Automatic driving path planning method based on artificial intelligence

Abstract

The invention provides an automatic driving path planning method based on artificial intelligence. The automatic driving path planning method based on artificial intelligence comprises the steps of a, obtaining real-time environment information which comprises but is not limited to lane information, obstacle information, traffic sign information and road condition information, b, carrying out preprocessing and data fusion on the obtained real-time environment information, and c, carrying out data fusion on the obtained real-time environment information, the data fusion process is carried out through Kalman filtering so as to reduce data noise and improve the precision of environmental perception. According to the automatic driving path planning method based on artificial intelligence, a Kalman filtering data fusion method is adopted, the calculation efficiency is optimized on the basis of environmental perception, and dependence on calculation resources is reduced. And meanwhile, the running time, the vehicle stability and the energy efficiency are balanced by using a multi-objective optimization method, so that the system can efficiently operate in an environment with a high real-time requirement.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence-based autonomous driving, and specifically to an artificial intelligence-based autonomous driving path planning method. Background Art

[0002] The artificial intelligence-based autonomous driving path planning method uses artificial intelligence technology to help the autonomous driving system plan a safe, optimal, and real-time driving path. This method is generally implemented through the following steps: First, the perception module is used to obtain real-time environmental information; then, the path planning module generates a path based on the perception data combined with the current vehicle state; finally, the control module controls the driving of the vehicle according to the planned path. Common path planning methods include graph-based algorithms and deep learning-based algorithms, among which the deep learning method can handle more complex and dynamic driving environments and has strong adaptability and robustness.

[0003] Although the artificial intelligence-based path planning method performs well in many aspects, there are still great challenges in terms of computational resource consumption. Deep learning algorithms usually require a large amount of training data and high computing power, which poses great pressure on the implementation of real-time path planning systems. Especially when dealing with large-scale traffic networks or high-complexity scenarios, the computational overhead may be too high, resulting in a decrease in the system response speed and even being unable to meet the real-time driving requirements. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides an artificial intelligence-based autonomous driving path planning method, which solves the problem of high computational resource consumption in the autonomous driving path planning method.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: The artificial intelligence-based autonomous driving path planning method includes a. Obtain real-time environmental information, where the real-time environmental information includes but is not limited to lane information, obstacle information, traffic sign information, and road surface condition information; b. Preprocess and fuse the obtained real-time environmental information. The data fusion process is carried out through Kalman filtering to reduce data noise and improve the accuracy of environmental perception; c. Generate a preliminary driving path according to the perception data and the current vehicle state, and the path planning module executes the calculation of path planning; d. Optimize the path based on real-time traffic signal information and road conditions. By combining traffic flow, the generated path is dynamically adjusted according to the real-time traffic signal information. The path optimization formula is: , where is the cost of the path segment, is the estimated travel time for each route segment, is the weighted coefficient of traffic flow, and the optimization goal is to minimize the overall travel time; e. Generate actual driving instructions based on the optimized path planning results, and control the vehicle's steering, acceleration, and braking actions through a control module. The control module adjusts vehicle behavior in real time based on the vehicle's current state and path planning results; f. During vehicle travel, the navigation system dynamically adjusts the route based on real-time sensor data and environmental changes. When an emergency is detected, the navigation system replans and adjusts the vehicle's route using a path correction algorithm. g. executing the dynamically adjusted path planning result in real time through the control module.

[0006] Preferably, the real-time environmental information is obtained from a sensor through a perception module, and the perception module includes a laser radar and a camera. The real-time environmental information is represented as ,in Representative Perception data, and each perception data contains the distance and relative speed between the vehicle and the surrounding environment.

[0007] Preferably, the path planning module aims to minimize the To the target state Path cost ,in is the coordinate position of the vehicle, is the vehicle orientation, and the path planning formula is: , in, is the speed of the vehicle, is the steering angle, and is the weighting coefficient to adjust the relative importance of speed and steering angle. The state transition of the vehicle is described by the following formula: , in is the vehicle wheelbase, is the steering angle.

[0008] Preferably, the real-time traffic signal information includes: traffic flow, traffic light status, and construction area. The real-time traffic signal information is updated in real time through a navigation system and used for route optimization.

[0009] Preferably, the control module generates control instructions for the vehicle in real time through a predictive control algorithm according to the path planning result, and adjusts the steering, acceleration, and braking actions. The control module further includes an obstacle avoidance module. When the vehicle approaches an obstacle, the control module automatically performs an avoidance operation.

[0010] Preferably, the dynamically adjusted path planning is described by the following formula: , where, is the path after dynamic adjustment, is the path correction amount based on real-time environmental changes.

[0011] Preferably, the path correction algorithm adopts a model-based predictive control method to correct the path according to real-time data and control the vehicle to adjust. The navigation system obtains global traffic data through a cloud computing platform and combines it with local real-time data for path planning.

[0012] Preferably, the path planning module adopts a multi-objective optimization method during the planning process to balance driving time, vehicle stability, and energy efficiency. The control module is adjusted by the deviation between the current orientation and the target orientation of the vehicle.

[0013] The present invention provides an autonomous driving path planning method based on artificial intelligence, having the following beneficial effects: This autonomous driving path planning method based on artificial intelligence optimizes the calculation efficiency on the basis of environmental perception and reduces the dependence on computing resources by adopting the Kalman filter data fusion method. At the same time, the multi-objective optimization method is used to balance driving time, vehicle stability, and energy efficiency, enabling the system to operate efficiently in an environment with high real-time requirements and avoiding the excessive dependence on computing power in traditional methods. In terms of real-time dynamic path optimization, by combining information such as real-time traffic flow, traffic signal status, and construction areas, the navigation system can perform path dynamic adjustment. Especially when emergencies occur, it can quickly correct the path to ensure the vehicle's rapid response and avoid the problem of unreasonable paths caused by ignoring traffic changes in traditional path planning methods.

[0014] The present invention also introduces an efficient obstacle avoidance ability. The system can perform timely avoidance when the vehicle approaches an obstacle to ensure safe driving. Combined with the model-based predictive control method, the control module can dynamically correct the path according to real-time perception data, further enhancing the system's response ability to emergencies. By integrating a cloud computing platform, the system can obtain real-time traffic signal information worldwide and combine it with local data for path planning, thereby improving navigation accuracy, optimizing the overall traffic flow, and ensuring the system's adaptability in different traffic environments. Especially in long-distance or complex urban environments, it can perform precise path correction. Brief Description of the Drawings

[0015] Figure 1 This is a schematic structural diagram of the present invention. Detailed Description of the Embodiments

[0016] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0017] Embodiment 1

[0018] As Figure 1 shown, the embodiment of the present invention provides an artificial intelligence-based autonomous driving path planning method, including: a. Obtain real-time environment information, which includes but is not limited to lane information, obstacle information, traffic sign information, and road surface condition information. The real-time environment information is obtained from sensors through a perception module, and the perception module includes lidar and cameras. The real-time environment information is represented as , where represents the th perception data, and each perception data includes the distance between the vehicle and the surrounding environment and the relative speed.

[0019] b. Preprocess and fuse the obtained real-time environment information. The data fusion process is carried out through Kalman filtering to reduce data noise and improve the accuracy of environmental perception.

[0020] c. Generate a preliminary driving path according to the perception data and the current vehicle state. The path planning module performs the calculation of path planning. The goal of the path planning module is to minimize the path cost from the current state to the target state , where is the coordinate position of the vehicle, is the vehicle orientation. The path planning formula: , where, is the speed of the vehicle, is the steering angle, and are weighting coefficients to adjust the relative importance of the speed and the steering angle. The state transition of the vehicle is described by the following formula: , where is the wheelbase of the vehicle is the steering angle. The path planning module adopts a multi-objective optimization method during the planning process to balance driving time, vehicle stability, and energy efficiency.

[0021] d. Optimize the path based on real-time traffic signal information and road conditions. Dynamically adjust the generated path by combining traffic flow and real-time traffic signal information. The path optimization formula is: , where is the cost of the path segment, is the estimated driving time of each path segment, is the weighting coefficient of traffic flow. The optimization goal is to minimize the overall driving time. Real-time traffic signal information includes: traffic flow, traffic signal status, construction areas. The real-time traffic signal information is updated in real time through the navigation system and used for path optimization.

[0022] e. Generate actual driving instructions according to the optimized path planning result, and control the steering, acceleration, and braking actions of the vehicle through the control module. The control module adjusts the vehicle behavior in real time according to the current state of the vehicle and the path planning result. The control module generates vehicle control instructions in real time through a model predictive control algorithm and adjusts the steering, acceleration, and braking actions. The control module also includes an obstacle avoidance module. When the vehicle approaches an obstacle, the control module automatically performs an avoidance operation. The control module is adjusted by the deviation between the current orientation of the vehicle and the target orientation.

[0023] f. During the vehicle driving process, the navigation system dynamically corrects the path according to real-time perception data and environmental changes. When an emergency is detected, the navigation system re-plans and adjusts the vehicle driving route through a path correction algorithm. The path correction algorithm adopts a model-based predictive control method to correct the path according to real-time data and control the vehicle to adjust. The navigation system obtains global traffic data through the cloud computing platform and combines local real-time data for path planning.

[0024] g. Execute the dynamically adjusted path planning result in real time through the control module to ensure that the vehicle always drives safely in a complex traffic environment. The dynamically adjusted path planning is described by the following formula: , where, is the path after dynamic adjustment, is the path correction amount based on real-time environmental changes.

[0025] Embodiment 2 Traditional autonomous driving path planning methods: a. Obtain real-time environment information, which includes lane information, obstacle information, traffic sign information, road surface condition information, etc. The real-time environment information is obtained from sensors through a perception module. In this embodiment, the original perception data is directly used without additional data preprocessing steps. The perception module includes lidar and cameras, and the obtained perception data is directly used for path planning calculation.

[0026] b. Generate a driving path based on the obtained real-time environment information and vehicle state. The path planning module calculates the path from the current state to the target state through traditional shortest path algorithms without the need for complex multi-objective optimization. During the path planning process, the direct distance and shortest driving time of the path are mainly concerned, while the stability of the vehicle and the balance of energy efficiency are ignored.

[0027] c. When performing path optimization, this embodiment does not combine real-time traffic signal information and road conditions, but uses static preset traffic data to correct the path. The optimization after path planning is mainly based on the preset traffic flow and fixed traffic signal schedule in advance. When emergencies occur, the system cannot be adjusted in time and cannot effectively respond to traffic changes.

[0028] d. After the path is generated, driving instructions are generated and executed through the control module. However, the control module in this embodiment only performs basic steering, acceleration, and braking control according to the planned path, without considering environmental changes or emergencies. The control module does not have an obstacle avoidance function, and the vehicle can only drive on the preset path and cannot cope with sudden obstacles during driving.

[0029] e. During driving, when the vehicle detects an emergency, the traditional method cannot perform path correction, and the vehicle may continue to drive according to the original path, increasing the safety risk. Path correction relies on manual intervention or inefficient system adjustment and cannot perform real-time dynamic optimization.

[0030] f. In this embodiment, the path planning is not combined with the cloud computing platform. The system cannot obtain real-time traffic data globally and cannot perform real-time path correction according to environmental changes. The adaptability of the system is poor, especially in complex traffic environments, and the path correction ability is limited.

[0031] g. Due to the lack of a dynamic adjustment mechanism in path planning, the vehicle may have problems such as unreasonable paths and unsafe driving in complex traffic environments and cannot ensure safe driving at all times.

[0032] Comparative analysis: In this embodiment, efficient data fusion, real-time path optimization, and dynamic adjustment are not performed during the path planning process, resulting in the system having weak capabilities in dealing with emergencies, complex traffic conditions, and obstacle avoidance, and unable to achieve efficient and safe autonomous driving functions.

[0033] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An artificial intelligence-based automatic driving path planning method, characterized in that, Including: a. Obtain real-time environment information, which includes but is not limited to lane information, obstacle information, traffic sign information, and road surface condition information; b. Preprocess and perform data fusion on the obtained real-time environment information, and the data fusion process is carried out through Kalman filtering; c. Generate a preliminary driving path according to the perception data and the current vehicle state, and the path planning module executes the calculation of path planning; d. Optimize the path based on real-time traffic signal information and road conditions. By combining traffic flow, the real-time traffic signal information dynamically adjusts the generated path. The path optimization formula is: ， wherein is the cost of the path segment, is the estimated travel time of each path segment, is the weighted coefficient of traffic flow, and the optimization objective is to minimize the overall travel time; e. Generate actual driving instructions according to the optimized path planning result, and control the steering, acceleration, and braking actions of the vehicle through the control module. The control module adjusts the vehicle behavior in real time according to the current state of the vehicle and the path planning result; f. During the vehicle driving process, the navigation system dynamically corrects the path according to real-time perception data and environmental changes. When an emergency is detected, the navigation system replans and adjusts the vehicle driving route through the path correction algorithm; g. Execute the dynamically adjusted path planning result in real time through the control module.

2. The method for autonomous driving path planning based on artificial intelligence according to claim 1, characterized in that: The real-time environmental information is obtained from sensors by the perception module, which includes lidar and cameras. The real-time environmental information is represented as , where represents the th perception data, and each perception data includes the distance between the vehicle and the surrounding environment and the relative speed.

3. The method for autonomous driving path planning based on artificial intelligence according to claim 1, characterized in that: The goal of the path planning module is to minimize the path cost from the current state to the target state , where is the coordinate position of the vehicle, is the vehicle orientation, and the path planning formula is: ， Among them, is the speed of the vehicle, is the steering angle, and are weighting coefficients to adjust the relative importance of speed and steering angle. The state transition of the vehicle is described by the following formula: ， Among them is the wheelbase of the vehicle, is the steering angle.

4. The method for autonomous driving path planning based on artificial intelligence according to claim 1, characterized in that: The real-time traffic signal information includes: traffic flow, traffic signal light status, and construction area. The real-time traffic signal information is updated in real time through the navigation system and is used for path optimization.

5. The method for autonomous driving path planning based on artificial intelligence according to claim 1, characterized in that: The control module generates vehicle control instructions in real time through the predictive control algorithm according to the path planning result, and adjusts the steering, acceleration, and braking actions. The control module also includes an obstacle avoidance module. When the vehicle approaches an obstacle, the control module automatically performs an avoidance operation.

6. The method for autonomous driving path planning based on artificial intelligence according to claim 1, wherein: The dynamically adjusted path planning is described by the following formula: ， Among them, is the path after dynamic adjustment, is the path correction amount based on real-time environmental changes.

7. The method for autonomous driving path planning based on artificial intelligence according to claim 1, wherein: The path correction algorithm adopts a model-based predictive control method to correct the path according to real-time data and control the vehicle to adjust. The navigation system obtains global traffic data through the cloud computing platform and combines local real-time data for path planning.

8. The method for autonomous driving path planning based on artificial intelligence according to claim 1, wherein: The path planning module adopts a multi-objective optimization method during the planning process to balance driving time, vehicle stability, and energy efficiency. The control module is adjusted by the deviation between the current orientation of the vehicle and the target orientation.

Images