An automatic start-stop system for intelligent connected vehicles

Abstract

The embodiment of the application relates to an automatic start-stop system of an intelligent networked vehicle, and the system comprises a self-vehicle perception module, a vehicle networking communication module, a data storage module, a strategy decision module and a start-stop execution module; the self-vehicle collection module is used for collecting self-vehicle state data; the self-vehicle perception module is used for continuously perceiving other traffic participants around the current vehicle; the vehicle networking communication module is used for receiving front road congestion data, front road signal lamp data and vehicle guiding data; the data storage module is used for storing self-vehicle state, target perception, congestion, signal and guiding data sets; the strategy decision module is used for judging whether an automatic start-stop condition is met according to the data sets on the storage module and sending a start-stop control instruction to the start-stop execution module when the judgment result is that the condition is met; and the start-stop execution module is used for controlling the current vehicle to perform corresponding start-stop operations according to the start-stop control instruction. The application can improve the energy-saving and emission-reducing performance of the vehicle and improve the start-stop configuration flexibility.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to an automatic start-stop system for intelligent connected vehicles. Background Technology

[0002] Automatic start-stop system is an energy-saving technology that automatically shuts off the engine when the vehicle is briefly stopped (such as at a red light or in traffic jam) and quickly restarts the engine when it needs to start moving again. Conventional vehicle start-stop systems primarily rely on vehicle status parameters (such as vehicle speed, brake pedal engagement, accelerator pedal engagement, clutch pedal engagement, and steering wheel angle) to determine trigger conditions, and this trigger condition determination process is fixed. Because most traditional cars lack intelligent sensing and network communication capabilities, they cannot combine environmental information (such as traffic light status and road congestion conditions) for comprehensive judgment.

[0003] With the maturity and development of vehicle-to-everything (V2X) and autonomous driving technologies, an increasing number of intelligent connected vehicles (ICVs) have emerged. These new vehicles possess environmental perception, collaborative control, and autonomous driving capabilities, and can obtain real-time environmental information (such as traffic light status, distance to vehicles ahead, and road congestion) through the V2X network. If an automatic start-stop scheme that integrates vehicle status parameters and environmental information can be customized for intelligent connected vehicles based on these characteristics, it will undoubtedly further improve the vehicle's energy-saving and emission-reduction performance and enhance the driving experience. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by providing an automatic start-stop system for intelligent connected vehicles. This system includes: a vehicle data acquisition module, a vehicle perception module, a vehicle-to-everything (V2X) communication module, a data storage module, a strategy decision-making module, and a start-stop execution module. Specifically, the vehicle data acquisition module continuously collects vehicle status data via the CAN bus and stores the collected data in the data storage module; the vehicle perception module connects to multiple types of onboard sensors and continuously perceives other traffic participants around the vehicle based on sensor data, storing the perceived data in the data storage module; the V2X communication module receives traffic environment information from a V2X platform or roadside V2X devices and stores it in the data storage module; the strategy decision-making module determines whether the automatic start-stop conditions are met based on the dataset in the data storage module according to two configurable judgment rules (automatic stop condition judgment rules and automatic start condition judgment rules), and sends a corresponding start-stop control command to the start-stop execution module when the judgment result is satisfied; the start-stop execution module controls the vehicle to perform corresponding start-stop operations based on the start-stop control command. This invention provides an automatic start-stop solution for intelligent connected vehicles that integrates vehicle status parameters and traffic environment information. This invention can not only further improve the energy-saving and emission-reduction performance of vehicles and enhance the driving experience, but also improve the configuration flexibility of the trigger condition judgment process.

[0005] To achieve the above objectives, embodiments of the present invention provide an automatic start-stop system for intelligent connected vehicles, the system comprising: a vehicle data acquisition module, a vehicle perception module, a vehicle-to-everything (V2X) communication module, a data storage module, a strategy decision-making module, and a start-stop execution module;

[0006] The system is installed on intelligent connected vehicles; the data storage module is connected to the vehicle acquisition module, the vehicle perception module, the vehicle network communication module and the strategy decision module respectively; the strategy decision module is connected to the start-stop execution module; the vehicle network communication module is connected to a remote vehicle network platform and a vehicle network roadside device on the road where the vehicle is located through a V2X communication network.

[0007] The vehicle data acquisition module is used to continuously acquire the current vehicle status data via the CAN bus and store the acquired data in the data storage module.

[0008] The vehicle sensing module is used to connect to multiple types of vehicle sensors, including cameras, lidar, millimeter-wave radar, and ultrasonic radar. These multiple types of vehicle sensors achieve hardware-level time synchronization based on the IEEE 1588PTP protocol.

[0009] The vehicle perception module is also used to continuously perceive other traffic participants around the current vehicle based on the sensor data of the multiple types of vehicle sensors and store the perception data into the data storage module.

[0010] The vehicle-to-everything (V2X) communication module is used to receive road congestion data, traffic light data, or vehicle guidance data sent by the V2X platform or the V2X roadside equipment and store them in the data storage module.

[0011] The data storage module is used to store vehicle status datasets, target perception datasets, congestion datasets, signal datasets, and guidance datasets;

[0012] The strategy decision module is used to determine whether the automatic start-stop conditions are met based on the dataset of the data storage module, and send the corresponding start-stop control command to the start-stop execution module when the determination result is met.

[0013] The start-stop execution module is used to control the current vehicle to perform corresponding start-stop operations according to the start-stop control command.

[0014] Preferably, the vehicle status dataset includes multiple vehicle status data; the vehicle status data includes timestamps, vehicle coordinates, vehicle speed, gear position, battery SOC, brake pedal engagement, clutch pedal engagement, accelerator pedal engagement, and steering wheel angle; the vehicle coordinates are in the world coordinate system.

[0015] The target perception dataset includes multiple target tracking sequences; the target tracking sequence includes multiple target perception data; the target perception data includes timestamps, target identifiers, target types, target coordinates, target shapes, relative distances, relative speeds, and target orientations; the target types include pedestrians, motor vehicles, and non-motor vehicles; the target shapes include height, width, and depth; all target identifiers and all target types are the same in each target tracking sequence;

[0016] The congestion dataset includes multiple road congestion data points ahead; the road congestion data ahead includes timestamps and road congestion indices.

[0017] The signal dataset includes multiple traffic light data points for the road ahead; the traffic light data for the road ahead includes timestamps, traffic light identifiers, traffic light coordinates, signal status, and remaining signal duration; the signal status includes red, green, and yellow lights.

[0018] The guidance dataset includes multiple vehicle guidance data sets; the vehicle guidance data is used to guide the current straight-line movement of the vehicle; the vehicle guidance data includes a timestamp and a guidance type; the guidance type includes acceleration, deceleration, constant speed, and stopping;

[0019] The start-stop control command includes command types; the command types include engine start and engine stop.

[0020] Preferably, the vehicle data acquisition module is specifically used when continuously acquiring the current vehicle status data via the CAN bus and storing the acquired data in the data storage module:

[0021] At a preset first acquisition frequency, the current time is periodically used as the current timestamp, and the latest vehicle coordinates, vehicle speed, gear position, battery SOC, brake pedal opening degree, clutch pedal opening degree, accelerator pedal opening degree, and steering wheel angle are collected via the CAN bus. The current timestamp and the collected vehicle coordinates, vehicle speed, gear position, battery SOC, brake pedal opening degree, accelerator pedal opening degree, and steering wheel angle are combined to form a new vehicle status data set, which is then stored in the vehicle status dataset.

[0022] Preferably, the vehicle perception module is specifically used when continuously perceiving other traffic participants around the current vehicle based on sensor data from the multiple types of onboard sensors and storing the perception data in the data storage module:

[0023] Step 41: In the perception module, a first buffer sequence is set to buffer the visual images periodically output by the camera, a second buffer sequence is set to buffer the dense point cloud periodically output by the lidar, a third buffer sequence is set to buffer the sparse point cloud periodically output by the millimeter-wave radar, and a fourth buffer sequence is set to buffer the ranging point set periodically output by the ultrasonic radar; and a corresponding first multimodal data group is formed by the time-aligned four types of sensor data in each of the four buffer sequences.

[0024] The first cache sequence is formed by sequentially sorting multiple first images; the second cache sequence is formed by sequentially sorting multiple first point clouds; the first point cloud includes multiple first scan points; the first scan point includes first point coordinates and a first reflection intensity; the third cache sequence is formed by sequentially sorting multiple second point clouds; the first point coordinates are coordinates in the lidar coordinate system; the second point cloud includes multiple second scan points; the second scan point includes second point coordinates and a first velocity; the second point coordinates are coordinates in the millimeter-wave radar coordinate system; the first velocity is the relative velocity of the current scan point relative to the current vehicle; the fourth cache sequence is formed by sequentially sorting multiple first ranging point sets; the first ranging point set includes multiple first ranging points; the first ranging point includes third point coordinates and a first distance; the third point coordinates are coordinates in the ultrasonic radar coordinate system; the first distance is the radial distance from the current third point coordinates to the ultrasonic radar installation position; the first multimodal data group consists of a set of time-aligned first images, first point clouds, second point clouds, and first ranging point sets;

[0025] Step 42: When a new first multimodal data group is obtained, the current first multimodal data group is taken as the current data group; target detection, target feature fusion and target tracking processing are performed based on the current data group, and the target perception dataset is refreshed based on the processing results.

[0026] Furthermore, the vehicle perception module is specifically used when performing target detection, target feature fusion, and target tracking processing based on the current data set and refreshing the target perception dataset based on the processing results:

[0027] Step 51: Take the first image, the first point cloud, the second point cloud, and the first ranging point set of the current data group as the corresponding current image, current laser point cloud, current millimeter-wave point cloud, and current ranging point set; and take the time point corresponding to the current data group as the current timestamp;

[0028] Step 52, and based on the preset visual target recognition model, perform target detection and classification recognition processing on the current image to obtain the corresponding first target detection box set; and convert the coordinates of the first center point of each first target detection box in the first target detection box set from image coordinate system coordinates to vehicle coordinate system coordinates, and convert the height and width of the first detection box of each first target detection box from pixel height and width to real-world height and width;

[0029] Step 53: Based on the preset point cloud target detection model, target detection processing is performed on the current laser point cloud to obtain the corresponding second target detection box set; and the coordinates of the second center point of each second target detection box in the second target detection box set are converted from the coordinates of the laser radar coordinate system to the coordinates of the vehicle coordinate system, and the orientation of each second detection box is converted from the orientation of the laser radar coordinate system to the orientation of the vehicle coordinate system.

[0030] Step 54, and convert the coordinates of each of the second points in the current millimeter-wave point cloud from millimeter-wave radar coordinates to vehicle coordinates;

[0031] Step 55: Based on the coordinates of the third point of each of the first ranging points in the current ranging point set and the first distance, calculate the corresponding second distance from the current ranging point to the edge of the current vehicle body; and convert the coordinates of each of the third points in the current ranging point set from the ultrasonic radar coordinate system to the vehicle coordinate system.

[0032] Step 56, and identify the second target detection boxes that match each of the first target detection boxes, and perform detection box attribute fusion based on the identification results to obtain the corresponding third target detection box;

[0033] Step 57, and based on the current millimeter-wave point cloud and the current ranging point set, perform relative motion attribute fusion on each of the third target detection boxes to generate the corresponding target perception data;

[0034] Step 58, and locate the corresponding target tracking sequence for each newly added target perception data in the target perception dataset and add each newly added perception data to its corresponding sequence.

[0035] More preferably, the visual target recognition model includes the YOLO series model, the Faster R-CNN model, and the SSD model; the first target detection box set includes multiple first target detection boxes; the first target detection box includes the first center point coordinates, the first detection box height, the first detection box width, and the first target type; the first target type includes pedestrians, motor vehicles, and non-motor vehicles;

[0036] The point cloud target recognition model includes the PointPillars model, the VoteNet model, and the SECOND model; the second target detection box set includes multiple second target detection boxes; the first target detection box includes the second center point coordinates, the second detection box height, the second detection box width, the second detection box depth, and the second detection box orientation;

[0037] The third target detection box includes the coordinates of the third center point, the height of the third detection box, the width of the third detection box, the depth of the third detection box, the orientation of the third detection box, and the type of the third target;

[0038] The vehicle perception module is specifically used when, during the process of recognizing the second target detection boxes matched by each of the first target detection boxes and fusing the detection box attributes based on the recognition results to obtain the corresponding third target detection box:

[0039] Each of the first target detection boxes is taken as the current detection box; and the Euclidean distance between the first center point coordinates of the current detection box and each of the second center point coordinates is calculated; and the second target detection box corresponding to the shortest Euclidean distance is taken as the current matching box; and the second center point coordinates, second detection box height, second detection box width, second detection box depth, and second detection box orientation of the current candidate box are taken as a set of corresponding third center point coordinates, third detection box height, third detection box width, third detection box depth, and third detection box orientation; and the first target type of the current detection box is taken as the corresponding third target type; and the current third center point coordinates, third detection box height, third detection box width, third detection box depth, third detection box orientation, and third target type are combined to form a corresponding third target detection box;

[0040] The vehicle perception module is specifically used when generating corresponding target perception data by fusing the relative motion attributes of each of the third target detection boxes based on the current millimeter-wave point cloud and the current ranging point set:

[0041] Each of the aforementioned third target detection boxes is taken as the current target box; the three-dimensional bounding region of the current target box in the vehicle coordinate system is taken as the current region; and the corresponding current first point set is composed of all the second scanning points in the current millimeter-wave point cloud whose second point coordinates are located in the current region, and the corresponding current second point set is composed of all the first ranging points in the current ranging point set whose third point coordinates are located in the current region; the average speed of all the first velocities in the current first point set is taken as the corresponding relative speed; and it is determined whether the current second point set is empty. If it is, the corresponding relative distance is estimated based on the third center point coordinates of the current target box; otherwise, the relative distance is calculated based on the third center point coordinates of the current second point set. The minimum of the two distances is taken as the corresponding relative distance; and the height, width, and depth of the third detection box of the current target box form a corresponding target shape; and a unique identifier is assigned to the current target box as the corresponding target identifier; and the third center point coordinates, third detection box orientation, and third target type of the current target box form a set of corresponding target coordinates, target orientation, and target type; and the current timestamp and a set of target identifiers, target types, target coordinates, target shape, relative distance, relative speed, and target orientation corresponding to the current target box form a new target perception data;

[0042] The vehicle perception module is specifically used when, in the target perception dataset, it locates the corresponding target tracking sequence for each newly added target perception data and adds each newly added perception data to its corresponding sequence:

[0043] Each newly added target perception data is taken as the corresponding current newly added data; and it is identified whether there is a target type in the target perception dataset that matches the target type of the target tracking sequence in the current newly added data;

[0044] If no target type in the target sensing dataset matches the target type of the currently added data, then an empty target tracking sequence is created in the target sensing dataset as the corresponding current sequence; and the currently added data is added to the current sequence.

[0045] If at least one target tracking sequence in the target perception dataset matches the target type of the currently added data, then each matching target tracking sequence is recorded as a corresponding search sequence; and the corresponding target of each search sequence is recorded as the corresponding search target; and using Kalman filtering, based on the historical data of each search sequence, the target coordinates and target shape of the corresponding search target at the current timestamp are predicted to obtain a set of corresponding predicted coordinates and predicted shapes to form a corresponding predicted target box; and the intersection-union ratio (IU) of the spatial regions of each predicted target box in the vehicle coordinate system with the corresponding spatial region of the currently added data is estimated; and whether the largest IU exceeds a preset IU threshold is identified; if so, the search sequence corresponding to the largest IU is taken as the corresponding current sequence, and the target identifier of the currently added data is reset to the target identifier corresponding to the current sequence, and the currently added data is added to the current sequence after the reset; if not, an empty target tracking sequence is created in the target perception dataset as the corresponding current sequence, and the currently added data is added to the current sequence.

[0046] Preferably, the vehicle perception module is further configured to periodically traverse all target tracking sequences in the target perception dataset; and during this traversal, the currently traversed target tracking sequence is taken as the current sequence; and to identify whether the time interval between the last timestamp of the current sequence and the current time exceeds a preset first duration; if so, the current sequence is deleted.

[0047] Preferably, the strategy decision module is specifically used when the automatic start / stop conditions are determined based on the dataset of the data storage module, and when the determination result is satisfied, the module sends a corresponding start / stop control command to the start / stop execution module:

[0048] Step 81: Periodically use the current time as the current timestamp according to the preset status query frequency;

[0049] Step 82, and take the latest vehicle status data in the vehicle status dataset as the current vehicle status; and take the vehicle coordinates, vehicle speed, gear, battery SOC, brake pedal opening degree, clutch pedal opening degree, accelerator pedal opening degree, and steering wheel angle of the current vehicle status as the corresponding current vehicle coordinates, current vehicle speed, current gear, current battery SOC, current brake pedal opening degree, current clutch pedal opening degree, current accelerator pedal opening degree, and current steering wheel angle;

[0050] Step 83: If the time interval between the latest timestamp of the congestion dataset and the current timestamp does not exceed a preset second duration, the latest road congestion index is used as the corresponding current congestion index.

[0051] Step 84: If the time interval between the latest timestamp of the bootstrap dataset and the current timestamp does not exceed the second duration, the latest bootstrap type is taken as the corresponding current bootstrap type.

[0052] Step 85, and based on the current vehicle coordinates, the current vehicle speed and the target perception dataset, identify whether there is a stationary target on the straight road in front of the current vehicle to obtain the corresponding stationary target identification result;

[0053] The results of the stationary target identification include whether a stationary target exists or not.

[0054] Step 86, and identify the current signal status and the remaining duration of the current signal based on the current vehicle coordinates and the signal dataset;

[0055] Step 87, and identify the current start-stop phase status of the vehicle;

[0056] The start-stop phase states include the start-up phase, the stop phase, and the non-start-stop phase;

[0057] Step 88: If the current start-stop phase is a non-start-stop phase, then according to the preset automatic shutdown condition judgment rules, based on the current vehicle speed, current gear, current battery SOC, current brake pedal opening degree, current steering wheel angle, stationary target recognition result, current congestion index, current signal status, current signal remaining duration, and current guidance type, the automatic shutdown condition satisfaction status judgment is performed to obtain the corresponding first judgment result; and when the first judgment result is satisfied, the start-stop control command with the command type set to engine shutdown is sent to the start-stop execution module; and the start-stop phase state is switched to the shutdown phase;

[0058] Step 89: If the current start-stop phase is in the shutdown phase, then according to the preset self-start condition judgment rules, the self-start condition is judged based on the current gear, the current clutch pedal opening degree, the current accelerator pedal opening degree, the current steering wheel angle, the current signal status, the remaining duration of the current signal, and the current guidance type to obtain the corresponding second judgment result; and if the second judgment result is satisfied, the start-stop control command with the command type set to engine start is sent to the start-stop execution module; and the start-stop phase state is switched to the start phase.

[0059] Step 90: If the current start-stop phase is in the start phase, then identify whether the current vehicle speed exceeds a preset first vehicle speed threshold; if so, then switch the start-stop phase to a non-start-stop phase.

[0060] Furthermore, the strategy decision module is specifically used to obtain the corresponding stationary target identification result when identifying whether there is a stationary target on the straight road ahead of the current vehicle based on the current vehicle coordinates, the current vehicle speed, and the target perception dataset:

[0061] Step 91: Take the target tracking sequence in the target perception dataset whose time interval between the latest timestamp and the current timestamp does not exceed the second duration as the corresponding current trajectory; and take the target corresponding to the current trajectory as the current target;

[0062] Step 92, and determine whether the Euclidean distance between the last two target coordinates of the current trajectory is less than a preset first distance threshold;

[0063] Step 93: If the Euclidean distance between the last two target coordinates of the current trajectory is less than the first distance threshold, then the current target is regarded as a corresponding stationary target.

[0064] Step 94: If the Euclidean distance between the last two target coordinates of the current trajectory is greater than or equal to the first distance threshold, then the current timestamp is used as the future start time, and the time obtained by adding the current timestamp to the preset third duration is used as the future end time. The future start time and the future end time form a corresponding future time period. Based on the current trajectory, the motion trajectory of the current target in the future time period is predicted to obtain a corresponding first predicted trajectory. Based on the current vehicle coordinates and the current vehicle speed, the motion trajectory of the current vehicle in the future time period is predicted to obtain a corresponding second predicted trajectory. The existence of a spatiotemporal intersection point between the first and second predicted trajectories in the future time period is identified. If the spatiotemporal intersection point exists, the trajectory segment after the spatiotemporal intersection point on the first predicted trajectory is further identified as a stationary trajectory segment. If it is confirmed that the trajectory segment after the spatiotemporal intersection point is a stationary trajectory segment, then the current target is regarded as a corresponding stationary target.

[0065] Step 95, and identify the total number of the obtained stationary targets; if the total number of stationary targets is 0, then set the corresponding stationary target identification result to "no stationary targets exist"; if the total number of stationary targets is greater than 0, then set the corresponding stationary target identification result to "stationary targets exist".

[0066] Furthermore, the strategy decision module is specifically used when identifying the current signal state and the remaining duration of the current signal based on the current vehicle coordinates and the signal dataset:

[0067] By querying a preset high-precision road map, it is confirmed whether the coordinates of the latest traffic light data ahead in the signal dataset are located ahead of the current vehicle's driving road. If confirmed, the latest traffic light data ahead in the signal dataset is taken as the current traffic light data, and the signal status of the current traffic light data is taken as the corresponding current signal status. Based on the timestamp of the current traffic light data, the remaining signal duration, and the current timestamp, the corresponding remaining signal duration is calculated as: remaining signal duration of the current traffic light data - (current timestamp - timestamp of the current traffic light data).

[0068] Furthermore, the automatic shutdown condition judgment rule is implemented based on a first condition judgment matrix; the matrix columns of the first condition judgment matrix correspond one-to-one with the current vehicle speed, the current gear, the current battery SOC, the current brake pedal opening degree, the current steering wheel angle, the stationary target recognition result, the current congestion index, the current signal status, the current signal remaining duration, and the current guidance type; each matrix row of the first condition judgment matrix corresponds to a first combination condition configuration; each matrix unit of the first condition judgment matrix includes a condition switch and a condition threshold; the condition switch includes two switch states: on and off; the condition threshold is a threshold range.

[0069] The judgment rule of the first condition judgment matrix is ​​as follows: if the judgment result of any of the first combination conditions is satisfied, it is considered that the self-shutdown condition state is satisfied, and the corresponding first judgment result is satisfied; if the judgment results of all the first combination conditions are not satisfied, it is considered that the self-shutdown condition state is not satisfied, and the corresponding first judgment result is not satisfied.

[0070] The judgment rule for each of the first combination conditions in the first condition judgment matrix is ​​as follows: if the condition judgment results of all matrix cells in the matrix row corresponding to the current first combination condition configuration that have the condition switch in the on state are satisfied, then the condition judgment result of the current first combination condition configuration is satisfied; if the condition judgment result of at least one matrix cell in the matrix row corresponding to the current first combination condition configuration that has the condition switch in the on state is not satisfied, then the condition judgment result of the current first combination condition configuration is not satisfied.

[0071] The judgment rule for each matrix unit in the first condition judgment matrix where the condition switch is on is as follows: the current vehicle speed, current gear, current battery SOC, current brake pedal opening degree, current steering wheel angle, stationary target recognition result, current congestion index, current signal status, current signal remaining duration, or current guidance type corresponding to the current matrix unit is taken as the current judgment object; if the current judgment object meets the condition threshold of the current matrix unit, the condition judgment result of the current matrix unit is satisfied; if the current judgment object does not meet the condition threshold of the current matrix unit, the condition judgment result of the current matrix unit is not satisfied.

[0072] The self-starting condition judgment rule is implemented based on a second condition judgment matrix; the matrix columns of the second condition judgment matrix correspond one-to-one with the current gear, the current clutch pedal opening degree, the current accelerator pedal opening degree, the current steering wheel angle, the current signal status, the current signal remaining duration, and the current guidance type; each matrix row of the second condition judgment matrix corresponds to a second combination condition configuration; each matrix unit of the second condition judgment matrix includes the condition switch and the condition threshold;

[0073] The judgment rule of the second condition judgment matrix is ​​as follows: if the judgment result of any of the second combination conditions is satisfied, it is considered that the self-starting condition state is satisfied, and the corresponding second judgment result is satisfied; if the judgment results of all the second combination conditions are not satisfied, it is considered that the self-starting condition state is not satisfied, and the corresponding second judgment result is not satisfied.

[0074] The judgment rule for each of the second combination conditions in the second condition judgment matrix is ​​as follows: if the condition judgment results of all matrix cells in the matrix row corresponding to the current second combination condition configuration that have the condition switch in the on state are satisfied, then the condition judgment result of the current second combination condition configuration is satisfied; if the condition judgment result of at least one matrix cell in the matrix row corresponding to the current second combination condition configuration that has the condition switch in the on state is not satisfied, then the condition judgment result of the current second combination condition configuration is not satisfied.

[0075] The judgment rule for each matrix unit in the second condition judgment matrix where the condition switch is in the on state is as follows: the current gear, the current clutch pedal opening degree, the current accelerator pedal opening degree, the current steering wheel angle, the current signal status, the remaining duration of the current signal, or the current guidance type corresponding to the current matrix unit are taken as the current judgment objects; if the current judgment object satisfies the condition threshold of the current matrix unit, the condition judgment result of the current matrix unit is satisfied; if the current judgment object does not satisfy the condition threshold of the current matrix unit, the condition judgment result of the current matrix unit is not satisfied.

[0076] Preferably, the start-stop execution module is specifically used when controlling the current vehicle to perform the corresponding start-stop operation according to the start-stop control command:

[0077] The type of the start-stop control command is identified; if the command type is engine start, the engine of the current vehicle is restarted; if the command type is engine stop, the current vehicle is controlled to decelerate and stop, and the engine of the current vehicle is stopped when the vehicle speed drops to 0.

[0078] This invention provides an automatic start-stop system for intelligent connected vehicles. As described above, the system includes: a vehicle data acquisition module, a vehicle perception module, a vehicle-to-everything (V2X) communication module, a data storage module, a strategy decision-making module, and a start-stop execution module. Specifically, the vehicle data acquisition module continuously collects vehicle status data via the CAN bus and stores the collected data in the data storage module; the vehicle perception module connects to multiple types of onboard sensors and continuously perceives other traffic participants around the vehicle based on sensor data, storing the perceived data in the data storage module; the V2X communication module receives traffic environment information from a V2X platform or roadside V2X devices and stores it in the data storage module; the strategy decision-making module determines whether the automatic start-stop conditions are met based on the dataset in the data storage module according to two configurable judgment rules (automatic stop condition judgment rules and automatic start condition judgment rules), and sends a corresponding start-stop control command to the start-stop execution module when the judgment result is met; the start-stop execution module controls the vehicle to perform corresponding engine start-stop operations according to the start-stop control command. The embodiments of the present invention not only improve the energy-saving and emission-reduction performance of vehicles and enhance the driving experience, but also increase the flexibility of the configuration of the trigger condition judgment process. Attached Figure Description

[0079] Figure 1 A modular structure diagram of an automatic start-stop system for an intelligent connected vehicle provided in an embodiment of the present invention;

[0080] Figure 2This is a schematic diagram of five types of datasets provided in an embodiment of the present invention;

[0081] Figure 3 A schematic diagram of the first condition judgment matrix provided in an embodiment of the present invention;

[0082] Figure 4 This is a schematic diagram of the second condition judgment matrix provided in an embodiment of the present invention. Detailed Implementation

[0083] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0084] An embodiment of the present invention provides an automatic start-stop system 1 for intelligent connected vehicles, such as... Figure 1 The module structure diagram of an automatic start-stop system for an intelligent connected vehicle provided in an embodiment of the present invention is shown, which mainly includes: a vehicle data acquisition module 11, a vehicle perception module 12, a vehicle network communication module 13, a data storage module 14, a strategy decision module 15, and a start-stop execution module 16.

[0085] The automatic start-stop system 1 of this invention is installed on an intelligent connected vehicle. The internal component connections of the automatic start-stop system 1 are as follows: the data storage module 14 is connected to the vehicle data acquisition module 11, the vehicle perception module 12, the vehicle-to-everything (V2X) communication module 13, and the strategy decision module 15; the strategy decision module 15 is connected to the start-stop execution module 16. The V2X communication module 13 is also connected to a remote V2X platform 2 and a V2X roadside device 3 on the road where the vehicle is located via a V2X communication network.

[0086] (I) Vehicle Data Acquisition Module 11:

[0087] The self-vehicle data acquisition module 11 of this embodiment of the invention is used to continuously acquire the current self-vehicle status data of the vehicle via the CAN bus and store the acquired data into the data storage module 14.

[0088] In a specific implementation of this invention, the vehicle data acquisition module 11 is specifically used to continuously acquire vehicle status data of the current vehicle via the CAN bus and store the acquired data in the data storage module 14: periodically using the current time as the current timestamp at a preset first acquisition frequency, and acquiring the latest vehicle coordinates, speed, gear, battery SOC, brake pedal opening degree, clutch pedal opening degree, accelerator pedal opening degree, and steering wheel angle via the CAN bus. A new vehicle status data set is then constructed from the current timestamp and the acquired vehicle coordinates, speed, gear, battery SOC, brake pedal opening degree, accelerator pedal opening degree, and steering wheel angle and stored in the vehicle status dataset. Here, the first acquisition frequency in this embodiment is a preset time frequency parameter.

[0089] (II) Vehicle Sensing Module 12:

[0090] The vehicle sensing module 12 in this embodiment of the invention is used to connect to multiple types of vehicle-mounted sensors. These multiple types of vehicle-mounted sensors include cameras, LiDAR, millimeter-wave radar, and ultrasonic radar. It should be noted that the multiple types of vehicle-mounted sensors in this embodiment of the invention achieve hardware-level time synchronization based on the IEEE 1588PTP protocol.

[0091] The vehicle perception module 12 is also used to continuously perceive other traffic participants around the current vehicle based on sensor data from multiple types of vehicle sensors and store the perception data in the data storage module 14.

[0092] The vehicle perception module 12 is also used to periodically traverse all target tracking sequences in the target perception dataset; and during this traversal, the target tracking sequence being traversed is taken as the current sequence; and whether the time interval between the last timestamp of the current sequence and the current time exceeds a preset first duration is identified; if so, the current sequence is deleted.

[0093] Here, the first duration is a pre-set time length parameter.

[0094] In another specific implementation of this invention, the vehicle perception module 12 is specifically used to continuously perceive other traffic participants around the current vehicle based on sensor data from multiple types of vehicle-mounted sensors and store the perception data in the data storage module 14:

[0095] Step A1: In the perception module, a first buffer sequence is set to buffer the visual images periodically output by the camera, a second buffer sequence is set to buffer the dense point cloud periodically output by the lidar, a third buffer sequence is set to buffer the sparse point cloud periodically output by the millimeter-wave radar, and a fourth buffer sequence is set to buffer the ranging point set periodically output by the ultrasonic radar; and a corresponding first multimodal data group is formed by the time-aligned four types of sensor data in each of the four buffer sequences.

[0096] The first cache sequence is composed of multiple first images ordered sequentially;

[0097] The second buffer sequence is formed by sequentially sorting multiple first point clouds; the first point cloud includes multiple first scan points; the first scan point includes the first point coordinates and the first reflection intensity; the third buffer sequence is formed by sequentially sorting multiple second point clouds; the first point coordinates are coordinates in the lidar coordinate system;

[0098] The second point cloud includes multiple second scanning points; each second scanning point includes second point coordinates and a first velocity; the second point coordinates are in the millimeter-wave radar coordinate system; the first velocity is the relative velocity of the current scanning point with respect to the current vehicle.

[0099] The fourth buffer sequence is formed by sequentially sorting multiple first ranging point sets; the first ranging point set includes multiple first ranging points; the first ranging point includes the coordinates of a third point and a first distance; the coordinates of the third point are the coordinates of the ultrasonic radar coordinate system; the first distance is the radial distance from the current third point coordinates to the installation position of the ultrasonic radar;

[0100] The first multimodal data set consists of a time-aligned first image, a first point cloud, a second point cloud, and a first ranging point set;

[0101] Step A2, and each time a new first multimodal data group is obtained, the current first multimodal data group is taken as the current data group; and target detection, target feature fusion and target tracking are performed based on the current data group, and the target perception dataset is refreshed based on the processing results.

[0102] In another specific implementation of this invention, the vehicle perception module 12 is specifically used when performing target detection, target feature fusion, and target tracking processing based on the current data set and refreshing the target perception dataset based on the processing results:

[0103] Step B1: Take the first image, first point cloud, second point cloud, and first ranging point set of the current data group as the corresponding current image, current laser point cloud, current millimeter wave point cloud, and current ranging point set; and take the time point corresponding to the current data group as the current timestamp.

[0104] Step B2 involves performing target detection and classification on the current image based on a pre-set visual target recognition model to obtain a set of corresponding first target detection boxes; and converting the coordinates of the first center point of each first target detection box in the set of first target detection boxes from image coordinate system coordinates to vehicle coordinate system coordinates, and converting the height and width of each first target detection box from pixel height and width to real-world height and width;

[0105] Here, the visual target recognition model in this embodiment of the invention includes the YOLO series model, the Faster R-CNN model, and the SSD model;

[0106] The first target detection box set in this embodiment of the invention includes multiple first target detection boxes; each first target detection box includes first center point coordinates, first detection box height, first detection box width, and first target type; the first target type includes pedestrians, motor vehicles, and non-motor vehicles;

[0107] Step B3 involves performing target detection processing on the current laser point cloud based on a preset point cloud target detection model to obtain the corresponding set of second target detection boxes; and converting the coordinates of the second center point of each second target detection box in the set of second target detection boxes from the coordinates of the laser radar coordinate system to the coordinates of the vehicle coordinate system, and converting the orientation of each second detection box from the orientation of the laser radar coordinate system to the orientation of the vehicle coordinate system.

[0108] Here, the point cloud target recognition model in this embodiment of the invention includes the PointPillars model, the VoteNet model, and the SECOND model;

[0109] The second target detection box set in this embodiment of the invention includes multiple second target detection boxes; the first target detection box includes second center point coordinates, second detection box height, second detection box width, second detection box depth, and second detection box orientation;

[0110] Step B4, and convert the coordinates of each second point in the current millimeter-wave point cloud from the millimeter-wave radar coordinate system to the vehicle coordinate system;

[0111] Step B5, and based on the coordinates of the third point of each first ranging point in the current ranging point set and the first distance, calculate the corresponding second distance from the current ranging point to the edge of the current vehicle body; and convert the coordinates of each third point in the current ranging point set from the ultrasonic radar coordinate system to the vehicle coordinate system.

[0112] Step B6 involves identifying the second target detection boxes matched by each first target detection box and fusing the detection box attributes based on the identification results to obtain the corresponding third target detection box;

[0113] The third target detection box includes the coordinates of the third center point, the height of the third detection box, the width of the third detection box, the depth of the third detection box, the orientation of the third detection box, and the type of the third target.

[0114] Step B7, and based on the current millimeter-wave point cloud and the current ranging point set, perform relative motion attribute fusion on each third target detection box to generate corresponding target perception data;

[0115] Step B8, and locate the corresponding target tracking sequence for each newly added target sensing data in the target sensing dataset and add each newly added sensing data to its corresponding sequence.

[0116] In another specific implementation of this invention, the vehicle perception module 12 is specifically used to identify the second target detection boxes matched by each first target detection box and to obtain the corresponding third target detection box by fusing the detection box attributes based on the identification results:

[0117] Step C1: Select each first target detection box as the current detection box;

[0118] Step C2, and calculate the Euclidean distance between the coordinates of the first center point of the current detection box and the coordinates of each second center point;

[0119] Step C3, and take the second target detection box corresponding to the shortest Euclidean distance as the current matching box;

[0120] Step C4, and take the second center point coordinates, second detection box height, second detection box width, second detection box depth, and second detection box orientation of the current candidate box as a set of corresponding third center point coordinates, third detection box height, third detection box width, third detection box depth, and third detection box orientation;

[0121] Step C5, and use the first target type of the current detection box as the corresponding third target type;

[0122] Step C6, and a corresponding third target detection box is formed by the current third center point coordinates, third detection box height, third detection box width, third detection box depth, third detection box orientation, and third target type.

[0123] In another specific implementation of this invention, the vehicle perception module 12 is specifically used to generate corresponding target perception data by fusing the relative motion attributes of each third target detection box based on the current millimeter-wave point cloud and the current ranging point set:

[0124] Step D1: Use each third object detection box as the current object box;

[0125] Step D2, and take the 3D bounding box region of the current target box in the vehicle coordinate system as the current region;

[0126] Step D3, and the corresponding current first point set is formed by the second scan points whose second point coordinates are in the current region in the current millimeter wave point cloud, and the corresponding current second point set is formed by the first ranging points whose third point coordinates are in the current region in the current ranging point set;

[0127] Step D4, and take the average velocity of all first velocities in the current first point set as the corresponding relative velocity;

[0128] Step D5, and identify whether the current second point set is empty. If it is, estimate the corresponding relative distance based on the coordinates of the third center point of the current target box. Otherwise, take the minimum value among all the second distances corresponding to the current second point set as the corresponding relative distance.

[0129] Step D6 involves constructing a target shape from the height, width, and depth of the third detection box of the current target bounding box; assigning a unique identifier to the current target bounding box as the corresponding target identifier; and using the coordinates of the third center point of the current target bounding box, the orientation of the third detection box, and the type of the third target as a set of corresponding target coordinates, target orientation, and target type.

[0130] Step D7, and a new target perception data is formed by the current timestamp and a set of target identifiers, target types, target coordinates, target shapes, relative distances, relative speeds, and target orientations corresponding to the current target bounding box.

[0131] In another specific implementation of this invention, the vehicle perception module 12 is specifically used to locate the corresponding target tracking sequence for each newly added target perception data in the target perception dataset and add each newly added perception data to its corresponding sequence:

[0132] Step E1: Take each newly added target perception data as the corresponding current newly added data; and identify whether there is a target type in the target perception dataset that matches the target type of the target tracking sequence in the current newly added data.

[0133] Step E2: If the target type of any target tracking sequence in the target perception dataset matches the target type of the newly added data, then create a new empty target tracking sequence in the target perception dataset as the corresponding current sequence; and add the newly added data to the current sequence.

[0134] Step E3: If the target type of at least one target tracking sequence in the target perception dataset matches the target type of the currently added data, then each matching target tracking sequence is recorded as the corresponding search sequence; and the corresponding target of each search sequence is recorded as the corresponding search target; and using Kalman filtering, based on the historical data of each search sequence, the target coordinates and target shape of the corresponding search target at the current timestamp are predicted to obtain a set of corresponding predicted coordinates and predicted shapes to form a corresponding predicted target box; and the intersection-union ratio (IU) of the spatial regions of each predicted target box in the vehicle coordinate system with the corresponding spatial regions of the currently added data is estimated; and whether the largest IU exceeds the preset IU threshold is identified; if so, the search sequence corresponding to the largest IU is taken as the corresponding current sequence, and the target identifier of the currently added data is reset to the target identifier corresponding to the current sequence, and the currently added data is added to the current sequence after the reset; if not, an empty target tracking sequence is created in the target perception dataset as the corresponding current sequence, and the currently added data is added to the current sequence.

[0135] Here, the crossover ratio threshold in this embodiment of the invention is a pre-set percentage threshold parameter.

[0136] (III) Vehicle-to-Everything (V2X) Communication Module 13:

[0137] The vehicle-to-everything (V2X) communication module 13 of this invention is used to receive traffic congestion data, traffic light data, or vehicle guidance data sent by the V2X platform 2 or the V2X roadside equipment 3 and store them in the data storage module 14.

[0138] (iv) Data storage module 14:

[0139] The data storage module 14 in this embodiment of the invention is used to store vehicle status dataset, target perception dataset, congestion dataset, signal dataset, and guidance dataset.

[0140] like Figure 2 As shown in the schematic diagram of the five types of datasets provided in the embodiments of the present invention, the vehicle status dataset of the embodiments of the present invention includes multiple vehicle status data; each vehicle status data includes timestamp, vehicle coordinates, vehicle speed, gear, battery SOC, brake pedal opening degree, clutch pedal opening degree, accelerator pedal opening degree, and steering wheel angle; wherein, the vehicle coordinates are world coordinate system coordinates.

[0141] like Figure 2As shown, the target perception dataset of this embodiment includes multiple target tracking sequences; each target tracking sequence corresponds to a traffic participant target. The target tracking sequence includes multiple target perception data; each target perception data includes a timestamp, target identifier, target type, target coordinates, target shape, relative distance, relative speed, and target orientation; target types include pedestrians, motor vehicles, and non-motor vehicles; target shapes include height, width, and depth; it should be noted that all target identifiers and all target types are the same in each target tracking sequence.

[0142] like Figure 2 As shown, the congestion dataset in this embodiment of the invention includes multiple road congestion data points ahead; each road congestion data point ahead includes a timestamp and a road congestion index.

[0143] like Figure 2 As shown, the signal dataset of this embodiment includes multiple traffic light data for the road ahead; each traffic light data includes a timestamp, traffic light identifier, traffic light coordinates, signal status, and remaining signal duration; wherein, the signal status includes red light, green light, and yellow light.

[0144] like Figure 2 As shown, the guidance dataset of this embodiment includes multiple vehicle guidance data; the vehicle guidance data of this embodiment is used to guide the current straight-line movement of the vehicle; each vehicle guidance data includes a timestamp and a guidance type; wherein, the guidance type includes acceleration, deceleration, constant speed, and stopping.

[0145] (V) Strategy Decision Module 15:

[0146] In this embodiment of the invention, the strategy decision module 15 is used to determine whether the automatic start-stop conditions are met based on the dataset of the data storage module 14, and when the determination result is met, send the corresponding start-stop control command to the start-stop execution module 16.

[0147] Here, the start-stop control command in this embodiment of the invention includes command types; the command types include engine start and engine stop.

[0148] In another specific implementation of this invention, the strategy decision module 15 is specifically used to determine whether the automatic start-stop conditions are met based on the dataset of the data storage module 14, and to send a corresponding start-stop control command to the start-stop execution module 16 when the determination result is met:

[0149] Step F1: Periodically use the current time as the current timestamp according to the preset status query frequency;

[0150] Here, the status query frequency in this embodiment of the invention is a pre-set time frequency parameter;

[0151] Step F2, and take the latest vehicle status data in the vehicle status dataset as the current vehicle status; and take the vehicle coordinates, speed, gear, battery SOC, brake pedal opening degree, clutch pedal opening degree, accelerator pedal opening degree, and steering wheel angle of the current vehicle status as the corresponding current vehicle coordinates, current speed, current gear, current battery SOC, current brake pedal opening degree, current clutch pedal opening degree, current accelerator pedal opening degree, and current steering wheel angle;

[0152] Step F3: If the time interval between the latest timestamp and the current timestamp in the congestion dataset does not exceed the preset second duration, the latest road congestion index is used as the corresponding current congestion index.

[0153] Here, the second duration in this embodiment of the invention is a pre-set time length parameter;

[0154] Step F4: If the time interval between the latest timestamp and the current timestamp of the bootstrap dataset does not exceed the second duration, the latest bootstrap type is taken as the corresponding current bootstrap type.

[0155] Step F5, and based on the current vehicle coordinates, current vehicle speed and target perception dataset, identify whether there is a stationary target on the straight road in front of the current vehicle to obtain the corresponding stationary target identification result;

[0156] The results of stationary target identification include whether a stationary target exists or not.

[0157] Step F6 involves identifying the current signal status and remaining duration of the current signal based on the current vehicle coordinates and signal dataset.

[0158] Step F7, and identify the current start-stop phase status of the vehicle;

[0159] Here, the start-stop phase states in this embodiment of the invention include three types of phase states: start-up phase, stop phase, and non-start-stop phase, and are initialized to the non-start-stop phase by default when the vehicle leaves the factory;

[0160] Step F8: If the current start-stop phase is not a start-stop phase, then according to the preset automatic shutdown condition judgment rules, the automatic shutdown condition is satisfied based on the current vehicle speed, current gear, current battery SOC, current brake pedal opening degree, current steering wheel angle, stationary target recognition result, current congestion index, current signal status, current signal remaining duration, and current guidance type to obtain the corresponding first judgment result; and when the first judgment result is satisfied, a start-stop control command with the command type set to engine shutdown is sent to the start-stop execution module 16; and the start-stop phase state is switched to the shutdown phase.

[0161] The first judgment result includes either "satisfied" or "not satisfied";

[0162] Step F9: If the current start-stop phase is in the shutdown phase, then according to the preset self-start condition judgment rules, the self-start condition is judged based on the current gear, current clutch pedal opening degree, current accelerator pedal opening degree, current steering wheel angle, current signal status, current signal remaining duration, and current guidance type to determine whether the self-start condition is met, and obtain the corresponding second judgment result; if the second judgment result is met, the start-stop control command with the command type set to engine start is sent to the start-stop execution module 16; and the start-stop phase state is switched to the start phase.

[0163] The second judgment result includes whether it satisfies or not.

[0164] Step F10: If the current start-stop phase is in the start phase, then identify whether the current vehicle speed exceeds the preset first vehicle speed threshold; if so, then switch the start-stop phase to the non-start-stop phase.

[0165] Here, the first vehicle speed threshold in this embodiment of the invention is a preset speed threshold.

[0166] In another specific implementation of this invention, the strategy decision module 15 is specifically used to identify whether there is a stationary target on the straight road ahead of the current vehicle based on the current vehicle coordinates, current vehicle speed, and target perception dataset, and to obtain the corresponding stationary target identification result:

[0167] Step G1: Take the target tracking sequence in the target perception dataset whose time interval between the latest timestamp and the current timestamp does not exceed the second duration as the corresponding current trajectory; and take the target corresponding to the current trajectory as the current target;

[0168] Step G2, and determine whether the Euclidean distance between the last two target coordinates of the current trajectory is less than a preset first distance threshold;

[0169] Here, the first distance threshold in this embodiment of the invention is a pre-set distance threshold parameter;

[0170] Step G3: If the Euclidean distance between the last two target coordinates of the current trajectory is less than the first distance threshold, then the current target is regarded as a corresponding stationary target.

[0171] Step G4: If the Euclidean distance between the last two target coordinates of the current trajectory is greater than or equal to the first distance threshold, then the current timestamp is used as the future start time, and the time obtained by adding the current timestamp to the preset third duration is used as the future end time. The future start time and the future end time constitute the corresponding future time period. Based on the current trajectory, the motion trajectory of the current target in the future time period is predicted to obtain the corresponding first predicted trajectory. Based on the current vehicle coordinates and the current vehicle speed, the motion trajectory of the current vehicle in the future time period is predicted to obtain the corresponding second predicted trajectory. The existence of a spatiotemporal intersection point between the first and second predicted trajectories in the future time period is identified. If a spatiotemporal intersection point exists, the trajectory segment after the spatiotemporal intersection point on the first predicted trajectory is further identified as a stationary trajectory segment. If it is confirmed that the trajectory segment after the spatiotemporal intersection point is a stationary trajectory segment, then the current target is regarded as a corresponding stationary target.

[0172] Here, the third duration in this embodiment of the invention is a pre-set time length parameter;

[0173] Step G5, and identify the total number of stationary targets obtained; if the total number of stationary targets is 0, then set the corresponding stationary target identification result to "no stationary target exists"; if the total number of stationary targets is greater than 0, then set the corresponding stationary target identification result to "stationary target exists".

[0174] In another specific implementation of this invention, the strategy decision module 15 is specifically used to identify the current signal state and the remaining duration of the current signal based on the current vehicle coordinates and the signal dataset:

[0175] By querying a preset high-precision road map, it is confirmed whether the coordinates of the latest traffic light data ahead in the signal dataset are located ahead of the current vehicle's driving road. If confirmed, the latest traffic light data ahead in the signal dataset is used as the current traffic light data, and the signal status of the current traffic light data is used as the corresponding current signal status. Based on the timestamp of the current traffic light data, the remaining signal duration, and the current timestamp, the corresponding remaining signal duration is calculated as: remaining signal duration of the current traffic light data - (current timestamp - timestamp of the current traffic light data).

[0176] The self-shutdown condition judgment rule in this embodiment of the invention is implemented based on the first condition judgment matrix. For example... Figure 3As shown in the schematic diagram of the first condition judgment matrix provided in the embodiment of the present invention, the matrix columns of the first condition judgment matrix correspond one-to-one with the current vehicle speed, current gear, current battery SOC, current brake pedal opening degree, current steering wheel angle, stationary target recognition result, current congestion index, current signal status, current signal remaining duration, and current guidance type; each matrix row of the first condition judgment matrix corresponds to a first combination condition configuration; each matrix unit of the first condition judgment matrix includes a condition switch and a condition threshold; the condition switch includes two switch states: on and off; the condition threshold is a threshold range.

[0177] The judgment rule of the first condition judgment matrix in this embodiment of the invention is as follows: if the condition judgment result of any first combination condition configuration is satisfied, it is considered that the self-shutdown condition state is satisfied, and the corresponding first judgment result is satisfied; if the condition judgment results of all first combination condition configurations are not satisfied, it is considered that the self-shutdown condition state is not satisfied, and the corresponding first judgment result is not satisfied.

[0178] The judgment rule for each first combination condition configuration of the first condition judgment matrix in this embodiment of the invention is as follows: if the condition judgment result of all matrix units in the matrix row corresponding to the current first combination condition configuration that have condition switches in the on state is satisfied, then the condition judgment result of the current first combination condition configuration is satisfied; if the condition judgment result of at least one matrix unit in the matrix row corresponding to the current first combination condition configuration that has condition switches in the on state is not satisfied, then the condition judgment result of the current first combination condition configuration is not satisfied.

[0179] The judgment rule for each matrix unit in the first condition judgment matrix of this invention that is in the on state is as follows: the current vehicle speed, current gear, current battery SOC, current brake pedal opening degree, current steering wheel angle, stationary target recognition result, current congestion index, current signal status, current signal remaining duration, or current guidance type corresponding to the current matrix unit is taken as the current judgment object; if the current judgment object meets the condition threshold of the current matrix unit, the condition judgment result of the current matrix unit is satisfied; if the current judgment object does not meet the condition threshold of the current matrix unit, the condition judgment result of the current matrix unit is not satisfied.

[0180] The self-starting condition judgment rule in this embodiment of the invention is implemented based on the second condition judgment matrix. For example... Figure 4As shown in the schematic diagram of the second condition judgment matrix provided in the embodiment of the present invention, the matrix columns of the second condition judgment matrix correspond one-to-one with the current gear, the current clutch pedal opening degree, the current accelerator pedal opening degree, the current steering wheel angle, the current signal status, the current signal remaining duration, and the current guidance type; each matrix row of the second condition judgment matrix corresponds to a second combination condition configuration; each matrix unit of the second condition judgment matrix includes a condition switch and a condition threshold.

[0181] The judgment rule of the second condition judgment matrix in this embodiment of the invention is as follows: if the condition judgment result of any second combination condition configuration is satisfied, it is regarded as the self-starting condition state is satisfied, and the corresponding second judgment result is satisfied; if the condition judgment results of all second combination condition configurations are not satisfied, it is regarded as the self-starting condition state is not satisfied, and the corresponding second judgment result is not satisfied.

[0182] The judgment rule for each second combination condition configuration of the second condition judgment matrix in this embodiment of the invention is as follows: if the condition judgment result of all matrix cells with condition switches in the matrix row corresponding to the current second combination condition configuration is satisfied, then the condition judgment result of the current second combination condition configuration is satisfied; if the condition judgment result of at least one matrix cell with condition switches in the matrix row corresponding to the current second combination condition configuration is not satisfied, then the condition judgment result of the current second combination condition configuration is not satisfied.

[0183] The judgment rule for each matrix unit in the second condition judgment matrix of this invention that is in the on state is as follows: the current gear, current clutch pedal opening degree, current accelerator pedal opening degree, current steering wheel angle, current signal status, current signal remaining duration, or current guidance type corresponding to the current matrix unit is taken as the current judgment object; if the current judgment object meets the condition threshold of the current matrix unit, the condition judgment result of the current matrix unit is satisfied; if the current judgment object does not meet the condition threshold of the current matrix unit, the condition judgment result of the current matrix unit is not satisfied.

[0184] (vi) Start and stop execution module 16:

[0185] The start-stop execution module 16 of this embodiment of the invention is used to control the current vehicle to perform corresponding start-stop operations according to the start-stop control command.

[0186] In another specific implementation of this invention, the start-stop execution module 16 is specifically used to control the current vehicle to perform corresponding start-stop operations according to the start-stop control command:

[0187] The system identifies the type of start-stop control command; if the command type is engine start, it restarts the current vehicle's engine; if the command type is engine stop, it controls the current vehicle to decelerate and stop, and stops the current vehicle's engine when the vehicle speed drops to 0.

[0188] This invention provides an automatic start-stop system for intelligent connected vehicles. As described above, the system includes: a vehicle data acquisition module, a vehicle perception module, a vehicle-to-everything (V2X) communication module, a data storage module, a strategy decision-making module, and a start-stop execution module. Specifically, the vehicle data acquisition module continuously collects vehicle status data via the CAN bus and stores the collected data in the data storage module; the vehicle perception module connects to multiple types of onboard sensors and continuously perceives other traffic participants around the vehicle based on sensor data, storing the perceived data in the data storage module; the V2X communication module receives traffic environment information from a V2X platform or roadside V2X devices and stores it in the data storage module; the strategy decision-making module determines whether the automatic start-stop conditions are met based on the dataset in the data storage module according to two configurable judgment rules (automatic stop condition judgment rules and automatic start condition judgment rules), and sends a corresponding start-stop control command to the start-stop execution module when the judgment result is met; the start-stop execution module controls the vehicle to perform corresponding engine start-stop operations according to the start-stop control command. The embodiments of the present invention not only improve the energy-saving and emission-reduction performance of vehicles and enhance the driving experience, but also increase the flexibility of the configuration of the trigger condition judgment process.

[0189] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0190] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0191] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An automatic start-stop system for intelligent connected vehicles, characterized in that, The system includes: a vehicle data acquisition module, a vehicle perception module, a vehicle-to-everything (V2X) communication module, a data storage module, a strategy decision-making module, and a start / stop execution module; The system is installed on intelligent connected vehicles; the data storage module is connected to the vehicle acquisition module, the vehicle perception module, the vehicle network communication module and the strategy decision module respectively; the strategy decision module is connected to the start-stop execution module; the vehicle network communication module is connected to a remote vehicle network platform and a vehicle network roadside device on the road where the vehicle is located through a V2X communication network. The vehicle data acquisition module is used to continuously acquire the current vehicle status data via the CAN bus and store the acquired data in the data storage module. The vehicle sensing module is used to connect to multiple types of vehicle sensors, including cameras, lidar, millimeter-wave radar, and ultrasonic radar. These multiple types of vehicle sensors achieve hardware-level time synchronization based on the IEEE 1588 PTP protocol. The vehicle perception module is also used to continuously perceive other traffic participants around the current vehicle based on the sensor data of the multiple types of vehicle sensors and store the perception data into the data storage module. The vehicle-to-everything (V2X) communication module is used to receive road congestion data, traffic light data, or vehicle guidance data sent by the V2X platform or the V2X roadside equipment and store them in the data storage module. The data storage module is used to store vehicle status datasets, target perception datasets, congestion datasets, signal datasets, and guidance datasets; The strategy decision module is used to determine whether the automatic start-stop conditions are met based on the dataset of the data storage module, and send the corresponding start-stop control command to the start-stop execution module when the determination result is met. The start-stop execution module is used to control the current vehicle to perform corresponding start-stop operations according to the start-stop control command. The vehicle status dataset includes multiple vehicle status data; the vehicle status data includes timestamp, vehicle coordinates, vehicle speed, gear, battery SOC, brake pedal opening degree, clutch pedal opening degree, accelerator pedal opening degree, and steering wheel angle; the vehicle coordinates are in the world coordinate system. The target perception dataset includes multiple target tracking sequences; the target tracking sequence includes multiple target perception data; the target perception data includes timestamps, target identifiers, target types, target coordinates, target shapes, relative distances, relative speeds, and target orientations; the target types include pedestrians, motor vehicles, and non-motor vehicles; the target shapes include height, width, and depth; all target identifiers and all target types are the same in each target tracking sequence; The congestion dataset includes multiple road congestion data points ahead; the road congestion data ahead includes timestamps and road congestion indices. The signal dataset includes multiple traffic light data points for the road ahead; the traffic light data for the road ahead includes timestamps, traffic light identifiers, traffic light coordinates, signal status, and remaining signal duration; the signal status includes red, green, and yellow lights. The guidance dataset includes multiple vehicle guidance data sets; the vehicle guidance data is used to guide the current straight-line movement of the vehicle; the vehicle guidance data includes a timestamp and a guidance type; the guidance type includes acceleration, deceleration, constant speed, and stopping; The start-stop control command includes command types; the command types include engine start and engine stop. The strategy decision module, during the process of determining whether the automatic start-stop conditions are met, identifies whether there is a stationary target on the straight road ahead of the vehicle based on the current vehicle coordinates, current vehicle speed, and the target perception dataset, and obtains the corresponding stationary target identification result. Specifically: Step 91: Take the target tracking sequence in the target perception dataset whose time interval between the latest timestamp and the current timestamp does not exceed the second duration as the corresponding current trajectory; and take the target corresponding to the current trajectory as the current target; Step 92, and determine whether the Euclidean distance between the last two target coordinates of the current trajectory is less than a preset first distance threshold; Step 93: If the Euclidean distance between the last two target coordinates of the current trajectory is less than the first distance threshold, then the current target is regarded as a corresponding stationary target. Step 94: If the Euclidean distance between the last two target coordinates of the current trajectory is greater than or equal to the first distance threshold, then the current timestamp is used as the future start time, and the time obtained by adding the current timestamp to the preset third duration is used as the future end time. The future start time and the future end time form a corresponding future time period. Based on the current trajectory, the motion trajectory of the current target in the future time period is predicted to obtain a corresponding first predicted trajectory. Based on the current vehicle coordinates and the current vehicle speed, the motion trajectory of the current vehicle in the future time period is predicted to obtain a corresponding second predicted trajectory. The existence of a spatiotemporal intersection point between the first and second predicted trajectories in the future time period is identified. If the spatiotemporal intersection point exists, the trajectory segment after the spatiotemporal intersection point on the first predicted trajectory is further identified as a stationary trajectory segment. If it is confirmed that the trajectory segment after the spatiotemporal intersection point is a stationary trajectory segment, then the current target is regarded as a corresponding stationary target. Step 95, and identify the total number of the obtained stationary targets; if the total number of stationary targets is 0, then set the corresponding stationary target identification result to "no stationary targets exist"; if the total number of stationary targets is greater than 0, then set the corresponding stationary target identification result to "stationary targets exist".

2. The automatic start-stop system for intelligent connected vehicles according to claim 1, characterized in that, The vehicle data acquisition module is specifically used when continuously acquiring the current vehicle status data via the CAN bus and storing the acquired data into the data storage module: At a preset first acquisition frequency, the current time is periodically used as the current timestamp, and the latest vehicle coordinates, vehicle speed, gear position, battery SOC, brake pedal opening degree, clutch pedal opening degree, accelerator pedal opening degree, and steering wheel angle are collected via the CAN bus. The current timestamp and the collected vehicle coordinates, vehicle speed, gear position, battery SOC, brake pedal opening degree, accelerator pedal opening degree, and steering wheel angle are combined to form a new vehicle status data set, which is then stored in the vehicle status dataset.

3. The automatic start-stop system for intelligent connected vehicles according to claim 1, characterized in that, The vehicle perception module is specifically used when continuously perceiving other traffic participants around the current vehicle based on sensor data from the multiple types of onboard sensors and storing the perception data in the data storage module: Step 41: In the perception module, a first buffer sequence is set to buffer the visual images periodically output by the camera, a second buffer sequence is set to buffer the dense point cloud periodically output by the lidar, a third buffer sequence is set to buffer the sparse point cloud periodically output by the millimeter-wave radar, and a fourth buffer sequence is set to buffer the ranging point set periodically output by the ultrasonic radar; and a corresponding first multimodal data group is formed by the time-aligned four types of sensor data in each of the four buffer sequences. The first cache sequence is formed by sequentially sorting multiple first images; The second cache sequence is formed by sequentially sorting multiple first point clouds; the first point cloud includes multiple first scan points; the first scan point includes first point coordinates and a first reflection intensity; the third cache sequence is formed by sequentially sorting multiple second point clouds; the first point coordinates are coordinates in the lidar coordinate system; The second point cloud includes multiple second scanning points; each second scanning point includes second point coordinates and a first velocity; the second point coordinates are coordinates in a millimeter-wave radar coordinate system; the first velocity is the relative velocity of the current scanning point relative to the current vehicle. The fourth cache sequence is formed by sequentially sorting multiple first ranging point sets; the first ranging point set includes multiple first ranging points; the first ranging point includes third point coordinates and a first distance; the third point coordinates are coordinates in the ultrasonic radar coordinate system; the first distance is the radial distance from the current third point coordinates to the ultrasonic radar installation position; The first multimodal data set consists of a time-aligned set of the first image, the first point cloud, the second point cloud, and the first ranging point set; Step 42: When a new first multimodal data group is obtained, the current first multimodal data group is taken as the current data group; target detection, target feature fusion and target tracking processing are performed based on the current data group, and the target perception dataset is refreshed based on the processing results.

4. The automatic start-stop system for intelligent connected vehicles according to claim 3, characterized in that, The vehicle perception module is specifically used when performing target detection, target feature fusion, and target tracking processing based on the current data set, and refreshing the target perception dataset based on the processing results: Step 51: Take the first image, the first point cloud, the second point cloud, and the first ranging point set of the current data group as the corresponding current image, current laser point cloud, current millimeter wave point cloud, and current ranging point set; And use the time point corresponding to the current data group as the current timestamp; Step 52, and based on the preset visual target recognition model, perform target detection and classification recognition processing on the current image to obtain the corresponding first target detection box set; The coordinates of the first center point of each first target detection box in the first target detection box set are converted from image coordinate system coordinates to vehicle coordinate system coordinates, and the height and width of the first detection box of each first target detection box are converted from pixel height and width to real-world height and width. Step 53, and perform target detection processing on the current laser point cloud based on the preset point cloud target recognition model to obtain the corresponding second target detection box set; The coordinates of the second center point of each second target detection box in the second target detection box set are converted from the coordinates of the lidar coordinate system to the coordinates of the vehicle coordinate system, and the orientation of each second detection box is converted from the orientation of the lidar coordinate system to the orientation of the vehicle coordinate system. Step 54, and convert the coordinates of each of the second points in the current millimeter-wave point cloud from millimeter-wave radar coordinates to vehicle coordinates; Step 55: Based on the coordinates of the third point of each of the first ranging points in the current ranging point set and the first distance, calculate the corresponding second distance from the current ranging point to the edge of the current vehicle body; and convert the coordinates of each of the third points in the current ranging point set from the ultrasonic radar coordinate system to the vehicle coordinate system. Step 56, and identify the second target detection boxes that match each of the first target detection boxes, and perform detection box attribute fusion based on the identification results to obtain the corresponding third target detection box; Step 57, and based on the current millimeter-wave point cloud and the current ranging point set, perform relative motion attribute fusion on each of the third target detection boxes to generate the corresponding target perception data; Step 58, and locate the corresponding target tracking sequence for each newly added target perception data in the target perception dataset and add each newly added perception data to its corresponding sequence.

5. The automatic start-stop system for intelligent connected vehicles according to claim 4, characterized in that, The visual target recognition model includes the YOLO series model, the Faster R-CNN model, and the SSD model; the first target detection box set includes multiple first target detection boxes; the first target detection box includes the first center point coordinates, the first detection box height, the first detection box width, and the first target type; the first target type includes pedestrians, motor vehicles, and non-motor vehicles; The point cloud target recognition model includes the PointPillars model, the VoteNet model, and the SECOND model; the second target detection box set includes multiple second target detection boxes; the second target detection box includes the second center point coordinates, the second detection box height, the second detection box width, the second detection box depth, and the second detection box orientation; The third target detection box includes the coordinates of the third center point, the height of the third detection box, the width of the third detection box, the depth of the third detection box, the orientation of the third detection box, and the type of the third target; The vehicle perception module is specifically used when, during the process of recognizing the second target detection boxes matched by each of the first target detection boxes and fusing the detection box attributes based on the recognition results to obtain the corresponding third target detection box: Each of the first target detection boxes is taken as the current detection box; and the Euclidean distance between the coordinates of the first center point of the current detection box and the coordinates of each of the second center points is calculated; and the second target detection box corresponding to the shortest Euclidean distance is taken as the current matching box. The second center point coordinates, second detection box height, second detection box width, second detection box depth, and second detection box orientation of the current matching box are used as a set of corresponding third center point coordinates, third detection box height, third detection box width, third detection box depth, and third detection box orientation; and the first target type of the current detection box is used as the corresponding third target type. A corresponding third target detection box is formed by the current coordinates of the third center point, the height of the third detection box, the width of the third detection box, the depth of the third detection box, the orientation of the third detection box, and the type of the third target. The vehicle perception module is specifically used when generating corresponding target perception data by fusing the relative motion attributes of each of the third target detection boxes based on the current millimeter-wave point cloud and the current ranging point set: Each of the aforementioned third target detection boxes is used as the current target box; The current target bounding box is defined as the 3D bounding box region in the vehicle coordinate system. A current first point set is formed by all second scan points in the current millimeter-wave point cloud whose second point coordinates are located within the current region. A current second point set is formed by all first ranging points in the current ranging point set whose third point coordinates are located within the current region. The average speed of all first velocities in the current first point set is used as the corresponding relative speed. The system identifies whether the current second point set is empty. If it is, the corresponding relative distance is estimated based on the third center point coordinates of the current target bounding box. Otherwise, the minimum value among all second distances corresponding to the current second point set is used as the corresponding relative distance. A corresponding target shape is formed by the third detection box height, third detection box width, and third detection box depth of the current target bounding box. A unique identifier is assigned to the current target bounding box as the corresponding target identifier. The third center point coordinates, third detection box orientation, and third target type of the current target bounding box are used as a set of corresponding target coordinates, target orientation, and target type. The current timestamp and a set of target identifiers, target type, target coordinates, target shape, relative distance, relative speed, and target orientation corresponding to the current target bounding box constitute a new target perception data. The vehicle perception module is specifically used when, in the target perception dataset, it locates the corresponding target tracking sequence for each newly added target perception data and adds each newly added perception data to its corresponding sequence: Each newly added target perception data is taken as the corresponding current newly added data; and it is identified whether there is a target type in the target perception dataset that matches the target type of the target tracking sequence in the current newly added data; If no target type in the target sensing dataset matches the target type of the currently added data, then an empty target tracking sequence is created in the target sensing dataset as the corresponding current sequence; and the currently added data is added to the current sequence. If the target type of at least one target tracking sequence in the target perception dataset matches the target type of the currently added data, then each matching target tracking sequence is recorded as a corresponding search sequence; and the corresponding target of each search sequence is recorded as the corresponding search target; and using Kalman filtering, based on the historical data of each search sequence, the target coordinates and target shape of the corresponding search target at the current timestamp are predicted to obtain a set of corresponding predicted coordinates and predicted shapes to form a corresponding predicted target box; The cross-intersection over union (CUI) ratio of the spatial regions of each predicted target bounding box in the vehicle coordinate system and the corresponding spatial regions of the newly added data is estimated; and the largest CUI ratio is identified as exceeding a preset CUI threshold. If so, the query sequence corresponding to the largest intersection-union ratio of the regions is taken as the corresponding current sequence, and the target identifier of the currently added data is reset to the target identifier corresponding to the current sequence. After the reset, the currently added data is added to the current sequence. If not, then create a new empty target tracking sequence in the target perception dataset as the corresponding current sequence, and add the newly added data to the current sequence.

6. The automatic start-stop system for intelligent connected vehicles according to claim 1, characterized in that, The vehicle perception module is also used to periodically traverse all the target tracking sequences in the target perception dataset; and during this traversal, the target tracking sequence currently being traversed is taken as the current sequence; and to identify whether the time interval between the last timestamp of the current sequence and the current time exceeds a preset first duration. If so, the current sequence is deleted.

7. The automatic start-stop system for intelligent connected vehicles according to claim 1, characterized in that, The strategy decision module is specifically used to determine whether the automatic start / stop conditions are met based on the dataset of the data storage module, and to send the corresponding start / stop control command to the start / stop execution module when the determination result is met: Step 81: Periodically use the current time as the current timestamp according to the preset status query frequency; Step 82, and take the latest vehicle status data in the vehicle status dataset as the current vehicle status; The vehicle coordinates, vehicle speed, gear, battery SOC, brake pedal opening degree, clutch pedal opening degree, accelerator pedal opening degree, and steering wheel angle of the current vehicle state are used as the corresponding current vehicle coordinates, current vehicle speed, current gear, current battery SOC, current brake pedal opening degree, current clutch pedal opening degree, current accelerator pedal opening degree, and current steering wheel angle. Step 83: If the time interval between the latest timestamp of the congestion dataset and the current timestamp does not exceed a preset second duration, the latest road congestion index is used as the corresponding current congestion index. Step 84: If the time interval between the latest timestamp of the bootstrap dataset and the current timestamp does not exceed the second duration, the latest bootstrap type is taken as the corresponding current bootstrap type. Step 85, and based on the current vehicle coordinates, the current vehicle speed and the target perception dataset, identify whether there is a stationary target on the straight road in front of the current vehicle to obtain the corresponding stationary target identification result; The results of the stationary target identification include whether a stationary target exists or not. Step 86, and identify the current signal status and the remaining duration of the current signal based on the current vehicle coordinates and the signal dataset; Step 87, and identify the current start-stop phase status of the vehicle; The start-stop phase states include the start-up phase, the stop phase, and the non-start-stop phase; Step 88: If the current start-stop phase is a non-start-stop phase, then according to the preset automatic shutdown condition judgment rules, based on the current vehicle speed, current gear, current battery SOC, current brake pedal opening degree, current steering wheel angle, stationary target recognition result, current congestion index, current signal status, current signal remaining duration, and current guidance type, the automatic shutdown condition satisfaction status judgment is obtained to obtain the corresponding first judgment result; and when the first judgment result is satisfied, the start-stop control command with the command type set to engine shutdown is sent to the start-stop execution module; and the start-stop phase state is switched to the shutdown phase; Step 89: If the current start-stop phase is in the shutdown phase, then according to the preset self-start condition judgment rules, the self-start condition is judged based on the current gear, the current clutch pedal opening degree, the current accelerator pedal opening degree, the current steering wheel angle, the current signal status, the remaining duration of the current signal, and the current guidance type to obtain the corresponding second judgment result; and if the second judgment result is satisfied, the start-stop control command with the command type set to engine start is sent to the start-stop execution module; and the start-stop phase state is switched to the start phase. Step 90: If the current start-stop phase is in the start phase, then identify whether the current vehicle speed exceeds a preset first vehicle speed threshold; if so, then switch the start-stop phase to a non-start-stop phase.

8. The automatic start-stop system for intelligent connected vehicles according to claim 7, characterized in that, The strategy decision module is specifically used when identifying the current signal state and the remaining duration of the current signal based on the current vehicle coordinates and the signal dataset: By querying a preset high-precision road map, it is confirmed whether the coordinates of the latest traffic light data ahead in the signal dataset are located ahead of the current vehicle's driving road. If confirmed, the latest traffic light data ahead in the signal dataset is taken as the current traffic light data, and the signal status of the current traffic light data is taken as the corresponding current signal status. Based on the timestamp of the current traffic light data, the remaining signal duration, and the current timestamp, the corresponding remaining signal duration is calculated as: remaining signal duration of the current traffic light data - (current timestamp - timestamp of the current traffic light data).

9. The automatic start-stop system for intelligent connected vehicles according to claim 7, characterized in that, The automatic shutdown condition judgment rule is implemented based on a first condition judgment matrix; the matrix columns of the first condition judgment matrix correspond one-to-one with the current vehicle speed, the current gear, the current battery SOC, the current brake pedal opening degree, the current steering wheel angle, the stationary target recognition result, the current congestion index, the current signal status, the current signal remaining duration, and the current guidance type; each matrix row of the first condition judgment matrix corresponds to a first combination condition configuration; each matrix unit of the first condition judgment matrix includes a condition switch and a condition threshold; the condition switch includes two switch states: on and off; the condition threshold is a threshold range; The judgment rule of the first condition judgment matrix is ​​as follows: if the judgment result of any of the first combination conditions is satisfied, it is considered that the self-shutdown condition state is satisfied, and the corresponding first judgment result is satisfied; if the judgment results of all the first combination conditions are not satisfied, it is considered that the self-shutdown condition state is not satisfied, and the corresponding first judgment result is not satisfied. The judgment rule for each of the first combination conditions in the first condition judgment matrix is ​​as follows: if the condition judgment results of all matrix cells in the matrix row corresponding to the current first combination condition configuration that have the condition switch in the on state are satisfied, then the condition judgment result of the current first combination condition configuration is satisfied; if the condition judgment result of at least one matrix cell in the matrix row corresponding to the current first combination condition configuration that has the condition switch in the on state is not satisfied, then the condition judgment result of the current first combination condition configuration is not satisfied. The judgment rule for each matrix unit in the first condition judgment matrix where the condition switch is in the on state is as follows: the current vehicle speed, the current gear, the current battery SOC, the current brake pedal opening degree, the current steering wheel angle, the stationary target recognition result, the current congestion index, the current signal status, the current signal remaining duration, or the current guidance type corresponding to the current matrix unit is taken as the current judgment object; If the current judgment object satisfies the condition threshold of the current matrix unit, then the condition judgment result of the current matrix unit is satisfied; if the current judgment object does not satisfy the condition threshold of the current matrix unit, then the condition judgment result of the current matrix unit is not satisfied. The self-starting condition judgment rule is implemented based on a second condition judgment matrix; the matrix columns of the second condition judgment matrix correspond one-to-one with the current gear, the current clutch pedal opening degree, the current accelerator pedal opening degree, the current steering wheel angle, the current signal status, the current signal remaining duration, and the current guidance type; each matrix row of the second condition judgment matrix corresponds to a second combination condition configuration; each matrix unit of the second condition judgment matrix includes the condition switch and the condition threshold; The judgment rule of the second condition judgment matrix is ​​as follows: if the judgment result of any of the second combination conditions is satisfied, it is considered that the self-starting condition state is satisfied, and the corresponding second judgment result is satisfied; if the judgment results of all the second combination conditions are not satisfied, it is considered that the self-starting condition state is not satisfied, and the corresponding second judgment result is not satisfied. The judgment rule for each of the second combination conditions in the second condition judgment matrix is ​​as follows: if the condition judgment results of all matrix cells in the matrix row corresponding to the current second combination condition configuration that have the condition switch in the on state are satisfied, then the condition judgment result of the current second combination condition configuration is satisfied; if the condition judgment result of at least one matrix cell in the matrix row corresponding to the current second combination condition configuration that has the condition switch in the on state is not satisfied, then the condition judgment result of the current second combination condition configuration is not satisfied. The judgment rule for each matrix unit in the second condition judgment matrix that is in the on state is as follows: the current gear, the current clutch pedal opening degree, the current accelerator pedal opening degree, the current steering wheel angle, the current signal state, the current signal remaining duration, or the current guidance type corresponding to the current matrix unit is taken as the current judgment object; If the current judgment object satisfies the condition threshold of the current matrix unit, then the condition judgment result of the current matrix unit is satisfied; if the current judgment object does not satisfy the condition threshold of the current matrix unit, then the condition judgment result of the current matrix unit is not satisfied.

10. The automatic start-stop system for intelligent connected vehicles according to claim 1, characterized in that, The start-stop execution module is specifically used when controlling the current vehicle to perform the corresponding start-stop operation according to the start-stop control command: The type of the start-stop control command is identified; if the command type is engine start, the engine of the current vehicle is restarted; if the command type is engine stop, the current vehicle is controlled to decelerate and stop, and the engine of the current vehicle is stopped when the vehicle speed drops to 0.

Images