An intelligent vehicle path prediction system for autonomous driving

Abstract

The application provides an intelligent vehicle path prediction system for automatic driving, comprising: a blind spot detection module, which is used for obtaining blind spot area information based on environment data collected by sensors of an automatic driving intelligent vehicle; a judgment module, which is used for judging whether a current driving path of the vehicle will pass through the blind spot area, and planning an avoidance path if yes; an avoidance path planning module, which is used for searching an avoidance path scheme for avoiding the blind spot area in a drivable area of the vehicle based on the blind spot area information; a path optimization module, which is used for optimizing the avoidance path scheme to obtain an optimal path; and a path prediction module, which is used for obtaining a local motion trajectory and speed planning of the vehicle based on the optimal path, and completing intelligent vehicle path prediction. The application effectively solves the problems of blind spot avoidance and path prediction planning of an automatic driving vehicle in a complex environment, improves driving safety and energy efficiency, and provides reliable technical support for the application of automatic driving technology in an actual road environment.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of automatic driving, and particularly relates to an intelligent vehicle path prediction system for automatic driving. BACKGROUND

[0002] The core technical challenge faced by an automatic driving vehicle in a complex road environment is how to effectively identify and respond to blind spot areas and make path predictions to ensure driving safety. The existence of blind spot areas makes it difficult for the vehicle to fully perceive the surrounding environment, which easily causes collision risks. However, simply avoiding all potential blind spot areas may cause the driving path of the vehicle to be too conservative, which affects driving efficiency. Therefore, the key problem is how to balance the cost and benefit of path adjustment under the premise of ensuring safety. This involves a series of complex processes such as accurate identification of blind spot areas, real-time matching with the driving path, generation and evaluation of alternative paths, etc. At the same time, energy efficiency needs to be considered to avoid unnecessary energy consumption caused by excessive adjustment. In a dynamic and variable road environment, how to achieve rapid response and smooth avoidance of blind spot areas is also a big challenge. In addition, different types of blind spots (such as static obstacles, dynamic vehicles, pedestrians, etc.) may require different avoidance strategies, and how to design a general and robust avoidance and path prediction method is also a problem to be solved. In general, the blind spot area avoidance problem of an automatic driving vehicle involves perception, decision-making, control and other aspects, and needs to seek the best balance between safety, efficiency and comfort. SUMMARY

[0003] In view of the deficiencies of the prior art, the application provides an intelligent vehicle path prediction system for automatic driving, which realizes detection of blind spot areas and efficient prediction of driving paths.

[0004] To achieve the above-mentioned purpose, the application provides the following solutions.

[0005] An intelligent vehicle path prediction system for automatic driving comprises a blind spot detection module, a judgment module, an avoidance path planning module, a path optimization module and a path prediction module.

[0006] The blind spot detection module is used to obtain blind spot area information based on environmental data collected by sensors of an automatic driving intelligent vehicle.

[0007] The judgment module is used to match the blind spot area information with current position and driving path information of the vehicle, judge whether the current driving path of the vehicle will pass through a blind spot area, and plan an avoidance path if so.

[0008] The avoidance path planning module is configured for searching an avoidance path scheme for avoiding the blind spot area in the drivable area of the vehicle based on the blind spot area information, in combination with a road topology structure and traffic rules, and by using a path search algorithm.

[0009] The path optimization module is configured for optimizing the avoidance path scheme to obtain an optimal path.

[0010] The path prediction module is configured for obtaining a local motion trajectory and a speed plan of the vehicle based on the optimal path, in combination with a vehicle dynamics model and real-time traffic condition information, to complete path prediction of the intelligent vehicle.

[0011] Preferably, the blind spot detection module comprises:

[0012] The data acquisition unit is configured for acquiring surrounding environment data based on a vehicle-mounted sensor to obtain obstacle information and road information.

[0013] The data format conversion unit is configured for converting the obstacle information and the road information into a digital image format and pre-processing the converted digital image.

[0014] The blind spot feature extraction unit is configured for segmenting the pre-processed digital image to obtain a blind spot area feature vector.

[0015] The obstacle detection unit is configured for classifying the blind spot area feature vector based on a support vector machine to determine whether there is an obstacle in the blind spot area and to obtain position information of the obstacle in the blind spot area.

[0016] The blind spot area information acquisition unit is configured for calculating a relative position and a distance of the obstacle in the blind spot area based on the position information of the obstacle and in combination with current driving state information of the vehicle to obtain the blind spot area information.

[0017] Preferably, the determination module comprises:

[0018] The first coordinate acquisition unit is configured for converting current position information and driving path information of the vehicle into first coordinate data in a unified coordinate system.

[0019] The second coordinate acquisition unit is configured for extracting position coordinate data of the blind spot area information to obtain second coordinate data.

[0020] The coordinate matching unit is configured for matching the first coordinate data with the second coordinate data to determine whether each position point on the driving path of the vehicle falls within the blind spot area.

[0021] The marking unit is configured for marking position coordinate data of an overlapping area as a target area that needs to be avoided if the determination result indicates that the driving path of the vehicle overlaps with the blind spot area.

[0022] Preferably, the avoidance path planning module comprises:

[0023] a path reference line acquisition unit configured to search a globally optimal path from the starting point to the destination based on the high-precision map, and take the globally optimal path as a path reference line;

[0024] a reference line smoothing unit configured to smooth the path reference line to obtain a smoothed path reference line;

[0025] an initial path acquisition unit configured to establish a Frenet coordinate system based on the smoothed path reference line, perform raster sampling on a current road environment, and search an initial path by using a search algorithm;

[0026] a path space acquisition unit configured to generate a drivable path space that avoids blind spot areas based on the initial path and the road environment, and obtain an avoidance path scheme based on the drivable path space.

[0027] Preferably, the path optimization module comprises:

[0028] a collision detection unit configured to determine whether the avoidance path scheme has a collision risk based on a collision detection algorithm, and screen a safe avoidance path;

[0029] an adjustment cost calculation unit configured to compare the safe avoidance path with an original driving path of the vehicle, obtain a path adjustment cost, and take the safe avoidance path as the optimal path when the path adjustment cost meets a preset threshold.

[0030] Preferably, the path optimization module comprises:

[0031] a traffic prediction unit configured to collect dynamic traffic information around the vehicle in real time based on a vehicle-mounted sensor, and fuse the dynamic traffic information with historical traffic data to obtain a traffic condition change trend;

[0032] a waypoint optimization unit configured to discretize the optimal path into waypoints, and perform constraint optimization on each waypoint by using a vehicle dynamics model and road curvature and slope;

[0033] a motion state prediction unit configured to predict a vehicle motion state based on the optimized waypoints as a reference track;

[0034] a path and speed prediction unit configured to perform vehicle motion trajectory planning and speed planning in a short time based on a vehicle motion state prediction result.

[0035] Preferably, the waypoints comprise vehicle position, speed and acceleration.

[0036] Preferably, the process of constructing the vehicle dynamics model is as follows:

[0037] Based on the vehicle actual position vector and the projection vector, a vehicle speed vector and a projection speed vector are obtained;

[0038] Based on the vehicle speed vector and the projection speed vector, a lateral speed of the vehicle in the Frenet coordinate system is calculated;

[0039] Based on the lateral speed, the vehicle yaw angle and the mass center side slip angle, an error set is obtained;

[0040] Based on the error set, a state quantity and a control quantity are obtained;

[0041] Based on the state quantity and the control quantity, the vehicle dynamics model is obtained.

[0042] Compared with the prior art, the beneficial effects of the present application are: through the sensor to collect environmental data, using image processing algorithm to identify blind spot area, and matching it with vehicle driving path. If there is overlap, the path search algorithm is used to generate avoidance path, and safety evaluation and energy consumption optimization are carried out, and the optimal path is selected. The present application also considers the path adjustment cost, executes the new path under the condition of meeting the threshold value. Finally, the motion trajectory and speed planning are carried out, and the avoidance control of the vehicle is realized. The blind spot avoidance and path prediction and planning problem of the automatic driving vehicle in complex environment are effectively solved, the driving safety and energy efficiency are improved, and reliable technical support is provided for the application of automatic driving technology in actual road environment. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the present application, the following briefly introduces the drawings needed to be used in the embodiments. Obviously, the drawings described in the following embodiments are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0044] Figure 1 The structure diagram of the intelligent vehicle path prediction system for automatic driving of the embodiment of the present application. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0046] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0047] Example 1

[0048] like Figure 1 As shown, an intelligent vehicle path prediction system for autonomous driving includes: a blind spot detection module, a judgment module, an avoidance path planning module, a path optimization module, and a path prediction module;

[0049] A blind spot detection module is configured to obtain blind spot area information based on environmental data collected by sensors of the autonomous driving intelligent vehicle. In a further embodiment, the blind spot detection module includes:

[0050] A data acquisition unit is used to collect surrounding environment data based on vehicle-mounted sensors to obtain obstacle information and road information;

[0051] A data format conversion unit, used to convert obstacle information and road information into a digital image format and pre-process the converted digital image;

[0052] A blind spot feature extraction unit is used to segment the pre-processed digital image to obtain a blind spot area feature vector;

[0053] The obstacle detection unit is used to classify the feature vector of the blind spot area based on the support vector machine, determine whether there is an obstacle in the blind spot area, and obtain the location information of the obstacle in the blind spot area;

[0054] The blind spot area information acquisition unit is used to calculate the relative position and distance of obstacles in the blind spot area based on the position information of the obstacle and the current driving state information of the vehicle to obtain the blind spot area information.

[0055] In the embodiment, the autonomous intelligent vehicle collects environmental data around the vehicle through various sensors such as laser radar, millimeter wave radar, camera, etc., converts the collected point cloud data, radar data, etc. into an RGB image format with a resolution of 1920x1080. The converted image is subjected to median filtering to remove salt and pepper noise, and histogram equalization algorithm is used to enhance the image contrast and improve the image quality. A semantic segmentation algorithm based on SegNet is used to divide the image into 12 semantic categories such as road, vehicle, pedestrian, building, etc., and extract the semantic mask of the blind spot area. Color histogram, GLCM texture feature, Hu invariant moment, etc. algorithms are used to extract the color, texture, shape, etc. features of the blind spot area, and a 36-dimensional blind spot area feature vector is constructed. A support vector machine algorithm is used to classify the blind spot area feature vector, to determine whether there is an obstacle in the blind spot area, and the classification accuracy is above 95%. According to the pixel coordinates of the obstacle in the blind spot area, combined with the speed, direction, IMU, etc. driving state information of the vehicle, the relative position and distance of the obstacle are calculated using the triangulation principle, and the range of the blind spot area is determined.

[0056] The judging module is configured to match the blind spot area information with the current position and driving path information of the vehicle, to determine whether the current driving path of the vehicle will pass through the blind spot area, and if so, to plan an avoidance path. In a further embodiment, the judging module comprises:

[0057] The first coordinate acquisition unit is configured to convert the current position information and driving path information of the vehicle into first coordinate data in a unified coordinate system.

[0058] The second coordinate acquisition unit is configured to extract the position coordinate data of the blind spot area information, to obtain second coordinate data.

[0059] The coordinate matching unit is configured to match the first coordinate data with the second coordinate data, to determine whether each position point on the driving path of the vehicle falls within the blind spot area.

[0060] The marking unit is configured to mark the position coordinate data of the overlapping area as a target area that needs to be avoided, if the determination result shows that the driving path of the vehicle overlaps with the blind spot area.

[0061] In the embodiment, when the current position information of the vehicle and the preset driving path information are acquired, the vehicle's longitude and latitude coordinates (e.g., 122.34 degrees east longitude and 35.67 degrees north latitude) can be acquired by using a vehicle-mounted GPS positioning system, which are converted into coordinate points in a unified plane rectangular coordinate system (e.g., X = 123.5 meters and Y = 56.79 meters) to form a first coordinate data set of the vehicle driving path. Meanwhile, the center point coordinates (e.g., X = 23.46 meters and Y = 67.80 meters) of the blind spot region are extracted by using the position information of the blind spot region identified in the previous step to obtain a second coordinate data set. The first coordinate data set and the second coordinate data set are spatially matched, the Euclidean distance from each coordinate point on the vehicle driving path to the center point of the blind spot region is calculated using a nearest neighbor algorithm, and if the distance is less than the radius (e.g., 50 meters) of the blind spot region, it is determined that the coordinate point falls within the blind spot region and needs to be avoided. Cluster analysis is performed on the coordinate points determined to need to be avoided to obtain a plurality of target regions that need to be avoided.

[0062] The avoidance path planning module is configured to search for an avoidance path scheme that avoids the blind spot region in the drivable region of the vehicle based on the blind spot region information, in combination with the road topology structure and the traffic rules, and by using a path search algorithm. In a further embodiment, the avoidance path planning module comprises:

[0063] The path reference line acquisition unit is configured to search for a globally optimal path from the starting point to the destination based on the high-precision map, and to take the globally optimal path as the path reference line.

[0064] The reference line smoothing unit is configured to smooth the path reference line to obtain a smoothed path reference line. In the embodiment, the smoothing optimization problem is solved from three aspects, including optimization variables, optimization objectives, and constraint conditions. The optimization variables are a series of discrete points to be smoothed, the optimization objectives are a smoothness cost function constructed based on the geometric relationship between the discrete points, including a smoothness cost, a geometric similarity cost, and a length cost, and the constraint conditions are the numerical range between the difference and the sum of the original path coordinates and the relative original path point offset threshold value.

[0065] The initial path acquisition unit is configured to establish a Frenet coordinate system based on the smoothed path reference line, perform raster sampling on the current road environment, and search for an initial path by using a search algorithm.

[0066] The path space acquisition unit is configured to generate a drivable path space that avoids the blind spot region based on the initial path and the road environment, and to obtain an avoidance path scheme based on the drivable path space.

[0067] The path optimization module is configured to optimize the avoidance path scheme to obtain an optimal path. In a further embodiment, the path optimization module:

[0068] a collision detection unit configured to determine whether the evasive path plan has a collision risk based on a collision detection algorithm, and to filter out safe evasive paths; in this embodiment, a time-based dynamic collision detection (TDCD) algorithm is used, which can predict the relative position and speed of the vehicle and other obstacles within the next few seconds, thereby evaluating the collision probability; assuming that on a certain path, the minimum distance between the vehicle and a pedestrian predicted by the TDCD algorithm is 5 meters, which is much lower than the safety threshold of 2 meters, the path is therefore marked as high-risk. Then, a safety margin analysis is performed on all paths marked as low-risk, for example, the average safety distance on each path is calculated, and the shortest parking distance of the vehicle in an emergency is calculated. Through these data, paths with higher safety can be further filtered out. After determining the safe and feasible paths, these paths are further sorted, taking into account factors including path length and evasion efficiency.

[0069] an adjustment cost calculation unit configured to compare the safe evasive path with the original driving path of the vehicle, and to obtain a path adjustment cost, and when the path adjustment cost meets a preset threshold, the safe evasive path is taken as the optimal path.

[0070] In this embodiment, the path data of the recommended evasive path and the original planning path of the vehicle is obtained, including the starting point, ending point, passing point, road segment distance, road segment speed limit and other information of the path. A dynamic programming algorithm is used to calculate the driving time of the recommended path and the original planning path. By considering factors such as road segment distance and speed limit, the predicted driving time of the two paths is obtained. According to the power system parameters of the vehicle and the road segment information, an energy consumption model is used to calculate the predicted energy consumption of the recommended path and the original planning path. Considering factors such as road segment slope, speed and acceleration, the energy consumption values of the two paths are determined. The driving time of the recommended path is compared with the driving time of the original planning path, and the increment of driving time is calculated. At the same time, the energy consumption of the recommended path is compared with the energy consumption of the original planning path, and the increment of energy consumption is obtained. The increments of driving time and energy consumption are weighted and summed to obtain a comprehensive adjustment cost. By setting the weight coefficients of time and energy consumption, the two increment values are normalized, and the weighted sum is calculated as the comprehensive adjustment cost. It is judged whether the comprehensive adjustment cost exceeds a preset cost threshold. If it exceeds the threshold, the original planning path is maintained; if it does not exceed the threshold, the recommended path is taken as the new driving path.

[0071] a path prediction module configured to obtain a local motion trajectory and speed planning of the vehicle based on the optimal path, in combination with a vehicle dynamics model and real-time traffic condition information, and to complete intelligent vehicle path prediction.

[0072] Further embodiments are directed to a path optimization module comprising:

[0073] A traffic prediction unit is configured to collect dynamic traffic information around the vehicle in real time based on a vehicle-mounted sensor, and fuse the dynamic traffic information with historical traffic data to obtain a traffic condition change trend.

[0074] A waypoint optimization unit is configured to discretize the optimal path into waypoints, and perform constraint optimization on each waypoint by using a vehicle dynamics model and road curvature and slope.

[0075] A motion state prediction unit is configured to predict a vehicle motion state based on the optimized waypoints as a reference trajectory.

[0076] A path and speed prediction unit is configured to perform vehicle motion trajectory planning and speed planning in a short time based on the vehicle motion state prediction result.

[0077] In a further implementation, the waypoints include vehicle position, speed, and acceleration.

[0078] In a further implementation, the process of constructing the vehicle dynamics model includes:

[0079] Based on the actual vehicle position vector and the projection vector, a vehicle speed vector and a projection speed vector are obtained.

[0080] Based on the vehicle speed vector and the projection speed vector, a lateral speed of the vehicle in the Frenet coordinate system is calculated.

[0081] Based on the lateral speed, the vehicle yaw angle, and the center of mass side slip angle, an error set is obtained; in this embodiment, the error set includes lateral error, heading error, position error, and speed error.

[0082] Based on the error set, a state quantity and a control quantity are obtained; the state quantity is the transpose of a matrix formed by the error set, and the control quantity is the transpose of a matrix formed by the vehicle acceleration change and the front wheel steering angle.

[0083] Based on the state quantity and the control quantity, the vehicle dynamics model is obtained.

[0084] The above-described embodiments are merely descriptions of the preferred modes of the present application, and are not intended to limit the scope of the present application. Various modifications and improvements to the technical solutions of the present application made by those of ordinary skill in the art without departing from the design spirit of the present application shall fall within the scope of protection of the present application as defined by the claims.

Claims

1. An intelligent vehicle path prediction system for autonomous driving, characterized in that: include: Blind spot detection module, judgment module, avoidance path planning module, path optimization module and path prediction module; The blind spot detection module is used to obtain blind spot area information based on environmental data collected by sensors of the autonomous driving intelligent vehicle; The judgment module is used to match the blind spot area information with the vehicle's current position and driving path information to determine whether the vehicle's current driving path will pass through the blind spot area, and if so, plan an avoidance path; The avoidance path planning module is used to search for an avoidance path solution that avoids the blind spot area within the vehicle's drivable area based on the blind spot area information, combined with the road topology and traffic rules, and using a path search algorithm; The path optimization module is used to optimize the avoidance path solution to obtain the optimal path; The path prediction module is used to obtain the vehicle's local motion trajectory and speed planning based on the optimal path, combined with the vehicle dynamics model and real-time traffic status information, to complete intelligent vehicle path prediction; The blind spot detection module includes: A data acquisition unit is used to collect surrounding environment data based on vehicle-mounted sensors to obtain obstacle information and road information; a data format conversion unit, configured to convert the obstacle information and road information into a digital image format and pre-process the converted digital image; a blind spot feature extraction unit, configured to segment the pre-processed digital image to obtain a blind spot area feature vector; An obstacle detection unit is used to classify the feature vector of the blind spot area based on a support vector machine, determine whether there is an obstacle in the blind spot area, and obtain the position information of the obstacle in the blind spot area; A blind spot area information acquisition unit is used to calculate the relative position and distance of the obstacle in the blind spot area based on the position information of the obstacle and the current driving state information of the vehicle to obtain the blind spot area information; The path optimization module includes: Traffic prediction unit, used to collect dynamic traffic information around the vehicle in real time based on on-board sensors, and fuse the dynamic traffic information with historical traffic data to obtain traffic condition change trends; a waypoint optimization unit, configured to discretize the optimal path into waypoints, and perform constraint optimization on each waypoint using a vehicle dynamics model and road curvature and slope; A motion state prediction unit, used to predict the vehicle motion state based on the optimized waypoints as reference tracks; The path and speed prediction unit is used to plan the vehicle's motion trajectory and speed in a short period of time based on the vehicle's motion state prediction results.

2. The intelligent vehicle path prediction system for autonomous driving according to claim 1, characterized in that: The judgment module includes: A first coordinate acquisition unit, configured to convert the vehicle's current position information and driving path information into first coordinate data in a unified coordinate system; A second coordinate acquisition unit, configured to extract the position coordinate data of the blind spot area information to obtain second coordinate data; a coordinate matching unit, configured to match the first coordinate data with the second coordinate data to determine whether each position point on the vehicle's driving path falls within a blind spot area; The marking unit marks the position coordinate data of the overlapping area as a target area that needs to be avoided if the judgment result shows that the vehicle's driving path overlaps with the blind spot area.

3. The intelligent vehicle path prediction system for autonomous driving according to claim 1, characterized in that: The avoidance path planning module includes: A path reference line acquisition unit is used to search for a global optimal path from a starting point to a destination based on a high-precision map, and use the global optimal path as a path reference line; A reference line smoothing unit, configured to smooth the path reference line to obtain a smoothed path reference line; An initial path acquisition unit is used to establish a Frenet coordinate system based on the smooth path reference line, perform raster sampling on the current road environment, and search for an initial path using a search algorithm; The path space acquisition unit is used to generate a drivable path space that avoids the blind spot area based on the initial path and the road environment; and obtain an avoidance path solution based on the drivable path space.

4. The intelligent vehicle path prediction system for autonomous driving according to claim 1, characterized in that: The path optimization module: A collision detection unit, configured to determine whether the avoidance path plan has a collision risk based on a collision detection algorithm, and select a safe avoidance path; The adjustment cost calculation unit is used to compare the safe avoidance path with the original driving path of the vehicle to obtain a path adjustment cost. When the path adjustment cost meets a preset threshold, the safe avoidance path is used as the optimal path.

5. The intelligent vehicle path prediction system for autonomous driving according to claim 1, characterized in that: The waypoints include vehicle position, speed, and acceleration.

6. The intelligent vehicle path prediction system for autonomous driving according to claim 1, characterized in that: The process of constructing the vehicle dynamics model is as follows: Obtaining a vehicle velocity vector and a projected velocity vector based on the vehicle's actual position vector and the projected vector; Calculating the lateral velocity of the vehicle in a Frenet coordinate system based on the vehicle velocity vector and the projected velocity vector; Obtaining an error set based on the lateral velocity, the vehicle yaw angle, and the center of mass sideslip angle; Based on the error set, a state variable and a control variable are obtained; The vehicle dynamics model is obtained based on the state variable and the control variable.

Images