An AGV adaptive turning obstacle avoidance device based on artificial intelligence

Abstract

The present invention provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, which includes: a road condition detection module, which is used to collect a target road image, and pre-process the target road image to determine the boundary line of the target road and the location information of obstacles; a path planning module, which is used to determine the turning type of the AGV based on the boundary line of the target road and the location information of obstacles, and plan the turning path of the AGV based on the turning type; an adaptive turning module, which is used to determine the turning speed of the AGV based on the planning result, and complete the AGV adaptive turning obstacle avoidance based on the turning speed. By analyzing the target road image, the road condition and the location of obstacles on the road are accurately obtained, which provides convenience for determining the turning type and turning path of the AGV, and at the same time, the turning speed of the AGV is set according to the turning path, which improves the safety factor and reliability factor of the AGV adaptive turning obstacle avoidance.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an AGV self-adaptive turning obstacle avoidance device based on artificial intelligence. Background Art

[0002] At present, an automated guided vehicle, or AGV for short, refers to a transport vehicle equipped with an electromagnetic or optical automatic guidance device that can travel along a specified guidance path and has safety protection and various transfer functions;

[0003] However, in the prior art, when the AGV turns, a turning reminder point is usually set on the guide medium before the turning section to remind the vehicle that there is a turning section ahead. When there is an obstacle in front, it is impossible to perform obstacle avoidance operations during the turning process. At the same time, the set turning type is fixed and the current turning type cannot be analyzed according to the current driving conditions. When turning, the speed must be reduced to zero before changing the forward direction. The operation process is complicated and time-consuming.

[0004] Therefore, the present invention provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, which is used to accurately obtain the road conditions and the positions of obstacles on the road by analyzing the target road image, thereby providing convenience for determining the turning type and turning path of the AGV, and at the same time setting the turning speed of the AGV according to the turning path, thereby improving the safety factor and reliability factor of the AGV adaptive turning obstacle avoidance. Summary of the invention

[0005] The present invention provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, which is used to accurately obtain the road conditions and the positions of obstacles on the road by analyzing the target road image, thereby providing convenience for determining the turning type and turning path of the AGV, and at the same time setting the turning speed of the AGV according to the turning path, thereby improving the safety factor and reliability factor of the AGV adaptive turning obstacle avoidance.

[0006] The present invention provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, comprising:

[0007] A road condition detection module is used to collect target road images and pre-process the target road images to determine the boundary lines of the target road and the location information of obstacles;

[0008] A path planning module, used to determine the turning type of the AGV based on the boundary line of the target road and the location information of the obstacles, and plan the turning path of the AGV based on the turning type;

[0009] The adaptive turning module is used to determine the turning speed of the AGV based on the planning results, and to complete the adaptive turning obstacle avoidance of the AGV based on the turning speed.

[0010] Preferably, an AGV adaptive turning obstacle avoidance device based on artificial intelligence and a road condition detection module include:

[0011] An image acquisition unit, used for acquiring an initial road image of a target road based on a fisheye camera, and determining a current shooting area of ​​the target road based on the initial road image, wherein the fisheye cameras are respectively arranged at the front, rear, left and right of the AGV;

[0012] An angle adjustment unit, used to correct the shooting angle of the fisheye camera based on the current shooting area, and shoot the target road image based on the correction result to obtain a target road image set;

[0013] An image screening unit, configured to screen the target road image set based on preset rules, wherein the preset rules include an evaluation of the clarity of the target road image and an assessment of the target road boundary in the target road image;

[0014] The image screening unit is further used to retain the target road images in the target road image set whose clarity is greater than or equal to the preset clarity and whose target road boundaries are within the target image area, and to eliminate the target road images in the target road image set that do not meet the preset rules, so as to obtain the target road image set to be analyzed.

[0015] Preferably, an AGV adaptive turning obstacle avoidance device based on artificial intelligence, an image screening unit, comprises:

[0016] An image compression subunit is used to obtain the target road image set to be analyzed, and compress and package the target road image set to be analyzed to obtain a data packet to be transmitted;

[0017] The image transmission subunit is used to place the data packet to be transmitted in a data transmission queue, and send the data packet to be transmitted to the AGV processing chip based on the data transmission queue.

[0018] Preferably, an AGV adaptive turning obstacle avoidance device based on artificial intelligence and a road condition detection module include:

[0019] An image processing unit, used to obtain the collected target road image and divide the target road image into N image blocks;

[0020] The image processing unit is used to set reference pixels in each image block respectively, and determine the range coefficient of the pixel to be denoised based on the component value of the reference pixel, and at the same time, determine the spatial distance between the pixel to be denoised and the reference pixel, and determine the spatial domain coefficient of the pixel to be denoised based on the spatial distance;

[0021] A denoising unit, configured to determine a noise sub-image in each image block based on the range coefficient and the spatial domain coefficient, and to splice the noise sub-images to obtain an image to be denoised;

[0022] The denoising unit is further used to perform denoising on the image to be denoised, and fuse the denoised image with the target road image to obtain a final target road image.

[0023] Preferably, an AGV adaptive turning obstacle avoidance device based on artificial intelligence, a road condition detection module, further includes:

[0024] An image segmentation unit, used to acquire a target road image and determine a road direction of the target road based on the target road image;

[0025] The image segmentation unit is used to determine a target processing section by a road expansion method based on the road direction, and determine a brightness threshold and a color threshold of an image area corresponding to the target processing section;

[0026] The image segmentation unit is used to segment the target road image based on the brightness threshold and the color threshold to obtain a target image corresponding to a target processing section;

[0027] An image color processing unit, used to perform binarization processing on the target image to obtain a binarized image corresponding to the target image, and determine the distribution attributes of white pixels and black pixels in the target image based on the binarized image, wherein the white pixels are non-road areas and the black pixels are target road areas;

[0028] A road boundary determination unit, configured to determine a boundary line between the white pixel points and the black pixel points based on a distribution attribute, and determine a boundary line of the target processing road section based on the boundary line;

[0029] A coordinate system construction unit, used to construct a driving coordinate system based on the right boundary line of the target processing section as a reference line, the AGV as an origin, the cross-sectional direction of the target processing section as a horizontal axis, and the forward direction of the AGV as a vertical axis;

[0030] a position determination unit, configured to determine a relative distance between an obstacle and the AGV and an azimuth of the obstacle relative to the AGV based on the target image, and determine a coordinate value of the obstacle in the vehicle coordinate system based on the relative distance and the azimuth;

[0031] The position determination unit is further used to determine the imaging ratio of the target image, and based on the imaging ratio, convert the coordinate value of the obstacle in the driving coordinate system to obtain the geographical location information of the obstacle on the target road.

[0032] Preferably, an AGV adaptive turning obstacle avoidance device based on artificial intelligence, a path planning module, comprises:

[0033] A width analysis unit, used to construct a road surface coordinate system, obtain a boundary line of the target road, and place the boundary line in the road surface coordinate system, wherein the boundary line includes left and right boundary lines;

[0034] The width analysis unit is used to traverse the valid boundary point pairs of the boundary line based on the road surface coordinate system, and determine the width of the target road based on the coordinate values ​​of the valid boundary point pairs in the road surface coordinate system;

[0035] A road condition analysis unit, configured to obtain a road sign of the target road, and determine curve information of the target road based on the road sign, wherein the curve information includes curve shape parameters and slope parameters;

[0036] The road condition analysis unit is used to determine the path distance between the starting point and the end point of the AGV traveling through the current curve and the curvature of the current curve based on the curve information;

[0037] A turning type determination unit, configured to determine the turning type of the AGV based on the width of the target road, the path distance between the starting point and the end point of the current curve, and the curvature of the current curve, and to obtain the wheelbase information of the AGV;

[0038] A turning angle determination unit, configured to determine a minimum turning angle required for the AGV to perform the turning type based on the wheelbase information of the AGV, and to determine a turning key point of the AGV in performing the turning type based on the minimum turning angle and the position information of the obstacle;

[0039] a path planning unit, configured to determine a lateral deviation between the AGV and the obstacle based on the turning key point, and compare the lateral deviation with a preset threshold;

[0040] If the lateral deviation is less than the preset threshold, it is determined that the AGV cannot turn smoothly, and the turning key point is adjusted until the lateral deviation is greater than or equal to the preset threshold;

[0041] Otherwise, it is determined that the AGV can turn smoothly, and the turning key point and the minimum turning angle are recorded to complete the planning of the turning path of the AGV.

[0042] Preferably, an AGV adaptive turning obstacle avoidance device based on artificial intelligence, a turning angle determination unit, comprises:

[0043] An inflection point determination subunit, used for obtaining a turning key point of the AGV when executing the turning type and a curvature of a current curve, and determining a relative angle between a next inflection point of the turning key point and the turning key point based on the curvature of the current curve;

[0044] The inflection point determination subunit is used to determine a target position of the next inflection point based on the relative angle, and continue to determine the next inflection point based on the target position until all inflection points on the current curve are determined, wherein the inflection point is not unique;

[0045] The route planning subunit is used to generate a turning path according to the positions of all the turning points based on a preset curve generator, so as to complete the planning of the AGV's adaptive turning path.

[0046] Preferably, an AGV adaptive turning obstacle avoidance device based on artificial intelligence, a turning angle determination unit, comprises:

[0047] A turning angle range determination subunit, configured to determine a maximum turning angle of the AGV based on the wheelbase information of the AGV and the curvature of the current curve;

[0048] a distance analysis subunit, configured to perform a pre-simulated turn on the AGV based on the maximum turning angle, and determine, based on the pre-simulated turn, whether a lateral deviation between the AGV and the obstacle at the maximum turning angle is greater than or equal to a preset threshold;

[0049] If it is greater than or equal to the preset threshold, the obtained maximum turning angle is determined to be qualified, and the turning angle range of the AGV when automatically turning to avoid obstacles is determined based on the maximum turning angle and the minimum turning angle;

[0050] Otherwise, the maximum turning angle is decreased until the lateral deviation is greater than or equal to the preset threshold, thereby completing the determination of the turning range of the AGV when automatically turning to avoid obstacles.

[0051] Preferably, an AGV adaptive turning obstacle avoidance device based on artificial intelligence, an adaptive turning module, comprises:

[0052] A speed analysis unit, used to obtain a planning result, and determine the curve radius of the current curve and the turning angle corresponding to the AGV when executing the turning type based on the planning result;

[0053] The speed analysis unit is used to determine the center of gravity of the AGV based on the wheelbase and wheel track of the AGV, and determine the safety centripetal force corresponding to the AGV when executing the turning type based on the curve radius, the turning angle and the center of gravity of the AGV;

[0054] The speed determination unit is used to calculate the wheel moving speed of the AGV when turning based on the safety centripetal force, and convert the wheel moving speed into the moving speed at the center of gravity to complete the determination of the turning speed of the AGV.

[0055] Preferably, an AGV adaptive turning obstacle avoidance device based on artificial intelligence, a speed determination unit, comprises:

[0056] A monitoring subunit, used to obtain in real time the force conditions of the four wheels of the AGV when turning, and determine the safe turning speed of the AGV based on the force conditions;

[0057] a comparison subunit, configured to compare the current turning speed of the AGV with the safe turning speed;

[0058] A speed adjustment subunit, configured to determine that there is a safety hazard in the AGV turning when the current turning speed is greater than the safe turning speed, and determine an adjustment strategy for the AGV turning speed based on the force condition;

[0059] The speed adjustment subunit is used to determine an adjustment threshold for the turning speed based on the adjustment strategy, and adjust the turning speed of the AGV based on the adjustment threshold;

[0060] The monitoring subunit is used to continue monitoring the force conditions of the four wheels of the AGV when turning when the current turning speed is less than or equal to the safe turning speed until the AGV completes adaptive turning and obstacle avoidance.

[0061] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings.

[0062] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0064] Figure 1 This is a structural diagram of an AGV adaptive turning obstacle avoidance device based on artificial intelligence in an embodiment of the present invention;

[0065] Figure 2 This is a structural diagram of a road condition detection module in an AGV adaptive turning obstacle avoidance device based on artificial intelligence in an embodiment of the present invention;

[0066] Figure 3 The structure diagram of an adaptive turning module in an AGV adaptive turning obstacle avoidance device based on artificial intelligence in an embodiment of the present invention. DETAILED DESCRIPTION

[0067] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0068] Embodiment 1:

[0069] This embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, such as Figure 1 As shown, including:

[0070] A road condition detection module is used to collect target road images and pre-process the target road images to determine the boundary lines of the target road and the location information of obstacles;

[0071] A path planning module, used to determine the turning type of the AGV based on the boundary line of the target road and the location information of the obstacles, and plan the turning path of the AGV based on the turning type;

[0072] The adaptive turning module is used to determine the turning speed of the AGV based on the planning results, and to complete the adaptive turning obstacle avoidance of the AGV based on the turning speed.

[0073] In this embodiment, the target road image refers to the current driving section of the AGV, which may be the interior of a factory building or a production workshop, etc.

[0074] In this embodiment, the preprocessing refers to denoising the target road image and correcting the image color deviation, etc., so as to facilitate the determination of the width of the target road and the boundary condition of the target road.

[0075] In this embodiment, the boundary line of the target road refers to the left and right boundaries of the target road.

[0076] In this embodiment, the position information of the obstacle refers to the position of the obstacle in the target road, for example, it may be the edge of the road or other areas.

[0077] In this embodiment, the turn types include right-angle turns and arc turns.

[0078] The beneficial effects of the above technical solution are: by analyzing the target road image, the road conditions and the locations of obstacles on the road can be accurately obtained, which provides convenience for determining the turning type and turning path of the AGV. At the same time, the turning speed of the AGV is set according to the turning path, which improves the safety factor and reliability factor of the AGV adaptive turning obstacle avoidance.

[0079] Embodiment 2:

[0080] Based on the above embodiment 1, this embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, such as Figure 2 As shown, the road condition detection module includes:

[0081] An image acquisition unit, used for acquiring an initial road image of a target road based on a fisheye camera, and determining a current shooting area of ​​the target road based on the initial road image, wherein the fisheye cameras are respectively arranged at the front, rear, left and right of the AGV;

[0082] An angle adjustment unit, used to correct the shooting angle of the fisheye camera based on the current shooting area, and shoot the target road image based on the correction result to obtain a target road image set;

[0083] An image screening unit, configured to screen the target road image set based on preset rules, wherein the preset rules include an evaluation of the clarity of the target road image and an assessment of the target road boundary in the target road image;

[0084] The image screening unit is further used to retain the target road images in the target road image set whose clarity is greater than or equal to the preset clarity and whose target road boundaries are within the target image area, and to eliminate the target road images in the target road image set that do not meet the preset rules, so as to obtain the target road image set to be analyzed.

[0085] In this embodiment, the initial road image refers to an image obtained after the fisheye camera shoots the target road without adjusting the shooting angle, in order to determine the current shooting angle of the fisheye camera.

[0086] In this embodiment, correcting the shooting angle of the fisheye camera based on the current shooting area means that when the area shot by the fisheye camera does not contain the target road or the target road is incomplete, the shooting angle of the fisheye camera is adjusted to ensure that the fisheye camera can capture the complete road.

[0087] In this embodiment, the preset rules include the evaluation of the clarity and the assessment of the target road boundary in the target road image.

[0088] In this embodiment, the preset definition is set in advance and is used to measure whether the definition of the image meets the expected requirements.

[0089] The beneficial effect of the above technical solution is: by collecting images of the target road and screening the collected target road images, it is ensured that the obtained target road images are reliable and effective, which provides a guarantee for accurately analyzing the boundary information of the target road and the location of obstacles, and ensures that the AGV can accurately perform adaptive turning and obstacle avoidance.

[0090] Embodiment 3:

[0091] Based on the above-mentioned embodiment 2, this embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, and an image screening unit, including:

[0092] An image compression subunit is used to obtain the target road image set to be analyzed, and compress and package the target road image set to be analyzed to obtain a data packet to be transmitted;

[0093] The image transmission subunit is used to place the data packet to be transmitted in a data transmission queue, and send the data packet to be transmitted to the AGV processing chip based on the data transmission queue.

[0094] In this embodiment, the data packet to be transmitted refers to a data packet obtained by compressing the target road image set to be analyzed, which is convenient for transmission.

[0095] In this embodiment, the data transmission queue refers to a carrier for transmitting data, which carries data packets from one party to another.

[0096] In this embodiment, the AGV processing chip is set in advance in the AGV, for example, it can be a single-chip microcomputer chip.

[0097] The beneficial effect of the above technical solution is that by compressing and packaging the obtained target road image to be analyzed and transmitting it to the processing chip in the AGV, it is convenient for the processing chip to analyze the obtained image in time, which provides convenience and guarantee for improving the efficiency of AGV adaptive turning and obstacle avoidance.

[0098] Embodiment 4:

[0099] Based on the above embodiment 1, this embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, and a road condition detection module, including:

[0100] An image processing unit, used to obtain the collected target road image and divide the target road image into N image blocks;

[0101] The image processing unit is used to set reference pixels in each image block respectively, and determine the range coefficient of the pixel to be denoised based on the component value of the reference pixel, and at the same time, determine the spatial distance between the pixel to be denoised and the reference pixel, and determine the spatial domain coefficient of the pixel to be denoised based on the spatial distance;

[0102] A denoising unit, configured to determine a noise sub-image in each image block based on the range coefficient and the spatial domain coefficient, and to splice the noise sub-images to obtain an image to be denoised;

[0103] The denoising unit is further used to perform denoising on the image to be denoised, and fuse the denoised image with the target road image to obtain a final target road image.

[0104] In this embodiment, the reference pixel points are selected according to the conditions of the pixel points in different image blocks, and the reference pixel points have no color deviation and the like.

[0105] In this embodiment, the range coefficient refers to the value of the current color of the pixel to be denoised.

[0106] In this embodiment, the spatial domain coefficient refers to the distance between the pixel to be denoised and the reference pixel, so as to facilitate the measurement of the color deviation between the reference pixel and the pixel to be denoised.

[0107] In this embodiment, the noise sub-image refers to the noise existing in each image block, and the noise is separately generated into a corresponding image, so as to facilitate denoising processing.

[0108] The beneficial effect of the above technical solution is that by denoising the acquired target road image, accurate and effective target road data can be obtained, thereby providing a guarantee for accurately determining the boundary conditions of the target road and the positions of obstacles in the target road.

[0109] Embodiment 5:

[0110] On the basis of the above-mentioned embodiment 1, this embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, a road condition detection module, and further includes:

[0111] An image segmentation unit, used to acquire a target road image and determine a road direction of the target road based on the target road image;

[0112] The image segmentation unit is used to determine a target processing section by a road expansion method based on the road direction, and determine a brightness threshold and a color threshold of an image area corresponding to the target processing section;

[0113] The image segmentation unit is used to segment the target road image based on the brightness threshold and the color threshold to obtain a target image corresponding to a target processing section;

[0114] An image color processing unit, used to perform binarization processing on the target image to obtain a binarized image corresponding to the target image, and determine the distribution attributes of white pixels and black pixels in the target image based on the binarized image, wherein the white pixels are non-road areas and the black pixels are target road areas;

[0115] A road boundary determination unit, configured to determine a boundary line between the white pixel points and the black pixel points based on a distribution attribute, and determine a boundary line of the target processing road section based on the boundary line;

[0116] A coordinate system construction unit, used to construct a driving coordinate system based on the right boundary line of the target processing section as a reference line, the AGV as an origin, the cross-sectional direction of the target processing section as a horizontal axis, and the forward direction of the AGV as a vertical axis;

[0117] a position determination unit, configured to determine a relative distance between an obstacle and the AGV and an azimuth of the obstacle relative to the AGV based on the target image, and determine a coordinate value of the obstacle in the vehicle coordinate system based on the relative distance and the azimuth;

[0118] The position determination unit is further used to determine the imaging ratio of the target image, and based on the imaging ratio, convert the coordinate value of the obstacle in the driving coordinate system to obtain the geographical location information of the obstacle on the target road.

[0119] In this embodiment, the target road direction refers to the extension direction of the target road in the target road image captured by the fisheye camera, that is, the forward direction of the AGV.

[0120] In this embodiment, the road expansion method refers to traversing the extension direction of the AGV from the location of the AGV to determine the section of the target section that needs to be analyzed and processed, that is, the curved road section in the target section.

[0121] In this embodiment, the target processing section refers to a section of a certain length or having special shape parameters that needs to be analyzed for road conditions.

[0122] In this embodiment, the brightness threshold refers to the brightness value of the image area corresponding to the target processing section, including the brightness values ​​corresponding to different pixel points.

[0123] In this embodiment, the color threshold refers to the color values ​​of different pixels in the image area corresponding to the target processing section.

[0124] In this embodiment, the target image refers to an image obtained by cutting and segmenting an image area corresponding to a target processing road section in a target road image.

[0125] In this embodiment, the distribution attribute refers to the distribution of white pixels and black pixels in the target image, so as to facilitate the determination of the boundary line of the target processing section according to the distribution of white pixels and black pixels.

[0126] In this embodiment, the reference line refers to taking the right boundary line of the target processing section as the reference direction of the coordinate system.

[0127] In this embodiment, the cross-sectional direction of the target section refers to the width direction of the target processing section.

[0128] The beneficial effects of the above technical solution are: by analyzing the target road image, the boundary line of the target processing section can be accurately judged, and at the same time, a driving coordinate system is constructed according to the boundary line of the target processing section, so as to accurately lock the position of obstacles in the target processing section. By analyzing and processing the target road image, the accuracy of determining the boundary line of the target road and the position of obstacles is improved, which provides a guarantee for planning the turning path of the AGV and increases the safety factor of the AGV adaptive turning obstacle avoidance.

[0129] Embodiment 6:

[0130] Based on the above embodiment 1, this embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, and a path planning module, including:

[0131] A width analysis unit, used to construct a road surface coordinate system, obtain a boundary line of the target road, and place the boundary line in the road surface coordinate system, wherein the boundary line includes left and right boundary lines;

[0132] The width analysis unit is used to traverse the valid boundary point pairs of the boundary line based on the road surface coordinate system, and determine the width of the target road based on the coordinate values ​​of the valid boundary point pairs in the road surface coordinate system;

[0133] A road condition analysis unit, configured to obtain a road sign of the target road, and determine curve information of the target road based on the road sign, wherein the curve information includes curve shape parameters and slope parameters;

[0134] The road condition analysis unit is used to determine the path distance between the starting point and the end point of the AGV traveling through the current curve and the curvature of the current curve based on the curve information;

[0135] A turning type determination unit, configured to determine the turning type of the AGV based on the width of the target road, the path distance between the starting point and the end point of the current curve, and the curvature of the current curve, and to obtain the wheelbase information of the AGV;

[0136] A turning angle determination unit, configured to determine a minimum turning angle required for the AGV to perform the turning type based on the wheelbase information of the AGV, and to determine a turning key point of the AGV in performing the turning type based on the minimum turning angle and the position information of the obstacle;

[0137] a path planning unit, configured to determine a lateral deviation between the AGV and the obstacle based on the turning key point, and compare the lateral deviation with a preset threshold;

[0138] If the lateral deviation is less than the preset threshold, it is determined that the AGV cannot turn smoothly, and the turning key point is adjusted until the lateral deviation is greater than or equal to the preset threshold;

[0139] Otherwise, it is determined that the AGV can turn smoothly, and the turning key point and the minimum turning angle are recorded to complete the planning of the turning path of the AGV.

[0140] In this embodiment, the effective boundary point pair refers to parallel points in two boundary lines in the target road, and the width of the target road is conveniently determined by the two points.

[0141] In this embodiment, the road sign refers to a road type used to mark the target road, for example, the road slope, curvature, and turning radius of a curve.

[0142] In this embodiment, the starting point refers to the starting turning point of the curve in the target road, that is, the turning point where the target road completes the curve in a straight line.

[0143] In this embodiment, the end point refers to the point where the road condition of the target road changes from a curve to a straight line.

[0144] In this embodiment, the curvature of the curve refers to the curvature and radius of the curve.

[0145] In this embodiment, the turn types include right-angle turns, curved turns, S-turns, and the like.

[0146] In this embodiment, the minimum turning angle refers to the deflection angle of the AGV wheels when passing through the current curve from a straight-line driving state, wherein the AGV can increase the turning angle according to actual conditions when turning.

[0147] In this embodiment, the turning key point refers to the point where the AGV needs to rotate its wheels when passing through the current curve.

[0148] In this embodiment, the lateral deviation refers to the distance between the AGV and the obstacle.

[0149] In this embodiment, the preset threshold is set in advance and is used to measure the safe distance between the AGV and the obstacle.

[0150] The beneficial effects of the above technical solution are: by constructing a road coordinate system, the width of the target road and the curvature of the curve can be judged, and then the turning type of the AGV can be effectively analyzed. The turning key points of the AGV when turning and the distance between the AGV and the obstacles can be determined by the turning type, so as to effectively plan the turning path of the AGV, improve the intelligence and safety of the AGV's adaptive turning obstacle avoidance, and ensure that the AGV can turn smoothly.

[0151] Embodiment 7:

[0152] Based on the above-mentioned embodiment 6, this embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, and the turning angle determination unit includes:

[0153] An inflection point determination subunit, used for obtaining a turning key point of the AGV when executing the turning type and a curvature of a current curve, and determining a relative angle between a next inflection point of the turning key point and the turning key point based on the curvature of the current curve;

[0154] The inflection point determination subunit is used to determine a target position of the next inflection point based on the relative angle, and continue to determine the next inflection point based on the target position until all inflection points on the current curve are determined, wherein the inflection point is not unique;

[0155] The route planning subunit is used to generate a turning path according to the positions of all the turning points based on a preset curve generator, so as to complete the planning of the AGV's adaptive turning path.

[0156] In this embodiment, the next turning point refers to the point where the AGV changes from the turning key point to the next wheel deflection angle when executing the turning type.

[0157] In this embodiment, the relative angle refers to the azimuth relationship between the next turning point and the turning key point.

[0158] In this embodiment, the target position refers to the actual geographical location of the next turning point on the target road.

[0159] In this embodiment, the preset curve generator is set in advance and is used to generate a corresponding path according to each inflection point, and the path passes through all the inflection points.

[0160] The beneficial effect of the above technical solution is: by accurately judging the position of the next turning point of the turning key point according to the turning key point and the curvature of the current curve, until the judgment of all turning points is completed, it provides convenience for accurately generating the corresponding turning path, ensures that the AGV can successfully complete automatic turning and obstacle avoidance, and improves the reliability of automatic turning and obstacle avoidance.

[0161] Embodiment 8:

[0162] Based on the above-mentioned embodiment 6, this embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, and the turning angle determination unit includes:

[0163] A turning angle range determination subunit, configured to determine a maximum turning angle of the AGV based on the wheelbase information of the AGV and the curvature of the current curve;

[0164] a distance analysis subunit, configured to perform a pre-simulated turn on the AGV based on the maximum turning angle, and determine, based on the pre-simulated turn, whether a lateral deviation between the AGV and the obstacle at the maximum turning angle is greater than or equal to a preset threshold;

[0165] If it is greater than or equal to the preset threshold, the obtained maximum turning angle is determined to be qualified, and the turning angle range of the AGV when automatically turning to avoid obstacles is determined based on the maximum turning angle and the minimum turning angle;

[0166] Otherwise, the maximum turning angle is decreased until the lateral deviation is greater than or equal to the preset threshold, thereby completing the determination of the turning range of the AGV when automatically turning to avoid obstacles.

[0167] In this embodiment, the maximum turning angle refers to the maximum deflection angle allowed for the wheels of the AGV when passing through the current curve. If the deflection angle of the wheels is less than the minimum deflection angle or greater than the maximum deflection angle, the AGV cannot pass through the current curve smoothly.

[0168] In this embodiment, the pre-simulated turning refers to simulating the turning process of the AGV according to the maximum turning angle.

[0169] In this embodiment, the turning angle range refers to the range in which the deflection angle of the wheel can be controlled when the AGV passes through the current curve.

[0170] The beneficial effect of the above technical solution is: by analyzing the maximum turning angle of the AGV based on the wheelbase information of the AGV and the curvature of the current curve, the angle range within which the wheels of the AGV can be deflected when passing through the current curve can be accurately determined, providing angle adjustment guarantee for the AGV to pass through the current curve smoothly, while ensuring that the AGV can pass through the current curve smoothly.

[0171] Embodiment 9:

[0172] Based on the above embodiment 1, this embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, such as Figure 3 As shown, the adaptive turning module includes:

[0173] A speed analysis unit, used to obtain a planning result, and determine the curve radius of the current curve and the turning angle corresponding to the AGV when executing the turning type based on the planning result;

[0174] The speed analysis unit is used to determine the center of gravity of the AGV based on the wheelbase and wheel track of the AGV, and determine the safety centripetal force corresponding to the AGV when executing the turning type based on the curve radius, the turning angle and the center of gravity of the AGV;

[0175] The speed determination unit is used to calculate the wheel moving speed of the AGV when turning based on the safety centripetal force, and convert the wheel moving speed into the moving speed at the center of gravity to complete the determination of the turning speed of the AGV.

[0176] In this embodiment, the curve radius of the curve refers to the radius of the circle to which the current curve belongs.

[0177] In this embodiment, the safety centripetal force refers to the centripetal force corresponding to the AGV not rolling over when turning.

[0178] In this embodiment, the wheel moving speed is converted into the moving speed at the center of gravity by using a professional conversion formula.

[0179] In this embodiment, the wheel moving speed of the AGV when turning is calculated based on the safety centripetal force, and the wheel moving speed is converted into the moving speed at the center of gravity, including:

[0180] The current curve width and the wheelbase of the AGV are obtained, the minimum turning radius of the AGV is calculated based on the current curve width and the wheelbase of the AGV, and the safe speed of the AGV when turning is calculated based on the minimum turning radius. The specific steps include:

[0181] The minimum turning radius of the AGV is calculated according to the following formula:

[0182]

[0183] Wherein, R represents the minimum turning radius of the AGV; L represents the wheelbase of the AGV; K represents the front overhang length of the AGV; D represents the width of the AGV; S represents the width of the current curve; M represents the safe distance from the outer boundary when the AGV turns; θ represents the maximum turning angle of the front wheel of the AGV, and the value range is (0°, 34°); Q represents the safe distance between the AGV and obstacles when turning;

[0184] The safe speed of the AGV when turning is calculated according to the following formula:

[0185]

[0186] Wherein, V represents the safe speed of the AGV when turning; R represents the minimum turning radius of the AGV; g represents the acceleration of gravity, which is 9.8m / s 2 ; represents the roll angle of the current curve relative to the horizontal direction; H represents the height of the center of gravity of the AGV relative to the ground; τ represents the error factor, and its value range is (0.02, 0.05);

[0187] Monitor the current travel speed of the AGV in real time, and compare the current travel speed with the safe speed;

[0188] If the current driving speed is less than or equal to the safe speed, it is determined that the AGV will not be in danger of rolling over when turning;

[0189] Otherwise, it is determined that the AGV is in danger of rolling over when turning, and an alarm is sent to the management terminal, while the current driving speed of the AGV is reduced until the driving speed is less than or equal to the safe speed.

[0190] The beneficial effect of the above technical solution is: by accurately analyzing the centripetal force of the AGV passing through the current curve according to the wheelbase, wheelbase and curve radius of the AGV, the safe speed corresponding to the AGV passing through the current curve is determined, thereby improving the driving safety of the AGV and providing a guarantee for ensuring that the AGV can complete adaptive turning and obstacle avoidance.

[0191] Embodiment 10:

[0192] On the basis of the above-mentioned embodiment 9, this embodiment provides an AGV adaptive turning obstacle avoidance device based on artificial intelligence, and a speed determination unit, including:

[0193] A monitoring subunit, used to obtain in real time the force conditions of the four wheels of the AGV when turning, and determine the safe turning speed of the AGV based on the force conditions;

[0194] a comparison subunit, configured to compare the current turning speed of the AGV with the safe turning speed;

[0195] A speed adjustment subunit, configured to determine that there is a safety hazard in the AGV turning when the current turning speed is greater than the safe turning speed, and determine an adjustment strategy for the AGV turning speed based on the force condition;

[0196] The speed adjustment subunit is used to determine an adjustment threshold for the turning speed based on the adjustment strategy, and adjust the turning speed of the AGV based on the adjustment threshold;

[0197] The monitoring subunit is used to continue monitoring the force conditions of the four wheels of the AGV when turning when the current turning speed is less than or equal to the safe turning speed until the AGV completes adaptive turning and obstacle avoidance.

[0198] In this embodiment, the safe turning speed refers to the maximum driving speed allowed when the AGV passes through the current curve.

[0199] In this embodiment, the adjustment strategy is determined according to the curve type of the current curve and the current driving speed of the AGV, and is used to adjust the current driving speed of the AGV to ensure that the AGV can pass the current curve smoothly.

[0200] In this embodiment, the adjustment threshold refers to an adjustment value for the current turning speed of the AGV.

[0201] The beneficial effect of the above technical solution is: by real-time monitoring of the force conditions of the AGV wheels and determining the safe turning speed of the AGV based on the force conditions, the current turning speed of the AGV is analyzed, and the turning speed of the AGV is adjusted when the turning speed is greater than the safe turning speed, thereby improving the turning safety of the AGV.

[0202] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An AGV adaptive turning obstacle avoidance device based on artificial intelligence, characterized in that: include: A road condition detection module is used to collect target road images and pre-process the target road images to determine the boundary lines of the target road and the location information of obstacles; A path planning module, used to determine the turning type of the AGV and plan the turning path of the AGV based on the turning type; An adaptive turning module, used to determine the turning speed of the AGV based on the planning result, and to complete the AGV adaptive turning obstacle avoidance based on the turning speed; Among them, the path planning module includes: A width analysis unit, used to construct a road surface coordinate system, obtain a boundary line of the target road, and place the boundary line in the road surface coordinate system, wherein the boundary line includes left and right boundary lines; The width analysis unit is used to traverse the valid boundary point pairs of the boundary line based on the road surface coordinate system, and determine the width of the target road based on the coordinate values ​​of the valid boundary point pairs in the road surface coordinate system; A road condition analysis unit, configured to obtain a road sign of the target road, and determine curve information of the target road based on the road sign, wherein the curve information includes curve shape parameters and slope parameters; The road condition analysis unit is used to determine the path distance between the starting point and the end point of the AGV traveling through the current curve and the curvature of the current curve based on the curve information; A turning type determination unit, configured to determine the turning type of the AGV based on the width of the target road, the path distance between the starting point and the end point of the current curve, and the curvature of the current curve, and to obtain the wheelbase information of the AGV; A turning angle determination unit, configured to determine a minimum turning angle required for the AGV to perform the turning type based on the wheelbase information of the AGV, and to determine a turning key point of the AGV in performing the turning type based on the minimum turning angle and the position information of the obstacle; a path planning unit, configured to determine a lateral deviation between the AGV and the obstacle based on the turning key point, and compare the lateral deviation with a preset threshold; If the lateral deviation is less than the preset threshold, it is determined that the AGV cannot turn smoothly, and the turning key point is adjusted until the lateral deviation is greater than or equal to the preset threshold; Otherwise, it is determined that the AGV can turn smoothly, and the turning key point and the minimum turning angle are recorded to complete the planning of the turning path of the AGV.

2. The AGV adaptive turning obstacle avoidance device based on artificial intelligence according to claim 1 is characterized in that: Road condition detection module, including: An image acquisition unit, used for acquiring an initial road image of a target road based on a fisheye camera, and determining a current shooting area of ​​the target road based on the initial road image, wherein the fisheye cameras are respectively arranged at the front, rear, left and right of the AGV; An angle adjustment unit, used to correct the shooting angle of the fisheye camera based on the current shooting area, and shoot the target road image based on the correction result to obtain a target road image set; An image screening unit, configured to screen the target road image set based on preset rules, wherein the preset rules include an evaluation of the clarity of the target road image and an assessment of the target road boundary in the target road image; The image screening unit is further used to retain the target road images in the target road image set whose clarity is greater than or equal to the preset clarity and whose target road boundaries are within the target image area, and to eliminate the target road images in the target road image set that do not meet the preset rules, so as to obtain the target road image set to be analyzed.

3. The AGV adaptive turning obstacle avoidance device based on artificial intelligence according to claim 2 is characterized in that: Image screening unit, including: An image compression subunit is used to obtain the target road image set to be analyzed, and compress and package the target road image set to be analyzed to obtain a data packet to be transmitted; The image transmission subunit is used to place the data packet to be transmitted in a data transmission queue, and send the data packet to be transmitted to the AGV processing chip based on the data transmission queue.

4. The AGV adaptive turning obstacle avoidance device based on artificial intelligence according to claim 1 is characterized in that: Road condition detection module, including: An image processing unit, used to obtain the collected target road image and divide the target road image into N image blocks; The image processing unit is used to set reference pixels in each image block respectively, and determine the range coefficient of the pixel to be denoised based on the component value of the reference pixel, and at the same time, determine the spatial distance between the pixel to be denoised and the reference pixel, and determine the spatial domain coefficient of the pixel to be denoised based on the spatial distance; A denoising unit, configured to determine a noise sub-image in each image block based on the range coefficient and the spatial domain coefficient, and to splice the noise sub-images to obtain an image to be denoised; The denoising unit is further used to perform denoising on the image to be denoised, and fuse the denoised image with the target road image to obtain a final target road image.

5. The AGV adaptive turning obstacle avoidance device based on artificial intelligence according to claim 1 is characterized in that: The road condition detection module also includes: An image segmentation unit, used to acquire a target road image and determine a road direction of the target road based on the target road image; The image segmentation unit is used to determine a target processing section by a road expansion method based on the road direction, and determine a brightness threshold and a color threshold of an image area corresponding to the target processing section; The image segmentation unit is used to segment the target road image based on the brightness threshold and the color threshold to obtain a target image corresponding to a target processing section; An image color processing unit, used to perform binarization processing on the target image to obtain a binarized image corresponding to the target image, and determine the distribution attributes of white pixels and black pixels in the target image based on the binarized image, wherein the white pixels are non-road areas and the black pixels are target road areas; A road boundary determination unit, configured to determine a boundary line between the white pixel points and the black pixel points based on a distribution attribute, and determine a boundary line of the target processing road section based on the boundary line; A coordinate system construction unit, used to construct a driving coordinate system based on the right boundary line of the target processing section as a reference line, the AGV as an origin, the cross-sectional direction of the target processing section as a horizontal axis, and the forward direction of the AGV as a vertical axis; a position determination unit, configured to determine a relative distance between an obstacle and the AGV and an azimuth of the obstacle relative to the AGV based on the target image, and determine a coordinate value of the obstacle in the vehicle coordinate system based on the relative distance and the azimuth; The position determination unit is further used to determine the imaging ratio of the target image, and based on the imaging ratio, convert the coordinate value of the obstacle in the driving coordinate system to obtain the geographical location information of the obstacle on the target road.

6. The AGV adaptive turning obstacle avoidance device based on artificial intelligence according to claim 1 is characterized in that: A turning angle determination unit, comprising: An inflection point determination subunit, used for obtaining a turning key point of the AGV when executing the turning type and a curvature of a current curve, and determining a relative angle between a next inflection point of the turning key point and the turning key point based on the curvature of the current curve; The inflection point determination subunit is used to determine a target position of the next inflection point based on the relative angle, and continue to determine the next inflection point based on the target position until all inflection points on the current curve are determined, wherein the inflection point is not unique; The route planning subunit is used to generate a turning path according to the positions of all the turning points based on a preset curve generator, so as to complete the planning of the AGV's adaptive turning path.

7. The AGV adaptive turning obstacle avoidance device based on artificial intelligence according to claim 1 is characterized in that: A turning angle determination unit, comprising: A turning angle range determination subunit, configured to determine a maximum turning angle of the AGV based on the wheelbase information of the AGV and the curvature of the current curve; a distance analysis subunit, configured to perform a pre-simulated turn on the AGV based on the maximum turning angle, and determine, based on the pre-simulated turn, whether a lateral deviation between the AGV and the obstacle at the maximum turning angle is greater than or equal to a preset threshold; If it is greater than or equal to the preset threshold, the obtained maximum turning angle is determined to be qualified, and the turning angle range of the AGV when automatically turning to avoid obstacles is determined based on the maximum turning angle and the minimum turning angle; Otherwise, the maximum turning angle is decreased until the lateral deviation is greater than or equal to the preset threshold, thereby completing the determination of the turning range of the AGV when automatically turning to avoid obstacles.

8. The AGV adaptive turning obstacle avoidance device based on artificial intelligence according to claim 1 is characterized in that: Adaptive turning module, including: A speed analysis unit, used to obtain a planning result, and determine the curve radius of the current curve and the turning angle corresponding to the AGV when executing the turning type based on the planning result; The speed analysis unit is used to determine the center of gravity of the AGV based on the wheelbase and wheel track of the AGV, and determine the safety centripetal force corresponding to the AGV when executing the turning type based on the curve radius of the curve, the turning angle and the center of gravity of the AGV; The speed determination unit is used to calculate the wheel movement speed of the AGV when turning based on the safety centripetal force, and convert the wheel movement speed into the movement speed at the center of gravity to complete the determination of the turning speed of the AGV.

9. The AGV adaptive turning obstacle avoidance device based on artificial intelligence according to claim 8, characterized in that: A speed determination unit, comprising: A monitoring subunit, used to obtain in real time the force conditions of the four wheels of the AGV when turning, and determine the safe turning speed of the AGV based on the force conditions; a comparison subunit, configured to compare the current turning speed of the AGV with the safe turning speed; A speed adjustment subunit, configured to determine that there is a safety hazard in the AGV turning when the current turning speed is greater than the safe turning speed, and determine an adjustment strategy for the AGV turning speed based on the force condition; The speed adjustment subunit is used to determine an adjustment threshold for the turning speed based on the adjustment strategy, and adjust the turning speed of the AGV based on the adjustment threshold; The monitoring subunit is used to continue monitoring the force conditions of the four wheels of the AGV when turning when the current turning speed is less than or equal to the safe turning speed until the AGV completes adaptive turning and obstacle avoidance.

Images