Intelligent obstacle avoidance method and system for flood drainage robot

Abstract

This invention relates to the field of robot motion technology, specifically to an intelligent obstacle avoidance method and system for a flood drainage robot. The method involves controlling a lidar sensor and a sonar sensor to collect data from a target area under preset acquisition conditions, obtaining a first point cloud dataset and a second point cloud dataset, and then determining the data error scale for each acquisition point. Based on the data error scale for each acquisition point, a target error growth coefficient is obtained. Data processing is performed on the target point cloud data in the first point cloud dataset to obtain at least one obstacle in the target point cloud data. Based on the motion trajectory corresponding to each obstacle and the target error growth coefficient, a track threat coefficient corresponding to each obstacle is determined. Obstacle avoidance planning is performed based on the track threat coefficient to obtain the target obstacle avoidance path for the flood drainage robot. This invention achieves intelligent obstacle avoidance planning, effectively improving the reliability and efficiency of obstacle avoidance function during robot movement.

Description

Technical Field

[0001] The present invention relates to the field of robot motion technology, and in particular to an intelligent obstacle avoidance method and system for a flood drainage robot. Background Art

[0002] Drainage robots are widely used in flood prevention and drainage work in cities and rural areas. The drainage robot drives autonomously according to the preset direction and track. When it arrives at the designated area, it immediately activates the drainage system and effectively extracts the accumulated water in the area through the built-in pipes. However, given the high diversity and complexity of the working environment, including the dynamic changes of water flow and potential movable obstacles, the intelligent obstacle avoidance function of the drainage robot may face many interference factors when it goes to the predetermined drainage area.

[0003] Although existing drainage robots are equipped with a variety of environmental monitoring sensors that can deeply analyze and monitor the surrounding environment in real time to achieve intelligent obstacle avoidance, there are still some technical difficulties in practical applications. In particular, when the water environment is significantly different and changes dramatically, the single monitoring signal that the robot relies on is easily subject to strong interference, which seriously affects the robot's recognition accuracy of obstacles. Due to this interference, the robot may not be able to make accurate obstacle avoidance adjustments in a complex and changing environment, thus affecting the reliability and efficiency of its obstacle avoidance function. Summary of the invention

[0004] In order to solve the technical problem that when a drainage robot moves in a complex underwater terrain environment, it is susceptible to significant interference and cannot accurately predict targets and avoid obstacles, the purpose of the present invention is to provide an intelligent obstacle avoidance method and system for a drainage robot. The technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present invention provides an intelligent obstacle avoidance method for a flood drainage robot, which is applied to an intelligent obstacle avoidance system of the flood drainage robot. The intelligent obstacle avoidance system of the flood drainage robot includes a laser radar sensor and a sonar detection sensor. The method includes:

[0006] Control the laser radar sensor and the sonar detection sensor that meet the preset collection conditions to collect data from multiple collection points in the target area to obtain a first point cloud data set and a second point cloud data set, wherein the first point cloud data set is point cloud data collected in multiple batches by the laser radar sensor within a collection time period, and the second point cloud data set is point cloud data collected in multiple batches by the sonar detection sensor within a collection time period;

[0007] Determining a data error scale of each of the plurality of acquisition points according to the first point cloud data set and the second point cloud data set;

[0008] According to the data error scale of each acquisition point, a target error growth coefficient is obtained, wherein the target error growth coefficient is used to evaluate the trend of the overall data error scale corresponding to the multiple acquisition points changing with distance;

[0009] Performing data processing on target point cloud data in the first point cloud data set to obtain at least one obstacle in the target point cloud data, wherein the target point cloud data is point cloud data in a target acquisition time period;

[0010] Determining a track threat coefficient corresponding to each obstacle in the at least one obstacle according to the motion trajectory corresponding to each obstacle and the target error growth coefficient;

[0011] Obstacle avoidance planning is performed according to the track threat coefficient corresponding to each obstacle to obtain a target obstacle avoidance path for the drainage robot.

[0012] In a second aspect, an embodiment of the present invention provides an intelligent obstacle avoidance system for a flood drainage robot, comprising a memory and a processor, wherein the memory is connected to the processor, and the processor is used to execute one or more computer programs stored in the memory, and when the processor executes the one or more computer programs, the intelligent obstacle avoidance system of the flood drainage robot implements the intelligent obstacle avoidance method for the flood drainage robot as described in the first aspect.

[0013] In a third aspect, an embodiment of the present invention provides an intelligent obstacle avoidance device for a flood drainage robot, which is applied to an intelligent obstacle avoidance system for the flood drainage robot. The intelligent obstacle avoidance system for the flood drainage robot includes a laser radar sensor and a sonar detection sensor. The device includes:

[0014] A collection unit, used for controlling the laser radar sensor and the sonar detection sensor that meet preset collection conditions, to collect data from multiple collection points in a target area, to obtain a first point cloud data set and a second point cloud data set, wherein the first point cloud data set is point cloud data collected by the laser radar sensor, and the second point cloud data set is point cloud data collected by the sonar detection sensor, wherein the first point cloud data set is point cloud data collected by the laser radar sensor at multiple collection points in multiple batches within a collection time period, and the second point cloud data set is point cloud data collected by the sonar detection sensor at multiple collection points in multiple batches within a collection time period;

[0015] a determining unit, configured to determine a data error scale of each of the plurality of acquisition points according to the first point cloud data set and the second point cloud data set;

[0016] The determination unit is further configured to obtain a target error growth coefficient according to the data error scale of each acquisition point, wherein the target error growth coefficient is used to evaluate the trend of the overall data error scale corresponding to the multiple acquisition points changing with distance;

[0017] a processing unit, configured to perform data processing on target point cloud data in the first point cloud data set to obtain at least one obstacle in the target point cloud data, wherein the target point cloud data is point cloud data in a target acquisition time period;

[0018] The determining unit is further configured to determine a track threat coefficient corresponding to each obstacle in the at least one obstacle according to a motion trajectory corresponding to each obstacle and the target error growth coefficient;

[0019] The planning unit is used to perform obstacle avoidance planning according to the track threat coefficient corresponding to each obstacle to obtain a target obstacle avoidance path of the drainage robot.

[0020] In a fourth aspect, an electronic device is provided, comprising a memory and a processor, wherein the memory is connected to the processor, and the processor is used to execute one or more computer programs stored in the memory, and when the processor executes the one or more computer programs, the electronic device implements the intelligent obstacle avoidance method for the drainage robot as described in the first aspect.

[0021] In a fifth aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the program instructions are executed by a processor, the processor executes the intelligent obstacle avoidance method for the drainage robot as described in the first aspect.

[0022] The present invention has the following beneficial effects: the present invention can collect more comprehensive and accurate data in different environments and conditions through the combined use of laser radar sensors and sonar detection sensors. The laser radar sensor provides high-precision direct-view distance measurement, while the sonar detection sensor can work effectively in an environment with obstructed vision or underwater, which enables the drainage robot to efficiently collect data in a complex and changeable environment; further, by combining the analysis and processing of the first point cloud data set and the second point cloud data set, the data error scale is accurately evaluated, and obstacles can be effectively identified and located; further, by calculating the target error growth coefficient and the track threat coefficient of the obstacle, the system can accurately evaluate the actual threat level of the obstacle to the drainage robot, and perform effective obstacle avoidance planning based on the actual threat level, ensuring that the drainage robot can safely and efficiently perform drainage tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0024] Figure 1 A schematic diagram of a framework of an intelligent obstacle avoidance system for a flood drainage robot provided by an embodiment of the present invention;

[0025] Figure 2 A schematic flow chart of an intelligent obstacle avoidance method for a flood drainage robot provided by an embodiment of the present invention;

[0026] Figure 3 A schematic structural diagram of an intelligent obstacle avoidance device for a flood drainage robot provided by one embodiment of the present invention;

[0027] Figure 4 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the intelligent obstacle avoidance method and system of a drainage robot proposed by the present invention, its specific implementation, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0029] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0030] The specific scheme of the intelligent obstacle avoidance method and system for a flood drainage robot provided by the present invention is described in detail below with reference to the accompanying drawings.

[0031] See also Figure 1 , which shows a schematic framework diagram of an intelligent obstacle avoidance system for a flood drainage robot provided by an embodiment of the present invention.

[0032] The intelligent obstacle avoidance system 10 of the flood drainage robot provided by the present invention includes a laser radar sensor 20 and a sonar detection sensor 30.

[0033] Among them, the LiDAR sensor 20 measures the distance to surrounding objects by emitting a laser beam. When the laser beam hits an object, it is reflected back to the sensor, and the LiDAR sensor 20 determines the distance to the object by calculating the round-trip time of the light. Therefore, the position, shape and size of the object can be accurately measured, thereby generating a high-precision three-dimensional map. The main advantage of the LiDAR sensor 20 is its high resolution and ability to work under various lighting conditions, including direct sunlight. It can provide a detailed view of the surrounding environment, helping the drainage robot understand its environment.

[0034] Among them, the sonar detection sensor 30 evaluates the distance to the object by emitting sound waves and measuring the time it takes for the sound waves to be reflected back after colliding with the object. Similar to the lidar sensor 20, this "echolocation" method allows the robot to perceive the surrounding environment and identify potential obstacles. The sonar detection sensor 30 is particularly suitable for underwater environments or other environments where the lidar sensor 20 may not be so effective. It is not sensitive to line of sight obstacles such as fog and smoke, and can remain functional in visually restricted environments.

[0035] In addition to the two main sensors mentioned above, the intelligent obstacle avoidance system of the drainage robot may also include other sensors, such as cameras, infrared sensors, ultrasonic sensors, etc., to provide more comprehensive perception capabilities. These sensors can complement each other and improve the robustness and reliability of the system.

[0036] It can be seen that the intelligent obstacle avoidance system of the drainage robot can ensure that the robot can autonomously navigate in a complex and changing environment and avoid collisions to achieve stable operation, thereby improving the efficiency and safety of drainage work.

[0037] See also Figure 2 , which shows a flow chart of an intelligent obstacle avoidance method for a drainage robot provided by an embodiment of the present invention, the method comprising the following steps:

[0038] S10. Control the laser radar sensor and the sonar detection sensor that meet the preset collection conditions to collect data from multiple collection points in the target area to obtain a first point cloud data set and a second point cloud data set, wherein the first point cloud data set is point cloud data collected in multiple batches by the laser radar sensor within a collection time period, and the second point cloud data set is point cloud data collected in multiple batches by the sonar detection sensor within a collection time period.

[0039] The above-mentioned preset acquisition conditions refer to a set of specific conditions that the laser radar sensor and the sonar detection sensor meet, and the set of specific conditions may include but are not limited to: the time base of the laser radar sensor and the sonar detection sensor are consistent and the laser radar sensor and the sonar detection sensor are converted to the same coordinate system. When both conditions are met, it can be determined that the laser radar sensor and the sonar detection sensor have met the preset acquisition conditions.

[0040] Among them, before the data collection process is carried out in the target area, the drainage robot needs to be in a stationary state before data collection can be carried out.

[0041] Among them, the target area is the spatial range where the lidar sensor and the sonar detection sensor work together and cover, and the sensor collects data at multiple collection points in the target area.

[0042] The target area is usually predefined and can be part of the land, seabed or any specific environment, or the potential collection area is explored and evaluated using maps, satellite images, on-site surveys and other methods to determine the final target area. This step requires comprehensive consideration of terrain, environmental characteristics and drainage mission requirements. During the data collection process of the target area, the lidar sensor will continuously emit laser pulses and receive the reflected signals to obtain the three-dimensional information of the target area and obtain the first point cloud data set.

[0043] During the data collection process of the target area, the sonar detection sensor will continuously emit sound waves and receive reflected signals, thereby obtaining three-dimensional information of the target area and obtaining a second point cloud data set.

[0044] Point cloud data is collected by lidar sensors and sonar detection sensors. It contains a series of three-dimensional space points, each of which has location information. These points are distributed in space to form a three-dimensional grid that describes the target area.

[0045] The collection point refers to the specific location point captured by the sensor in the target area during the collection time period.

[0046] The collection time period refers to the time period from when the sensor starts to collect data to when it stops collecting data. During this time period, the sensor may collect data multiple times to obtain more comprehensive environmental information.

[0047] The first point cloud data set includes point cloud data collected in multiple batches at multiple collection points within a collection time period.

[0048] The second point cloud data set includes point cloud data collected in multiple batches at multiple collection points within a collection time period.

[0049] It can be seen that in this embodiment, by controlling the lidar and sonar sensors to achieve the preset collection conditions, it can ensure that the acquired first point cloud data set and the second point cloud data set have sufficient quality and accuracy, thereby improving the obstacle avoidance capability and operating efficiency of the drainage robot.

[0050] In one embodiment, before the control of the laser radar sensor and the sonar detection sensor to achieve the preset collection condition performs data collection, the method further includes: performing time debugging on the laser radar sensor and the sonar detection sensor to obtain that the monitoring frequencies of the laser radar sensor and the sonar detection sensor are the same time reference; constructing a target three-dimensional rectangular coordinate system with the current position of the drainage robot as the origin; performing coordinate conversion on the coordinate system of the laser radar sensor and the coordinate system of the sonar detection sensor to obtain the laser radar sensor and the sonar detection sensor in the target three-dimensional rectangular coordinate system, so that the first point cloud data set corresponding to the laser radar sensor and the second point cloud data set corresponding to the sonar detection sensor are in the same coordinate system; when it is detected that the monitoring frequencies of the laser radar sensor and the sonar detection sensor are the same time reference, and the laser radar sensor and the sonar detection sensor have completed the coordinate conversion, it is determined that the laser radar sensor and the sonar detection sensor meet the preset collection condition.

[0051] Among them, the time debugging of the lidar sensor and the sonar detection sensor is to ensure that the monitoring frequencies of the two sensors are synchronized, that is, to ensure that the data of the target area collected each time belongs to the same time node to ensure the synchronization and consistency of the data.

[0052] Among them, the coordinate system conversion refers to converting the coordinate system of the lidar sensor and the sonar detection sensor into a target three-dimensional rectangular coordinate system with the current position of the drainage robot as the origin.

[0053] Among them, the target three-dimensional rectangular coordinate system provides a unified spatial reference for the collected point cloud dataset, so that the data from different sensors can be accurately represented and compared in the same spatial framework, that is, the point cloud data of all target areas are uniformly aligned in the target three-dimensional rectangular coordinate system to ensure accurate matching and comprehensive analysis of the data.

[0054] It can be seen that in this embodiment, the data collected by the lidar sensor and the sonar detection sensor are ensured to be accurate and reliable, and can be effectively integrated together to provide accurate environmental information for subsequent obstacle avoidance planning.

[0055] S20: Determine a data error scale of each of the multiple acquisition points according to the first point cloud data set and the second point cloud data set.

[0056] The data error scale refers to the error size of point cloud data at the same or adjacent positions in the data collected by different sensors (such as lidar sensors and sonar detection sensors). Specifically, the data error scale reflects the spatial consistency and accuracy of point cloud data collected by different sensors. The data error scale of each collection point can be obtained by comparing the sonar detection data corresponding to the same collection point with the adjacent lidar detection data in detail and taking into account the distance factor.

[0057] Among them, according to the size of the data error scale, the accuracy and reliability of each collection point can be evaluated. Points with larger data error scales may indicate inaccurate measurements. In addition, by identifying collection points with larger data error scales, the system can also optimize its data collection strategy, such as by increasing the collection frequency in certain areas or by using higher-precision sensors to improve data quality.

[0058] It can be seen that in this embodiment, by determining the data error scale of each collection point, the system can more accurately evaluate the position and size of obstacles in the environment, thereby improving the accuracy and efficiency of obstacle avoidance path planning.

[0059] In one embodiment, determining the data error scale of each of the multiple acquisition points based on the first point cloud data set and the second point cloud data set includes: obtaining the total number of target acquisition batches in the first point cloud data set, wherein the total number of target acquisition batches is equal to the total number of acquisition batches in the second point cloud data set; obtaining a first curvature value corresponding to a first acquisition point among the multiple acquisition points in the first point cloud data set; obtaining a second curvature value of the first acquisition point in the second point cloud data set; performing error size calculation on the total number of target acquisition batches, the first curvature value, and the second curvature value according to a first preset formula to obtain the data error scale of the first acquisition point; performing error size calculation on each of the multiple acquisition points to obtain the data error scale of each of the multiple acquisition points.

[0060] Among them, the total number of target collection batches refers to the total number of data batches collected by the lidar sensor and the sonar detection sensor respectively within the collection time period, and the total number of collection batches of collection points in the first point cloud dataset is the same as the total number of collection batches in the second point cloud dataset to ensure data consistency.

[0061] The curvature value is used to measure the curvature of the surface or edge in the point cloud data. In the first point cloud data set and the second point cloud data set, each acquisition point has a corresponding curvature value, and the curvature value can reflect the shape and direction of the acquisition point.

[0062] The curvature value can be obtained by calculating the derivative of the acquisition point in a specific direction, which may be a normal direction perpendicular to the acquisition point, such as calculating the principal curvature of the acquisition point or fitting the local point cloud data using the least squares method. The specific calculation method of the curvature is not uniquely limited here.

[0063] Among them, the first curvature value is the curvature value corresponding to the first collection point in the point cloud data collected by the laser radar sensor; the second curvature value is the curvature value of the position corresponding to the first collection point in the point cloud data collected by the sonar detection sensor.

[0064] The first preset formula may be as follows:

[0065] ;

[0066] In the above formula, Indicates The data error scale of the collection points of each LiDAR sensor; Indicates the data collection batches obtained by the drainage robot in a stationary state; Indicates the first The curvature size corresponding to the collection point of each lidar sensor; Indicates The curvature of the collection point of the lidar sensor in the data collected by the sonar detection sensor; Indicates the curvature difference of a single acquisition point in the same batch; Indicates The difference between the sum of the distances between the calculated curvature related points corresponding to the collection points of the lidar sensors and the drainage robot and the sum of the distances between the calculated curvature related points in the sonar detection sensor.

[0067] when The larger the difference, the lower the credibility of the result. The smaller the difference, the higher the reliability of the calculation. The curvature errors of the current single acquisition point in all acquisition batches are accumulated and averaged, and the final result is the data error scale of the single acquisition point.

[0068] in, When the sum of the distances between the calculated curvature related points corresponding to the collection point and the drainage robot is equal to the sum of the distances between the calculated curvature related points in the sonar detection sensor, the difference is 0, that is, when the curvature calculation results of the lidar sensor and the sonar detection sensor for the same collection point are exactly the same, the difference value can be zero. However, due to the differences in the working principles and environmental adaptability of lidar sensors and sonar detection sensors, their measurement results of the geometric morphology (such as curvature) of the same object surface may be slightly to significantly different. Although it is rare in practice for the difference value to be completely zero, it is possible to achieve this result under certain conditions. This may occur when the two sensors very accurately measure a very smooth surface or a surface with almost the same geometric characteristics in some special cases.

[0069] Among them, the curvature difference is used to evaluate the degree of curvature difference between the same acquisition point (or adjacent acquisition points) under different sensors, thereby estimating the data error.

[0070] The same error size calculation as the first acquisition point is performed on all acquisition points to obtain the data error scale of each acquisition point, which involves traversing all acquisition points and applying the same error size calculation method.

[0071] It can be seen that in this embodiment, by jointly analyzing the data of the lidar sensor and the data of the sonar detection sensor, the error status of the lidar scanning collection point can be accurately evaluated.

[0072] S30. Obtain a target error growth coefficient according to the data error scale of each acquisition point, wherein the target error growth coefficient is used to evaluate a trend of the overall data error scale corresponding to the multiple acquisition points changing with distance.

[0073] Among them, the target error growth coefficient is a coefficient calculated based on the data error scale of all collection points. It is used to evaluate the trend of the overall data error scale corresponding to multiple collection points changing with distance. That is, the target error growth coefficient can represent the law of change of the data error scale with distance, so as to better evaluate the accuracy and reliability of the data.

[0074] If the target error growth coefficient is low, it means that the error grows slowly and the system is more stable at long distances. If the target error growth coefficient is high, it means that the error grows faster with increasing distance, which may affect the recognition accuracy of long-distance targets. Therefore, the data can be corrected in a distance-related manner to compensate for the growth of the error. For example, in the processing of long-distance targets, compensation can be made by increasing the data sampling frequency or using more sensors.

[0075] In one embodiment, the target error growth coefficient is obtained according to the data error scale of each collection point, including: constructing a target three-dimensional rectangular coordinate system with the current position of the drainage robot as the origin, and obtaining the first distance between each collection point and the drainage robot according to the target three-dimensional rectangular coordinate system; obtaining the distance error scale coefficient corresponding to each collection point according to the data error scale of each collection point and the first distance between each collection point and the drainage robot; and calculating the distance error scale coefficients corresponding to all collection points to obtain the target error growth coefficient.

[0076] Among them, the target three-dimensional rectangular coordinate system is used to describe and locate the drainage robot and each collection point in the target area.

[0077] Among them, the first distance refers to the straight-line distance between each collection point and the drainage robot, and this distance is calculated based on the target three-dimensional rectangular coordinate system.

[0078] Among them, the calculation process of calculating the distance error scale coefficient corresponding to all the collection points can be expressed as follows:

[0079] ;

[0080] In the above formula, It represents the target error growth coefficient of the acquisition point under unit distance growth; Represents the total number of collection points of the lidar sensor. Indicates The data error scale of the collection points of the LiDAR sensor is Indicates The first distance between the collection point of the laser radar sensor and the drainage robot; Indicates The ratio of the data error scale of the collection point of the laser radar sensor to the corresponding first distance, that is, The distance error scale coefficient of the acquisition points of each lidar sensor; By accumulating and averaging the distance error scale coefficients of all acquisition points, the overall target error growth coefficient is obtained. Represents the normalization function.

[0081] It can be seen that in this embodiment, the sensor data error of the drainage robot at different distances can be accurately evaluated, and the target error growth coefficient can be used to optimize the system design and data processing to improve the accuracy of the overall data.

[0082] S40. Process the target point cloud data in the first point cloud data set to obtain at least one obstacle in the target point cloud data, where the target point cloud data is point cloud data in a target acquisition time period.

[0083] Among them, the target collection time period is the last moment of the static observation stage. The point cloud data at the last moment of the static observation stage can best reflect the actual situation of the current environment, because at this stage, the impact of the sensor's movement on the data is minimized.

[0084] The data processing may include but is not limited to: pre-processing such as denoising and filtering, point cloud segmentation and obstacle recognition, which is not the only limitation here.

[0085] Specifically, denoising can remove noise in point cloud data through filtering algorithms, such as using statistical filtering, radius filtering and other methods to remove abnormal points; point cloud segmentation can use algorithms such as RANSAC (Random Sample Consensus) to identify and separate fixed objects such as the ground and planes, or apply a regional growing algorithm to the point cloud data after removing background information to segment the point cloud data into multiple clusters, each cluster representing a possible obstacle; obstacle identification can include further analysis of point cloud data clusters using clustering algorithms (such as Euclidean clustering or density-based clustering algorithms), each cluster may represent an obstacle, and the clustering algorithm groups the point cloud data to help identify the outline and location of the obstacle; obstacle identification can also extract geometric features for each cluster, such as volume, shape, density, etc. Based on these features, it can be identified whether it is an obstacle, and the specific characteristics of the sensor and the expected characteristics of the obstacle can be combined to filter out point cloud data clusters that do not conform to reality.

[0086] It can be seen that in this embodiment, by processing the target point cloud data in the first point cloud data set, at least one obstacle in the target point cloud data can be effectively identified and extracted, and selecting the data at the last moment of the static observation stage as the target collection time period can ensure that the extracted obstacle information is more accurate, because the point cloud data at this time can best reflect the actual situation of the current environment.

[0087] S50. Determine a track threat coefficient corresponding to each obstacle in the at least one obstacle according to the motion trajectory corresponding to each obstacle and the target error growth coefficient.

[0088] The motion trajectory of each obstacle refers to the path of each obstacle in space that changes with time. The motion trajectory can be the moving path of the obstacle or the position sequence of the obstacle that changes with time.

[0089] The motion trajectory may use the obstacle position and time information in the first point cloud data or the second point cloud data to establish the motion trajectory of each obstacle, and the motion trajectory involves continuous tracking of the position of the obstacle at different time points.

[0090] The track threat coefficient is used to evaluate the potential threat level of each obstacle to the drainage robot's driving track. The track threat coefficient can help the drainage robot's intelligent obstacle avoidance system decide how to adjust its path to avoid collision with obstacles.

[0091] It can be seen that in this embodiment, the track threat coefficient corresponding to each obstacle is determined. The track threat coefficient is used to evaluate the threat level of each obstacle to the path of the drainage robot, thereby helping the drainage robot to make decisions and plans in a dynamic environment to avoid collisions or other potential safety issues.

[0092] S60. Perform obstacle avoidance planning according to the track threat coefficient corresponding to each obstacle to obtain a target obstacle avoidance path for the drainage robot.

[0093] Among them, obstacle avoidance planning refers to planning a safe path according to obstacles and other relevant information in the target area during the movement of the drainage robot, so that the drainage robot can avoid obstacles and reach the target location.

[0094] Among them, obstacle avoidance planning can use artificial potential field method for path planning.

[0095] Among them, the target obstacle avoidance path can be a safe movement path generated by combining the gravitational field and the repulsive field.

[0096] In one embodiment, obstacle avoidance planning is performed according to the track threat coefficient corresponding to each obstacle to obtain a target obstacle avoidance path of the drainage robot, including: obtaining the position information of each obstacle; calculating the position information of each obstacle and the position information of the drainage robot according to a fourth preset formula to obtain a gravitational field value corresponding to each obstacle; calculating the position information of each obstacle and the position information of the drainage robot according to a fifth preset formula using the track threat coefficient corresponding to each obstacle to obtain a repulsive field value corresponding to each obstacle; obtaining an obstacle avoidance path for each obstacle based on the gravitational field value corresponding to each obstacle and the repulsive field value corresponding to each obstacle; and obtaining the target obstacle avoidance path of the drainage robot by combining the obstacle avoidance paths of all obstacles.

[0097] The position information of each obstacle can be obtained by constructing a target three-dimensional rectangular coordinate system based on the current position of the drainage robot as the origin to obtain the position information of each obstacle.

[0098] Among them, the repulsive field value reflects the repulsive force of obstacles on the drainage robot, which is used to drive the drainage robot away from obstacles and avoid collision.

[0099] Among them, the fourth preset formula is as follows:

[0100] ;

[0101] The above formula, Indicates the location of the drainage robot, Represents a flood drainage robot Location and obstacles The Euclidean distance between two points is the straight-line distance between two points. The vector direction is from the position of the drainage robot to the position of the obstacle. Represents the proportional gain coefficient, which controls the strength of the gravitational field. The value is 1. Represents the repulsive field value, n represents the nth obstacle; for obstacles It is the hypothetical target point when avoiding a single obstacle. The target point is the shortest distance point to return to the elliptical orbit after avoiding the obstacle. When the obstacle avoidance trajectory is unknown, the maximum obstacle avoidance diameter is used for limitation, that is, the maximum probability elliptical range diameter of the obstacle plus the distance between the drainage robot and the obstacle.

[0102] The gravitational field value is a square function of the distance from the drainage robot to the obstacle. The greater the distance, the smaller the gravitational field value. The purpose of the gravitational field is to guide the drainage robot to approach the obstacle avoidance target point.

[0103] Among them, the fifth preset formula is as follows:

[0104] ;

[0105] In the above formula, Indicates the location of the obstacle. Indicates that the repulsive force proportional gain coefficient is 1, represents a vector, the direction is from the obstacle to the drainage robot, and the magnitude is the Euclidean distance between the drainage robot and the obstacle. is a constant, indicating the maximum impact range of obstacles on the drainage robot. Set to the maximum active range diameter of the predicted probability ellipse for obstacles , represents the gravitational field value, r represents the radius of the predicted probability ellipse, Indicates the maximum radius; is the repulsive force formula of the artificial potential field, Indicates The greater the threat coefficient of the track, the greater the repulsion between the robot and the obstacle. It means that the repulsive force increases as the threat coefficient of the obstacle increases.

[0106] In the process of obtaining the obstacle avoidance path of each obstacle according to the gravitational field value corresponding to each obstacle and the repulsive field value corresponding to each obstacle, the gravitational field value and the repulsive field value corresponding to each obstacle can be added to obtain the resultant potential field value corresponding to each obstacle.

[0107] Among them, by calculating the resultant potential field, the influence range of each obstacle and the obstacle avoidance strategy that the drainage robot should adopt can be determined, so the drainage robot can approach the target while avoiding obstacles.

[0108] Among them, in the process of combining the obstacle avoidance paths of all obstacles to obtain the target obstacle avoidance path of the drainage robot, it is equivalent to combining the resultant potential field values ​​of all obstacles. In the resultant potential field, the drainage robot will automatically calculate the gravity value and repulsion value of the current position during the movement, and avoid obstacles and move towards the target by adjusting the direction of movement. Due to the properties of the potential field function, the robot will generate a collision-free path under the combined action of gravity and repulsion.

[0109] It can be seen that in this embodiment, by obtaining the position information of each obstacle and based on the calculation of the gravitational field and the repulsive field, the obstacle avoidance path of each obstacle can be accurately generated, thereby effectively integrating the obstacle avoidance paths of all obstacles, optimizing the target obstacle avoidance path of the drainage robot, and improving its navigation ability and safety in complex environments.

[0110] The present invention has the following beneficial effects: the present invention can collect more comprehensive and accurate data in different environments and conditions through the combined use of laser radar sensors and sonar detection sensors. The laser radar sensor provides high-precision direct-view distance measurement, while the sonar detection sensor can work effectively in an environment with obstructed vision or underwater, which enables the drainage robot to efficiently collect data in a complex and changeable environment; further, by combining the analysis and processing of the first point cloud data set and the second point cloud data set, the data error scale is accurately evaluated, and obstacles can be effectively identified and located; further, by calculating the target error growth coefficient and the track threat coefficient of the obstacle, the system can accurately evaluate the actual threat level of the obstacle to the drainage robot, and perform effective obstacle avoidance planning based on the actual threat level, ensuring that the drainage robot can safely and efficiently perform drainage tasks.

[0111] In one embodiment, determining the orbit threat coefficient corresponding to each obstacle according to the motion trajectory corresponding to each obstacle in the at least one obstacle and the target error growth coefficient includes: determining the spatial velocity fluctuation scale of each obstacle according to the motion trajectory corresponding to each obstacle in the at least one obstacle and the target error growth coefficient; constructing a target error ellipsoid area according to the spatial velocity fluctuation scale of each obstacle; and determining the orbit threat coefficient corresponding to each obstacle according to the target error ellipsoid area.

[0112] Among them, the spatial speed fluctuation scale refers to the speed fluctuation degree of the obstacle in three-dimensional space. The spatial speed fluctuation scale includes the speed fluctuation scale of the X-axis, the speed fluctuation scale of the Y-axis, and the speed fluctuation scale of the Z-axis. In the three-dimensional rectangular coordinate system, the direction from due west to due east can be set as the X-axis, the direction from due south to due north can be set as the Y-axis, and the height can be set as the Z-axis.

[0113] The target error ellipsoid is constructed based on the velocity fluctuation scale of each obstacle in three-dimensional space to describe the uncertainty and range of motion of the obstacle in space. The target error ellipsoid represents the possible position range of the obstacle, with its center at the current position of the obstacle.

[0114] The three-axis length of the target error ellipsoid area represents the position uncertainty of the obstacle in each direction. For example, if the velocity fluctuation scale on the X-axis is large, the target error ellipsoid area will be longer in the X-axis direction, which means that the position uncertainty in the east-west direction (X-axis) is high.

[0115] Among them, by constructing the target error ellipsoid area, the potential threat of obstacles to the path of the drainage robot can be evaluated. If the preset path of the drainage robot overlaps with the target error ellipsoid area, the path may need to be adjusted to avoid the potential collision area.

[0116] It can be seen that in this embodiment, the motion characteristics and error range of each obstacle are analyzed to construct the error ellipsoid area, and the track threat coefficient is further calculated, which helps to accurately evaluate the potential threat of obstacles to the path of the drainage robot and provide a more reliable obstacle avoidance strategy for the drainage robot.

[0117] In one embodiment, the determining of the spatial velocity fluctuation scale of each obstacle according to the motion trajectory corresponding to each obstacle in the at least one obstacle and the target error growth coefficient includes: obtaining the motion trajectory corresponding to each obstacle in the at least one obstacle within the acquisition time period; extracting data on the motion trajectory corresponding to each obstacle in a target three-dimensional rectangular coordinate system to obtain the velocity value of each acquisition batch in the X-axis direction, the velocity value of each acquisition batch in the Y-axis direction, and the velocity value of each acquisition batch in the Z-axis direction of each obstacle in the target three-dimensional rectangular coordinate system, wherein the target three-dimensional rectangular coordinate system is a three-dimensional rectangular coordinate system constructed with the current position of the drainage robot as the origin; according to The second preset formula is used to calculate the target error growth coefficient and the speed value of each acquisition batch in the X-axis direction to obtain the speed fluctuation scale in the X-axis direction; according to the second preset formula, the target error growth coefficient and the speed value of each acquisition batch in the Y-axis direction are calculated to obtain the speed fluctuation scale in the Y-axis direction; according to the second preset formula, the target error growth coefficient and the speed value of each acquisition batch in the Z-axis direction are calculated to obtain the speed fluctuation scale in the Z-axis direction; the spatial speed fluctuation scale of each obstacle is obtained by combining the speed fluctuation scale in the X-axis direction, the speed fluctuation scale in the Y-axis direction, and the speed fluctuation scale in the Z-axis direction.

[0118] In the process of obtaining the corresponding motion trajectory of each obstacle during the acquisition time period, the position data of the obstacle can be obtained through laser radar, sonar or other sensors, and the motion path of the obstacle can be calculated.

[0119] The speed value refers to the instantaneous speed of the obstacle in the X-axis, Y-axis and Z-axis directions, and the speed value can be obtained by differentiating the position data of the obstacle.

[0120] The data extraction process is to extract the speed data of each obstacle in the three directions of X, Y, and Z in the target three-dimensional rectangular coordinate system. Specifically, the speed values ​​of the obstacles in the directions of X, Y, and Z in each acquisition batch are extracted.

[0121] Among them, the speed fluctuation scale refers to the degree of speed change of an obstacle in a certain direction in three-dimensional space, reflecting the instability of the speed in that direction.

[0122] The second preset formula is as follows: The following is a calculation formula for the speed fluctuation scale in the X-axis direction:

[0123] ;

[0124] In the above formula, Indicates The speed fluctuation scale of an obstacle on the X-axis, Indicates the data collection batches obtained by the drainage robot in a stationary state; Indicates the first The speed value of the obstacle in the tth batch in the X direction; Indicates The average velocity of the targets in the target area in the X direction; Indicates The average distance between obstacles and the drainage robot, It represents the target error growth coefficient of the acquisition point under unit distance growth; It indicates the difference degree of the speed at each moment. When the corresponding difference degree is larger, it means that the speed fluctuation of the obstacle is larger. We obtain the overall fluctuation size of the wooden template area and normalize it to obtain the speed fluctuation scale of the obstacle. It indicates the acquisition error scale of the corresponding obstacle. As the distance increases, the speed of the corresponding obstacle will cause a larger speed fluctuation scale due to data deviation. Therefore, the speed fluctuation scale in the X-axis direction is obtained by removing the overall error scale from the corresponding result.

[0125] Among them, in the second preset formula In, due to Indicates the data collection batch collected by the drainage robot in a stationary state. If the drainage robot does not collect any data, then =0; due to Indicates The average distance between obstacles and the drainage robot, if If it is zero, it means that the obstacle coincides with the drainage robot; Represents the target error growth coefficient of the acquisition point under unit distance growth. If 0, indicating no error growth. The speed fluctuation scale in the Y-axis direction can be calculated by replacing X in the second preset formula with Y; the speed fluctuation scale in the Z-axis direction can be calculated by replacing X in the second preset formula with Z. The speed fluctuation scale in a single direction is calculated according to the second preset formula.

[0126] Among them, the overall speed fluctuation scale of each obstacle in three-dimensional space is obtained by combining the speed fluctuation scales in the X, Y, and Z directions, which reflects the overall speed fluctuation of the obstacle in space.

[0127] It can be seen that in this embodiment, by acquiring and analyzing the motion trajectory and speed fluctuation scale of each obstacle in the target three-dimensional rectangular coordinate system, the dynamic change range of the obstacle in the three-dimensional space can be accurately evaluated, thereby providing an important reference for the precise obstacle avoidance path planning of the drainage robot.

[0128] In one embodiment, constructing a target error ellipsoid area according to the spatial speed fluctuation scale of each obstacle includes: calculating the acquisition time period with the speed fluctuation scale in the X-axis direction, the speed fluctuation scale in the Y-axis direction, and the speed fluctuation scale in the Z-axis direction in the spatial speed fluctuation scale, respectively, to obtain a first probability radius set in the X-axis direction, a second probability radius set in the Y-axis direction, and a third probability radius set in the Z-axis direction in the acquisition time period; constructing a target error ellipsoid area according to the first probability radius set, the second probability radius set, and the third probability radius set, wherein the target error ellipsoid area represents the operating area of ​​the drainage robot.

[0129] Among them, obtain each time point t in the acquisition time period, calculate the product of the speed fluctuation scale of the obstacle in the X direction and each time point, and obtain the first probability radius in the X-axis direction corresponding to each time point ; Calculate the product of the speed fluctuation scale of the obstacle in the Y direction and each time point to obtain the second probability radius in the Y axis direction corresponding to each time point ; Calculate the product of the speed fluctuation scale of the obstacle in the Z direction and each time point to obtain the third probability radius in the Z axis direction corresponding to each time point .

[0130] Among them, the first probability radius of each time point in the collection time period is summarized to obtain a first probability radius set; the second probability radius of each time point in the collection time period is summarized to obtain a second probability radius set; the third probability radius of each time point in the collection time period is summarized to obtain a third probability radius set.

[0131] Among them, a three-dimensional target error ellipsoid area is constructed by using the probability radius set of each direction calculated above and combining the position estimation of the obstacle at different time points in the acquisition time period.

[0132] It can be seen that in this embodiment, the probability radius is calculated by combining the spatial velocity fluctuation scale with time, and the target error ellipsoid area is constructed based on this, so as to accurately predict the dynamic range of obstacles in the future time period, which can further help the drainage robot to perform obstacle avoidance planning in complex environments and ensure safe operation.

[0133] In one embodiment, determining the track threat coefficient corresponding to each obstacle according to the target error ellipsoid area includes: obtaining a preset track of the drainage robot; judging whether each obstacle has an intersection with the preset track; if no intersection exists, determining that the track threat coefficient of the obstacle that does not currently have an intersection is 0; or, if an intersection exists, determining that the obstacle that has an intersection is a first obstacle; judging whether the first obstacle is in an intersection time period according to the target error ellipsoid area; if not, determining that the track threat coefficient of the first obstacle that is not in the intersection time period is 0; or, if yes, determining that the first obstacle in the intersection time period is a second obstacle; calculating the operating parameters corresponding to the second obstacle according to a third preset formula to obtain the track threat coefficient of the second obstacle.

[0134] The preset track may be set manually or according to a path without obstacles, and is not limited here.

[0135] Among them, if the trajectory of an obstacle has no intersection with the preset track, it can be determined that the obstacle has no effect on the current path, and its track threat coefficient is set to 0.

[0136] The first obstacle is an obstacle whose trajectory intersects with the preset track.

[0137] Among them, if the first obstacle is within the time period that intersects with the track of the drainage robot, the obstacle is called the second obstacle.

[0138] The third preset formula can be expressed as follows:

[0139] ;

[0140] In the above formula, Indicates The first obstacle The threat factor at that moment, It means the drainage robot. Indicates The obstacle is in The position coordinates at a moment, Represents the calculation of the drainage robot and the The collision distance of obstacles; Indicates The maximum radius of the obstacle's activity range at a given moment, It represents the derivative of the maximum radius. The derivative result is 0 or a positive number. The larger it is, the larger the range of obstacles is. The greater the threat to the waterway is. Finally, the threat coefficient of obstacles to the movement track of the drainage robot at a single moment is obtained by normalizing the scale as a whole. Indicates all the time The track threat factor of obstacles is It represents the corresponding fault tolerance time, that is, the time range of the intersection of the error ellipse of the drainage robot reaching the obstacle and the obstacle. The threat coefficient of a single obstacle at each moment within all the fault tolerance times is accumulated and averaged to obtain the track threat coefficient of a single obstacle.

[0141] The track threat coefficient of the second obstacle is the track threat coefficient of the obstacle that represents the threat, and the track threat coefficient corresponding to each obstacle can be obtained by adding other obstacles whose track threat coefficients are judged to be 0. Therefore, the track threat coefficient corresponding to each obstacle is equal to the track threat coefficient of each second obstacle + 0.

[0142] It can be seen that in this embodiment, by judging whether the obstacle intersects with the preset track of the drainage robot, its potential threat to the path is further analyzed. For obstacles with intersection, its threat coefficient at different times is calculated to evaluate its overall threat to the movement track of the drainage robot. This ensures that the drainage robot can accurately consider the impact of each obstacle on its running trajectory when performing path planning and obstacle avoidance, thereby improving its safety and efficiency in complex environments.

[0143] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0144] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0145] It should be noted that for all calculation formulas in this embodiment, when there is an extreme situation that causes the denominator of the fraction in the formula to be zero, a preset value is added to the denominator of the fraction. The preset value can be 0.01 as an empirical value to avoid the situation where the denominator of the fraction is zero.

[0146] As another aspect of the embodiment of the present invention, the embodiment of the present invention provides an intelligent obstacle avoidance device for a drainage robot. The intelligent obstacle avoidance device for the drainage robot can be a software module, which includes several instructions stored in a memory, and the processor can access the memory and call the instructions for execution to complete the intelligent obstacle avoidance method for the drainage robot described in the above embodiments.

[0147] See also Figure 3 , Figure 3 FIG. 1 is a schematic diagram of the structure of an intelligent obstacle avoidance device of a flood drainage robot provided by an embodiment of the present invention. Figure 3 As shown, the intelligent obstacle avoidance device 300 of the drainage robot includes:

[0148] The acquisition unit 301 is used to control the laser radar sensor and the sonar detection sensor that meet the preset acquisition conditions to perform data acquisition on multiple acquisition points in the target area to obtain a first point cloud data set and a second point cloud data set, wherein the first point cloud data set is point cloud data collected in multiple batches by the laser radar sensor within a collection time period, and the second point cloud data set is point cloud data collected in multiple batches by the sonar detection sensor within a collection time period;

[0149] A determining unit 302, configured to determine a data error scale of each of the plurality of acquisition points according to the first point cloud data set and the second point cloud data set;

[0150] The determining unit 302 is further configured to obtain a target error growth coefficient according to the data error scale of each acquisition point, wherein the target error growth coefficient is used to evaluate the trend of the overall data error scale corresponding to the multiple acquisition points changing with distance;

[0151] A processing unit 303 is used to perform data processing on the target point cloud data in the first point cloud data set to obtain at least one obstacle in the target point cloud data, wherein the target point cloud data is point cloud data in a target acquisition time period;

[0152] The determining unit 302 is further configured to determine a track threat coefficient corresponding to each obstacle in the at least one obstacle according to the motion trajectory corresponding to each obstacle and the target error growth coefficient;

[0153] The planning unit 304 is used to perform obstacle avoidance planning according to the track threat coefficient corresponding to each obstacle to obtain a target obstacle avoidance path of the drainage robot.

[0154] The present invention has the following beneficial effects: the present invention can collect more comprehensive and accurate data in different environments and conditions through the combined use of laser radar sensors and sonar detection sensors. The laser radar sensor provides high-precision direct-view distance measurement, while the sonar detection sensor can work effectively in an environment with obstructed vision or underwater, which enables the drainage robot to efficiently collect data in a complex and changeable environment; further, by combining the analysis and processing of the first point cloud data set and the second point cloud data set, the data error scale is accurately evaluated, and obstacles can be effectively identified and located; further, by calculating the target error growth coefficient and the track threat coefficient of the obstacle, the system can accurately evaluate the actual threat level of the obstacle to the drainage robot, and perform effective obstacle avoidance planning based on the actual threat level, ensuring that the drainage robot can safely and efficiently perform drainage tasks.

[0155] In one embodiment, in determining the data error scale of each of the multiple acquisition points based on the first point cloud data set and the second point cloud data set, the determining unit 302 is further used to: obtain the total number of target acquisition batches in the first point cloud data set, wherein the total number of target acquisition batches is equal to the total number of acquisition batches in the second point cloud data set; obtain a first curvature value corresponding to a first acquisition point among the multiple acquisition points in the first point cloud data set; obtain a second curvature value of the first acquisition point in the second point cloud data set; perform error size calculation on the total number of target acquisition batches, the first curvature value, and the second curvature value according to a first preset formula to obtain the data error scale of the first acquisition point; perform error size calculation on each of the multiple acquisition points to obtain the data error scale of each of the multiple acquisition points.

[0156] In one embodiment, in obtaining the target error growth coefficient based on the data error scale of each collection point, the determination unit 302 is further used to: construct a target three-dimensional rectangular coordinate system with the current position of the drainage robot as the origin, and obtain the first distance between each collection point and the drainage robot according to the target three-dimensional rectangular coordinate system; obtain the distance error scale coefficient corresponding to each collection point according to the data error scale of each collection point and the first distance between each collection point and the drainage robot; calculate the distance error scale coefficients corresponding to all collection points to obtain the target error growth coefficient.

[0157] In one embodiment, in determining the orbit threat coefficient corresponding to each obstacle according to the motion trajectory corresponding to each obstacle in the at least one obstacle and the target error growth coefficient, the determination unit 302 is further used to: determine the spatial velocity fluctuation scale of each obstacle according to the motion trajectory corresponding to each obstacle in the at least one obstacle and the target error growth coefficient; construct a target error ellipsoid area according to the spatial velocity fluctuation scale of each obstacle; and determine the orbit threat coefficient corresponding to each obstacle according to the target error ellipsoid area.

[0158] In one embodiment, in determining the spatial velocity fluctuation scale of each obstacle according to the motion trajectory corresponding to each obstacle in the at least one obstacle and the target error growth coefficient, the determination unit 302 is further used to: obtain the motion trajectory corresponding to each obstacle in the at least one obstacle within the collection time period; extract data on the motion trajectory corresponding to each obstacle in the target three-dimensional rectangular coordinate system to obtain the velocity value of each collection batch in the X-axis direction, the velocity value of each collection batch in the Y-axis direction, and the velocity value of each collection batch in the Z-axis direction of each obstacle in the target three-dimensional rectangular coordinate system, wherein the target three-dimensional rectangular coordinate system is a three-dimensional rectangular coordinate system constructed with the current position of the drainage robot as the origin. angular coordinate system; according to the second preset formula, the target error growth coefficient and the speed value of each collection batch in the X-axis direction are calculated to obtain the speed fluctuation scale in the X-axis direction; according to the second preset formula, the target error growth coefficient and the speed value of each collection batch in the Y-axis direction are calculated to obtain the speed fluctuation scale in the Y-axis direction; according to the second preset formula, the target error growth coefficient and the speed value of each collection batch in the Z-axis direction are calculated to obtain the speed fluctuation scale in the Z-axis direction; combining the speed fluctuation scale in the X-axis direction, the speed fluctuation scale in the Y-axis direction, and the speed fluctuation scale in the Z-axis direction, the spatial speed fluctuation scale of each obstacle is obtained.

[0159] In one embodiment, in the target error ellipsoid area constructed according to the spatial speed fluctuation scale of each obstacle, the determination unit 302 is further used to: calculate the acquisition time period with the speed fluctuation scale in the X-axis direction, the speed fluctuation scale in the Y-axis direction, and the speed fluctuation scale in the Z-axis direction in the spatial speed fluctuation scale, respectively, to obtain a first probability radius set in the X-axis direction, a second probability radius set in the Y-axis direction, and a third probability radius set in the Z-axis direction in the acquisition time period; construct a target error ellipsoid area according to the first probability radius set, the second probability radius set, and the third probability radius set, and the target error ellipsoid area represents the operating area of ​​the drainage robot.

[0160] In one embodiment, in determining the track threat coefficient corresponding to each obstacle according to the target error ellipsoid area, the determination unit 302 is further used to: obtain the preset track of the drainage robot; determine whether each obstacle has an intersection with the preset track; if no intersection exists, determine that the track threat coefficient of the obstacle that does not currently have an intersection is 0; or, if an intersection exists, determine that the obstacle that has an intersection is a first obstacle; determine whether the first obstacle is in a time period of intersection according to the target error ellipsoid area; if not, determine that the track threat coefficient of the first obstacle that is not in the time period of intersection is 0; or, if yes, determine that the first obstacle in the time period of intersection is a second obstacle; calculate the operating parameters corresponding to the second obstacle according to a third preset formula to obtain the track threat coefficient of the second obstacle.

[0161] In one embodiment, in performing obstacle avoidance planning according to the track threat coefficient corresponding to each obstacle to obtain the target obstacle avoidance path of the drainage robot, the planning unit 304 is further used to: obtain the position information of each obstacle; calculate the position information of each obstacle and the position information of the drainage robot according to the fourth preset formula to obtain the gravitational field value corresponding to each obstacle; calculate the position information of each obstacle and the position information of the drainage robot according to the fifth preset formula using the track threat coefficient corresponding to each obstacle to obtain the repulsive field value corresponding to each obstacle; obtain the obstacle avoidance path of each obstacle according to the gravitational field value corresponding to each obstacle and the repulsive field value corresponding to each obstacle; and obtain the target obstacle avoidance path of the drainage robot by combining the obstacle avoidance paths of all obstacles.

[0162] In one embodiment, before the control of the laser radar sensor and the sonar detection sensor that meet the preset collection conditions performs data collection, the collection unit 301 is further used to: perform time debugging on the laser radar sensor and the sonar detection sensor to obtain that the monitoring frequencies of the laser radar sensor and the sonar detection sensor are the same time reference; construct a target three-dimensional rectangular coordinate system with the current position of the drainage robot as the origin; perform coordinate conversion on the coordinate system of the laser radar sensor and the coordinate system of the sonar detection sensor to obtain the laser radar sensor and the sonar detection sensor in the target three-dimensional rectangular coordinate system, so that the first point cloud data set corresponding to the laser radar sensor and the second point cloud data set corresponding to the sonar detection sensor are in the same coordinate system; when it is detected that the monitoring frequencies of the laser radar sensor and the sonar detection sensor are the same time reference, and the laser radar sensor and the sonar detection sensor have completed the coordinate conversion, it is determined that the laser radar sensor and the sonar detection sensor meet the preset collection conditions.

[0163] It should be noted that the intelligent obstacle avoidance device of the above-mentioned drainage robot can execute the intelligent obstacle avoidance method of the drainage robot provided in the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not fully described in the embodiment of the intelligent obstacle avoidance device of the drainage robot, please refer to the intelligent obstacle avoidance method of the drainage robot provided in the embodiment of the present invention.

[0164] See also Figure 4 , Figure 4 Schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Figure 4 As shown, the electronic device 400 includes a processor 401 and a memory 402. The processor 401 is in communication connection with the memory 402.

[0165] The processor 401 is configured to support the electronic device to perform the corresponding functions in the intelligent obstacle avoidance method of the drainage robot in the above method embodiment. The processor 401 can be a central processing unit (CPU), a network processor (NP), a hardware chip or any combination thereof. The above hardware chip can be an application specific integrated circuit (ASIC), a programmable logic device (PLD) or a combination thereof. The above PLD can be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL) or any combination thereof.

[0166] Specifically, the processor 401 may include a sending card, a receiving card and a driving chip.

[0167] The memory 402 is used to store program codes, etc. The memory 402 may include a volatile memory (VM), such as a random access memory (RAM); the memory 402 may also include a non-volatile memory (NVM), such as a read-only memory (ROM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD); the memory 402 may also include a combination of the above-mentioned types of memory.

[0168] An embodiment of the present invention also provides a computer-readable storage medium, which stores a computer program. The computer program includes program instructions, and when the program instructions are executed by a computer, the computer executes the intelligent obstacle avoidance method for the drainage robot as described in the above embodiments.

[0169] A person skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0170] The above disclosure is only the preferred embodiment of the present invention, and certainly cannot be used to limit the scope of rights of the present invention.

Claims

1. An intelligent obstacle avoidance method for a flood drainage robot, characterized in that: An intelligent obstacle avoidance system applied to a flood drainage robot, the intelligent obstacle avoidance system of the flood drainage robot comprising a laser radar sensor and a sonar detection sensor, and the method comprising: Control the laser radar sensor and the sonar detection sensor that meet the preset collection conditions to collect data from multiple collection points in the target area to obtain a first point cloud data set and a second point cloud data set, wherein the first point cloud data set is point cloud data collected in multiple batches by the laser radar sensor within a collection time period, and the second point cloud data set is point cloud data collected in multiple batches by the sonar detection sensor within a collection time period; Determining a data error scale of each of the plurality of acquisition points according to the first point cloud data set and the second point cloud data set; According to the data error scale of each acquisition point, a target error growth coefficient is obtained, wherein the target error growth coefficient is used to evaluate the trend of the overall data error scale corresponding to the multiple acquisition points changing with distance; Performing data processing on target point cloud data in the first point cloud data set to obtain at least one obstacle in the target point cloud data, wherein the target point cloud data is point cloud data in a target acquisition time period; Determining a track threat coefficient corresponding to each obstacle in the at least one obstacle according to the motion trajectory corresponding to each obstacle and the target error growth coefficient; Obstacle avoidance planning is performed according to the track threat coefficient corresponding to each obstacle to obtain a target obstacle avoidance path of the drainage robot; Determining the data error scale of each of the plurality of acquisition points according to the first point cloud data set and the second point cloud data set includes: Acquire a total number of target acquisition batches in the first point cloud dataset, where the total number of target acquisition batches is equal to the total number of acquisition batches in the second point cloud dataset; Acquire a first curvature value corresponding to a first acquisition point among the multiple acquisition points in the first point cloud data set; Acquire a second curvature value of the first acquisition point in the second point cloud data set; According to a first preset formula, an error size calculation is performed on the total number of the target acquisition batches, the first curvature value, and the second curvature value to obtain a data error scale of the first acquisition point; An error size calculation is performed on each of the multiple acquisition points to obtain a data error scale of each of the multiple acquisition points.

2. The intelligent obstacle avoidance method for a flood drainage robot according to claim 1, characterized in that: The target error growth coefficient is obtained according to the data error scale of each acquisition point, including: Taking the current position of the drainage robot as the origin, constructing a target three-dimensional rectangular coordinate system; According to the target three-dimensional rectangular coordinate system, obtaining a first distance between each collection point and the drainage robot; Obtaining a distance error scale coefficient corresponding to each collection point according to the data error scale of each collection point and the first distance between each collection point and the drainage robot; The distance error scale coefficients corresponding to all the acquisition points are calculated to obtain the target error growth coefficient.

3. The intelligent obstacle avoidance method for a flood drainage robot according to claim 1, characterized in that: The determining, according to the motion trajectory corresponding to each obstacle of the at least one obstacle and the target error growth coefficient, a track threat coefficient corresponding to each obstacle includes: Determining a spatial velocity fluctuation scale of each obstacle according to a motion trajectory corresponding to each obstacle in the at least one obstacle and the target error growth coefficient; According to the spatial velocity fluctuation scale of each obstacle, a target error ellipsoid region is constructed; The orbit threat coefficient corresponding to each obstacle is determined according to the target error ellipsoid area.

4. The intelligent obstacle avoidance method for a flood drainage robot according to claim 3, characterized in that: The determining, according to the motion trajectory corresponding to each obstacle in the at least one obstacle and the target error growth coefficient, the spatial velocity fluctuation scale of each obstacle comprises: Acquire a motion trajectory corresponding to each obstacle of the at least one obstacle within the acquisition time period; In the target three-dimensional rectangular coordinate system, data is extracted for the motion trajectory corresponding to each obstacle to obtain the speed value of each collected batch in the X-axis direction, the speed value of each collected batch in the Y-axis direction, and the speed value of each collected batch in the Z-axis direction of each obstacle in the target three-dimensional rectangular coordinate system, wherein the target three-dimensional rectangular coordinate system is a three-dimensional rectangular coordinate system constructed with the current position of the drainage robot as the origin; According to a second preset formula, the target error growth coefficient and the speed value of each acquisition batch in the X-axis direction are calculated to obtain a speed fluctuation scale in the X-axis direction; According to the second preset formula, the target error growth coefficient and the speed value of each acquisition batch in the Y-axis direction are calculated to obtain the speed fluctuation scale in the Y-axis direction; According to the second preset formula, the target error growth coefficient and the speed value of each acquisition batch in the Z-axis direction are calculated to obtain the speed fluctuation scale in the Z-axis direction; The spatial speed fluctuation scale of each obstacle is obtained by combining the speed fluctuation scale in the X-axis direction, the speed fluctuation scale in the Y-axis direction, and the speed fluctuation scale in the Z-axis direction.

5. The intelligent obstacle avoidance method for a flood drainage robot according to claim 4, characterized in that: The target error ellipsoid region is constructed according to the spatial velocity fluctuation scale of each obstacle, including: The acquisition time period is calculated respectively with the speed fluctuation scale in the X-axis direction, the speed fluctuation scale in the Y-axis direction, and the speed fluctuation scale in the Z-axis direction in the spatial speed fluctuation scale to obtain a first probability radius set in the X-axis direction, a second probability radius set in the Y-axis direction, and a third probability radius set in the Z-axis direction in the acquisition time period; According to the first probability radius set, the second probability radius set and the third probability radius set, a target error ellipsoid area is constructed, and the target error ellipsoid area represents the operating area of ​​the drainage robot.

6. The intelligent obstacle avoidance method for a flood drainage robot according to claim 3, characterized in that: Determining the track threat coefficient corresponding to each obstacle according to the target error ellipsoid area includes: Obtaining a preset track of the drainage robot; Determining whether each obstacle intersects with the preset track; If there is no intersection, the track threat coefficient of the obstacle that does not currently have an intersection is determined to be 0; or, If there is an intersection, the obstacle with the intersection is determined to be the first obstacle; Determining whether the first obstacle is in an intersection time period according to the target error ellipsoid area; If not, the track threat coefficient of the first obstacle in the time period not in the intersection is determined to be 0; or, If so, the first obstacle in the time period of the intersection is determined to be the second obstacle; The operating parameters corresponding to the second obstacle are calculated according to a third preset formula to obtain a track threat coefficient of the second obstacle.

7. The intelligent obstacle avoidance method for a flood drainage robot according to claim 1, characterized in that: The obstacle avoidance planning is performed according to the track threat coefficient corresponding to each obstacle to obtain the target obstacle avoidance path of the drainage robot, including: Obtaining the location information of each obstacle; According to a fourth preset formula, the position information of each obstacle and the position information of the drainage robot are calculated to obtain a gravitational field value corresponding to each obstacle; According to the fifth preset formula, the track threat coefficient corresponding to each obstacle is used to calculate the position information of each obstacle and the position information of the drainage robot to obtain the repulsive field value corresponding to each obstacle; Obtaining an obstacle avoidance path for each obstacle according to the gravitational field value corresponding to each obstacle and the repulsive field value corresponding to each obstacle; The obstacle avoidance paths of all obstacles are combined to obtain the target obstacle avoidance path of the flood drainage robot.

8. The intelligent obstacle avoidance method for a flood drainage robot according to claim 1, characterized in that: Before the laser radar sensor and the sonar detection sensor that control the preset collection conditions to collect data, the method further includes: Performing time adjustment on the laser radar sensor and the sonar detection sensor to obtain that the monitoring frequencies of the laser radar sensor and the sonar detection sensor are the same time reference; Taking the current position of the drainage robot as the origin, constructing a target three-dimensional rectangular coordinate system; Performing coordinate conversion on the coordinate system of the laser radar sensor and the coordinate system of the sonar detection sensor to obtain the laser radar sensor and the sonar detection sensor in the target three-dimensional rectangular coordinate system, so that the first point cloud data set corresponding to the laser radar sensor and the second point cloud data set corresponding to the sonar detection sensor are in the same coordinate system; When it is detected that the monitoring frequencies of the laser radar sensor and the sonar detection sensor are the same time reference, and the laser radar sensor and the sonar detection sensor have completed the coordinate conversion, it is determined that the laser radar sensor and the sonar detection sensor meet the preset collection conditions.

9. An intelligent obstacle avoidance system for a flood drainage robot, characterized in that: It includes a memory and a processor, the memory is connected to the processor, the processor is used to execute one or more computer programs stored in the memory, and when the processor executes the one or more computer programs, the intelligent obstacle avoidance system of the drainage robot implements the intelligent obstacle avoidance method of the drainage robot as described in any one of claims 1-8.

Images