Multi-level hybrid obstacle avoidance control method adaptive to environmental complexity

Abstract

The invention discloses a multi-level hybrid obstacle avoidance control method adaptive to environment complexity, and relates to the technical field of robot control, and the method comprises the steps: obtaining the global map information of an inspection environment, dividing a map into grid units, marking static obstacles, and calculating a global path of an inspection robot from a starting position to a target position through an A * algorithm; position and speed information of a dynamic obstacle is obtained in real time through a laser radar, a gravitational field model and a repulsive force field model are established, a gravitational field is generated by a target point, and a repulsive force field is generated by the detected dynamic obstacle. According to the invention, by establishing an obstacle classification mechanism based on the distance threshold, quantitative evaluation of the threat degree is realized, and the problem of low control efficiency caused by adopting a unified processing strategy for all obstacles in a traditional method is avoided.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a multi-level hybrid obstacle avoidance control method that is adaptive to environmental complexity. Background Art

[0002] Obstacle avoidance control of mobile robots in complex dynamic environments has always been a research focus and difficulty. Traditional obstacle avoidance methods mainly include artificial potential field method, dynamic window method and speed obstacle method. The artificial potential field method achieves obstacle avoidance by constructing gravitational field and repulsive field. It has the advantages of simple calculation and strong real-time performance, but it is easy to fall into local minimum. The dynamic window method performs trajectory planning under the robot kinematic constraints, which can ensure the executability of control instructions, but its obstacle avoidance performance is significantly affected by the speed sampling resolution. The speed obstacle method performs trajectory planning based on speed-time space, which can effectively handle dynamic obstacles, but the computational complexity is high and is not conducive to real-time control. When dealing with multi-obstacle scenes, these methods often adopt a unified obstacle avoidance strategy, which makes it difficult to perform differentiated control according to the threat level of the obstacles.

[0003] Existing obstacle avoidance control methods generally have the following shortcomings: First, there is a lack of quantitative assessment of the threat level of obstacles, which leads to overly conservative or aggressive obstacle avoidance control; second, the obstacle avoidance control strategy is relatively fixed, and it is difficult to adopt differentiated control for obstacles of different threat levels; third, the collision risk warning mechanism is not perfect, and it is impossible to identify potential collision risks in time and respond. These problems limit the robot's obstacle avoidance performance in complex dynamic environments, affecting its actual application effect. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the present invention provides a multi-level hybrid obstacle avoidance control method that is adaptive to environmental complexity and can solve the problems mentioned in the background technology.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a multi-level hybrid obstacle avoidance control method that is adaptive to environmental complexity, comprising: obtaining global map information of the inspection environment, dividing the map into grid units and marking static obstacles, and using the A* algorithm to calculate the global path of the inspection robot from the starting position to the target position;

[0007] The position and velocity information of dynamic obstacles are acquired in real time by laser radar, and a gravitational field and a repulsive field model are established, wherein the gravitational field is generated by the target point and the repulsive field is generated by the detected dynamic obstacles;

[0008] The movement direction and speed of the inspection robot are calculated according to the synthetic force field of the gravitational field and the repulsive field. When a dynamic obstacle is detected, the local path is updated and the inspection robot is controlled to move along the updated path.

[0009] As a preferred solution of the multi-level hybrid obstacle avoidance control method for adaptive environment complexity described in the present invention, in which: in the gravitational field and repulsive field model, a dynamic gravitational field function is constructed by the obstacle motion trend term:

[0010]

[0011] Where α is the velocity direction influence factor, v obs is the obstacle velocity vector, is the unit vector direction from the robot to the target point; k att is the gravitational field gain coefficient; p is the current position coordinate of the robot; p goal is the target position coordinate;

[0012] The obstacle motion trend item is adjusted by determining the angle θ between the obstacle motion direction and the line connecting the robot to the target point:

[0013] If θ<π / 2, then α=α0cosθ; if θ≥π / 2, then α=0;

[0014] Among them, α0 is the benchmark impact factor;

[0015] The adaptive repulsive field function is dynamically adjusted according to the motion characteristics of the obstacle:

[0016]

[0017] Where γ is the weight coefficient of motion trend, and the relative speed of the obstacle Calculated; k rep is the repulsive field gain coefficient, d is the distance from the robot to the obstacle, d0 is the influence range of the repulsive field, p obs is the obstacle position coordinate.

[0018] As a preferred solution of the multi-level hybrid obstacle avoidance control method for adaptive environment complexity described in the present invention, wherein: in the calculation of the synthetic force field, the environment in which the robot is located is judged;

[0019] If it is located in a single obstacle environment, the resultant force is calculated directly:

[0020] F total =F att +F rep

[0021] If the environment is full of obstacles, the weight is calculated based on the threat level of the obstacles:

[0022]

[0023] Where λ is the distance attenuation coefficient, β is the speed influence coefficient, and n is the number of obstacles; is the relative velocity vector of the robot relative to the i-th obstacle;

[0024] Calculate the weighted resultant force:

[0025]

[0026] Determine the robot's movement speed based on the direction and magnitude of the resultant force:

[0027]

[0028] where k v is the speed adjustment coefficient, v max is the maximum permissible speed; k v is the speed adjustment coefficient, v max is the maximum permissible speed.

[0029] As a preferred solution of the multi-level hybrid obstacle avoidance control method for adaptive environment complexity described in the present invention, the local path update process includes constructing an obstacle motion prediction model. Specifically, if the obstacle motion direction change rate is less than a threshold value ω th , then the uniform linear motion model is used for prediction; if the obstacle movement direction change rate is greater than or equal to the threshold ω th , then the curve fitting prediction model is adopted;

[0030] Perform path conflict detection. Specifically, if the minimum distance between the predicted trajectory and the current path is less than the safety threshold d safe , and the time reserved is less than t safe , then the path update is triggered; otherwise, the current path remains unchanged;

[0031] A spline curve is generated in the obstacle avoidance area. The curve parameters are determined by the current posture and target posture of the robot. The optimal path is selected by minimizing the weighted sum of the curvature integral and the path length.

[0032] As a preferred solution of the multi-level hybrid obstacle avoidance control method for adaptive environment complexity described in the present invention, the fusion strategy of the global path and the local obstacle avoidance is realized by judging the environment complexity.

[0033] In the path planning process, if the environmental complexity index C e satisfy:

[0034]

[0035] where vi is the speed of the ith obstacle, d i is the distance to the ith obstacle, n is the number of obstacles within the field of view; C th is the environment complexity threshold, then switch to local obstacle avoidance mode:

[0036]

[0037] where t p is the prediction time domain, v robot is the robot velocity calculated from the synthetic force field; P current is the current position of the robot;

[0038] If C e ≤C th , then keep the global path following mode:

[0039]

[0040] Where h(p) is the heuristic distance from the path point to the target point; P global is a set of global path points.

[0041] As a preferred solution of the multi-level hybrid obstacle avoidance control method for adaptive environment complexity described in the present invention, the environment perception layer classifies obstacles by distance judgment. Specifically,

[0042] If the obstacle distance is less than d min , classifying it as a high-threat obstacle;

[0043] If the obstacle distance is between d min With d max between them, classifying them as potential threatening obstacles;

[0044] The motion control layer constrains the robot speed according to the obstacle threat level.

[0045]

[0046] where v cmd is the expected speed, k s is the safety factor; d is the current obstacle distance.

[0047] As a preferred solution of the multi-level hybrid obstacle avoidance control method for adaptive environment complexity described in the present invention, it also includes a collision warning mechanism for judging according to the time margin:

[0048]

[0049] When the collisionWhen it is less than the safety time threshold, the deceleration and obstacle avoidance control is triggered; is the unit vector from the robot to the obstacle.

[0050] To further solve the above technical problems, the present invention provides the following technical solutions: A system for calibrating the angle of a circular coal yard stacker and reclaimer based on microwave recognition, comprising: an environment perception module for classifying obstacles as high-threat obstacles or potential-threat obstacles based on distance threshold judgment;

[0051] The speed control module is used to constrain the robot speed according to the obstacle threat level;

[0052] The early warning monitoring module is used to determine whether to trigger deceleration and obstacle avoidance control based on the time margin.

[0053] A computer device includes a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the multi-level hybrid obstacle avoidance control method that is adaptive to the complexity of the environment as described above are implemented.

[0054] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the multi-level hybrid obstacle avoidance control method that is adaptive to the complexity of the environment as described above are implemented.

[0055] The beneficial effects of the present invention are as follows: by establishing an obstacle classification mechanism based on a distance threshold, a quantitative assessment of the threat level is achieved, avoiding the problem of low control efficiency caused by the use of a unified processing strategy for all obstacles in the traditional method; by introducing key parameters such as obstacle distance and robot speed into the dynamic speed constraint formula, a quantitative relationship between the threat level and the control strategy is established, providing a more accurate speed regulation mechanism compared to the simple linear speed constraint in the prior art; by introducing a collision warning criterion based on a time margin, the relative motion characteristics of the obstacle and the robot are taken into account, overcoming the hysteresis that may be caused by relying solely on the distance threshold judgment, and improving the accuracy and timeliness of the warning. While ensuring safety, the control efficiency is improved, enabling the robot to better adapt to complex dynamic environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0057] Figure 1 is a flow chart of the method in the present invention;

[0058] Figure 2 This is a diagram of a computer device in the present invention. DETAILED DESCRIPTION

[0059] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0060] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0061] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a multi-level hybrid obstacle avoidance control method that is adaptive to environmental complexity.

[0062] Figure 1 The overall flow chart of a multi-level hybrid obstacle avoidance control method with adaptive environment complexity is shown, including: S1: obtaining global map information of the inspection environment, dividing the map into grid units and marking static obstacles, and using the A* algorithm to calculate the global path of the inspection robot from the starting position to the target position;

[0063] S2: Obtain the position and velocity information of dynamic obstacles in real time through laser radar, and establish a gravitational field and repulsive field model, where the gravitational field is generated by the target point and the repulsive field is generated by the detected dynamic obstacles;

[0064] S3: Calculate the movement direction and speed of the inspection robot according to the synthetic force field of the gravitational field and the repulsive field, update the local path when a dynamic obstacle is detected, and control the inspection robot to move along the updated path.

[0065] It should be noted that this method, through the organic combination of global path planning and local obstacle avoidance, focuses on solving the technical problems of low path planning efficiency, single obstacle avoidance strategy, and poor real-time performance of traditional methods in complex dynamic environments. The specific implementation steps are as follows:

[0066] In step S1, the global map information of the inspection environment is first obtained. The environment is divided into grid units of size 0.1m×0.1m using a gridding method. The size is selected based on the size of the inspection robot (0.6m×0.8m) and the minimum turning radius (0.8m). When marking static obstacles, the obstacle expansion coefficient of 1.2 is considered to ensure that the planned path maintains a sufficiently safe distance from the static obstacles. On this basis, the improved A* algorithm is used to calculate the global path. The improvement is mainly reflected in the design of the heuristic function, which takes into account both the shortest distance and avoiding areas with dense obstacles.

[0067] In step S2, a laser radar is used to obtain dynamic obstacle information in real time, with a scanning frequency of 10 Hz, a ranging range of 0.1-30 m, and an angular resolution of 0.25°. An improved potential field model is established based on the detected obstacle information. The characteristics of this model are that both the gravitational field and the repulsive field take into account the motion characteristics of the obstacle, and the predictability and adaptability of obstacle avoidance are improved by introducing velocity coupling terms and motion trend weights.

[0068] In step S3, the system calculates the motion instructions based on the synthetic force field and updates the local path. A hierarchical processing strategy is adopted: collision detection and emergency obstacle avoidance are performed at a frequency of 5Hz to ensure safety; trajectory smoothing and local optimization are performed at a frequency of 2Hz to ensure the continuity of the motion trajectory. The local path is generated using a cubic spline curve, which ensures that the curvature continuity reaches the C2 level and significantly improves the smoothness of the robot's movement.

[0069] Through the implementation of the above steps, this solution has shown significant advantages in complex dynamic environments: the global path planning time is reduced by 44% compared with the traditional A* algorithm; the dynamic obstacle avoidance success rate is increased by 11.4 percentage points; and the trajectory smoothness is increased by 40%. These improvements make this solution particularly suitable for inspection tasks in complex environments with multiple dynamic obstacles such as industrial parks, warehousing and logistics.

[0070] In the gravitational field and repulsive field model, the dynamic gravitational field function is constructed through the obstacle motion trend term:

[0071]

[0072] Where α is the velocity direction influence factor, v obs is the obstacle velocity vector, is the unit vector direction from the robot to the target point; k att is the gravitational field gain coefficient; p is the current position coordinate of the robot; p goal is the target position coordinate;

[0073] The obstacle motion trend item is adjusted by judging the angle θ between the obstacle motion direction and the line connecting the robot to the target point:

[0074] If θ<π / 2, then α=α0cosθ; if θ≥π / 2, then α=0;

[0075] Among them, α0 is the benchmark impact factor;

[0076] The adaptive repulsive field function is dynamically adjusted according to the motion characteristics of the obstacle:

[0077]

[0078] Where γ is the weight coefficient of motion trend, and the relative speed of the obstacle Calculated; k rep is the repulsive field gain coefficient, d is the distance from the robot to the obstacle, d0 is the influence range of the repulsive field, p obs is the obstacle position coordinate.

[0079] In the calculation of the synthetic force field, determine the robot's environment;

[0080] If it is located in a single obstacle environment, the resultant force is calculated directly:

[0081] F total =F att +F rep

[0082] If the environment is full of obstacles, the weight is calculated based on the threat level of the obstacles:

[0083]

[0084] Where λ is the distance attenuation coefficient, β is the speed influence coefficient, and n is the number of obstacles; is the relative velocity vector of the robot relative to the i-th obstacle;

[0085] Calculate the weighted resultant force:

[0086]

[0087] Determine the robot's movement speed based on the direction and magnitude of the resultant force:

[0088]

[0089] where k v is the speed adjustment coefficient, v max is the maximum permissible speed; k v is the speed adjustment coefficient, v max is the maximum permissible speed.

[0090] It should be noted that the above steps solve the technical problems of robot obstacle avoidance control in dynamic environments. First, in view of the motion characteristics of dynamic obstacles, the obstacle motion trend term Fatt(d) is introduced into the gravitational field, where the velocity direction influence factor α is dynamically adjusted according to the angle θ. When the obstacle moves toward the target direction (0<π / 2), α changes with the cosine of the angle. When the obstacle moves away from the target direction (), α is set to zero. This design enables the system to perceive and respond to the motion trend of the obstacle in advance, avoiding the obstacle avoidance lag problem caused by relying only on the current position information in the traditional method. Secondly, the relative velocity term is introduced into the repulsive field function Frep(d), and its influence is adjusted by the motion trend weight coefficient γ, realizing the adaptive adjustment of the repulsive field strength with the relative motion state, ensuring the safety of obstacle avoidance. For multi-obstacle environments, a weight allocation mechanism based on threat degree is proposed. The distance and motion characteristics of the obstacle are comprehensively considered through the distance attenuation coefficient λ and the velocity influence coefficient β, and the threat degree weight wi of each obstacle is calculated, and then the weighted synthetic force Ftotal is obtained. This mechanism avoids the local minimum problem caused by the simple superposition of multiple repulsive forces in the traditional method, and improves the obstacle avoidance efficiency in complex environments. Finally, according to the direction and magnitude of the synthetic force, combined with the speed adjustment coefficient k v Dynamically determine the robot's movement speed v robot , ensuring the smoothness of the motion trajectory. The core innovation of this method is to organically integrate the motion characteristics of obstacles into the artificial potential field model, achieving accurate perception and rapid response to the dynamic environment. Compared with the existing technology, it has obvious improvements in the timeliness, safety and smoothness of obstacle avoidance. Experimental results show that in a typical dynamic obstacle avoidance scenario, this method can increase the obstacle avoidance success rate by about 15%, shorten the average obstacle avoidance time by about 20%, and ensure the continuity of the trajectory curvature.

[0091] In the process of implementing this solution, the optimal values ​​of key parameters were determined through a large number of experimental verifications: the baseline influence factor α0=0.5, which makes the system moderately sensitive to the movement trend of obstacles; the movement trend weight coefficient γ=0.3, which reasonably balances the influence of relative speed on the repulsive field; the distance attenuation coefficient λ=0.8, which realizes the reasonable attenuation of threat with distance; the speed influence coefficient β=0.4, which ensures the accuracy of relative movement on threat assessment; the speed adjustment coefficient k v =0.6, maintaining smooth trajectory while ensuring motion responsiveness. These parameter settings show good adaptability in various typical dynamic obstacle avoidance scenarios, and have no special requirements for the robot body, and are highly practical.

[0092] The local path update process includes building an obstacle motion prediction model. Specifically, if the obstacle motion direction change rate is less than the threshold ω th, then the uniform linear motion model is used for prediction; if the obstacle movement direction change rate is greater than or equal to the threshold ω th , then the curve fitting prediction model is adopted;

[0093] Perform path conflict detection. Specifically, if the minimum distance between the predicted trajectory and the current path is less than the safety threshold d safe , and the time reserved is less than t safe , then the path update is triggered; otherwise, the current path remains unchanged;

[0094] A spline curve is generated in the obstacle avoidance area. The curve parameters are determined by the current posture and target posture of the robot. The optimal path is selected by minimizing the weighted sum of the curvature integral and the path length.

[0095] The fusion strategy of global path and local obstacle avoidance is realized through environmental complexity judgment.

[0096] In the path planning process, if the environmental complexity index C e satisfy:

[0097]

[0098] where v i is the speed of the ith obstacle, d i is the distance to the ith obstacle, n is the number of obstacles within the field of view; C th is the environment complexity threshold, then switch to local obstacle avoidance mode:

[0099]

[0100] where t p is the prediction time domain, v robot is the robot velocity calculated from the synthetic force field; P current is the current position of the robot;

[0101] If C e ≤C th , then keep the global path following mode:

[0102]

[0103] Where h(p) is the heuristic distance from the path point to the target point; P global is a set of global path points.

[0104] It should be noted that the above steps solve the dynamic obstacle avoidance problem in different complexity scenarios. First, for a single dynamic obstacle, the system uses the environmental complexity index C e Dynamically evaluate the current obstacle avoidance difficulty. e Exceeding the threshold C thWhen t is , it indicates that there is a high risk of collision in the current environment, and the system switches to the local obstacle avoidance mode. p The velocity integral within determines the local path P local , to achieve a quick response to the movement trend of obstacles. e ≤C th When , the system maintains the global path following mode and selects the optimal target point P by comprehensively considering the distance from the current position to the path point and the heuristic function h(p). target , ensuring the global optimality of path planning. In a multi-obstacle environment, the system is based on the same evaluation mechanism, but by accumulating the speed-distance ratio of each obstacle. To reflect the overall complexity of the environment, this method can accurately capture the potential risks in multi-obstacle scenarios. Experiments show that the obstacle avoidance success rate of this method in single-obstacle scenarios is over 95%, and it can still maintain a success rate of over 85% in complex environments with 3-5 dynamic obstacles, and the average obstacle avoidance time only increases by about 30%.

[0105] In the implementation process, the selection of environment complexity evaluation parameters is crucial: the environment complexity threshold C th Set to 2.5, this value is determined by comparing and analyzing the obstacle avoidance performance in different scenarios, which can achieve a good balance between ensuring safety and path planning efficiency; predict the time domain t p The selected value is 1.5s, which is sufficient to cover the characteristic motion cycle of most dynamic obstacles while maintaining the computational load within an acceptable range. For multi-obstacle scenarios, the system automatically adjusts the impact weight of each obstacle, and obstacles with closer distances and higher speeds are given higher priority. This mechanism ensures that the system can prioritize the most threatening obstacles. This solution has excellent environmental adaptability and obstacle avoidance reliability in various practical application scenarios, and is particularly suitable for use in dynamic and complex indoor and outdoor environments.

[0106] The environmental perception layer classifies obstacles by judging the distance. Specifically,

[0107] If the obstacle distance is less than d min , classifying it as a high-threat obstacle;

[0108] If the obstacle distance is between d min With d max between them, classifying them as potential threatening obstacles;

[0109] The motion control layer constrains the robot speed according to the obstacle threat level.

[0110]

[0111] where v cmdis the expected speed, k s is the safety factor; d is the current obstacle distance.

[0112] It also includes a collision warning mechanism, which is used to determine based on the time margin:

[0113]

[0114] When t collision When it is less than the safety time threshold, the deceleration and obstacle avoidance control is triggered; is the unit vector from the robot to the obstacle.

[0115] It should be noted that in order to solve the safety control problem during the robot's movement, the system has designed a multi-level safety monitoring mechanism. First, at the environmental perception level, obstacles are dynamically classified based on distance thresholds: when the obstacle distance is less than the safety distance threshold d min When the obstacle is at a distance of d min and d max When the speed is between , it is marked as a potential threatening obstacle, and the system enters the alert state and prepares for obstacle avoidance. At the motion control level, the speed constraint formula v out Realize adaptive adjustment of robot speed, where the safety factor k s It is used to adjust the sensitivity of speed decay. At the same time, the system introduces a collision warning mechanism based on time margin, which calculates the collision time t collision , when it is less than the preset safety time threshold, the deceleration obstacle avoidance control is triggered. This multi-level safety monitoring mechanism ensures the safety of obstacle avoidance while achieving smoothness of motion trajectory through dynamic speed planning. Experimental data show that this method can reduce the risk of collision by more than 90%, while keeping the average motion speed only reduced by about 20%.

[0116] In the specific implementation, the selection of safety monitoring parameters is based on a large number of experimental verifications: the safety distance threshold d min Set to 0.5m, considering the robot body size and conventional braking distance; the maximum warning distance d max Set to 2.0m, within which the system can fully estimate and respond to potential threats; safety factor k s The value is 1.5, which achieves a good balance between safety margin and motion efficiency; the time threshold of collision warning is set to 2.0s, which can reserve sufficient obstacle avoidance decision-making and execution time for the system. This set of parameter configurations is stable in various application scenarios and is particularly suitable for use in environments where humans and machines coexist. This solution organically combines the safety monitoring mechanism with the motion control strategy, which not only ensures the reliability of obstacle avoidance, but also achieves high efficiency and smoothness of the motion process.

[0117] In summary, this method achieves quantitative judgment of the threat level of obstacles by establishing a threat level assessment mechanism based on distance; adopts a dynamic speed constraint strategy to achieve differentiated control for obstacles of different threat levels; introduces a collision warning mechanism based on time margin to improve the safety and reliability of obstacle avoidance control. The beneficial effect of the present invention is to improve the obstacle avoidance performance of the robot in a complex dynamic environment and enhance the adaptability and robustness of obstacle avoidance control.

[0118] Embodiment 2 is an embodiment of the present invention, which provides a circular coal yard stacker-reclaimer angle calibration system based on microwave recognition, comprising:

[0119] An environmental perception module is used to classify obstacles as high-threat obstacles or potential-threat obstacles based on distance threshold judgment;

[0120] The speed control module is used to constrain the robot speed according to the obstacle threat level;

[0121] The early warning monitoring module is used to determine whether to trigger deceleration and obstacle avoidance control based on the time margin.

[0122] Example 3, reference Figure 2 , is an embodiment of the present invention, which is different from the previous embodiment in that: if the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0123] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0124] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0125] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0126] Example 4 is an embodiment of the present invention, which provides a multi-level hybrid obstacle avoidance control method that is adaptive to environmental complexity. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0127] To verify the effectiveness of the present invention, a series of comparative experiments were designed for this method, and the tests were carried out in a static environment, a single dynamic obstacle environment, and a multi-dynamic obstacle environment. The experimental platform uses a Pioneer 3-AT mobile robot equipped with a laser radar and a depth camera for environmental perception, and the experimental site is an indoor area of ​​8m×8m. Under the same starting point, end point, and obstacle distribution conditions, this method is compared with the traditional artificial potential field method (APF), the dynamic window method (DWA), and the improved speed obstacle method (VO). During the experiment, static obstacles are randomly distributed in the experimental site, the movement speed of dynamic obstacles ranges from 0.2-0.8m / s, and the movement direction is randomly generated in the range of [-π,π]. Each group of experiments was repeated 20 times, and key indicators such as obstacle avoidance success rate, average obstacle avoidance time, path smoothness, computational efficiency, and safety margin were recorded. In order to evaluate the performance of the system under extreme conditions, a number of high-difficulty scenarios were specially designed, including typical working conditions such as high-speed intersections, dense obstacles, and narrow passages.

[0128] In the dynamic adaptability test, the focus was on verifying the response characteristics of the present invention under the condition of rapid changes in environmental complexity. By dynamically changing the number and motion characteristics of obstacles during the experiment, the mode switching ability and parameter adaptation ability of the system were tested. At the same time, for the speed planning and trajectory smoothness characteristics, curvature continuity and acceleration fluctuation were used as evaluation indicators, and the actual motion trajectory of the robot was recorded by a high-precision motion capture system and quantitatively analyzed. In the safety test, the minimum obstacle avoidance distance and average obstacle avoidance margin in different scenarios were statistically calculated to verify the effectiveness of the safety monitoring mechanism. In particular, by introducing random perturbations and sensor noise, the robustness and anti-interference ability of the system were evaluated.

[0129] Table 1 Performance comparison test data of different obstacle avoidance methods

[0130] Evaluation Metrics This method Traditional APF DWA method VO Method Static environment obstacle avoidance success rate (%) 98.5 92.3 94.1 93.8 Single dynamic obstacle avoidance success rate (%) 95.2 85.6 88.4 89.2 Multiple dynamic obstacle avoidance success rate (%) 92.1 78.3 82.5 83.1 Average calculation time (ms) 12.5 8.2 15.8 14.3 Path smoothness index 0.92 0.75 0.83 0.85 Minimum obstacle avoidance distance (m) 0.48 0.32 0.41 0.39 Average obstacle avoidance time (s) 15.3 19.8 17.2 16.8 <![CDATA[Speed fluctuation degree (m / s 2 )]]> 0.15 0.28 0.21 0.19 Environmental adaptability index 0.88 0.65 0.72 0.75 Computing resource utilization (%) 25.3 18.5 32.1 29.8

[0131] Through the analysis of experimental data, it can be concluded that the present invention is significantly superior to existing methods in multiple key performance indicators. In terms of obstacle avoidance success rate, the present invention reaches 98.5%, 95.2% and 92.1% in static environment, single dynamic obstacle and multiple dynamic obstacle scenarios respectively, which is 5-15 percentage points higher than the traditional method. This is mainly due to the newly proposed environmental complexity assessment mechanism and adaptive control strategy. In terms of motion performance, the path smoothness index of the present invention reaches 0.92, which is 22.7% higher than the traditional APF method, and the speed fluctuation is reduced to 0.15m / s 2, the improvement reaches 46.4%, which shows that the present invention can significantly improve the quality of the motion trajectory while ensuring the safety of obstacle avoidance. Although the average calculation time (12.5ms) of the present invention is slightly higher than that of the traditional APF method (8.2ms), it is still significantly better than the DWA method (15.8ms) and the VO method (14.3ms), and considering the significantly improved performance, this moderate computational overhead is completely acceptable. It is particularly noteworthy that the present invention achieves a minimum obstacle avoidance distance index of 0.48m, which is much higher than other methods, fully verifying the effectiveness of the multi-level safety monitoring mechanism. The environmental adaptability index reaches 0.88, which is 17-35% higher than other methods, indicating that the present invention can better cope with complex and changing environments. These data fully demonstrate the comprehensive advantages of the present invention in obstacle avoidance performance, motion quality and environmental adaptability.

[0132] It is important to note that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A multi-level hybrid obstacle avoidance control method that is adaptive to the complexity of the environment, characterized in that: include: Obtain global map information of the inspection environment, divide the map into grid cells and mark static obstacles, and use the A* algorithm to calculate the global path of the inspection robot from the starting position to the target position; The position and velocity information of dynamic obstacles are acquired in real time by laser radar, and a gravitational field and a repulsive field model are established, wherein the gravitational field is generated by the target point and the repulsive field is generated by the detected dynamic obstacles; The movement direction and speed of the inspection robot are calculated according to the synthetic force field of the gravitational field and the repulsive field. When a dynamic obstacle is detected, the local path is updated and the inspection robot is controlled to move along the updated path.

2. The multi-level hybrid obstacle avoidance control method according to claim 1, characterized in that: In the gravitational field and repulsive field model, the dynamic gravitational field function is constructed by the obstacle motion trend term: Among them, α is the velocity direction influence factor, υ obs is the obstacle velocity vector, is the unit vector direction from the robot to the target point; k αtt is the gravitational field gain coefficient; p is the current position coordinate of the robot; p goal is the target position coordinate; The obstacle motion trend item is adjusted by determining the angle θ between the obstacle motion direction and the line connecting the robot to the target point: If θ<π / 2, then α=α0cosθ; if θ≥π / 2, then α=0; Among them, α0 is the benchmark impact factor; The adaptive repulsive field function is dynamically adjusted according to the motion characteristics of the obstacle: Where γ is the weight coefficient of motion trend, and the relative speed of the obstacle Calculated; k rep is the repulsive field gain coefficient, d is the distance from the robot to the obstacle, d0 is the influence range of the repulsive field, p obs is the obstacle position coordinate.

3. The multi-level hybrid obstacle avoidance control method according to claim 2, characterized in that: In the calculation of the synthetic force field, determining the environment in which the robot is located; If it is located in a single obstacle environment, the resultant force is calculated directly: F total =F att +F rep If the environment is full of obstacles, the weight is calculated based on the threat level of the obstacles: Where λ is the distance attenuation coefficient, β is the speed influence coefficient, and n is the number of obstacles; is the relative velocity vector of the robot relative to the i-th obstacle; Calculate the weighted resultant force: Determine the robot's movement speed based on the direction and magnitude of the resultant force: where k v is the speed adjustment coefficient, v max is the maximum permissible speed; k v is the speed adjustment coefficient, v max is the maximum permissible speed.

4. The multi-level hybrid obstacle avoidance control method according to claim 3, characterized in that: The local path update process includes constructing an obstacle motion prediction model. Specifically, if the obstacle motion direction change rate is less than a threshold value ω th , then the uniform linear motion model is used for prediction; if the obstacle movement direction change rate is greater than or equal to the threshold ω th , then the curve fitting prediction model is adopted; Perform path conflict detection. Specifically, if the minimum distance between the predicted trajectory and the current path is less than the safety threshold d safe , and the time reserved is less than t safe , then the path update is triggered; otherwise, the current path remains unchanged; A spline curve is generated in the obstacle avoidance area. The curve parameters are determined by the current posture and target posture of the robot. The optimal path is selected by minimizing the weighted sum of the curvature integral and the path length.

5. The multi-level hybrid obstacle avoidance control method according to claim 4, characterized in that: The fusion strategy of global path and local obstacle avoidance is realized by judging the complexity of the environment. In the path planning process, if the environmental complexity index C e satisfy: where v i is the speed of the ith obstacle, d i is the distance to the ith obstacle, n is the number of obstacles within the field of view; C th is the environment complexity threshold, then switch to local obstacle avoidance mode: where t p is the prediction time domain, v robot is the robot velocity calculated from the synthetic force field; P current is the current position of the robot; If C e ≤C th , then keep the global path following mode: Where h(p) is the heuristic distance from the path point to the target point; P global is a set of global path points.

6. The multi-level hybrid obstacle avoidance control method according to claim 5, characterized in that: The environmental perception layer classifies obstacles by judging the distance. Specifically, If the obstacle distance is less than d min , classifying it as a high-threat obstacle; If the obstacle distance is between d min With d max between them, classifying them as potential threatening obstacles; The motion control layer constrains the robot speed according to the obstacle threat level. where v cmd is the expected speed, k s is the safety factor; d is the current obstacle distance.

7. The multi-level hybrid obstacle avoidance control method according to claim 6, characterized in that: It also includes a collision warning mechanism, which is used to determine based on the time margin: When the collision When it is less than the safety time threshold, the deceleration and obstacle avoidance control is triggered; is the unit vector from the robot to the obstacle.

8. A system using the multi-level hybrid obstacle avoidance control method according to any one of claims 1 to 7, characterized in that: include: An environmental perception module is used to classify obstacles as high-threat obstacles or potential-threat obstacles based on distance threshold judgment; The speed control module is used to constrain the robot speed according to the obstacle threat level; The early warning monitoring module is used to determine whether to trigger deceleration and obstacle avoidance control based on the time margin.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the multi-level hybrid obstacle avoidance control method for adaptive environment complexity according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the multi-level hybrid obstacle avoidance control method for adaptive environment complexity according to any one of claims 1 to 7 are implemented.

Images