A behavior control-based bionic robot motion control method

Abstract

The application discloses a kind of based on behavior control's bionic robot action control method, it is related to technical bionic robot control field, including the following steps: establishing global terrain coordinate system, completing multi-sensor calibration, robot starts autonomous navigation task, vision sensor starts to collect surrounding environment image;Image is handled, and the color, texture feature in image is extracted;Interference object identification module identifies the interference object therein, and extracts its feature information;Interference index analysis module calculates interference index according to the feature information of interference object;Terrain environment complexity evaluation module real-time collection terrain environment information, evaluates the complexity of terrain environment;Action control decision module comprehensively interference index and terrain environment complexity evaluation result, makes action control decision, and sends control instruction to motion execution mechanism, the application solves the problem of bionic robot in complex environment motion intelligence and safety.

Description

Technical Field

[0001] This invention relates to the field of biomimetic robot control technology, specifically to a biomimetic robot motion control method based on behavior control. Background Technology

[0002] With the continuous development of robotics technology, biomimetic robots are increasingly being used in disaster relief, complex terrain inspection, military reconnaissance, and field operations. Compared to traditional wheeled or tracked mobile robots, biomimetic robots mimic the limb structure and movement patterns of animals, possessing stronger terrain adaptability and flexibility, making them particularly suitable for tasks in complex environments such as rugged, uneven, and obstacle-filled terrains. However, in complex natural environments, robots face multiple challenges, such as obstacle avoidance, path selection, and terrain perception and adaptation. Simply relying on pre-defined rules or path planning algorithms is insufficient to achieve intelligent responses to the environment.

[0003] Currently, some biomimetic robots employ path planning combined with obstacle avoidance control for navigation and control. However, this approach often neglects comprehensive consideration of multiple dimensions, including the type and degree of interference in the environment, as well as the complexity of the terrain itself. For example, when a robot encounters obstacles such as grass, rocks, or fallen leaves, the system often only makes a simple judgment based on the presence or absence of the obstacle, lacking in-depth analysis of the obstacle's specific characteristics and selection of appropriate response strategies. Furthermore, most current biomimetic robot control strategies are based on rigid control logic, lacking biological-like behavioral decision-making capabilities. This makes it difficult to dynamically adjust action strategies according to environmental changes, resulting in low efficiency and safety in uncertain environments. For instance, when encountering obstructed vision or potential pits, traditional robots lack exploratory actions and sensory linkage mechanisms, posing risks such as accidental stepping or falls. Therefore, there is an urgent need for a biomimetic robot control method with behavioral decision-making capabilities, capable of identifying the characteristics of environmental obstacles and assessing their degree of interference, and dynamically adjusting action control strategies based on terrain complexity analysis. Summary of the Invention

[0004] The purpose of this invention is to provide a biomimetic robot motion control method based on behavior control, so as to solve the problems mentioned in the background art.

[0005] To address the aforementioned technical problems, this invention provides the following technical solution: a biomimetic robot motion control method based on behavior control, comprising the following steps:

[0006] Step S1: Establish a global terrain coordinate system, complete multi-sensor calibration, start the robot's autonomous navigation task, and the visual sensor begins to collect images of the surrounding environment;

[0007] Step S2: Process the acquired image to extract color and texture features;

[0008] Step S3: The interference identification module identifies the interference and extracts its feature information;

[0009] Step S4: The interference index analysis module calculates the interference index based on the characteristic information of the interfering object;

[0010] Step S5: The terrain environment complexity assessment module collects terrain environment information in real time and assesses the complexity of the terrain environment;

[0011] Step S6: The motion control decision module integrates the interference index and the terrain complexity assessment results to make motion control decisions and sends control commands to the motion actuator;

[0012] Step S7: The motion actuator drives the joint motors of the robot's four limbs according to the control command to achieve the corresponding actions;

[0013] Step S8: During the robot's movement, repeat the above steps continuously, adjust the motion control strategy in real time, and ensure that the robot can safely and accurately complete the autonomous navigation task.

[0014] According to the above technical solution, the specific methods for processing the acquired image and extracting color and texture features in step S2 include:

[0015] Step S21: Convert the color image captured by the camera into a grayscale image. Use the weighted average method for grayscale processing. The calculation expression is: Gray = 0.299R + 0.587G + 0.114B, where R, G, and B are the pixel values ​​of the red, green, and blue channels in the color image, respectively, and Gray is the pixel value of the grayscale image.

[0016] Step S22: Denoise the grayscale image using a Gaussian filtering algorithm. Let the Gaussian kernel be G(x,y), and its calculation expression is: Where σ is the standard deviation of the Gaussian kernel, and (x,y) are the coordinates of the pixel. The Gaussian kernel is convolved with the image to obtain the denoised image.

[0017] Step S23: Convert the grayscale image to the HSV color space. The conversion formula is as follows:

[0018] V = max(R,G,B);

[0019]

[0020] Step S24: Perform histogram statistics on the hue components in the HSV color space, divide the hue value range into several regions, count the number of pixels in each region, obtain the color histogram, and extract the main color features in the image based on the color histogram.

[0021] Step S25: For each pixel in the image, select a 3×3 neighborhood centered on it. Use the gray value of the center pixel of the neighborhood as a threshold and compare it with the gray values ​​of the surrounding 8 neighboring pixels. If the gray value of the neighboring pixel is greater than or equal to the gray value of the center pixel, then the position of the neighboring pixel is marked as 1; otherwise, it is marked as 0. Repeat this step until each pixel has an 8-bit binary number, which is converted into a decimal number and used as the LBP value of the pixel.

[0022] Step S26: Statistically analyze the distribution of different LBP values ​​in the image to obtain the LBP histogram; and extract the image texture features based on the LBP histogram.

[0023] According to the above technical solution, in step S3, the specific method for the interference identification module to identify the interference and extract its feature information is as follows:

[0024] Step S31: Obtain the image after image preprocessing, color feature extraction, and texture feature extraction;

[0025] Step S32: Establish a template database of interfering objects, and organize and summarize the samples of each template in the template database under different angles and lighting conditions;

[0026] Step S33: Using a template matching algorithm, the image is divided into multiple sub-regions. Each sub-region in the recognition image is compared with a template in the template library. The similarity between the sub-region and the template is calculated. The similarity is measured using the normalized cross-correlation coefficient, and the formula is as follows: Where I(x,y) represents the pixel value of the sub-region. Let T(x,y) be the average pixel value of the sub-region, and T(x,y) be the pixel value of the template. The average pixel value of the template; when the similarity exceeds the threshold G. u When the sub-region is identified as containing interfering objects, the interfering objects are initially classified based on the similarity matching results.

[0027] Step S34: For the identified interference areas, extract their feature information; the feature information includes the color features, texture features, and position and size features of the interference in the image.

[0028] According to the above technical solution, in step S4, the specific method by which the interference index analysis module calculates the interference index based on the characteristic information of the interfering object is as follows:

[0029] Step S41: Assign corresponding weights to each feature based on the degree of influence of the interfering object on the robot's autonomous navigation task;

[0030] Step S42: Calculate the interference index using the formula... The calculated interference exponent Q is given by ω, where Q represents the interference exponent. i f represents the weight of the i-th feature. i Let represent the value of the i-th feature, and n represent the total number of features;

[0031] Step S43: Evaluate the degree of influence of the interfering object based on the calculated interference index Q; set the interference index threshold Q0 and classify the interfering object into different levels. When Q≤Q0, it is judged as low interference, and when Q>Q0, it is judged as high interference.

[0032] According to the above technical solution, in step S5, the terrain environment complexity assessment module collects terrain environment information in real time, and the specific workflow for assessing the complexity of the terrain environment includes the following steps:

[0033] Step S51: Obtain the image information processed in the previous steps, including the color histogram, LBP histogram, and interference identification results;

[0034] Step S52: Extract the color histogram entropy H according to the preset algorithm. c The complexity index W of the texture corresponding to the LBP histogram f ;

[0035] Step S53: Using the terrain environment complexity index formula F d =αH c +βW f The terrain environment complexity index F was calculated. d , where α and β are the control parameters for the color histogram entropy and the complexity index of the texture corresponding to the LBP histogram, respectively;

[0036] Step S54: Determine the terrain and environmental complexity index F d >F d0 When the current situation is judged to be a complex terrain environment, the terrain environment complexity index F is used. d ≤F d0 At that time, it is determined that the current terrain environment is simple.

[0037] According to the above technical solution, the specific action control decision in step S6 is as follows:

[0038] When the interference index is high and the terrain is complex, the robot chooses to detour.

[0039] When the interference index is low and the terrain is simple, the robot can directly cross or step on the interfering objects.

[0040] When the interference index is low but the terrain is complex, further analysis is needed to determine whether the interference objects may be blocking potential pit hazards. If potential pit hazards are identified, proceed with caution; otherwise, proceed normally.

[0041] According to the above technical solution, the specific method for analyzing and determining whether interference objects obscure potential pit hazards is as follows:

[0042] Step A: Based on the depth map data from the visual sensor, the gradient thresholding method is used to extract areas of abrupt changes in terrain elevation. If the depth difference between adjacent pixels exceeds ΔD, it is marked as a suspected pit boundary.

[0043] Step B: Combine infrared thermal imaging data to analyze areas of abnormal surface temperature. If the temperature gradient distribution differs from the surrounding terrain by more than ΔT, it is identified as a potential pit.

[0044] Step C: Project the detected pit features onto the global terrain coordinate system and record the pit center coordinate set {C}. k (x k ,y k ,z k )}, where k = 1, 2, ..., n;

[0045] Step D: Calculate the three-dimensional spatial distance between adjacent pits. Generate a distance matrix by traversing all pit combinations;

[0046] Step E: Calculate the pit distribution density index based on the distance matrix When ρ>ρ0, it is necessary to consider the potential pitting hazard caused by interference objects blocking the view, where ρ0 is a preset pitting distribution density threshold.

[0047] According to the above technical solution, the method of handling the situation cautiously if a potential pitting hazard is determined to exist is as follows:

[0048] Control the robot's feet to make tentative contact with the edge of the obstacle;

[0049] Real-time monitoring of ground reaction force fluctuations using a six-dimensional force sensor, if the vertical stiffness coefficient K v <K v0 This indicates a risk of false support.

[0050] When the surface vibration spectrum is detected by combining a piezoelectric thin film sensor, and the shift of the dominant frequency component Δf > Δf max Confirm the presence of the pit.

[0051] Compared with the prior art, the beneficial effects achieved by the present invention are: the present invention,

[0052] (1) By introducing a joint evaluation mechanism of interference index and terrain complexity index, the robot can dynamically adjust its action strategy according to environmental changes, and realize diverse behavior choices such as crossing, detouring or tentative passage, which significantly reduces the risk of misjudging obstacles or falling.

[0053] (2) By using image processing techniques such as Gaussian filtering, HSV color space analysis, and LBP texture coding, more discriminative visual features are extracted, which effectively improves the accuracy of identifying small obstacles or disguised interference.

[0054] (3) By integrating multi-dimensional perception information such as image color, texture feature extraction, interference identification and feature analysis, and terrain environment complexity assessment, the bionic robot is given a complete control loop of "perception-analysis-decision-execution", which significantly improves its adaptability in unknown and complex environments. Attached Figure Description

[0055] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0056] In the attached diagram:

[0057] Figure 1 This is a flowchart of a biomimetic robot motion control method based on behavior control according to the present invention. Detailed Implementation

[0058] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0059] Please see Figure 1 This invention provides a technical solution: a biomimetic robot motion control method based on behavior control, comprising the following steps:

[0060] Step S1: Establish a global terrain coordinate system, complete multi-sensor calibration, start the robot's autonomous navigation task, and the visual sensor begins to collect images of the surrounding environment;

[0061] Step S2: Process the acquired image to extract color and texture features;

[0062] Specific methods for processing acquired images and extracting color and texture features include:

[0063] Step S21: Convert the color image captured by the camera into a grayscale image. A weighted average method is used for grayscale processing, calculated as: Gray = 0.299R + 0.587G + 0.114B, where R, G, and B are the pixel values ​​of the red, green, and blue channels in the color image, respectively, and Gray is the pixel value of the grayscale image. This reduces the amount of image data while preserving the main brightness information, facilitating subsequent processing.

[0064] Step S22: Denoise the grayscale image using a Gaussian filtering algorithm. Let the Gaussian kernel be G(x,y), and its calculation expression is: Where σ is the standard deviation of the Gaussian kernel, and (x,y) are the coordinates of the pixel. The Gaussian kernel is convolved with the image to obtain the denoised image.

[0065] Step S23: Convert the grayscale image to the HSV color space. The HSV color space consists of three components: hue, saturation, and lightness. Compared to the RGB color space, it better matches human color perception. The conversion formula is as follows:

[0066] V = max(R,G,B);

[0067]

[0068] Step S24: Perform histogram statistics on the hue components in the HSV color space, divide the hue value range into several regions, count the number of pixels in each region, obtain the color histogram, and extract the main color features in the image based on the color histogram.

[0069] Step S25: For each pixel in the image, select a 3×3 neighborhood centered on it. Use the gray value of the center pixel of the neighborhood as a threshold and compare it with the gray values ​​of the surrounding 8 neighboring pixels. If the gray value of the neighboring pixel is greater than or equal to the gray value of the center pixel, then the position of the neighboring pixel is marked as 1; otherwise, it is marked as 0. Repeat this step until each pixel has an 8-bit binary number, which is converted into a decimal number and used as the LBP value of the pixel.

[0070] Step S26: Statistically analyze the distribution of different LBP values ​​in the image to obtain the LBP histogram; and extract the image texture features based on the LBP histogram.

[0071] By introducing grayscale processing, Gaussian filtering for noise reduction, HSV color space conversion, and a texture feature extraction method based on Local Binary Pattern (LBP), a complete, stable, and efficient image preprocessing and feature extraction workflow is constructed. This method can not only accurately extract color and texture information from images in complex environments, but also has good noise resistance and computational efficiency, significantly improving the perception accuracy and robustness of bionic robots. Compared with existing image processing technologies, the HSV color space used in this invention is closer to human visual perception, and the LBP texture descriptor has a stronger ability to express local features. This enables the system to maintain stable operation when facing challenges such as changes in lighting and complex terrain, and provides a solid data foundation for subsequent behavioral decisions, demonstrating significant practical value and promising prospects for widespread application.

[0072] Step S3: The interference identification module identifies the interference and extracts its feature information;

[0073] The specific method by which the interference identification module identifies interference and extracts its feature information is as follows:

[0074] Step S31: Obtain the image after image preprocessing, color feature extraction, and texture feature extraction;

[0075] Step S32: Establish a template database of interfering objects, and organize and summarize the samples of each template in the template database under different angles and lighting conditions;

[0076] Step S33: Using a template matching algorithm, the image is divided into multiple sub-regions. Each sub-region in the recognition image is compared with a template in the template library. The similarity between the sub-region and the template is calculated. The similarity is measured using the normalized cross-correlation coefficient, and the formula is as follows: Where I(x,y) represents the pixel value of the sub-region. Let T(x,y) be the average pixel value of the sub-region, and T(x,y) be the pixel value of the template. The average pixel value of the template; when the similarity exceeds the threshold G. u When the sub-region is identified as containing interfering objects, the interfering objects are initially classified according to the similarity matching results, such as large rocks, plastic bags, dead branches and leaves, etc.

[0077] Step S34: For the identified interference areas, extract their feature information; the feature information includes the color features of the interference (obtained from the color histogram), texture features (obtained from the LBP histogram), and the position and size features of the interference in the image;

[0078] By constructing a database of interference templates and combining preprocessing results of color and texture features, a template matching algorithm is introduced for region-level similarity analysis, thereby effectively identifying the types and distribution characteristics of interference in images. Using the normalized cross-correlation coefficient (NCC) as a similarity metric, interference can be accurately identified under complex environments with varying lighting and angles, and preliminarily classified into categories such as large rocks, plastic bags, or withered branches and leaves. Based on this, further feature information such as color, texture, location, and size of the interference is extracted, providing accurate and comprehensive environmental cognition for the obstacle avoidance behavior and path planning of bionic robots in complex environments. Compared to traditional methods based on single features or coarse detection, the method of this invention has higher recognition accuracy and environmental adaptability, significantly enhancing the ability of bionic robots to handle interference factors in dynamic natural environments and improving their intelligence level.

[0079] Step S4: The interference index analysis module calculates the interference index based on the characteristic information of the interfering object;

[0080] The interference index analysis module calculates the interference index based on the characteristic information of the interfering object using the following method:

[0081] Step S41: Assign corresponding weights to each feature based on the degree of influence of the interfering object on the robot's autonomous navigation task;

[0082] Step S42: Calculate the interference index using the formula... The calculated interference exponent Q is given by ω, where Q represents the interference exponent. i f represents the weight of the i-th feature. i Let represent the value of the i-th feature, and n represent the total number of features;

[0083] Step S43: Evaluate the degree of influence of the interfering object based on the calculated interference index Q; set the interference index threshold Q0 and classify the interfering object into different levels. When Q≤Q0, it is judged as low interference, and when Q>Q0, it is judged as high interference.

[0084] Step S5: The terrain environment complexity assessment module collects terrain environment information in real time and assesses the complexity of the terrain environment;

[0085] The terrain environment complexity assessment module collects terrain environment information in real time. The specific workflow for assessing the complexity of the terrain environment includes the following steps:

[0086] Step S51: Obtain the image information processed in the previous steps, including the color histogram, LBP histogram, and interference identification results;

[0087] Step S52: Extract the color histogram entropy H according to the preset algorithm. cThe complexity index W of the texture corresponding to the LBP histogram f ;

[0088] Step S53: Using the terrain environment complexity index formula F d =αH c +βW f The terrain environment complexity index F was calculated. d , where α and β are the control parameters for the color histogram entropy and the complexity index of the texture corresponding to the LBP histogram, respectively;

[0089] Step S54: Determine the terrain and environmental complexity index F d >F d0 When the current situation is judged to be a complex terrain environment, the terrain environment complexity index F is used. d ≤F d0 When assessing terrain complexity, the system first determines if the current terrain environment is simple. Then, it integrates color histograms and texture feature (LBP histogram) information from the image, introducing two quantitative indicators: color entropy and texture complexity. A weighted calculation model is constructed to form a terrain complexity index. This index, by setting weight parameters α and β, comprehensively reflects the multidimensional complexity of the current terrain environment. By comparing it with a threshold, it achieves intelligent judgment of environmental complexity. This method can assess the changes and complexity of the terrain in which the robot is located in real time, providing an important basis for adjusting subsequent motion control strategies and significantly improving the robot's adaptability and autonomous behavior decision-making ability in dynamic and complex environments.

[0090] Step S6: The motion control decision module integrates the disturbance index and the terrain complexity assessment results to make motion control decisions and sends control commands to the motion actuator;

[0091] The specific action control decision is as follows:

[0092] When the interference index is high and the terrain is complex, the robot chooses to detour.

[0093] When the interference index is low and the terrain is simple, the robot can directly cross or step on the interfering objects.

[0094] When the interference index is low but the terrain is complex, further analysis is needed to determine whether the interference objects may be blocking potential pit hazards. If potential pit hazards are identified, proceed with caution; otherwise, proceed normally.

[0095] The specific method for analyzing and determining whether interference objects may obscure the potential pitting hazard is as follows:

[0096] Step A: Based on the depth map data from the visual sensor, the gradient thresholding method is used to extract areas of abrupt changes in terrain elevation. If the depth difference between adjacent pixels exceeds ΔD, it is marked as a suspected pit boundary.

[0097] Step B: Combine infrared thermal imaging data to analyze areas of abnormal surface temperature. If the temperature gradient distribution differs from the surrounding terrain by more than ΔT, it is identified as a potential pit.

[0098] Step C: Project the detected pit features onto the global terrain coordinate system and record the pit center coordinate set {C}. k (x k ,y k ,z k )}, where k = 1, 2, ..., n;

[0099] Step D: Calculate the three-dimensional spatial distance between adjacent pits. Generate a distance matrix by traversing all pit combinations;

[0100] Step E: Calculate the pit distribution density index based on the distance matrix When ρ>ρ0, it is necessary to consider the potential pitting hazard caused by interference objects blocking the view, where ρ0 is a preset pitting distribution density threshold.

[0101] If a potential pitting hazard is identified, the following cautious approach should be taken:

[0102] Control the robot's feet to make tentative contact with the edge of the obstacle;

[0103] Real-time monitoring of ground reaction force fluctuations using a six-dimensional force sensor, if the vertical stiffness coefficient K v <K v0 This indicates a risk of false support.

[0104] When the surface vibration spectrum is detected by combining a piezoelectric thin film sensor, and the shift of the dominant frequency component Δf > Δf max The system confirms the existence of pits in real time. For motion control decision-making, an intelligent strategy system based on a joint assessment of disturbance index and terrain complexity is constructed. To this end, the system extracts abrupt terrain change areas through visual depth maps, detects surface temperature anomalies using infrared thermal imaging, identifies pit locations, performs spatial analysis, and calculates pit distribution density to determine occlusion risk. Once a potential pit is identified, the robot executes a cautious handling strategy, including tentative foot contact with the ground, six-dimensional force sensors to detect support stability, and piezoelectric film sensors to monitor the surface vibration spectrum, to confirm the existence of false supports or pits, effectively improving the robot's stability and safety in uncertain terrain.

[0105] Step S7: The motion actuator drives the robot's limb joint motors according to the control instructions to achieve corresponding actions, such as walking, crossing, and circling.

[0106] Step S8: During the robot's movement, repeat the above steps continuously, adjust the motion control strategy in real time, and ensure that the robot can safely and accurately complete the autonomous navigation task.

[0107] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0108] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0109] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0110] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A biomimetic robot motion control method based on behavior control, characterized in that: Includes the following steps: Step S1: Establish a global terrain coordinate system, complete multi-sensor calibration, start the robot's autonomous navigation task, and the visual sensor begins to collect images of the surrounding environment; Step S2: Process the acquired image to extract color and texture features; Step S3: The interference identification module identifies the interference and extracts its feature information; Step S4: The interference index analysis module calculates the interference index based on the characteristic information of the interfering object; Step S5: The terrain environment complexity assessment module collects terrain environment information in real time and assesses the complexity of the terrain environment; Step S6: The motion control decision module integrates the disturbance index and the terrain complexity assessment results to make motion control decisions and sends control commands to the motion actuator; Step S7: The motion actuator drives the joint motors of the robot's four limbs according to the control command to achieve the corresponding actions; Step S8: During the robot's movement, repeat the above steps continuously and adjust the motion control strategy in real time to ensure that the robot can safely and accurately complete the autonomous navigation task; In step S4, the interference index analysis module calculates the interference index based on the characteristic information of the interfering object using the following specific method: Step S41: Assign corresponding weights to each feature based on the degree of influence of the interfering object on the robot's autonomous navigation task; Step S42: Calculate the interference index using the formula... Calculated interference index ,in Indicates the interference index. This represents the weight of the i-th feature. Let represent the value of the i-th feature, and n represent the total number of features; Step S43: Based on the calculated interference index The impact of interfering objects is assessed; an interference index threshold is set. Interference is classified into different levels, when It is judged to be low interference, when At that time, it was judged as high interference; In step S5, the terrain environment complexity assessment module collects terrain environment information in real time. The specific workflow for assessing the complexity of the terrain environment includes the following steps: Step S51: Obtain the image information processed in the previous steps, including the color histogram, LBP histogram, and interference identification results; Step S52: Extract color histogram entropy according to the preset algorithm. The complexity index of the texture corresponding to the LBP histogram ; Step S53: Using the terrain environment complexity index formula The terrain environment complexity index was calculated. ,in , These are the control parameters for the complexity indices of the textures corresponding to the color histogram entropy and the LBP histogram, respectively. Step S54: Determine the complexity index of the local terrain environment When the current situation is judged to be complex, the local terrain complexity index is used. At that time, it is determined that the current terrain environment is simple.

2. The biomimetic robot motion control method based on behavior control according to claim 1, characterized in that: In step S2, the specific methods for processing the acquired image and extracting color and texture features from the image include: Step S21: Convert the color image captured by the camera into a grayscale image. The grayscale conversion is performed using a weighted average method, and the calculation expression is as follows: ,in These are the pixel values ​​of the red, green, and blue channels in a color image, respectively. These are the pixel values ​​of a grayscale image; Step S22: Use a Gaussian filtering algorithm to reduce noise in the grayscale image, setting the Gaussian kernel as... Its calculation expression is: ,in The standard deviation of the Gaussian kernel. Using the pixel coordinates, a Gaussian kernel is convolved with the image to obtain the denoised image; Step S23: Convert the grayscale image to the HSV color space. The conversion formula is as follows: ； ； ； Step S24: Perform histogram statistics on the hue components in the HSV color space, divide the hue value range into several regions, count the number of pixels in each region, obtain the color histogram, and extract the main color features in the image based on the color histogram. Step S25: For each pixel in the image, select a [pixel name] centered on it. The gray value of the center pixel of the neighborhood is used as a threshold and compared with the gray values ​​of the surrounding 8 neighboring pixels. If the gray value of a neighboring pixel is greater than or equal to the gray value of the center pixel, the position of the neighboring pixel is marked as 1; otherwise, it is marked as 0. This step is repeated until each pixel has an 8-bit binary number, which is then converted into a decimal number as the LBP value of that pixel. Step S26: Statistically analyze the distribution of different LBP values ​​in the image to obtain the LBP histogram; and extract the image texture features based on the LBP histogram.

3. The biomimetic robot motion control method based on behavior control according to claim 1, characterized in that: In step S3, the specific method for the interference identification module to identify the interference and extract its feature information is as follows: Step S31: Obtain the image after image preprocessing, color feature extraction, and texture feature extraction; Step S32: Establish a template database of interfering objects, and organize and summarize the samples of each template in the template database under different angles and lighting conditions; Step S33: Using a template matching algorithm, the image is divided into multiple sub-regions. Each sub-region in the recognition image is compared with a template in the template library. The similarity between the sub-region and the template is calculated. The similarity is measured using the normalized cross-correlation coefficient, and the formula is as follows: ,in The pixel value of the sub-region The average pixel value of the sub-region. The pixel values ​​of the template. The average pixel value of the template; when the similarity exceeds a threshold. When the sub-region is identified as containing interfering objects, the interfering objects are initially classified based on the similarity matching results. Step S34: Extract feature information from the identified interference areas. The feature information includes the color features, texture features, and position and size features of the interference in the image.

4. The biomimetic robot motion control method based on behavior control according to claim 1, characterized in that: The specific action control decision in step S6 is as follows: When the interference index is high and the terrain is complex, the robot chooses to detour. When the interference index is low and the terrain is simple, the robot can directly cross or step on the interfering objects. When the interference index is low but the terrain is complex, further analysis is needed to determine whether the interference objects may be blocking potential pit hazards. If potential pit hazards are identified, proceed with caution; otherwise, proceed normally.

5. The biomimetic robot motion control method based on behavior control according to claim 4, characterized in that: The specific method for analyzing and determining whether interference objects may obscure the potential pitting hazard is as follows: Step A: Based on the depth map data from the visual sensor, the gradient thresholding method is used to extract areas of abrupt changes in terrain elevation. If the depth difference between adjacent pixels exceeds [a certain threshold], [the region is considered a transitional region]. Marked as the boundary of a suspected pit; Step B: Combine infrared thermal imaging data to analyze areas of abnormal surface temperature. If the temperature gradient distribution differs from the surrounding terrain by more than [a certain percentage], [further analysis is needed]. It was determined to be a potential pit; Step C: Project the detected pit features onto the global terrain coordinate system and record the pit center coordinate set. where k = 1, 2, ..., n; Step D: Calculate the three-dimensional spatial distance between adjacent pits. Iterate through all pit combinations to generate a distance matrix; Step E: Calculate the pit distribution density index based on the distance matrix ,when When making a judgment, it is necessary to consider the potential pitting hazard caused by obstructions. This is the preset threshold for the density of pit distribution.

6. The biomimetic robot motion control method based on behavior control according to claim 4, characterized in that: If a potential pitting hazard is identified, the following cautious approach should be taken: Control the robot's feet to make tentative contact with the edge of the obstacle; Real-time monitoring of ground reaction force fluctuations using a six-dimensional force sensor, if the vertical stiffness coefficient... This indicates a risk of false support. Combining piezoelectric thin-film sensors to detect the surface vibration spectrum, when the dominant frequency component shifts... Confirm the presence of the pit.

Images