Intelligent engine control method and device based on collision risk assessment

Abstract

The invention provides an intelligent engine control method and device based on collision risk assessment. The method comprises the following steps: acquiring obstacle information on a road in front of a vehicle in real time by using various sensors; the obstacle information is preprocessed, an occupation network model is utilized to process the preprocessed data, the occupation network model is used for dividing a sensing area into voxel grids, and feature extraction and occupation prediction are carried out on each voxel to predict the position and the type of an obstacle; calculating collision time by using a predetermined dynamic collision time algorithm based on obstacle prediction information output by the occupancy network model; and adjusting the running state and running parameters of the engine based on a comparison result between the collision time and a preset threshold range of the collision time. The fuel oil leakage and vehicle spontaneous combustion risks can be reduced, and the safety performance of the vehicle is improved.

Description

Technical Field

[0001] The present application relates to the field of new energy vehicle technology, and in particular to an intelligent engine control method and device based on collision risk assessment. Background Art

[0002] With the development of intelligent driving technology, smart cars have made continuous progress in improving safety, especially by integrating sensors and control systems to improve the active safety performance of vehicles. In order to deal with sudden collision accidents, automatic emergency braking (AEB) and engine control systems are usually used in existing technologies to reduce or avoid damage caused by collisions. However, existing engine shutdown control technology still has certain limitations.

[0003] Prior art 1 provides a collision fuel cut-off control system for range-extended vehicles, which cuts off the fuel supply to the range-extender engine after a vehicle collision through hard-wired connections of a collision sensor, an airbag controller, and an engine controller. However, this solution can only trigger the fuel cut-off after a collision and cannot effectively prevent the risk of fuel leakage before a collision.

[0004] The second existing technology integrates camera and millimeter wave radar technology to improve the accuracy of obstacle recognition and control the opening and closing of the engine throttle to prevent the failure of automatic emergency braking caused by the driver's mistaken stepping on the accelerator pedal. However, this technology only controls the engine throttle, and the engine is still in operation when there is a risk of collision. It is impossible to stop the fuel supply in time, and there is still a risk of fuel leakage and spontaneous combustion after the collision.

[0005] Therefore, the existing technology cuts off the fuel supply only after the vehicle collides, and fuel leakage and fire may occur when the vehicle structure is severely damaged. The intelligent driving system does not actively shut down the engine before the risk of collision occurs, resulting in the risk of fuel leakage and spontaneous combustion after the collision. Summary of the invention

[0006] In view of this, an embodiment of the present application provides an intelligent engine control method and device based on collision risk assessment to solve the problem in the prior art that the engine's fuel supply is cut off only after a vehicle collision occurs, which easily leads to fuel leakage and the risk of vehicle spontaneous combustion, and reduces the vehicle's safety performance.

[0007] According to a first aspect of an embodiment of the present application, an intelligent engine control method based on collision risk assessment is provided, comprising: using a plurality of sensors to collect obstacle information on a road ahead of a vehicle in real time, wherein the obstacle information includes the speed and acceleration of the obstacle; preprocessing the obstacle information, and processing the preprocessed data using an occupancy network model, wherein the occupancy network model is used to divide a sensing area into a voxel grid, and perform feature extraction and occupancy prediction on each voxel to predict the location and type of the obstacle; based on the obstacle prediction information output by the occupancy network model, calculating the collision time using a predetermined dynamic collision time algorithm, wherein the dynamic collision time algorithm is used to dynamically adjust collision time calculation parameters according to environmental conditions and vehicle status; and adjusting the operating state and operating parameters of the engine based on a comparison result between the collision time and a preset threshold range of the collision time.

[0008] According to a second aspect of an embodiment of the present application, an intelligent engine control device based on collision risk assessment is provided, comprising: an acquisition module configured to use a variety of sensors to collect obstacle information on a road ahead of a vehicle in real time, wherein the obstacle information includes the speed and acceleration of the obstacle; a processing module configured to pre-process the obstacle information and process the pre-processed data using an occupancy network model, wherein the occupancy network model is used to divide the perception area into voxel grids, perform feature extraction and occupancy prediction on each voxel to predict the location and type of the obstacle; a calculation module configured to calculate the collision time based on the obstacle prediction information output by the occupancy network model using a predetermined dynamic collision time algorithm, wherein the dynamic collision time algorithm is used to dynamically adjust the collision time calculation parameters according to environmental conditions and vehicle status; a control module configured to adjust the operating state and operating parameters of the engine based on the comparison result between the collision time and a preset threshold range of the collision time.

[0009] According to a third aspect of an embodiment of the present application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the steps of the above method are implemented when the processor executes the computer program.

[0010] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores a computer program, and when the computer program is executed by a processor, the steps of the above method are implemented.

[0011] At least one of the above technical solutions adopted in the embodiments of the present application can achieve the following beneficial effects:

[0012] By using a variety of sensors to collect obstacle information on the road ahead of the vehicle in real time, the obstacle information includes the speed and acceleration of the obstacle; preprocessing the obstacle information, and processing the preprocessed data using an occupancy network model, wherein the occupancy network model is used to divide the sensing area into voxel grids, and perform feature extraction and occupancy prediction on each voxel to predict the location and type of the obstacle; based on the obstacle prediction information output by the occupancy network model, the collision time is calculated using a predetermined dynamic collision time algorithm, wherein the dynamic collision time algorithm is used to dynamically adjust the collision time calculation parameters according to environmental conditions and vehicle status; based on the comparison result between the collision time and the preset collision time threshold range, the operating state and operating parameters of the engine are adjusted. This application can reduce the risk of fuel leakage and vehicle spontaneous combustion and improve the safety performance of the vehicle. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0014] Figure 1 It is a schematic diagram of the overall architecture of the intelligent engine control system provided in the embodiment of the present application;

[0015] Figure 2 is a flow chart of an intelligent engine control method based on collision risk assessment provided in an embodiment of the present application;

[0016] Figure 3 is a structural schematic diagram of an intelligent engine control device based on collision risk assessment provided in an embodiment of the present application;

[0017] Figure 4 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0019] It should be understood that the various steps described in the method implementation of the present application can be performed in different orders and / or performed in parallel. In addition, the method implementation may include additional steps and / or omit the steps shown. The scope of the present application is not limited in this respect.

[0020] The term "including" and its variations used in this document are open inclusions, that is, "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one other embodiment"; the term "some embodiments" means "at least some embodiments". Relevant definitions of other terms will be given in the description below. It should be noted that the concepts of "first", "second", etc. mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0021] It should be noted that the modifications of "one" and "plurality" mentioned in the present application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0022] In the field of smart cars, especially in the context of active engine shutdown technology, current technical solutions mainly focus on cutting off the engine fuel supply in certain situations to prevent the vehicle from continuing to run. However, these existing technologies still have some defects, resulting in risks in actual use.

[0023] The active safety system of smart cars is designed to detect the environment around the vehicle through a variety of sensors (such as collision sensors, cameras, millimeter-wave radars, etc.), and control the vehicle's drive system (such as the engine) according to different situations to reduce the occurrence of accidents or reduce the severity of accidents. At present, the automatic emergency braking (AEB) and active safety technologies of smart car systems can trigger some safety mechanisms when the vehicle collides, such as cutting off the oil line and shutting down the engine. However, these technologies still do not fully consider the solution of actively shutting down the engine before a collision is about to occur or in potential danger, which may bring some hidden dangers.

[0024] The technical solution of the prior art one is a collision fuel cut-off control system for extended-range vehicles. It uses a collision sensor and an airbag controller, relying on a hard-wired connection, to trigger the cut-off of the fuel supply after a vehicle collision. Specifically, after the collision sensor detects a vehicle collision, it sends a signal to the airbag controller, which then transmits the signal to the range extender engine controller, and finally the range extender engine controller cuts off the fuel supply to the engine. The system will only cut off the fuel supply when the collision has occurred and the vehicle body structure is severely damaged. Since the fuel supply is not shut down in advance, fuel leakage may occur after a collision, increasing the risk of vehicle fire and spontaneous combustion.

[0025] The technical solution of the second prior art improves the recognition accuracy of obstacles ahead by integrating the camera and millimeter-wave radar technology. When a potential danger is detected, the system controls the engine's throttle to prevent the vehicle from continuing to accelerate. This technology can effectively reduce serious accidents caused by the driver accidentally stepping on the accelerator pedal. However, this solution only controls the engine's throttle and does not completely shut down the engine. Although the system controls the throttle when a danger is detected, the engine remains in operation and the fuel supply is not cut off, which may still lead to the risk of fuel leakage and fire after a collision.

[0026] Therefore, the existing technology has obvious deficiencies in dealing with vehicle collision risks, which are mainly reflected in the following two aspects:

[0027] 1) Risk of fuel leakage after a collision: Existing technologies cut off the fuel supply to the engine only after a collision occurs. However, by then the vehicle may have already suffered serious damage, increasing the risk of fuel leakage or even fire.

[0028] 3) Inability to proactively prevent: Existing technologies fail to proactively shut down the engine before a collision occurs or when a potential risk is identified. Although the system can control the opening and closing of the engine throttle, the engine itself is still running, failing to completely eliminate safety hazards after an accident.

[0029] The main problem with the current existing technology is that the control of the engine fuel supply is still mainly post-event, that is, it reacts only after the collision occurs, and fails to fully consider the preventive strategy, that is, actively shutting down the engine and cutting off the fuel supply when the collision risk is identified, thereby reducing the possibility of fuel leakage and vehicle spontaneous combustion. This defect exposes the limitations of the existing technology when facing vehicle safety issues, and it is urgently needed to improve overall safety.

[0030] In response to the above-mentioned problems existing in the prior art, the present application provides an intelligent engine (range extender) control method based on collision risk assessment, aiming to solve the deficiency in the prior art that the engine cannot be actively shut down and the fuel supply is cut off before a collision.

[0031] The following is an explanation of the architecture of the intelligent engine control system involved in the actual scenario of the technical solution of the present application in conjunction with the accompanying drawings. Figure 1 Schematic diagram of the overall architecture of the intelligent engine control system provided by the embodiment of the present application. Figure 1 As shown, the overall architecture and implementation method of the intelligent engine control system may specifically include the following contents:

[0032] The technical solution of this application monitors the road conditions ahead in real time by installing a variety of intelligent assisted driving sensors, including cameras, lidars, and millimeter-wave radars, in front of the vehicle. The sensors can accurately identify key data such as the speed and acceleration of the target ahead, and send this information to the intelligent driving control system. The intelligent driving control system uses an occupancy network model to process sensor data. First, the perceived 3D space is divided into a voxel grid, and features are extracted for each voxel. Then, occupancy prediction is performed based on these features to determine the location and type of obstacles.

[0033] After data processing is completed, the intelligent driving control system uses a dynamic time-to-collision (TTC) calculation algorithm to evaluate the collision time between the vehicle and the obstacle based on the processed obstacle information. The dynamic TTC algorithm can dynamically adjust the calculation parameters of the collision time according to environmental conditions (such as road conditions, weather conditions) and vehicle status (such as current speed, acceleration), thereby more accurately predicting the risk of collision. If there are multiple targets ahead, the system can calculate the TTC of each target at the same time, and give priority to the target with the shortest TTC to better deal with the risk of multi-target collision in complex traffic environments.

[0034] Based on the calculated TTC value, the intelligent driving control system adopts a phased control strategy: when TTC is at a higher threshold (such as above 3.6 seconds), the system sends a collision warning signal to the driver and begins to gradually reduce the engine speed; when TTC drops to the intermediate threshold (such as between 1.6 seconds and 3.6 seconds), the system activates the automatic emergency braking (AEB) function, while further reducing the engine speed and preparing to stop the fuel supply; when TTC is at the shortest threshold (such as less than 1.6 seconds), the system immediately takes full braking, while the engine stops running and cuts off the fuel supply to avoid fuel leakage and fire risks after a collision.

[0035] An intelligent engine control method and device based on collision risk assessment provided by an embodiment of the present application will be described in detail below in conjunction with the accompanying drawings and embodiments.

[0036] Figure 2 is a flow chart of an intelligent engine control method based on collision risk assessment provided by an embodiment of the present application. Figure 2As shown, the intelligent engine control method based on collision risk assessment may specifically include:

[0037] S201, using a variety of sensors to collect obstacle information on the road ahead of the vehicle in real time, where the obstacle information includes the speed and acceleration of the obstacle;

[0038] S203, preprocessing the obstacle information, and processing the preprocessed data using an occupancy network model, wherein the occupancy network model is used to divide the sensing area into a voxel grid, perform feature extraction and occupancy prediction on each voxel, so as to predict the location and type of the obstacle;

[0039] S203, calculating the collision time using a predetermined dynamic collision time algorithm based on the obstacle prediction information output by the occupancy network model, wherein the dynamic collision time algorithm is used to dynamically adjust collision time calculation parameters according to environmental conditions and vehicle status;

[0040] S204: Based on the comparison result between the collision time and the preset collision time threshold range, the operating state and operating parameters of the engine are adjusted.

[0041] In some embodiments, before using multiple sensors to collect obstacle information on the road ahead of the vehicle in real time, the method further includes:

[0042] The laser radar, multi-function camera and millimeter-wave radar are used to identify the objects in front of the vehicle respectively, and the first target signal output by the laser radar, the second target signal output by the multi-function camera and the third target signal output by the millimeter-wave radar are obtained; the first target signal, the second target signal and the third target signal are subjected to multi-band signal fusion, and the obstacle is determined according to the multi-band signal fusion result.

[0043] Specifically, for the intelligent driving system of hybrid vehicles, in order to more accurately identify obstacles on the road ahead of the vehicle and achieve precise control, before using multiple sensors to collect obstacle information in real time, the system first identifies the objects in front through different types of sensors and performs multi-band signal fusion to ensure the reliability and accuracy of obstacle detection.

[0044] First, the sensors installed forward on the vehicle include LiDAR, multi-function cameras, and millimeter-wave radar. Each of these three sensors has different working principles and performance advantages: LiDAR has high-precision morphological recognition capabilities and can accurately measure the shape and distance of objects; multi-function cameras can capture visual information such as the color and texture of objects; millimeter-wave radar is good at penetrating environmental interference in adverse weather conditions such as rain and fog, and provides stable object detection capabilities. These sensors can identify different types of targets in front in real time, such as vehicles, pedestrians, two-wheeled vehicles, and special-shaped objects.

[0045] Furthermore, the laser radar, the multi-function camera and the millimeter-wave radar output the first target signal, the second target signal and the third target signal respectively, which represent the detection results of each sensor on the target object. In order to improve the accuracy of the system, the system further adopts multi-band signal fusion technology to generate a comprehensive target detection result by fusing the detection signals of the above sensors. The multi-band signal fusion process can effectively utilize the characteristics of each sensor. For example, the millimeter-wave radar can cope with low visibility environments, while the laser radar and the camera can provide high-precision morphological information. The fused detection results can more accurately identify the type and location of the target object, especially in complex environments (such as insufficient light, rainy and foggy weather, etc.), to ensure effective monitoring of obstacles.

[0046] Furthermore, the sensor system can further expand its types and combinations. For example, on the basis of LiDAR, millimeter-wave radar and camera, ultrasonic sensors and infrared sensors can also be added. Ultrasonic sensors can be used to detect obstacles at close range, while infrared sensors can detect the thermal radiation signals of objects under low light conditions. Through the fusion of these different types of sensors, the system can improve its adaptability to different scenes and environments, and accurately identify obstacles under various complex conditions.

[0047] In addition, in order to enhance the response speed and detection accuracy of the system, a multimodal fusion algorithm based on deep learning can also be introduced. By learning the correlation between different sensor data, the algorithm can more intelligently fuse the output signals of various sensors and improve the accuracy and real-time performance of obstacle recognition. At the same time, advanced sensor data processing algorithms such as Kalman filters can also be used to filter noise signals and reduce data fluctuations, thereby improving the system's prediction accuracy of the target's position and motion state.

[0048] Through the above multi-sensor fusion and data processing methods, the system can quickly and accurately identify obstacles ahead under different environmental conditions and provide accurate data support for subsequent intelligent driving control. This fusion method effectively reduces false alarms and missed alarms, improves the reliability and response speed of the system, and enables vehicles to drive safely even in complex road conditions.

[0049] In some embodiments, the obstacle information is pre-processed, including:

[0050] De-noising the obstacle information so as to remove the noise and interference signals in the sensor data using filtering algorithms;

[0051] The obstacle information collected by different sensors is time-series aligned so that the obstacle information can be synchronously processed at the same time point based on the data collection timestamp of each sensor;

[0052] Detect and filter abnormal data of obstacle information in order to identify and eliminate abnormal data.

[0053] Specifically, the intelligent driving system pre-processes obstacle information from multiple sensors to ensure the accuracy and consistency of the data, thereby providing a reliable basis for subsequent decision-making and control. In one example, the pre-processing process of obstacle information includes the following:

[0054] First, the system receives obstacle information from multiple sensors such as cameras, lidars, and millimeter-wave radars. Since the data collected by sensors in different environments and conditions may contain noise and interference signals, the system first denoises these data. Specifically, the data of each sensor is processed using a filtering algorithm to eliminate noise and irrelevant signals in the data. The filtering algorithm can select a suitable filtering method based on the characteristics of different sensors, such as using a low-pass filter to eliminate high-frequency noise in high-speed signals, ensuring that the obstacle information output by each sensor has a high signal-to-noise ratio, thereby improving the accuracy of the data.

[0055] Secondly, since there may be slight differences in the timestamps of obstacle information collected by multiple sensors, the intelligent driving system also needs to align the data from different sensors. In order to ensure that the data from various sensors can be processed synchronously at the same time point, the system aligns the information output by each sensor based on the data collection timestamp of each sensor. This process ensures data consistency by matching the timestamps of each sensor, so that subsequent obstacle identification and collision risk assessment can be completed based on accurate information at the same time.

[0056] Furthermore, after completing denoising and timing alignment, the system also detects and filters abnormal data for obstacle information. Abnormal data may be caused by sensor failure, environmental interference or other unpredictable factors, so it must be eliminated before the data enters the intelligent driving control system. The system identifies and eliminates abnormal data that may affect subsequent decisions by setting a reasonable range for data, checking data consistency and monitoring sensor status. This process ensures that the data processed by the system is valid and reliable obstacle information, avoiding misjudgments and erroneous responses caused by abnormal data.

[0057] Through the above preprocessing steps, the system can ensure that obstacle information from different sensors is highly accurate and consistent after cleaning, alignment and filtering, providing reliable data support for subsequent collision time calculation and vehicle control. This data preprocessing mechanism can effectively improve the stability and response speed of the system in complex traffic environments and changing external conditions.

[0058] In some embodiments, the pre-processed data is processed using the occupancy network model, including:

[0059] The occupancy network model is used to divide the perceived three-dimensional space into a fixed-size voxel grid, where each voxel represents a specific area in the three-dimensional space and the voxel is used as the basic unit of algorithm processing;

[0060] Based on the multi-view image data collected by multiple sensors, feature extraction is performed on each voxel to obtain the spatial features corresponding to each voxel;

[0061] Based on the extracted spatial features, the occupancy state of each voxel is predicted using the occupancy network model to determine the probability of each voxel being occupied and the type of object occupying the space, and to divide the three-dimensional space into occupied areas and free areas;

[0062] The data collected by multiple sensors from different perspectives are fused, and the occupancy network model predicts the occupancy state of the voxel grid in real time based on the fused multi-perspective information, and the classification results are adjusted according to environmental changes.

[0063] Specifically, the intelligent driving system processes the pre-processed sensor data by introducing an occupancy network model to identify obstacles in front of the vehicle and perform spatial division and occupancy state prediction on them. In one example, the present embodiment processes the pre-processed data using an occupancy network model, including the following:

[0064] First, the system obtains three-dimensional data of the road environment ahead through multiple sensors (such as lidar, camera, and millimeter-wave radar). In order to effectively process this data, the occupancy network model first divides the perceived three-dimensional space into a fixed-size voxel grid. Each voxel represents a specific area in the three-dimensional space and represents the basic unit of system analysis. This spatial division method can structure complex three-dimensional scenes and facilitate subsequent data processing.

[0065] Next, the system extracts features from each voxel. Based on multi-view image data from multiple sensors (including lidar data, visual information captured by cameras, and environmental information from millimeter-wave radars), relevant features are extracted for each voxel. These features may include color, depth, texture, distance, etc., which are used to describe the spatial information contained in each voxel. These features provide basic data support for subsequent classification and occupancy prediction.

[0066] Furthermore, after feature extraction is completed, the system uses the occupancy network model to predict the occupancy state of each voxel. Specifically, based on the extracted spatial features, the system calculates the occupancy probability of each voxel, that is, determines whether the voxel is occupied by an object and identifies possible object types. Through this process, the system can divide the entire three-dimensional space into occupied areas and free areas, and preliminarily classify the object types in the occupied areas. This is a typical three-dimensional space segmentation task, which aims to help the system identify vehicles, pedestrians, obstacles, etc. ahead.

[0067] Furthermore, in order to further improve the accuracy of occupancy prediction, the system also uses multi-view fusion technology to fuse multi-view data from different sensors. Multiple cameras and sensors provide scene information from different angles. The algorithm can fuse this information to reduce the occlusion or perspective limitation problems that may be caused by a single perspective, thereby understanding the surrounding scene more comprehensively and accurately. Through this multi-view fusion, the system can better identify obscured or partially hidden objects.

[0068] Finally, the system uses the occupancy network model for dynamic updates in real-time scenarios of autonomous driving. As the vehicle drives, the sensors continuously acquire new environmental information, and the occupancy network updates its prediction of the occupancy state of the voxel grid in real time based on the new data. This dynamic update process ensures that the system can promptly reflect changes in the environment. For example, when a new obstacle enters in front of the vehicle, the system can immediately update the prediction results and adjust the classification information of the obstacle. Through this real-time adjustment, the system can better cope with complex driving environments and improve driving safety.

[0069] It should be noted that the application advantages of the occupied network model compared to the traditional BEV model are as follows:

[0070] 1. Overcoming the long-tail problem: The occupancy network can predict unlabeled objects in the dataset (such as mounds, stones, etc.), thereby better overcoming the long-tail problem in detection.

[0071] 3. Fine-grained representation: Compared with algorithms that directly output 3D boxes of objects, occupancy networks can give finer-grained shape representations and provide richer detailed structural information.

[0072] 3. Open set object detection: The occupancy network can predict objects that do not belong to the dataset but exist in the scene, realizing open set object detection.

[0073] Through the method of the present embodiment described above, the system utilizes the occupancy network model to achieve efficient perception and prediction of the three-dimensional space around the vehicle, providing high-precision environmental information for the vehicle's intelligent driving decisions.

[0074] In some embodiments, based on the obstacle prediction information output by the occupancy network model, the collision time is calculated using a predetermined dynamic collision time algorithm, including:

[0075] Based on the obstacle prediction information output by the occupancy network model, determine the relative speed, relative acceleration and relative distance between the current vehicle and the obstacle;

[0076] Based on the relative speed, relative acceleration and relative distance, the collision time is calculated using the dynamic collision time algorithm. The collision time is used to represent the estimated time when the current vehicle collides with the obstacle in the current state.

[0077] When calculating the collision time, the collision time calculation parameters are dynamically adjusted based on environmental conditions and vehicle status. The environmental conditions include real-time road conditions and weather conditions, and the vehicle status includes real-time driving parameters of the vehicle.

[0078] Specifically, the intelligent driving system calculates the collision time (TTC) between the vehicle and the obstacle ahead using a dynamic collision time algorithm based on the obstacle information output by the occupancy network model. In one example, the process of calculating the collision time using the dynamic collision time algorithm in this embodiment may include the following:

[0079] First, the system obtains relevant information about the obstacle ahead through the occupancy network model, including the relative speed, relative acceleration, and relative distance of the obstacle. These parameters are the key basis for collision time calculation. Through multi-view fusion and feature extraction of the occupancy network model, the system can accurately determine the spatial relationship and motion state between the obstacle and the current vehicle.

[0080] Next, based on the above information, the system uses a dynamic time-to-collision algorithm to calculate the time to collision TTC. TTC is the time when the vehicle and the obstacle are expected to collide under the current conditions. Specifically, the system first calculates the preliminary time to collision based on the relative speed and relative acceleration of the obstacle and the vehicle, combined with the relative distance between the two. This calculation process usually uses a physical model to predict the change in the relative position of the two and determine whether the two are likely to collide in the future.

[0081] Furthermore, in order to ensure the accuracy of the collision time, the system introduces a dynamic adjustment mechanism, which adjusts the TTC calculation parameters according to environmental conditions and vehicle status. Specifically, it includes:

[0082] Environmental conditions: The system will correct the TTC calculation based on the current road conditions (such as road slope, wetness) and weather conditions (such as rain, snow, fog, etc.). For example, on a slippery road, the system will appropriately shorten the TTC to trigger the early warning and response mechanism.

[0083] Vehicle status: The system also monitors the vehicle's driving parameters in real time, such as current speed, acceleration, steering angle, etc., and adjusts the collision time calculation according to the vehicle's status. If the vehicle is accelerating or turning, the system will dynamically adjust the TTC calculation to adapt to the changing driving conditions.

[0084] The following is a detailed description of the calculation process of the collision time TTC with a specific example. After receiving the information sent by the sensor, the intelligent driving system will immediately make a judgment based on the vehicle's speed v3 and acceleration a3, and calculate the collision time TTC according to the dynamic collision time algorithm:

[0085]

[0086] Where v1 represents the speed of the obstacle, a1 represents the acceleration of the obstacle, v3 represents the speed of the current vehicle, a3 represents the acceleration of the current vehicle, and D rel It represents the relative distance between the current vehicle and the obstacle, and d0 represents the safe parking distance (set to 0.5m-3m).

[0087]

[0088] Where TTC represents the collision time, v rel represents the relative speed between the two vehicles, a rel Relative acceleration, D rel Indicates the relative distance between the current vehicle and the obstacle, v rel represents the relative speed between the current vehicle and the obstacle, a rel Indicates the relative acceleration between the current vehicle and the obstacle.

[0089] In addition, in some other embodiments, the system can further optimize TTC according to the driver's driving habits. For example, if the system detects that the driver usually reacts slowly (such as braking late), the threshold of TTC can be shortened, so that the system can take measures earlier when potential dangers are detected, thereby avoiding accidents.

[0090] Through this dynamic adjustment mechanism, the system can adapt to complex road environments and vehicle conditions in real time, ensuring more accurate calculation of collision time. In some complex traffic scenarios, such as urban roads or intersections, the system can calculate TTC based on multiple targets at the same time, and give priority to the most dangerous target, that is, the target with the shortest TTC value, to ensure that the vehicle can take appropriate preventive measures at any time.

[0091] Through the method of the present embodiment described above, the intelligent driving system can accurately predict the risk of collision between the vehicle and obstacles, and take corresponding control measures at the appropriate time, including deceleration, braking or engine shutdown, thereby improving the safety of the vehicle.

[0092] In some embodiments, when calculating the collision time, the method further includes:

[0093] When there are multiple obstacles in front of the vehicle, the dynamic collision time algorithm is used to simultaneously calculate the collision time of each obstacle, and the obstacles are prioritized according to the collision time of each obstacle. The obstacle with the shortest collision time is given the highest priority, and the corresponding collision risk level is determined according to the collision time of each obstacle, so that the corresponding control signal can be output according to the collision risk level.

[0094] Specifically, when the intelligent driving system detects that there are multiple obstacles in front of the vehicle, the system will use the dynamic collision time algorithm to calculate the collision time (TTC) of each obstacle and prioritize each obstacle according to its collision time. In one example, this embodiment may also include the following when calculating the collision time:

[0095] First, the system collects information about obstacles ahead through a variety of sensors installed in the front (such as lidar, camera, and millimeter-wave radar). For each obstacle, the system obtains its relative speed, relative acceleration, and relative distance. Based on these parameters, the system uses a dynamic collision time algorithm to calculate the TTC of each obstacle. TTC represents the estimated time when the vehicle collides with the obstacle under the current vehicle state.

[0096] Furthermore, when there are multiple obstacles in front of the vehicle, the system will calculate the TTC of each obstacle at the same time and prioritize these obstacles. The priority is determined based on the TTC value of each obstacle. Specifically, the obstacle with the shortest TTC is considered by the system to be the most potentially dangerous target and is therefore given the highest priority. The system regards this obstacle as the object that currently needs to be handled with priority, ensuring that the vehicle can respond to the most urgent collision risks in a timely manner.

[0097] On this basis, the system further determines the collision risk level based on the TTC of each obstacle. The collision risk level is usually divided into multiple levels, such as high risk, medium risk and low risk, and the specific classification depends on the range of TTC values. The shorter the TTC value, the higher the collision risk, and the more quickly the vehicle needs to take corresponding control measures. For example, when the TTC value is within a certain threshold range (such as less than 1.6 seconds), the system will mark the obstacle as a high-risk target and immediately output the corresponding control signal so that the system can take automatic emergency braking or other intervention measures.

[0098] For example, in other embodiments, the system can also flexibly handle multiple obstacles according to the complexity of the environment. For example, on a congested urban road or intersection, there may be multiple vehicles, pedestrians or other obstacles ahead at the same time. The intelligent driving system will calculate the collision time of all obstacles and determine their priority based on the TTC of each obstacle. At the same time, by comprehensively considering the distance, relative speed and relative acceleration of each obstacle, the most dangerous obstacle will be selected for priority response.

[0099] Once the system determines the highest priority obstacle and its collision risk level, the system will immediately output the corresponding control signal. These control signals may include deceleration, braking or engine shutdown to ensure that the vehicle avoids possible collision risks in time. For lower risk targets, the system can take relatively mild response measures, such as only issuing a warning signal or slightly slowing down.

[0100] Through this multi-target collision assessment mechanism, the system can quickly make the best driving decisions based on the TTC value of each obstacle in a complex traffic environment, thereby effectively improving vehicle safety and reducing potential collision risks.

[0101] In some embodiments, based on the comparison result between the collision time and the threshold range of the preset collision time, the operating state and operating parameters of the engine are adjusted, including:

[0102] When the collision time is within a first threshold range, a collision warning signal is generated, the collision warning signal is sent to a vehicle display device and an audible alarm is sounded, and a first control signal is sent to an engine controller so that the engine controller reduces the engine speed according to the first control signal;

[0103] When the collision time is within the second threshold range, the automatic emergency braking function is triggered, the current vehicle is controlled to brake and decelerate, and a second control signal is sent to the engine controller, so that the engine controller stops the engine and cuts off the fuel supply according to the second control signal;

[0104] When the collision time is within the third threshold range, maximum deceleration braking is performed on the current vehicle, and a third control signal is sent to the engine controller, so that the engine controller stops the engine running and cuts off the fuel supply according to the third control signal.

[0105] Specifically, the intelligent driving system compares the calculated collision time TTC with a preset collision time threshold range, and determines a corresponding intelligent engine control strategy based on the comparison result. In one example, the embodiment may include the following contents based on the comparison result between the collision time and the threshold:

[0106] First, the intelligent driving system calculates the collision time TTC between the vehicle and the obstacle ahead in real time and compares the TTC with three preset threshold ranges. These threshold ranges are used to determine different degrees of collision risk and trigger different control strategies accordingly.

[0107] When TTC is in the first threshold range (1.6 seconds < TTC ≤ 3.6 seconds): At this stage, the system believes that there is a risk of collision ahead, but it has not yet reached an emergency situation that requires immediate braking. The intelligent driving system generates a collision warning signal (FCW), sends this signal to the vehicle's display device (such as the instrument panel), and alerts the driver to potential dangers ahead through a sound alarm. At the same time, the system also sends a control signal to the engine controller, requiring the engine controller to reduce the engine speed to reduce the inertial impact during a collision. The preprocessing at this stage is intended to prepare for possible subsequent emergencies.

[0108] When TTC is in the second threshold range (0.6 seconds < TTC ≤ 1.6 seconds): In this range, the system detects that the risk of a forward collision has increased significantly and more urgent intervention measures are required. Therefore, the intelligent driving system triggers the automatic emergency braking function (AEB) and brakes the vehicle to decelerate, usually at a deceleration of 0.6g. At the same time, the system sends a control signal to the engine controller, requiring the engine to stop immediately and cut off the fuel supply. This operation is intended to reduce the vehicle's power output to avoid fuel leakage caused by the engine continuing to run when a collision occurs, thereby reducing the risk of fuel fires.

[0109] When TTC is in the third threshold range (TTC ≤ 0.6 seconds): At this stage, the vehicle is in extreme danger of an imminent collision, and the system needs to take the strongest braking measures. The intelligent driving system triggers full braking and performs emergency braking on the vehicle at a maximum deceleration of 1g. At the same time, the system continues to send control signals to the engine controller to ensure that the engine is completely stopped and the fuel supply is completely cut off to minimize the risk of fuel leakage and fire after the collision. At this time, the intelligent driving system reduces the impact of the collision as much as possible by fully braking and stopping the engine to protect the safety of the occupants and the vehicle.

[0110] When TTC is greater than 3.6 seconds: In this case, the system believes that there is no immediate risk of collision ahead, so it does not output any control signal, the vehicle maintains normal driving status, and the driver does not need to take any emergency action.

[0111] Therefore, in summary, the intelligent driving system compares the real-time calculated collision time TTC with the preset threshold range, and adopts corresponding control strategies according to different TTC values, including generating collision warnings, triggering automatic emergency braking, reducing engine speed, and stopping engine operation, thereby ensuring that the vehicle can effectively respond to collision risks and reduce potential dangers.

[0112] In some embodiments, the intelligent driving system can also dynamically adjust the collision time threshold (TTC) according to the driver's behavior pattern and environmental conditions to improve the system's responsiveness and adapt to the needs of individual drivers. The specific implementation process of this embodiment is as follows:

[0113] First, the intelligent driving system monitors the driver's operating habits and reaction time through sensors and control modules. This includes recording the driver's reaction speed to braking, steering, accelerator and other operations in different situations. For example, the system can monitor the driver's reaction time to step on the brakes in an emergency and continuously collect such data. In some embodiments, the system can also collect the driver's operating habits under different weather and road conditions in order to establish a complete driving behavior database.

[0114] Based on this data, the system optimizes the collision time threshold by analyzing the driver's behavior patterns. Specifically, the system will adjust the TTC trigger threshold based on the driver's reaction speed and operating habits. For example, if the system detects that a driver usually reacts slowly to emergencies, the system will trigger the TTC warning appropriately in advance and set the TTC threshold to a shorter time range so that a collision warning signal can be sent to the driver at an earlier time. This dynamic adjustment ensures that the system responds reasonably according to the individual characteristics of the driver, improving the safety and personalization of the system.

[0115] Furthermore, under different driving conditions, the system will dynamically adjust the TTC threshold according to external environmental factors. For example, in rainy, snowy or slippery roads, since the vehicle's braking distance and reaction time will increase accordingly, the system can automatically adjust the TTC threshold to shorten the warning trigger time and ensure that the vehicle can respond to potential collision risks in a timely manner on slippery roads. Similarly, when the vehicle is in a high-speed driving state, the system will further shorten the TTC trigger time according to the vehicle's speed and surrounding environment (such as traffic flow) to deal with potential dangers in high-speed situations.

[0116] In order to further enhance the intelligence and personalization of the system, the system can also record and learn the driver's long-term driving habits through machine learning algorithms. By analyzing a large amount of driving data, the system can continuously optimize the setting of the TTC threshold. For example, the system can dynamically adjust the trigger point of the collision warning based on the driver's long-term reaction time change trend, making the system more adaptable to the driver's changes. In addition, the system can also optimize the triggering time of emergency braking based on historical driving data to ensure that the most appropriate safety warning and response measures can be provided in each driving scenario.

[0117] Over time, the intelligent driving system gradually improves the accuracy of its collision risk assessment through continuous learning and adjustment. This self-learning mechanism not only improves the system's adaptability to the driver's individual behavior patterns, but also ensures the accuracy and stability of the system under various environmental conditions. Through this mechanism, the system can provide drivers with a safer and more personalized driving experience, and continuously optimize its prediction ability and response speed in long-term use.

[0118] Through the method of the above embodiment, the system realizes dynamic TTC threshold adjustment based on the driver's operating habits and environmental conditions, ensuring that the vehicle can accurately warn and take appropriate control measures under different driving conditions, effectively improving driving safety and user experience.

[0119] In addition to the technical solutions provided in the above embodiments, the technical solutions of the present application can also further enhance the safety and reliability of the system by improving aspects such as the coordinated control, coordinated perception, and fault prediction mechanism of the engine and braking system.

[0120] First, the coordinated control of the engine and the braking system

[0121] In order to improve the energy utilization efficiency of the system during emergency braking, the intelligent driving system can introduce a braking energy recovery mechanism. When the system triggers automatic emergency braking (AEB), in addition to decelerating through traditional mechanical brakes, the braking system can also recover the energy generated during braking and store it in the vehicle's energy storage device (such as a battery). This recovered energy can be used for subsequent vehicle starting or driving assistance, thereby reducing the vehicle's energy loss after emergency braking, improving energy efficiency and reducing the negative impact of braking on driving.

[0122] Furthermore, in order to improve the braking efficiency and safety of the vehicle, the system can combine motor reverse control to assist the vehicle in deceleration through reverse braking of the motor during emergency braking. This two-way braking control method combines the advantages of traditional chassis braking and motor braking, which can not only effectively shorten the braking distance, but also further control the speed of the vehicle through motor braking. Especially in high-load emergency braking scenarios, reverse braking of the motor can relieve the pressure of chassis braking, reduce the wear of the mechanical braking system, and enhance overall safety.

[0123] Second, collaborative perception and traffic infrastructure communication

[0124] In some embodiments, the intelligent driving system can expand its communication capabilities with surrounding road infrastructure. For example, the system can communicate with smart traffic lights or smart road signs in real time to obtain real-time information on the road ahead to further optimize collision prediction and TTC calculation. In high-risk areas such as intersections, the system communicates with traffic lights to learn about possible signal changes in advance and predict the risk of collision with side or oncoming vehicles, thereby dynamically adjusting the calculation and triggering threshold of TTC to ensure that the vehicle can respond effectively before potential risks arise.

[0125] For example, when the system receives a signal from a smart traffic light indicating that a red light is about to appear or a pedestrian is crossing the road, the system can predict the possible risk of collision in advance and adjust the vehicle's deceleration or braking strategy in real time based on this information. At the same time, communication with smart road signs can help the system obtain environmental factors such as road conditions, traffic congestion on the road ahead, and construction areas, further improving the system's ability to respond to complex driving scenarios.

[0126] Third, fault prediction mechanism

[0127] In order to improve the stability and reliability of the system, the technical solution of this application can also prevent system failures through a predictive maintenance mechanism. The intelligent driving system monitors the usage status of the engine and brake system and combines historical usage data to predict possible failures. For example, by monitoring the operating status of the engine (such as temperature, speed, oil pressure, etc.) and the performance of the brake system (such as brake pad wear, hydraulic system status, etc.), the system can determine the health status of these key components.

[0128] Combined with historical data analysis, the system can identify possible failure patterns and issue warnings or prompts before problems occur. Based on the prediction results, the vehicle can arrange maintenance or overhaul in advance to avoid system failure at critical moments. For example, when the system detects that the wear of the brake pads reaches a certain threshold, it can notify the driver in advance to replace them to prevent safety hazards caused by brake system failure during emergency braking. Through this predictive maintenance mechanism, the overall safety and reliability of the vehicle are further guaranteed.

[0129] Therefore, in summary, through the coordinated control of the engine and braking system, the coordinated perception and communication with the road infrastructure, and the introduction of a fault prediction mechanism, the system has been significantly improved in responding to emergencies, improving energy efficiency, and avoiding fault failures.

[0130] According to the technical solution provided in the embodiments of the present application, the technical solution of the present application has at least the following advantages:

[0131] When the system identifies obstacles such as vehicles, pedestrians, two-wheeled vehicles or special-shaped objects ahead and determines that there is a risk of collision, the system will trigger the forward collision warning (FCW) function signal. At this stage, the intelligent driving system sends a warning signal to make the engine (range extender) controller enter the first stage pre-preparation state and actively reduce the engine (range extender) speed. This process can effectively reduce the vehicle inertia and reduce the impact force of the vehicle before a collision, thereby reducing the risk of collision and improving driving safety.

[0132] When the risk of collision increases further and the automatic emergency braking (AEB) function is triggered, the system controls the vehicle to slow down through emergency braking while sending a signal to the engine (range extender) controller to immediately stop the engine (range extender) and cut off the fuel supply. Through this mechanism, the system can actively stop the engine when a collision is about to occur, avoiding the risk of fire caused by fuel leakage. This preventive measure greatly reduces the secondary dangers that may be caused by the accident, especially the risk of vehicle fire, and significantly improves the vehicle's safety assurance capabilities in collision accidents.

[0133] In addition, by optimizing the control of the engine (range extender), the technical solution of this application not only improves the safety of the vehicle, but also achieves good results in fuel economy. When the risk of collision occurs, the system can reduce the engine speed or stop the engine in advance, reducing unnecessary fuel consumption and thus improving the fuel efficiency of the vehicle. This intelligent control strategy can not only reduce the energy consumption of the vehicle in daily driving, but also achieve energy-saving and environmental protection effects, which meets the requirements of modern vehicles for environmental performance.

[0134] Therefore, the technical solution of the present application not only significantly improves the safety of the vehicle, especially avoiding the risk of fire caused by fuel leakage, through the precise coordination of the intelligent driving system and the engine controller, but also demonstrates excellent technical effects in terms of fuel efficiency and environmental protection performance.

[0135] The following is an embodiment of the device of the present application, which can be used to execute the embodiment of the method of the present application. For details not disclosed in the embodiment of the device of the present application, please refer to the embodiment of the method of the present application.

[0136] Figure 3 is a schematic diagram of the structure of an intelligent engine control device based on collision risk assessment provided by an embodiment of the present application. Figure 3 As shown, the intelligent engine control device based on collision risk assessment includes:

[0137] The acquisition module 301 is configured to use a variety of sensors to collect obstacle information on the road ahead of the vehicle in real time, wherein the obstacle information includes the speed and acceleration of the obstacle;

[0138] A processing module 302 is configured to pre-process the obstacle information and process the pre-processed data using an occupancy network model, wherein the occupancy network model is used to divide the sensing area into a voxel grid, perform feature extraction and occupancy prediction on each voxel to predict the location and type of the obstacle;

[0139] A calculation module 303 is configured to calculate the collision time using a predetermined dynamic collision time algorithm based on the obstacle prediction information output by the occupancy network model, wherein the dynamic collision time algorithm is used to dynamically adjust the collision time calculation parameters according to environmental conditions and vehicle status;

[0140] The adjustment module 304 is configured to adjust the operating state and operating parameters of the engine based on the comparison result between the collision time and a preset collision time threshold range.

[0141] In some embodiments, Figure 3Before using multiple sensors to collect obstacle information on the road in front of the vehicle in real time, the fusion module 305 uses a laser radar, a multi-function camera and a millimeter-wave radar to respectively identify objects in front of the vehicle, and obtains a first target signal output by the laser radar, a second target signal output by the multi-function camera and a third target signal output by the millimeter-wave radar; multi-band signal fusion is performed on the first target signal, the second target signal and the third target signal, and the obstacle is determined according to the multi-band signal fusion result.

[0142] In some embodiments, Figure 3 The processing module 302 performs denoising on the obstacle information so as to eliminate noise and interference signals in the sensor data by using a filtering algorithm; performs time alignment on the obstacle information collected by different sensors so as to synchronously process the obstacle information at the same time point based on the data collection timestamps of each sensor; and performs abnormal data detection and filtering on the obstacle information so as to identify and eliminate the abnormal data.

[0143] In some embodiments, Figure 3 The processing module 302 uses an occupancy network model to divide the perceived three-dimensional space into a voxel grid of a fixed size, wherein each voxel represents a specific area in the three-dimensional space, and the voxel is used as the basic unit of algorithm processing; based on the multi-view image data collected by multiple sensors, features are extracted for each voxel to obtain the spatial features corresponding to each voxel; based on the extracted spatial features, the occupancy state of each voxel is predicted using the occupancy network model to determine the probability of each voxel being occupied and the type of object occupying the space, and the three-dimensional space is divided into occupied areas and idle areas; data from different viewpoints collected by multiple sensors are fused, and the occupancy state prediction of the voxel grid by the occupancy network model is updated in real time based on the fused multi-view information, and the classification results are adjusted according to environmental changes.

[0144] In some embodiments, Figure 3 The calculation module 303 determines the relative speed, relative acceleration and relative distance between the current vehicle and the obstacle based on the obstacle prediction information output by the occupancy network model; based on the relative speed, relative acceleration and relative distance, a dynamic collision time algorithm is used to calculate the collision time, and the collision time is used to characterize the estimated time when the current vehicle collides with the obstacle in the current state; wherein, when calculating the collision time, the collision time calculation parameters are dynamically adjusted based on environmental conditions and vehicle states, the environmental conditions include real-time road conditions and weather conditions, and the vehicle state includes real-time driving parameters of the vehicle.

[0145] In some embodiments, Figure 3The calculation module 303 calculates the collision time of each obstacle simultaneously using a dynamic collision time algorithm when there are multiple obstacles in front of the vehicle, and prioritizes the obstacles according to the collision time of each obstacle, with the obstacle with the shortest collision time being the highest priority. The corresponding collision risk level is determined according to the collision time of each obstacle, so as to output a corresponding control signal according to the collision risk level.

[0146] In some embodiments, Figure 3 The adjustment module 304 generates a collision warning signal when the collision time is within a first threshold range, sends the collision warning signal to the vehicle display device and sounds an alarm, and sends a first control signal to the engine controller so that the engine controller reduces the engine speed according to the first control signal; when the collision time is within a second threshold range, the automatic emergency braking function is triggered to control the current vehicle to brake and decelerate, and sends a second control signal to the engine controller so that the engine controller stops the engine and cuts off the fuel supply according to the second control signal; when the collision time is within a third threshold range, the current vehicle is braked at maximum deceleration, and a third control signal is sent to the engine controller so that the engine controller stops the engine and cuts off the fuel supply according to the third control signal.

[0147] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0148] Figure 4 Schematic diagram of the structure of the electronic device 4 provided in the embodiment of the present application. Figure 4 As shown, the electronic device 4 of this embodiment includes: a processor 401, a memory 402, and a computer program 403 stored in the memory 402 and executable on the processor 401. When the processor 401 executes the computer program 403, the steps in the above-mentioned method embodiments are implemented. Alternatively, when the processor 401 executes the computer program 403, the functions of the modules / units in the above-mentioned device embodiments are implemented.

[0149] Exemplarily, the computer program 403 may be divided into one or more modules / units, which are stored in the memory 402 and executed by the processor 401 to complete the present application. The one or more modules / units may be a series of computer program instruction segments capable of completing specific functions, which are used to describe the execution process of the computer program 403 in the electronic device 4.

[0150] The electronic device 4 may be a desktop computer, a notebook, a PDA, a cloud server, or other electronic device. The electronic device 4 may include, but is not limited to, a processor 401 and a memory 402. Those skilled in the art will appreciate that Figure 4 It is only an example of the electronic device 4 and does not constitute a limitation of the electronic device 4. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the electronic device may also include input and output devices, network access devices, buses, etc.

[0151] The processor 401 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.

[0152] The memory 402 may be an internal storage unit of the electronic device 4, for example, a hard disk or memory of the electronic device 4. The memory 402 may also be an external storage device of the electronic device 4, for example, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 4. Further, the memory 402 may also include both an internal storage unit and an external storage device of the electronic device 4. The memory 402 is used to store computer programs and other programs and data required by the electronic device. The memory 402 may also be used to temporarily store data that has been output or is to be output.

[0153] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0154] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0155] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0156] In the embodiments provided in the present application, it should be understood that the disclosed devices / computer equipment and methods can be implemented in other ways. For example, the device / computer equipment embodiments described above are only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation. Multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0157] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0158] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0159] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. The computer program may include computer program code, which may be in source code form, object code form, executable file or some intermediate form. Computer-readable media may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0160] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. An intelligent engine control method based on collision risk assessment, characterized in that: include: Using multiple sensors to collect obstacle information on the road ahead of the vehicle in real time, wherein the obstacle information includes the speed and acceleration of the obstacle; Preprocessing the obstacle information and processing the preprocessed data using an occupancy network model, wherein the occupancy network model is used to divide the perception area into a voxel grid, perform feature extraction and occupancy prediction on each voxel to predict the location and type of the obstacle; Based on the obstacle prediction information output by the occupancy network model, a predetermined dynamic collision time algorithm is used to calculate the collision time, wherein the dynamic collision time algorithm is used to dynamically adjust the collision time calculation parameters according to environmental conditions and vehicle status; Based on the comparison result between the collision time and a preset threshold range of the collision time, the operating state and operating parameters of the engine are adjusted.

2. The method according to claim 1, characterized in that Before using a plurality of sensors to collect obstacle information on the road ahead of the vehicle in real time, the method further includes: Using a laser radar, a multi-function camera, and a millimeter-wave radar to respectively identify an object in front of the vehicle, and obtain a first target signal output by the laser radar, a second target signal output by the multi-function camera, and a third target signal output by the millimeter-wave radar; Multi-band signal fusion is performed on the first target signal, the second target signal, and the third target signal, and obstacles are determined according to the multi-band signal fusion result.

3. The method according to claim 1, characterized in that The preprocessing of the obstacle information includes: De-noising the obstacle information so as to eliminate noise and interference signals in the sensor data using a filtering algorithm; Performing time alignment on obstacle information collected by different sensors so as to synchronously process the obstacle information at the same time point based on the data collection timestamps of each sensor; The obstacle information is subjected to abnormal data detection and filtering so as to identify and eliminate the abnormal data.

4. The method according to claim 1, characterized in that: The method of processing the preprocessed data using the occupancy network model includes: Using an occupancy network model to divide the perceived three-dimensional space into a fixed-size voxel grid, where each voxel represents a specific area in the three-dimensional space, and the voxel is used as the basic unit of algorithm processing; Based on the multi-view image data collected by multiple sensors, feature extraction is performed on each voxel to obtain the spatial features corresponding to each voxel; Based on the extracted spatial features, an occupancy network model is used to predict the occupancy state of each voxel to determine the probability of each voxel being occupied and the type of object occupying the space, and the three-dimensional space is divided into an occupied area and an idle area; The data collected by multiple sensors from different perspectives are fused, and the occupancy network model predicts the occupancy state of the voxel grid in real time based on the fused multi-perspective information, and the classification results are adjusted according to environmental changes.

5. The method according to claim 1, characterized in that The calculating the collision time using a predetermined dynamic collision time algorithm based on the obstacle prediction information output by the occupancy network model includes: Determine the relative speed, relative acceleration and relative distance between the current vehicle and the obstacle based on the obstacle prediction information output by the occupancy network model; Based on the relative speed, relative acceleration and relative distance, a collision time is calculated using a dynamic collision time algorithm, wherein the collision time is used to represent an estimated time when the current vehicle collides with the obstacle in the current state; When calculating the collision time, the collision time calculation parameters are dynamically adjusted based on environmental conditions and vehicle status, the environmental conditions include real-time road conditions and weather conditions, and the vehicle status includes real-time driving parameters of the vehicle.

6. The method according to claim 5, characterized in that When calculating the collision time, the method further includes: When there are multiple obstacles in front of the vehicle, the dynamic collision time algorithm is used to simultaneously calculate the collision time of each obstacle, and the obstacles are prioritized according to the collision time of each obstacle, with the obstacle with the shortest collision time being the highest priority. The corresponding collision risk level is determined according to the collision time of each obstacle, so that a corresponding control signal is output according to the collision risk level.

7. The method according to claim 1, characterized in that The adjusting the operating state and operating parameters of the engine based on the comparison result between the collision time and the preset collision time threshold range includes: When the collision time is within a first threshold range, a collision warning signal is generated, the collision warning signal is sent to a vehicle display device and an audible alarm is sounded, and a first control signal is sent to an engine controller so that the engine controller reduces the engine speed according to the first control signal; When the collision time is within the second threshold range, the automatic emergency braking function is triggered, the current vehicle is controlled to brake and decelerate, and a second control signal is sent to the engine controller, so that the engine controller stops the engine and cuts off the fuel supply according to the second control signal; When the collision time is within the third threshold range, maximum deceleration braking is performed on the current vehicle, and a third control signal is sent to the engine controller, so that the engine controller stops the engine running and cuts off the fuel supply according to the third control signal.

8. An intelligent engine control device based on collision risk assessment, characterized in that: include: A collection module is configured to collect obstacle information on the road ahead of the vehicle in real time using a variety of sensors, wherein the obstacle information includes the speed and acceleration of the obstacle; A processing module is configured to pre-process the obstacle information and process the pre-processed data using an occupancy network model, wherein the occupancy network model is used to divide the perception area into a voxel grid, perform feature extraction and occupancy prediction on each voxel to predict the location and type of the obstacle; a calculation module configured to calculate the collision time using a predetermined dynamic collision time algorithm based on the obstacle prediction information output by the occupancy network model, wherein the dynamic collision time algorithm is used to dynamically adjust collision time calculation parameters according to environmental conditions and vehicle status; The adjustment module is configured to adjust the operating state and operating parameters of the engine based on a comparison result between the collision time and a preset threshold range of the collision time.

9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images