found, it is inferred that the robot is located in the current collection module 2 and the camera 22 is started to record the current light spot photo.

[0097] Step 3.3) Using image recognition software, the positions of the light spots are identified from the light spot photos recorded by the camera 22 to obtain the physical positions of the two points, front 1 and back 1.

[0098] Step 3.4) Load the coordinate system in the spot photo according to the number of acquisition module 2 and assign coordinate values ​​to the two points "front 1" and "back 1".

[0099] Step 3.5) Draw a new coordinate system with the line connecting the first 0 and the last 0 as the Y axis and the point where the last 0 is located as the origin. Then generate new coordinate values ​​based on the original coordinate values ​​of the first 1 and the last 1:

[0100] Actual front coordinate: front 1 (X 11 ,Y 11 )

[0101] Actual rear coordinate: rear 1 (X 12 ,Y 12 )

[0102] At the same time, the actual coordinates are calculated: 13 ,Y 13 ), , ;

[0103] Step 4) Calculate the center distance D using the following formula:

[0104] Center distance D=

[0105] Step 5) Calculate the offset angle θ using the following formula:

[0106] Offset angle θ=

[0107] Step 6) Evaluate the accuracy of the navigation programming and hardware coordination in this activity by counting the center distance D and offset angle θ after robot 1’s navigation action.

[0108] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as above in terms of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can, without departing from the scope of the technical solution of the present invention, make some changes or modifications to equivalent embodiments using the technical contents disclosed above. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A navigation coordinate point acquisition device for a quadruped robot (1), characterized by: The invention comprises a plurality of acquisition modules (2) arranged in a matrix, wherein the acquisition module (2) comprises a glass plate (20), a sealing shell (21), a camera (22) and an outer frame (24), wherein the outer frame (24) is in a cubic shape, the camera (22) is fixed to the bottom of the outer frame (24), the glass plate (20) is located on the upper surface of the outer frame (24), the camera (22) is provided with a lens (23), the lens (23) is aligned with the glass plate (20), the sealing shell (21) is used to seal the area between the lens (23) and the glass plate (20), and the lens (23) collects a light spot above the glass plate (20) to form a picture with the light spot.

2. The navigation coordinate point acquisition device for a quadruped robot (1) according to claim 1, characterized in that: The surface of the glass plate (20) is provided with a light-shielding film for filtering ambient light, and the light-shielding film has a light transmittance greater than 50%.

3. The navigation coordinate point acquisition device for a quadruped robot (1) according to claim 1, characterized in that: The collection module (2) further comprises a preset laser emitter (25), which is located below the glass plate (20). A weight sensing module is provided below the glass plate (20).

4. A method for evaluating navigation accuracy of a quadruped robot (1), comprising a starting position, a preset position and an actual position, characterized in that: The specific steps include: Step 1) The front end and the rear end of the robot (1) are divided into two picked coordinate points, namely the front coordinate and the rear coordinate, and the midpoint on the line connecting the front end and the rear end is taken to obtain the mid coordinate; Step 2) When the robot (1) is at the starting position, it plans its navigation path to the preset position, and the preset position generates preset coordinates, which correspond to the following three coordinates: Preset front coordinate: front 0 (X 01 ,Y 01 ) Preset rear coordinate: rear 0 (X 02 ,Y 02 ) Preset center coordinate: Center 0 (X 03 ,Y 03 ) The preset middle coordinate is the midpoint of the line connecting the preset front and rear coordinates; The robot (1) is enabled to move along a preset path to a preset position, and the evaluator marks the three coordinates of the preset position on a plane; Step 3) After the robot (1) moves, it follows the actual path and reaches the actual position. The evaluator marks the actual coordinates of the actual position on the plane. The actual coordinates correspond to the following three coordinates: Actual front coordinate: Front 1 (X 11 ,Y 11 ) Actual rear coordinate: rear 1 (X 12 ,Y 12 ) Actual mid-coordinate: mid-1 (X 13 ,Y 13 ), where the actual mid-coordinate is the midpoint of the line connecting the actual front and rear coordinates; Step 4) Collect center distance data: that is, measure or calculate the distance between center 0 and center 1 to obtain the center distance D between center 0 and center 1; Step 5) Collect the offset angle data: that is, measure or calculate the angle between the straight line formed by the first 0 and the last 0 and the straight line formed by the first 1 and the last 1 to obtain the offset angle θ; Step 6) Evaluate the accuracy of the navigation programming and hardware coordination in this activity by counting the center distance D and the offset angle θ after the robot (1) performs navigation actions; The robot (1) is provided with laser emitters at its head and tail, and the robot (1) is placed on the navigation coordinate point acquisition device of the quadruped robot (1) according to claim 2, and the laser emitters irradiate laser light onto the glass plate (20) of the acquisition module (2), forming light spots 1 in front and 1 in the back; The preset front and rear coordinates are obtained by navigation programming calculation, and the actual front and rear coordinates are obtained by shooting or taking pictures, and by collecting the light spot of the laser transmitter.

5. The method for evaluating navigation accuracy of a quadruped robot (1) according to claim 4, characterized in that: In step 4, the center distance D is calculated by the coordinate values ​​of center 0 and center 1 using a distance calculation formula. The calculation formula is as follows: Center distance D= .

6. The method for evaluating navigation accuracy of a quadruped robot (1) according to claim 4, characterized in that: In step 5, the offset angle θ is calculated by the angle calculation formula using the coordinate values ​​of front 0, back 0, front 1, and back 1. The calculation formula is as follows: Offset angle θ= .

7. The navigation accuracy evaluation method of a quadruped robot (1) according to claim 6, characterized in that: The process of obtaining the coordinate values ​​of 0 and 1 is as follows Step 2.1) The assessor collects the coordinates of the starting position and calculates the preset coordinates of the preset position (X 01 ,Y 01 ) and, after the preset coordinates are 0 (X 02 ,Y 02 ), and record the two coordinate points of the preset position in the plane; Step 3.1) Start the robot (1) and make it move to the actual position according to the navigation programming. The evaluator records the two coordinate points of the actual position; Step 3.2) Draw the coordinate system with the line connecting the preset front coordinate and the preset rear coordinate as the Y axis and the point where the preset rear coordinate is located as the origin. Calculate the preset center coordinate: Center 0 (X 03 ,Y 03 ) , , where Y 02 =0; Step 3.3) Check the actual front coordinate and the actual back coordinate in the coordinate system drawn in step 3.2) and record the two coordinate points of the actual position: front 1 (X 11 ,Y 11 ) and after 1 (X 12 ,Y 12 ); Calculate the actual coordinates: Center 1 (X 13 ,Y 13 ) ， 。 8. The method for evaluating navigation accuracy of a quadruped robot (1) according to any one of claims 4 to 7, characterized in that: The preset front and rear coordinates and the actual front and rear coordinates are both recorded by the evaluator drawing coordinate points in a plane.

9. The method for evaluating navigation accuracy of a quadruped robot (1) according to claim 4, characterized in that: The robot (1) is further provided with four laser emitters on its moving foot (10). The laser emitters on the head, tail and moving foot (10) use three different colors of lasers and form light spots of different colors, namely, front 1, rear 1 and foot light spots (11), on the surface of the glass plate (20). The lens (23) observes the light spots on the surface of the glass plate (20). When four foot light spots (11) appear, it is determined that the robot (1) is in the current acquisition module (2). The lens (23) acquires the image and supplements the coordinate system in the image through the processor. The actual coordinates of the front 1 and rear 1 light spots are calculated based on the positions of the light spots in the coordinate system.

Images