A remote human-machine co-driving system, method and device

Abstract

The present invention discloses a remote human-machine co-driving system, method and device, and relates to the field of human-machine co-driving technology. The remote human-machine co-driving method obtains the current two-dimensional position coordinates of the vehicle to be controlled on the road and the image data of the road, and performs identification and analysis to obtain the real-time vehicle distance value between the vehicle to be controlled and each fellow vehicle on the road, and performs judgment and analysis with the preset vehicle distance warning, and takes corresponding control measures based on the judgment and analysis. The present invention analyzes the distance between the vehicle to be controlled and each fellow vehicle in real time, and judges whether there is a potential driving risk according to the preset safety distance range. If a fellow vehicle enters the safety range, it will automatically identify and adjust the control strategy to avoid accidents. In addition, the safety of the vehicle and the distance warning vehicle can be evaluated in real time through the calculation of the driving safety index, thereby ensuring that the vehicle is always in a safe driving state.

Description

Technical Field

[0001] The present invention relates to the technical field of human-machine co-driving, and specifically to a remote human-machine co-driving system, method and device. Background Art

[0002] Remote human-machine co-driving means that during the vehicle's driving process, the autonomous driving system and remote workers work together through technical means such as on-board sensors, communication networks and intelligent algorithms to achieve collaborative sharing of driving tasks. The autonomous driving system is responsible for the basic control and perception of the vehicle, while the remote workers provide decision support through remote operation or supervision when encountering complex or uncertain situations. In this way, by combining remote workers with the autonomous driving system, the advantages of both parties can be fully utilized to improve driving safety and the intelligence level of the vehicle.

[0003] Prior art, such as a human-machine co-driving system, method and device disclosed in a patent application with announcement number: CN116931498B, the method includes: receiving a global trajectory route pre-planned by a cloud platform, parsing the global trajectory route into a global lateral motion trajectory and a global longitudinal motion trajectory, and sending the global lateral motion trajectory to an automatic driving module, and sending the global longitudinal motion trajectory to a remote driving module; generating a first control instruction for controlling the lateral motion of the unmanned vehicle according to the global lateral motion trajectory; receiving a second control instruction for remotely controlling the longitudinal motion of the unmanned vehicle generated by the remote driving module according to the global longitudinal motion trajectory; controlling the yaw angle of the unmanned vehicle based on the first control instruction by the lateral motion execution module to complete the lateral motion; controlling the throttle opening based on the second control instruction by the longitudinal motion execution module to complete the longitudinal driving, and completing the braking by the braking system. The present disclosure improves the driving safety of the unmanned vehicle.

[0004] Based on the above solution, it is found that the limitations of the existing technology include at least the following problems. First, the existing technology lacks real-time analysis and dynamic response mechanism for peer vehicles. It is difficult for the existing technology to evaluate and adjust the safe distance between the vehicle and the peer vehicle in real time. When the peer vehicle approaches or there is a potential risk of collision with the vehicle, there is a lack of timely safety warning and emergency response plan. Secondly, in the existing technology, the control mode conversion of the vehicle fails to take into account the comprehensive driving conditions and the actual driving index of the vehicle in the dynamic environment changes, which easily leads to the vehicle's lack of flexibility and accuracy in responding to the control mode switching in a complex environment, increasing the potential safety hazards during driving. Finally, the existing technology's analysis of the safety and control stability of the vehicle during driving lacks a comprehensive analysis of multi-dimensional information such as the vehicle's driving safety index, driving stability and the real-time approaching speed of peer vehicles, thereby ignoring the safety problems that arise in complex road conditions, which in turn affects its reliability and effectiveness in practical applications, limiting its applicability and safety in dynamic driving environments. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides a remote human-machine co-driving system, method and device, which solve the problems in the prior art that the prior art lacks real-time analysis and dynamic response mechanism for peer vehicles. The prior art makes it difficult to evaluate and adjust the safe distance between the vehicle and the peer vehicle in real time. When the peer vehicle approaches or there is a potential risk of collision with the vehicle, there is a lack of timely safety warning and emergency response plan. Secondly, in the prior art, the control mode conversion of the vehicle fails to take into account the comprehensive driving conditions and the actual driving index of the vehicle in dynamic environmental changes, which easily leads to the lack of flexibility and accuracy of the vehicle's response to the control mode switching in a complex environment, increasing the potential safety hazards during driving. Finally, the prior art lacks a comprehensive analysis of multi-dimensional information such as the vehicle's driving safety index, driving stability and the real-time approaching speed of peer vehicles in the analysis of the safety and control stability of the vehicle during driving, thereby ignoring the safety issues arising in complex road conditions, which in turn affects its reliability and effectiveness in practical applications, and limits its applicability and safety in dynamic driving environments.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: a remote human-machine co-driving method, comprising the following steps: obtaining the current two-dimensional position coordinates of the vehicle to be controlled on the road and the road image data, and performing identification and analysis to obtain a number of peer vehicles on the road to be controlled and a real-time vehicle distance value with each peer vehicle, wherein the road image data is specifically the pixel value and two-dimensional coordinates of each pixel point in the road image; the real-time vehicle distance value of the vehicle to be controlled on the road to be controlled and each peer vehicle is judged and analyzed with a preset vehicle distance warning range; if the real-time vehicle distance value of each peer vehicle on the road to be controlled is higher than the preset vehicle distance warning range, then the driving instruction data received several times on the road to be controlled and the actual control driving data of the vehicle are continuously obtained in real time, and a comprehensive analysis is performed to obtain the actual driving index of the vehicle to be controlled each time received on the road to be controlled, and a trend analysis is performed to obtain a comprehensive driving change index of the vehicle to be controlled on the road to be controlled, wherein the index Let data be specifically the instruction value of each operation instruction, and the actual control driving data of the vehicle be specifically the actual operation value of the vehicle to be controlled for each operation instruction; the comprehensive driving change index of the vehicle to be controlled on the road is judged and analyzed with the preset comprehensive driving change index threshold, and corresponding control measures are taken based on the judgment and analysis results; when there is a real-time vehicle distance value between the vehicle to be controlled and the peer vehicle on the road that is in the preset vehicle distance warning range, the corresponding peer vehicle is marked as a distance warning vehicle, and the vehicle condition data of the vehicle to be controlled at several time points on the road and the adjacent vehicle condition data of the distance warning vehicle at several time points are continuously acquired in real time, and a comprehensive analysis is performed to obtain the driving safety index of the vehicle to be controlled on the road, the vehicle condition data are specifically the vehicle speed value and the vehicle acceleration value, and the adjacent vehicle condition data are specifically the distance warning vehicle speed value and the distance warning vehicle acceleration value; the driving safety index of the vehicle to be controlled on the road is compared and analyzed with the preset driving safety interval, and corresponding vehicle control measures are taken based on the comparison and analysis results.

[0007] Furthermore, the specific steps for obtaining the real-time vehicle distance value between the vehicle to be controlled and each of its peers on the road are as follows: the pixel value and the two-dimensional coordinates of each pixel point in the road image of the vehicle to be controlled on the road are respectively input into the target recognition model to obtain the two-dimensional position coordinates of several peers of the vehicle to be controlled on the road; and the two-dimensional position coordinates of each peer of the vehicle to be controlled on the road are converted into the world coordinate system to obtain the world two-dimensional position coordinates of each peer of the vehicle to be controlled on the road; and the current two-dimensional position coordinates of the vehicle to be controlled on the road and the world two-dimensional position coordinates of each peer of the vehicle to be controlled on the road are respectively analyzed to obtain the real-time vehicle distance value between the vehicle to be controlled and each peer on the road.

[0008] Furthermore, the specific steps for obtaining the comprehensive driving variation index of the vehicle to be controlled on the road are as follows: comprehensively analyze the command value of each operation command received by the vehicle to be controlled on the road in combination with the actual operation value of the vehicle to be controlled for each operation command, and obtain the command deviation index of each operation command received by the vehicle to be controlled on the road; and comprehensively analyze the command deviation index of each operation command received by the vehicle to be controlled on the road in order to obtain the actual driving index of the vehicle received on the road; perform trend analysis on the actual driving index of the vehicle received each time by the vehicle to be controlled on the road in order to obtain several groups of vehicle actual driving index change rates of the vehicle to be controlled on the road, and conduct comprehensive analysis to obtain the comprehensive driving variation index of the vehicle to be controlled on the road.

[0009] Furthermore, the formula for calculating the actual driving index change rate and the comprehensive driving change index of several groups of vehicles to be controlled on the road is as follows: ;in, The first The actual driving index change rate of the group of vehicles, is a natural constant, The first The actual driving index of the vehicle received, The first The actual driving index of the vehicle received, is the comprehensive driving change index of the vehicle to be controlled on the road, The first The weighted coefficient of the actual driving index change rate of the group of vehicles, The first The actual driving index change rate of the group of vehicles, The first The weighted coefficient of the actual driving index change rate of the group of vehicles, and , , is the number of receptions, ,and , is the number of groups of the actual vehicle driving index change rate.

[0010] Furthermore, the specific steps for obtaining the driving safety index of the vehicle to be controlled on the road are as follows: perform difference analysis on the vehicle speed value and vehicle acceleration value of the vehicle to be controlled at each time point on the road, and the distance warning vehicle speed value and distance warning vehicle acceleration value at each time point of the distance warning vehicle, and obtain the relative speed difference and relative acceleration difference of the vehicle to be controlled and the distance warning vehicle at each time point on the road, and perform comprehensive analysis to obtain the driving stability index of the vehicle to be controlled and the distance warning vehicle on the road; perform mean analysis on the relative speed difference and relative acceleration difference of the vehicle to be controlled and the distance warning vehicle at each time point on the road, and obtain the mean value of the relative speed difference and relative acceleration difference of the vehicle to be controlled and the distance warning vehicle on the road; read the speed of the vehicle to be controlled A comprehensive analysis is performed on the current two-dimensional position coordinates on the road and the two-dimensional position coordinates of the distance warning vehicle to obtain the relative angle between the vehicle to be controlled and the distance warning vehicle on the road; the real-time vehicle distance value between the vehicle to be controlled and the distance warning vehicle on the road is read, and a comprehensive analysis is performed on the relative angle, the average relative speed difference, and the average relative acceleration difference between the vehicle to be controlled and the distance warning vehicle on the road to obtain the collision risk index between the vehicle to be controlled and the distance warning vehicle on the road; the collision risk index and the driving stability index of the vehicle to be controlled and the distance warning vehicle on the road are standardized; the collision risk index and the driving stability index of the vehicle to be controlled and the distance warning vehicle on the road after the standardized processing are comprehensively analyzed to obtain the driving safety index of the vehicle to be controlled on the road.

[0011] Furthermore, the formula for calculating the driving stability index of the vehicle to be controlled on the road and the distance warning vehicle and the driving safety index of the vehicle to be controlled on the road is as follows: ;in, is the driving stability index of the vehicle to be controlled on the road and the distance warning vehicle, The distance between the vehicle to be controlled and the warning vehicle on the road The relative velocity difference at each time point is is the speed coefficient stored in the database, The distance between the vehicle to be controlled and the warning vehicle on the road The relative acceleration difference at each time point is is the acceleration coefficient stored in the database, and , is a natural constant, is the driving safety index of the vehicle to be controlled on the road, is the collision risk index of the vehicle to be controlled on the road after the standardized processing and the distance warning vehicle. is the risk factor stored in the database, It is the driving stability index of the vehicle to be controlled on the road after the standardized processing and the distance warning vehicle. is the stability coefficient stored in the database, , , is the number of time points.

[0012] Furthermore, the specific steps of taking corresponding control measures based on the comparison and analysis results are as follows: if the driving safety index of the vehicle to be controlled on the road is higher than the preset upper limit of the driving safety interval, the first control measure is executed; if the driving safety index of the vehicle to be controlled on the road is within the preset driving safety interval, the current vehicle speed value of the vehicle to be controlled on the road and the current distance warning vehicle speed value of the distance warning vehicle are obtained on the road for comprehensive analysis to obtain an automatic control data set of the vehicle to be controlled on the road, and the second control measure is executed based on the automatic control data set of the vehicle to be controlled on the road, the automatic control data set includes the current command acceleration value, the current command speed value, and the current command steering wheel deflection angle value; if the driving safety index of the vehicle to be controlled on the road is lower than the preset lower limit of the driving safety interval, the third control measure is executed.

[0013] Furthermore, the specific steps for obtaining the automatic control data set of the vehicle to be controlled on the road are as follows: perform difference analysis on the current control vehicle speed value of the control vehicle on the road and the current approach speed value of the distance warning vehicle to obtain the current relative speed difference between the control vehicle and the distance warning vehicle on the road; read the real-time vehicle distance value between the vehicle to be controlled and the distance warning vehicle on the road, and obtain the maximum safe distance value of the vehicle to be controlled on the road, and perform comprehensive analysis on the maximum safe distance value of the vehicle to be controlled on the road, the real-time vehicle distance value of the distance warning vehicle, and the current relative speed difference to obtain to the current command acceleration value of the vehicle to be controlled on the road; to perform difference analysis on the current control vehicle speed value and the current command acceleration value of the vehicle to be controlled on the road, to obtain the current command speed value of the vehicle to be controlled on the road; to obtain the wheelbase value of the vehicle to be controlled, and to perform comprehensive analysis with the maximum safety distance value on the road and the real-time vehicle distance value of the distance warning vehicle, to obtain the current command steering wheel deflection angle value of the vehicle to be controlled on the road; and to mark the current command acceleration value, the current command speed value, and the current command steering wheel deflection angle value of the vehicle to be controlled on the road as an automatic control data set.

[0014] A remote human-machine co-driving system comprises: a data acquisition module, an identification and analysis module, a judgment and analysis module, a first execution module, and a second execution module; the data acquisition module is used to acquire the current two-dimensional position coordinates of the vehicle to be controlled on the road and the image data of the road; the identification and analysis module is used to identify and analyze the acquired current two-dimensional position coordinates of the vehicle to be controlled on the road and the image data of the road, and obtain a number of peer vehicles of the vehicle to be controlled on the road, and a real-time vehicle distance value with each peer vehicle, wherein the road image data is specifically each of the two-dimensional position coordinates of the vehicle to be controlled on the road image. the pixel value and two-dimensional coordinates of each pixel point; the judgment and analysis module is used to judge and analyze the real-time vehicle distance value between the vehicle to be controlled and each peer vehicle on the road and the preset vehicle distance warning range respectively; the first execution module is used to continuously obtain the driving instruction data received by the vehicle to be controlled on the road several times and the actual control driving data of the vehicle in real time when the real-time vehicle distance value of each peer vehicle on the road is higher than the preset vehicle distance warning range, and perform comprehensive analysis to obtain the actual vehicle driving index received by the vehicle to be controlled on the road each time, and perform trend analysis to obtain the comprehensive driving change index of the vehicle to be controlled on the road, the instruction data is specifically the instruction value of each operation instruction, the actual control driving data of the vehicle is specifically the actual operation value of the vehicle to be controlled for each operation instruction, and the comprehensive driving change index of the vehicle to be controlled on the road is judged and analyzed with the preset comprehensive driving change index threshold, and corresponding control measures are taken based on the judgment and analysis results; the second execution module is used to, when the real-time vehicle distance value of the vehicle to be controlled on the road and each peer vehicle is in the preset vehicle distance warning range, the corresponding The traveling vehicle is marked as a distance warning vehicle, and the vehicle condition data of the vehicle to be controlled at several time points on the road and the adjacent vehicle condition data of the distance warning vehicle at several time points are obtained in real time and continuously, and a comprehensive analysis is performed to obtain the driving safety index of the vehicle to be controlled on the road. The vehicle condition data specifically include the vehicle speed value and the vehicle acceleration value, and the adjacent vehicle condition data specifically include the distance warning vehicle speed value and the distance warning vehicle acceleration value. The driving safety index of the vehicle to be controlled on the road is compared and analyzed with the preset driving safety interval, and corresponding vehicle control measures are taken based on the comparison and analysis results.

[0015] A remote human-machine co-driving device comprises: a data acquisition unit, a peer vehicle identification unit, a judgment unit, a first acquisition unit, a first analysis unit, a first measure execution unit, a vehicle marking unit, a second acquisition unit, a second analysis unit, and a second measure execution unit; the data acquisition unit is used to acquire the current two-dimensional position coordinates of the vehicle to be controlled on the road and the road image data, and the road image data is specifically the pixel value and two-dimensional coordinates of each pixel point in the road image; the peer vehicle identification unit is used to identify and analyze the acquired road image data of the vehicle to be controlled on the road, and obtain a plurality of peer vehicles of the vehicle to be controlled on the road, and the real-time vehicle status of each peer vehicle; The first acquisition unit is used to continuously acquire the travel instruction data received several times on the road by the vehicle to be controlled, and the actual control driving data of the vehicle, when the real-time vehicle distance value of each of the vehicles on the road by the vehicle to be controlled is higher than the preset vehicle distance warning range. The instruction data is specifically the instruction value of each operation instruction, and the actual control driving data of the vehicle is specifically the actual operation value of the vehicle to be controlled for each operation instruction. The first analysis unit is used to continuously acquire the travel instruction data received several times on the road by the vehicle to be controlled, and the actual control driving data of the vehicle, when the real-time vehicle distance value of each of the vehicles on the road by the vehicle to be controlled is higher than the preset vehicle distance warning range. The command data and the actual control driving data of the vehicle are comprehensively analyzed to obtain the actual driving index of the vehicle to be controlled each time on the road, and a trend analysis is performed to obtain the comprehensive driving change index of the vehicle to be controlled on the road; the first measure execution unit is used to judge and analyze the comprehensive driving change index of the vehicle to be controlled on the road with a preset comprehensive driving change index threshold, and take corresponding control measures based on the judgment and analysis results; the vehicle marking unit is used to mark the corresponding peer vehicle as a distance warning vehicle when the real-time vehicle distance value between the vehicle to be controlled and the peer vehicle on the road is within a preset vehicle distance warning range; the second acquisition unit is used to continuously acquire the vehicle to be controlled in real time. the vehicle condition data of the vehicle to be controlled at several time points on the road and the adjacent vehicle condition data at several time points of the distance warning vehicle, the vehicle condition data specifically include the vehicle speed value and the vehicle acceleration value, and the adjacent vehicle condition data specifically include the distance warning vehicle speed value and the distance warning vehicle acceleration value; the second analysis unit is used to comprehensively analyze the vehicle condition data of the vehicle to be controlled at several time points on the road and the adjacent vehicle condition data of the distance warning vehicle at several time points, so as to obtain the driving safety index of the vehicle to be controlled on the road; the second measure execution unit compares and analyzes the driving safety index of the vehicle to be controlled on the road with the preset driving safety interval, and takes corresponding vehicle control measures based on the comparison and analysis results.

[0016] The present invention has the following beneficial effects:

[0017] (1) The remote human-machine co-driving method analyzes the distance between the vehicle to be controlled and each of the accompanying vehicles in real time, and determines whether there is a potential driving risk based on a preset safe distance range. If a accompanying vehicle enters the safe range, it will automatically identify and adjust the control strategy to avoid accidents. In addition, the safety of the vehicle and the distance warning vehicle can be evaluated in real time through the calculation of the driving safety index, thereby ensuring that the vehicle is always in a safe driving state.

[0018] (2) This remote human-machine co-driving method can dynamically adjust according to the vehicle's driving data and safety by switching between automatic control and manual intervention mechanisms, avoiding the limitations of a single control method. When there is no threat from other vehicles, the vehicle can be efficiently and automatically controlled to improve driving efficiency. When driving conditions are complex or potential risks arise, it can intelligently switch to manual control to enhance driving flexibility and safety.

[0019] (3) The remote human-machine co-driving method can accurately calculate the driving safety index by comprehensively analyzing the vehicle condition data and the vehicle condition data of the nearby warning vehicle, and switch between automatic and manual control modes based on the driving safety index, thereby making the remote automatic driving more flexible and intelligent in complex environments. Through real-time monitoring and adjustment, the remote automatic driving can continuously optimize its behavior during driving, thereby ensuring a smoother, more stable and efficient driving process.

[0020] (4) The remote human-machine co-driving system, through the close collaboration of the data acquisition module, the identification and analysis module and the judgment and analysis module, can accurately identify the relative position and safe distance between the vehicle to be controlled and the other vehicles in real time, and make judgments and analyses in a timely manner, so as to make decisions quickly according to the real-time traffic conditions, thereby effectively optimizing the control measures. Under different road conditions, by real-time monitoring of the vehicle's condition data and the vehicle condition data of the distance warning vehicle, it can quickly identify potential risks and automatically adjust the control strategy, thereby improving the system's adaptability and flexibility to complex traffic scenes. At the same time, through the dynamic coordination of the first execution module and the second execution module, the system can adjust the control strategy in a timely manner under complex or emergency situations, thereby ensuring that the driving process always remains in the safest and most efficient state.

[0021] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The present invention is a flow chart of a remote human-machine co-driving method.

[0023] Figure 2 The present invention is a flowchart of the specific steps for obtaining a comprehensive driving change index of a vehicle to be controlled on a road in a remote human-machine co-driving method.

[0024] Figure 3 This is a block diagram of a remote human-machine co-driving system of the present invention. DETAILED DESCRIPTION

[0025] The embodiments of the present application solve the problem that the prior art lacks real-time analysis and dynamic response mechanism for peer vehicles through a remote human-machine co-driving system, method and device. The prior art makes it difficult to evaluate and adjust the safe distance between the vehicle and the peer vehicle in real time. When the peer vehicle approaches or there is a potential risk of collision with the vehicle, there is a lack of timely safety warning and emergency response plan. Secondly, in the prior art, the control mode conversion of the vehicle fails to take into account the comprehensive driving conditions and the actual driving index of the vehicle in dynamic environmental changes, which easily leads to the lack of flexibility and accuracy of the vehicle's response to the control mode switching in a complex environment, increasing the potential safety hazards during driving. Finally, the prior art analysis of the safety and control stability of the vehicle during driving lacks a comprehensive analysis of multi-dimensional information such as the vehicle's driving safety index, driving stability and the real-time approaching speed of the peer vehicle, thereby ignoring the safety issues arising in complex road conditions, which in turn affects its reliability and effectiveness in actual applications, and limits its applicability and safety in dynamic driving environments.

[0026] The overall idea of ​​the problem in the embodiment of this application is as follows:

[0027] First, the current two-dimensional position coordinates of the vehicle to be controlled on the road and the image data of the road are obtained, and identification and analysis are performed to obtain a number of peer vehicles on the road to be controlled and the real-time vehicle distance value with each peer vehicle. The real-time vehicle distance is compared with a preset safety distance range. If there is no peer vehicle within the preset safety distance range, the driving instructions and vehicle driving data are continuously analyzed, the comprehensive driving change index is calculated, and compared with the threshold to decide whether to automatically control or switch to manual control. If there is a peer vehicle, it is marked as an adjacent vehicle, and the driving safety index is calculated for the adjacent vehicle condition data and the vehicle's own vehicle condition data; corresponding control is performed according to the index result.

[0028] See also Figure 1, an embodiment of the present invention provides a technical solution: a remote human-machine co-driving method, comprising the following steps: obtaining the current two-dimensional position coordinates of the vehicle to be controlled on the road (based on the world two-dimensional coordinate system) and the road image data, the road image data specifically being the pixel value and two-dimensional coordinates (with the upper left corner of the image as the origin, the horizontal direction as the X axis, and the vertical direction as the Y axis) of each pixel point in the road image, and performing identification and analysis to obtain a number of peer vehicles on the road of the vehicle to be controlled, and a real-time vehicle distance value with each peer vehicle (i.e., the real-time vehicle distance value between the vehicle to be controlled and the vehicle to be controlled); the distance between the vehicle to be controlled and each peer vehicle on the road of the vehicle to be controlled is obtained; The real-time vehicle distance value is judged and analyzed with the preset vehicle distance warning range (i.e., a rectangular range with a length of 80 meters and a width of 5 meters, and the center point of the rectangular range is the center point of the vehicle to be controlled); if the real-time vehicle distance value of each fellow vehicle of the vehicle to be controlled on the road is higher than the preset vehicle distance warning range, the driving instruction data received several times by the vehicle to be controlled on the road and the actual control driving data of the vehicle (after receiving the driving instruction) are continuously obtained in real time, the instruction data is specifically the instruction value of each operation instruction, and the actual control driving data of the vehicle is specifically the actual operation value of the vehicle to be controlled for each operation instruction, and a comprehensive analysis is performed to obtain the vehicle to be controlled. The actual driving index of the vehicle to be controlled is received each time on the road, and trend analysis is performed to obtain the comprehensive driving change index of the vehicle to be controlled on the road; the comprehensive driving change index of the vehicle to be controlled on the road is judged and analyzed with the preset comprehensive driving change index threshold, and corresponding control measures are taken based on the judgment and analysis results (if the comprehensive driving change index of the vehicle to be controlled on the road is lower than the preset comprehensive driving change index threshold, automatic control continues; if the comprehensive driving change index of the vehicle to be controlled on the road is higher than or equal to the preset comprehensive driving change index threshold, manual control is switched); when there is a difference between the vehicle to be controlled and the other vehicles on the road When the real-time vehicle distance value of a vehicle is within a preset vehicle distance warning range, the corresponding traveling vehicle will be marked as a distance warning vehicle, and the vehicle condition data of the vehicle to be controlled at several time points on the road and the nearby vehicle condition data of the distance warning vehicle at several time points will be continuously acquired in real time. The vehicle condition data specifically include the vehicle speed value and the vehicle acceleration value, and the nearby vehicle condition data specifically include the distance warning vehicle speed value and the distance warning vehicle acceleration value. A comprehensive analysis will be performed to obtain the driving safety index of the vehicle to be controlled on the road; the driving safety index of the vehicle to be controlled on the road will be compared and analyzed with the preset driving safety interval, and corresponding vehicle control measures will be taken based on the comparison and analysis results.

[0029] Among them, the instruction value of each operation instruction represents the specific numerical values ​​of various instructions issued by the remote operator (including but not limited to vehicle speed value, vehicle acceleration value, vehicle steering wheel deflection angle value).

[0030] The actual value of the command corresponding to each operation command represents the actual driving data of the vehicle's automatic driving after the remote operator issues various commands (including but not limited to the vehicle's actual driving speed value, the vehicle's actual driving acceleration value, and the vehicle's actual driving steering wheel deflection angle value).

[0031] Specifically, the specific steps for obtaining the real-time vehicle distance value between the vehicle to be controlled and each of the peer vehicles on the road are as follows: the pixel value and the two-dimensional coordinates of each pixel point in the road image of the vehicle to be controlled on the road are respectively input into the target recognition model to obtain the two-dimensional position coordinates of several peer vehicles on the road of the vehicle to be controlled; and the two-dimensional position coordinates of each peer vehicle on the road of the vehicle to be controlled are converted into the world coordinate system to obtain the world two-dimensional position coordinates of each peer vehicle on the road of the vehicle to be controlled; and the current two-dimensional position coordinates of the vehicle to be controlled on the road and the world two-dimensional position coordinates of each peer vehicle on the road are respectively analyzed to obtain the real-time vehicle distance value between the vehicle to be controlled and each peer vehicle on the road.

[0032] The target recognition model is specifically a YOLO model, and the specific process of the two-dimensional position coordinates of several peer vehicles on the road to be controlled is as follows:

[0033] Extract features from the pixel value and two-dimensional coordinates of each pixel point in the road image of the vehicle to be controlled on the road input by the YOLO model to obtain a feature data set of the vehicle to be controlled on the road (including but not limited to texture features, color features, edge features, and corner point features);

[0034] Generate a road feature map (i.e., a high-dimensional matrix, where each position of the feature map represents a high-level feature of a specific area in the image) based on the feature data set of the vehicle to be controlled on the road.

[0035] Combined with the fully connected layer in the YOLO model, the road feature map of the vehicle to be controlled on the road is gridded to obtain several feature grids of the vehicle to be controlled on the road (that is, the area corresponding to each grid is calculated according to the size of the input image and the size of the road feature map), and prediction analysis is performed to obtain the boundary two-dimensional position coordinates of several peer vehicle prediction bounding boxes of each feature grid of the vehicle to be controlled on the road (with the upper left corner of the image as the origin, the horizontal direction as the X axis, and the vertical direction as the Y axis) and the corresponding confidence levels;

[0036] A suppression analysis is performed on the predicted bounding box of each fellow vehicle of each feature grid of the vehicle to be controlled on the road and the corresponding confidence score to obtain the predicted bounding boxes of several fellow vehicles of the vehicle to be controlled on the road, that is, each fellow vehicle of the vehicle to be controlled on the road, and the boundary two-dimensional position coordinates of the predicted bounding box of each fellow vehicle of each feature grid of the vehicle to be controlled on the road are marked as the two-dimensional position coordinates of each fellow vehicle of the vehicle to be controlled on the road (with the upper left corner of the image as the origin, the horizontal direction as the X-axis, and the vertical direction as the Y-axis).

[0037] It needs to be explained that the YOLO model includes input layer, convolution layer, pooling layer, fully connected layer, grid division layer, output layer, loss function and non-maximum suppression.

[0038] The input layer is used to accept image data. The input image will be resized to a fixed size to ensure that the image meets the input requirements of the model.

[0039] The convolution layer is used to extract features from the input image. By using the convolution kernel (filter) to scan each local area of ​​the image, the convolution layer can identify low-level features in the image (such as edges, colors, textures, shapes, etc.). As the depth of the network increases, the convolution layer can learn more and more complex and abstract features (such as the contours and textures of objects).

[0040] The pooling layer is used to reduce the dimension and compress the features generated by the convolutional layer while maintaining important features. A common pooling method is maximum pooling, which selects the maximum value of each local area, reduces overfitting, and retains the most significant features.

[0041] The fully connected layer is used to map the features obtained from the convolution and pooling layers to the final target output space by flattening the road feature map into a one-dimensional vector.

[0042] The grid division layer is used to divide the road feature map into multiple road grids. Each road grid also predicts the target category in the area and outputs the bounding box of the target, including the coordinates of the bounding box. In addition, each grid also predicts the confidence of the target (the probability of whether the target exists).

[0043] The output layer is used to output the final result of target detection based on the calculation results of the network, that is, the bounding box coordinates, target category probability, and confidence of each grid prediction.

[0044] The loss function is used to calculate the error between the model's predicted value and the true value, and to update the model's weights through the back-propagation algorithm.

[0045] Non-maximum suppression is used to retain the box with the highest score and suppress the boxes with high overlap based on the confidence score of each bounding box. By setting a threshold, the boxes with overlap greater than the threshold are removed and the optimal box, that is, the bounding box of the same vehicle, is retained.

[0046] The world coordinate system conversion processing is the world two-dimensional plane conversion processing, and the specific process is: obtain the intrinsic and extrinsic parameters of the shooting device, and based on the intrinsic and extrinsic parameters of the shooting device, calculate the camera's posture (through the rotation matrix RRR) and position (through the translation vector TTT), use the intrinsic and extrinsic parameters of the shooting device to convert the image coordinates into world coordinates, and the world coordinates here are three-dimensional coordinates, but in this implementation example, all vehicles are in the same horizontal plane, so the height value is not considered, so the height value defaults to 0, that is, the world coordinates are a two-dimensional coordinate system.

[0047] In this implementation, the YOLO model is used for target detection, so that the vehicle to be controlled can accurately identify and track the position of the vehicle on the road. The YOLO model uses an end-to-end convolutional neural network to quickly and accurately locate peer vehicles while processing images. It also extracts features from the convolutional layer and pooling layer and further analyzes the fully connected layer, so that it still has a high detection accuracy in complex road environments, which enables the vehicle system to respond promptly in a dynamic traffic environment to ensure driving safety. Secondly, by analyzing each pixel and converting it into the two-dimensional position coordinates of the vehicle, and then converting it through the world coordinate system, it can accurately calculate the relative position of each peer vehicle and the vehicle to be controlled, and update the position and safety distance of the peer vehicle in real time, thereby enhancing the perception of the traffic flow of the vehicle to be controlled, helping the autonomous driving system to make intelligent decisions, optimize the driving route, and avoid collision risks. Finally, the YOLO model detects and locates peer vehicles, combined with the non-maximum suppression (NMS) algorithm, effectively filtering out redundant target predictions, ensuring that the model only retains accurate peer vehicle information, thereby effectively reducing false positives and false negatives, and then being able to more accurately determine which vehicles need attention, which helps to derive the best driving strategy based on real-time data analysis, such as whether to accelerate, decelerate or change lanes, thereby improving driving safety.

[0048] Specifically, Figure 2As shown, the specific steps for obtaining the comprehensive driving change index of the vehicle to be controlled on the road are as follows: comprehensively analyze the command value of each operation command received by the vehicle to be controlled on the road in combination with the actual operation value of the vehicle for each operation command, and obtain the command deviation index of each operation command received by the vehicle to be controlled on the road; and comprehensively analyze the command deviation index of each operation command received by the vehicle to be controlled on the road in order to obtain the actual driving index of the vehicle received on the road; and perform trend analysis on the actual driving index of the vehicle received each time by the vehicle to be controlled on the road in order to obtain the change rate of several groups of actual driving indexes of the vehicle to be controlled on the road, and perform comprehensive analysis to obtain the comprehensive driving change index of the vehicle to be controlled on the road.

[0049] The formula for calculating the command deviation index and the vehicle actual driving index received each time by the vehicle to be controlled on the road is as follows: ;in, The first The first received The instruction deviation index of the operation instruction, The first The first received The instruction value of the operation instruction, The first The first received The actual running value of the operation instruction. The first The actual driving index of the vehicle received, The first The weighting coefficient of the received , is the number of receptions, , The number of operation instructions.

[0050] It needs to be explained that The specific acquisition process is: read the first The command deviation index of each operation command received for the first time is obtained, and the mean value is analyzed to obtain the first The average value of the command deviation index received for the first time is summed and analyzed to obtain the vehicle condition and value, and the vehicle to be controlled is The average value of the command deviation index received times is analyzed with the vehicle condition and value respectively, and the analysis result is used as the corresponding coefficient.

[0051] The formula for calculating the actual driving index change rate and the comprehensive driving change index of several groups of vehicles to be controlled on the road is as follows: ;in, The first The actual driving index change rate of the group of vehicles, is a natural constant, which is taken as 2.71 in this embodiment. The first The actual driving index of the vehicle received, The first The actual driving index of the vehicle received, is the comprehensive driving change index of the vehicle to be controlled on the road, The first The weighted coefficient of the actual driving index change rate of the group of vehicles, The first The actual driving index change rate of the group of vehicles, The first The weighted coefficient of the actual driving index change rate of the group of vehicles, and , , is the number of receptions, ,and , is the number of groups of the actual vehicle driving index change rate, .

[0052] It needs to be explained that , The specific acquisition process is: , The absolute value of the actual driving index change rate of the group of vehicles is summed and analyzed to obtain the The actual driving index change rate and value of the group vehicle, and the , The absolute values ​​of the actual driving index change rates of the vehicles in the first The actual driving index change rate and value of the group of vehicles are analyzed for proportion, and the proportion analysis result is the corresponding weighting coefficient.

[0053] A specific implementation example of calculating the comprehensive driving change index of the vehicle to be controlled on the road is as follows. The existing data is as follows: the vehicle to be controlled receives a total of 5 command operations on the road, and the actual vehicle driving indexes received 5 times on the road by the vehicle to be controlled are: 0.34, 0.25, 0.21, 0.31, and 0.35 respectively.

[0054] Substituting the above data into the formula of the actual driving index change rate and comprehensive driving change index of several groups of vehicles to be controlled on the road, we get:

[0055] Then the actual driving index change rate of the first group of vehicles to be controlled on the road = 2.71 （（0.25 -0.34） / 0.34） ≈0.77.

[0056] The actual driving index change rate of the second group of vehicles to be controlled on the road = 2.71 （（0.21-0.25） / 0.25） ≈0.85.

[0057] The actual driving index change rate of the third group of vehicles to be controlled on the road = 2.71 （（0.31-0.21） / 0.21） ≈1.60.

[0058] The actual driving index change rate of the fourth group of vehicles to be controlled on the road = 2.71 （（0.34-0.31） / 0.31） ≈1.10.

[0059] The comprehensive driving change index of the vehicle to be controlled on the road is ln(1+((1 / 3)*((0.47*0.77+0.53*0.85)+(0.35*0.85+0.65*1.60)+(0.59*1.60+0.41*1.10)))≈0.78.

[0060] In this implementation scheme, by comprehensively analyzing the deviation of each operation instruction and the changes in the actual driving state of the vehicle, the driving changes of the vehicle to be controlled are evaluated in real time and accurate feedback is given. The deviation index of each instruction and the actual driving index combined with trend analysis can effectively reflect the deviation of the vehicle during driving, so as to judge whether the control system has accurately executed the instruction and whether it needs to be adjusted. By calculating the rate of change of the driving index, the control strategy can be dynamically monitored and adjusted, thereby improving the system's responsiveness to changes in vehicle status. Secondly, this step adopts technologies such as weighted coefficients and trend analysis, so as to flexibly adjust the comprehensive driving change index according to the actual impact of different reception times and operation instructions. Through weighted analysis, the relative importance of various changes is dynamically balanced, so that the control system can be adaptively adjusted when facing different driving scenarios. Finally, this method can compare and analyze multiple instruction deviations and actual driving states, and optimize the vehicle's driving control strategy in real time through the calculation of trend changes and the formation of a comprehensive driving change index.

[0061] Specifically, the specific steps for obtaining the driving safety index of the vehicle to be controlled on the road are as follows: perform difference analysis on the vehicle speed value and vehicle acceleration value of the vehicle to be controlled at each time point on the road, and the distance warning vehicle speed value and distance warning vehicle acceleration value at each time point of the distance warning vehicle, and obtain the relative speed difference and relative acceleration difference between the vehicle to be controlled and the distance warning vehicle at each time point on the road, and perform comprehensive analysis to obtain the driving stability index of the vehicle to be controlled and the distance warning vehicle on the road; perform mean analysis on the relative speed difference and relative acceleration difference of the vehicle to be controlled and the distance warning vehicle at each time point on the road, and obtain the mean value of the relative speed difference and relative acceleration difference between the vehicle to be controlled and the distance warning vehicle on the road; read the relative speed difference and relative acceleration difference of the vehicle to be controlled on the road The current two-dimensional position coordinates on the vehicle and the two-dimensional position coordinates of the distance warning vehicle are comprehensively analyzed to obtain the relative angle between the vehicle to be controlled and the distance warning vehicle on the road; the real-time vehicle distance value between the vehicle to be controlled and the distance warning vehicle on the road is read, and the relative angle, the average relative speed difference, and the average relative acceleration difference between the vehicle to be controlled and the distance warning vehicle on the road are comprehensively analyzed to obtain the collision risk index between the vehicle to be controlled and the distance warning vehicle on the road; the collision risk index and the driving stability index of the vehicle to be controlled and the distance warning vehicle on the road are standardized (i.e., de-unitized); the collision risk index and the driving stability index of the vehicle to be controlled and the distance warning vehicle on the road after the standardized processing are comprehensively analyzed to obtain the driving safety index of the vehicle to be controlled on the road.

[0062] The current two-dimensional position coordinates include a current vertical two-dimensional coordinate value and a current horizontal two-dimensional coordinate value, and the two-dimensional position coordinates include a vertical two-dimensional coordinate value and a horizontal two-dimensional coordinate value.

[0063] The formula for calculating the relative angle between the vehicle to be controlled and each distance warning vehicle on the road is as follows: ;in, is the relative angle between the vehicle to be controlled and the warning vehicle on the road, is the vertical two-dimensional coordinate value of the warning vehicle, is the current vertical two-dimensional coordinate value of the vehicle to be controlled, is the horizontal two-dimensional coordinate value of the warning vehicle, is the current horizontal two-dimensional coordinate value of the vehicle to be controlled.

[0064] The discriminant formula of the collision risk index of the vehicle to be controlled on the road and each distance warning vehicle is as follows: ;in, is the collision risk index between the vehicle to be controlled and the distance warning vehicle on the road, is the average relative speed difference between the vehicle to be controlled and the warning vehicle on the road, is the relative angle between the vehicle to be controlled and the warning vehicle on the road, is the real-time vehicle distance between the vehicle to be controlled and the distance warning vehicle on the road. It is the average value of the relative acceleration difference between the vehicle to be controlled and the distance warning vehicle on the road.

[0065] The formula for calculating the driving stability index of the vehicle to be controlled on the road and the distance warning vehicle and the driving safety index of the vehicle to be controlled on the road is as follows: ;in, is the driving stability index of the vehicle to be controlled on the road and the distance warning vehicle, The distance between the vehicle to be controlled and the warning vehicle on the road The relative velocity difference at each time point is is the speed coefficient stored in the database, The distance between the vehicle to be controlled and the warning vehicle on the road The relative acceleration difference at each time point is is the acceleration coefficient stored in the database, and , is a natural constant, which is taken as 2.71 in this embodiment. is the driving safety index of the vehicle to be controlled on the road, is the collision risk index of the vehicle to be controlled on the road after the standardized processing and the distance warning vehicle. is the risk factor stored in the database, It is the driving stability index of the vehicle to be controlled on the road after the standardized processing and the distance warning vehicle. is the stability coefficient stored in the database, , , is the number of time points.

[0066] It needs to be explained that , The specific acquisition process is as follows: read the mean value of the relative speed difference and the mean value of the relative acceleration difference between the vehicle to be controlled and the distance warning vehicle on the road, and perform sum analysis to obtain a stable sum value, and perform proportion analysis on the mean value of the relative speed difference and the mean value of the relative acceleration difference between the vehicle to be controlled and the distance warning vehicle on the road, respectively, and use the analysis results as the corresponding coefficients.

[0067] , The specific acquisition process is: read the collision risk index and driving stability index of the vehicle to be controlled on the road with the distance warning vehicle after the standardized processing, and perform sum analysis to obtain the safety sum value, and perform proportion analysis on the collision risk index and driving stability index of the vehicle to be controlled on the road with the distance warning vehicle after the standardized processing and the safety sum value, and use the analysis results as the corresponding coefficients.

[0068] In this implementation scheme, by comprehensively analyzing multi-dimensional data such as the relative speed difference, relative acceleration difference, real-time vehicle distance value and relative angle between the vehicle to be controlled and the distance warning vehicle, it is possible to comprehensively evaluate the potential risks in the driving process of the vehicle to be controlled. The calculation of the relative speed difference, relative acceleration difference and collision risk index enables the system to accurately capture the relative movement with the distance warning vehicle, so as to better predict possible collision risks or unstable factors, optimize the vehicle driving strategy in real time, and thus improve driving safety. Secondly, by combining the analysis of the relative speed difference and relative acceleration difference of the vehicles, as well as the collision risk index and driving stability index obtained by further standardization processing, it is possible to more scientifically The collision risk is assessed accurately. The combined analysis of relative angle and real-time distance effectively reduces the interference of external factors on risk assessment, making the collision risk prediction more accurate and avoiding inefficient processing in complex driving scenarios. In addition, the weighted coefficient is used for comprehensive analysis, so that the risk factor and stability coefficient of each distance warning vehicle are dynamically adjusted in the calculation, so that the vehicle can adjust the risk assessment results according to different traffic conditions (such as vehicle speed, road conditions), so as to ensure that an accurate driving safety index can still be provided in a complex traffic environment. Finally, the comprehensive calculation of the driving safety index enables the system to monitor the potential risks in the driving process in real time, provide decision support for the vehicle to be controlled, and improve safety at the same time.

[0069] Specifically, the specific steps of taking corresponding control measures based on the comparison and analysis results are as follows: if the driving safety index of the vehicle to be controlled on the road is higher than the preset upper limit of the driving safety interval, the first control measure is executed (i.e., continue driving based on the received operation instruction); if the driving safety index of the vehicle to be controlled on the road is within the preset driving safety interval, the current vehicle speed value of the vehicle to be controlled on the road and the current distance warning vehicle speed value of the distance warning vehicle are obtained for comprehensive analysis to obtain an automatic control data set of the vehicle to be controlled on the road, and the second control measure (i.e., auxiliary control to prevent collision with the distance warning vehicle) is executed based on the automatic control data set of the vehicle to be controlled on the road, and the automatic control data set includes the current command acceleration value, the current command speed value, and the current command steering wheel deflection angle value; if the driving safety index of the vehicle to be controlled on the road is lower than the preset lower limit of the driving safety interval, the third control measure (i.e., switching to manual control) is executed.

[0070] In this implementation scheme, automated control measures are used to ensure the safety of the controlled vehicle during driving, reducing the need for manual intervention and improving the intelligence level of the driving experience. For example, when the driving safety index exceeds the preset range, the corresponding control measures are automatically triggered, and the system can make independent judgments and take corresponding actions. By comprehensively analyzing the current speed of the controlled vehicle and the distance from the warning vehicle, the system can adjust the control data set (such as acceleration, speed, and steering wheel angle, etc.) in real time, so as to flexibly respond to different traffic conditions. For potential dangers that are about to occur, the system can respond quickly to avoid collisions, thereby enhancing the real-time safety of driving. The system takes different control measures according to different driving safety index ranges: the first control measure: when the vehicle's driving safety index is high, continue driving to ensure the smooth operation of the vehicle; the second control measure: when When the safety index is within the normal range, auxiliary control is used to avoid collisions with other vehicles, which not only ensures the driving safety of the vehicle, but also avoids excessive intervention and the third control measure: when the driving safety index is lower than the lower limit, it switches to manual control to ensure that the staff can intervene and take action immediately. This plays the last role in safety protection. By real-time monitoring of the driving safety index and analyzing it based on multi-dimensional data such as vehicle status, speed, and distance, it can effectively identify potential dangers and take corresponding measures in time, which helps to reduce the risk of traffic accidents, especially in dense or complex road environments. Different control measures provide the system with flexible adjustment space and can adapt to different road conditions and driving environments. The system can dynamically adjust the control strategy based on real-time data, further improving the adaptability and intelligence of the system.

[0071] Specifically, the specific steps for obtaining the automatic control data set of the vehicle to be controlled on the road are as follows: perform difference analysis on the current control vehicle speed value of the control vehicle on the road and the current approach speed value of the distance warning vehicle to obtain the current relative speed difference between the control vehicle and the distance warning vehicle on the road; read the real-time vehicle distance value between the vehicle to be controlled and the distance warning vehicle on the road, and obtain the maximum safe distance value of the vehicle to be controlled on the road, and perform comprehensive analysis on the maximum safe distance value of the vehicle to be controlled on the road, the real-time vehicle distance value of the distance warning vehicle, and the current relative speed difference to obtain The current command acceleration value of the vehicle to be controlled on the road; performing difference analysis on the current control vehicle speed value and the current command acceleration value of the vehicle to be controlled on the road to obtain the current command speed value of the vehicle to be controlled on the road; obtaining the wheelbase value of the vehicle to be controlled, and performing comprehensive analysis with the maximum safety distance value on the road and the real-time vehicle distance value of the distance warning vehicle to obtain the current command steering wheel deflection angle value of the vehicle to be controlled on the road; and marking the current command acceleration value, current command speed value, and current command steering wheel deflection angle value of the vehicle to be controlled on the road as an automatic control data set.

[0072] The formula for calculating the current command acceleration value, current command speed value, and current command steering wheel deflection angle value of the vehicle to be controlled on the road is as follows: ;in, is the current command acceleration value of the vehicle to be controlled on the road, is the current relative speed difference between the vehicle to be controlled and the distance warning vehicle on the road, is the real-time vehicle distance between the vehicle to be controlled and the distance warning vehicle on the road. is the maximum safe distance value of the vehicle to be controlled on the road, is the current command speed value of the vehicle to be controlled on the road, is the current control speed value of the vehicle to be controlled on the road, is the current command steering wheel deflection angle value of the vehicle to be controlled on the road, is the proportional coefficient stored in the database, which is 2 in this implementation example. is the wheelbase value of the vehicle to be controlled.

[0073] In this implementation, the command acceleration, command speed and steering wheel deflection angle are calculated by combining information such as the current control speed, relative speed difference, vehicle distance and safety distance, so that the vehicle's driving state can be accurately controlled to ensure the stability and safety of the vehicle in complex traffic environments. Secondly, by automatically calculating the control instructions (acceleration, speed, steering angle), the intervention of staff is reduced, so that the vehicle can make adaptive adjustments autonomously in most cases, thereby improving the reliability of automatic driving. In addition, the acceleration and speed are intelligently adjusted through the relative speed difference and the real-time distance value to ensure that the vehicle's acceleration and deceleration process is smooth and comfortable, avoiding sudden acceleration or sudden braking, thereby enhancing safety. Finally, this method not only ensures a safe distance between the vehicle to be controlled and the approaching vehicle, but also intelligently optimizes the vehicle speed and steering angle to avoid frequent sudden braking and acceleration, thereby improving the smoothness of the entire road traffic and reducing traffic congestion.

[0074] See also Figure 3The embodiment of the present invention provides a technical solution: a remote human-machine co-driving system, comprising: a data acquisition module, an identification and analysis module, a judgment and analysis module, a first execution module, and a second execution module; the data acquisition module is used to obtain the current two-dimensional position coordinates of the vehicle to be controlled on the road and the image data of the road; the identification and analysis module is used to identify and analyze the current two-dimensional position coordinates of the vehicle to be controlled on the road and the image data of the road, and obtain a number of peer vehicles of the vehicle to be controlled on the road, and a real-time vehicle distance value with each peer vehicle, and the road image data has The first execution module is used to continuously acquire the driving instruction data received several times on the road by the vehicle to be controlled, as well as the actual control driving data of the vehicle, and perform comprehensive analysis to obtain the driving instruction data received each time on the road by the vehicle to be controlled, when the real-time vehicle distance value of each of the vehicles to be controlled on the road is higher than the preset vehicle distance warning range. The actual driving index of the vehicle to be controlled is obtained by performing trend analysis to obtain the comprehensive driving change index of the vehicle to be controlled on the road, the command data is specifically the command value of each operation command, and the actual vehicle control driving data is specifically the actual operation value of the vehicle to be controlled for each operation command. The comprehensive driving change index of the vehicle to be controlled on the road is judged and analyzed with the preset comprehensive driving change index threshold, and corresponding control measures are taken based on the judgment and analysis results; the second execution module is used for, when there is a real-time vehicle distance value between the vehicle to be controlled and the other vehicle on the road is within the preset vehicle distance warning range, The corresponding traveling vehicle is marked as a distance warning vehicle, and the vehicle condition data of the vehicle to be controlled at several time points on the road and the nearby vehicle condition data of the distance warning vehicle at several time points are obtained continuously in real time, and a comprehensive analysis is performed to obtain the driving safety index of the vehicle to be controlled on the road. The vehicle condition data specifically include the vehicle speed value and the vehicle acceleration value, and the nearby vehicle condition data specifically include the distance warning vehicle speed value and the distance warning vehicle acceleration value. The driving safety index of the vehicle to be controlled on the road is compared and analyzed with the preset driving safety interval, and corresponding vehicle control measures are taken based on the comparison and analysis results.

[0075] The embodiment of the present invention provides a technical solution: a remote human-machine co-driving device, comprising: a data acquisition unit, a peer vehicle identification unit, a judgment unit, a first acquisition unit, a first analysis unit, a first measure execution unit, a vehicle marking unit, a second acquisition unit, a second analysis unit, and a second measure execution unit; the data acquisition unit is used to acquire the current two-dimensional position coordinates of the vehicle to be controlled on the road and the road image data, and the road image data is specifically the pixel value and two-dimensional coordinates of each pixel point in the road image; the peer vehicle identification unit is used to identify and analyze the acquired road image data of the vehicle to be controlled on the road, and obtain a number of peer vehicles and the vehicle to be controlled on the road. a real-time vehicle distance value with each fellow vehicle; a judgment unit, used to judge and analyze the real-time vehicle distance value of the vehicle to be controlled on the road with each fellow vehicle and the preset vehicle distance warning range; a first acquisition unit, used to continuously acquire the travel instruction data received by the vehicle to be controlled on the road several times and the actual control driving data of the vehicle in real time when the real-time vehicle distance value of each fellow vehicle on the road is higher than the preset vehicle distance warning range, the instruction data is specifically the instruction value of each operation instruction, and the actual control driving data of the vehicle is specifically the actual operation value of the vehicle to be controlled for each operation instruction; a first analysis unit, used to continuously acquire the real-time vehicle distance value of the vehicle to be controlled on the road several times and the actual control driving data of the vehicle in real time when the real-time vehicle distance value of each fellow vehicle on the road is higher than the preset vehicle distance warning range. The received driving instruction data and the actual control driving data of the vehicle are comprehensively analyzed to obtain the actual driving index of the vehicle to be controlled each time on the road, and trend analysis is performed to obtain the comprehensive driving change index of the vehicle to be controlled on the road; the first measure execution unit is used to judge and analyze the comprehensive driving change index of the vehicle to be controlled on the road with a preset comprehensive driving change index threshold, and take corresponding control measures based on the judgment and analysis results; the vehicle marking unit is used to mark the corresponding vehicle as a distance warning vehicle when the real-time vehicle distance value between the vehicle to be controlled and the vehicle to be controlled on the road is within the preset vehicle distance warning range; the second acquisition unit is used to continuously acquire the vehicle to be controlled on the road in real time. the vehicle condition data of the vehicle to be controlled at several time points on the road and the adjacent vehicle condition data of the distance warning vehicle at several time points, the vehicle condition data specifically include the vehicle speed value and the vehicle acceleration value, and the adjacent vehicle condition data specifically include the distance warning vehicle speed value and the distance warning vehicle acceleration value; a second analysis unit, used to comprehensively analyze the vehicle condition data of the vehicle to be controlled at several time points on the road and the adjacent vehicle condition data of the distance warning vehicle at several time points, so as to obtain the driving safety index of the vehicle to be controlled on the road; a second measure execution unit, used to compare and analyze the driving safety index of the vehicle to be controlled on the road with the preset driving safety interval, and take corresponding vehicle control measures based on the comparison and analysis results.

[0076] In summary, this application has at least the following effects:

[0077] By analyzing the distance between the vehicle to be controlled and each of the accompanying vehicles in real time, and judging whether there is a potential driving risk based on the preset safe distance range, if a accompanying vehicle enters the safe range, it will automatically identify and adjust the control strategy to avoid accidents. In addition, through the calculation of the driving safety index, the safety of the vehicle and the distance warning vehicle can be evaluated in real time, thereby ensuring that the vehicle is always in a safe driving state.

[0078] By switching between automatic control and manual intervention mechanisms, it can dynamically adjust according to the vehicle's driving data and safety, avoiding the limitations of a single control method. When there is no threat from other vehicles, the vehicle can be efficiently and automatically controlled to improve driving efficiency. When driving conditions are complex or potential risks arise, it intelligently switches to manual control to enhance driving flexibility and safety.

[0079] By comprehensively analyzing the vehicle condition data and the nearby vehicle condition data of the distance warning vehicle, the driving safety index can be accurately calculated, and the automatic or manual control mode can be switched based on the driving safety index, making remote autonomous driving more flexible and intelligent in complex environments. Through real-time monitoring and adjustment, remote autonomous driving can continuously optimize its behavior during driving, thereby ensuring a smoother, more stable and efficient driving process.

[0080] Through the close collaboration of the data acquisition module, the identification and analysis module, and the judgment and analysis module, the relative position and safe distance between the vehicle to be controlled and the other vehicles can be identified in real time and accurately, and judgment and analysis can be made in a timely manner, so as to make decisions quickly according to the real-time traffic conditions, thereby effectively optimizing the control measures. In different road environments, by real-time monitoring of the vehicle's condition data and the proximity data of the distance warning vehicle, potential risks can be quickly identified and the control strategy can be automatically adjusted, thereby improving the system's adaptability and flexibility to complex traffic scenarios. At the same time, through the dynamic coordination of the first execution module and the second execution module, the system can adjust the control strategy in a timely manner in complex or emergency situations, thereby ensuring that the driving process always remains in the safest and most efficient state.

[0081] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0082] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A remote human-machine co-driving method, characterized in that: The following steps are involved: The current two-dimensional position coordinates of the vehicle to be controlled on the road and the image data of the road are obtained, and identification and analysis are performed to obtain a number of fellow vehicles on the road to be controlled and a real-time vehicle distance value to each fellow vehicle, wherein the road image data specifically includes the pixel value and two-dimensional coordinates of each pixel point in the road image; The real-time vehicle distance value between the vehicle to be controlled and each other vehicle on the road is judged and analyzed with the preset vehicle distance warning range; If the real-time vehicle distance value of each fellow vehicle of the to-be-controlled vehicle on the road is higher than the preset vehicle distance warning range, the driving instruction data received by the to-be-controlled vehicle on the road for several times and the vehicle actual control driving data are continuously acquired in real time, and a comprehensive analysis is performed to obtain the vehicle actual driving index received by the to-be-controlled vehicle on the road for each time, and a trend analysis is performed to obtain the comprehensive driving change index of the to-be-controlled vehicle on the road, wherein the instruction data is specifically the instruction value of each operation instruction, and the vehicle actual control driving data is specifically the actual operation value of the to-be-controlled vehicle for each operation instruction; The comprehensive driving change index of the vehicle to be controlled on the road is judged and analyzed with a preset comprehensive driving change index threshold, and corresponding control measures are taken based on the judgment and analysis results; When the real-time vehicle distance value between the to-be-controlled vehicle and the traveling vehicle on the road is within the preset vehicle distance warning range, the corresponding traveling vehicle is marked as a distance warning vehicle, and the vehicle condition data of the to-be-controlled vehicle at several time points on the road and the adjacent vehicle condition data of the distance warning vehicle at several time points are continuously acquired in real time, and a comprehensive analysis is performed to obtain the driving safety index of the to-be-controlled vehicle on the road, wherein the vehicle condition data specifically includes the vehicle speed value and the vehicle acceleration value, and the adjacent vehicle condition data specifically includes the distance warning vehicle speed value and the distance warning vehicle acceleration value; Compare and analyze the driving safety index of the vehicle to be controlled on the road with the preset driving safety interval, and take corresponding vehicle control measures based on the comparison and analysis results; The specific steps of obtaining the comprehensive driving change index of the vehicle to be controlled on the road are as follows: Comprehensively analyzing the command value of each operation command received by the vehicle to be controlled on the road and the actual operation value of the vehicle to be controlled for each operation command, to obtain the command deviation index of each operation command received by the vehicle to be controlled on the road; And comprehensively analyzing the instruction deviation index of each operation instruction received by the controlled vehicle on the road, to obtain the vehicle actual driving index received by the controlled vehicle on the road; Performing trend analysis on each received actual driving index of the vehicle to be controlled on the road, obtaining a plurality of groups of vehicle actual driving index change rates of the vehicle to be controlled on the road, and performing comprehensive analysis to obtain a comprehensive driving change index of the vehicle to be controlled on the road; The formula for calculating the actual driving index change rate and the comprehensive driving change index of several groups of vehicles to be controlled on the road is as follows: ; in, The first The actual driving index change rate of the group of vehicles, is a natural constant, The first The actual driving index of the vehicle received, The first The actual driving index of the vehicle received, is the comprehensive driving change index of the vehicle to be controlled on the road, The first The weighted coefficient of the actual driving index change rate of the group of vehicles, The first The actual driving index change rate of the group of vehicles, The first The weighted coefficient of the actual driving index change rate of the group of vehicles, and , , is the number of receptions, ,and , is the number of groups of the actual vehicle driving index change rate; The specific steps for obtaining the driving safety index of the vehicle to be controlled on the road are as follows: Perform difference analysis on the vehicle speed value and vehicle acceleration value of the to-be-controlled vehicle at each time point on the road, and the distance warning vehicle speed value and distance warning vehicle acceleration value at each time point of the distance warning vehicle, respectively, to obtain the relative speed difference and relative acceleration difference of the to-be-controlled vehicle and the distance warning vehicle at each time point on the road, and perform comprehensive analysis to obtain the driving stability index of the to-be-controlled vehicle and the distance warning vehicle on the road; The relative speed difference and relative acceleration difference of the controlled vehicle on the road at each time point with the distance warning vehicle are respectively analyzed to obtain the mean value of the relative speed difference and relative acceleration difference between the controlled vehicle and the distance warning vehicle on the road; Read the current two-dimensional position coordinates of the vehicle to be controlled on the road and the two-dimensional position coordinates of the distance warning vehicle for comprehensive analysis to obtain the relative angle between the vehicle to be controlled and the distance warning vehicle on the road; The real-time vehicle distance value between the vehicle to be controlled and the distance warning vehicle on the road is read, and the relative angle, relative speed difference average, and relative acceleration difference average of the vehicle to be controlled and the distance warning vehicle on the road are combined for comprehensive analysis to obtain the collision risk index between the vehicle to be controlled and the distance warning vehicle on the road; Standardize the collision risk index and driving stability index of the controlled vehicle on the road with the distance warning vehicle; Comprehensively analyze the collision risk index and driving stability index of the vehicle to be controlled on the road after the standardized processing and the distance warning vehicle, and obtain the driving safety index of the vehicle to be controlled on the road; The formula for calculating the driving stability index of the vehicle to be controlled on the road and the distance warning vehicle and the driving safety index of the vehicle to be controlled on the road is as follows: ； in, is the driving stability index of the vehicle to be controlled on the road and the distance warning vehicle, The distance between the vehicle to be controlled and the warning vehicle on the road The relative velocity difference at each time point is is the speed coefficient stored in the database, The distance between the vehicle to be controlled and the warning vehicle on the road The relative acceleration difference at each time point is is the acceleration coefficient stored in the database, and , is a natural constant, is the driving safety index of the vehicle to be controlled on the road, is the collision risk index of the vehicle to be controlled on the road after the standardized processing and the distance warning vehicle. is the risk factor stored in the database, It is the driving stability index of the vehicle to be controlled on the road after the standardized processing and the distance warning vehicle. is the stability coefficient stored in the database, , , is the number of time points.

2. The remote human-machine co-driving method according to claim 1, characterized in that: The specific steps for obtaining the real-time vehicle distance value between the vehicle to be controlled and each other vehicle on the road are as follows: The pixel value and two-dimensional coordinates of each pixel point in the road image of the vehicle to be controlled on the road are respectively input into the target recognition model to obtain the two-dimensional position coordinates of several vehicles traveling with the vehicle to be controlled on the road; and performing a world coordinate system conversion process on the two-dimensional position coordinates of each of the vehicles traveling along the road to be controlled, so as to obtain the world two-dimensional position coordinates of each of the vehicles traveling along the road to be controlled; The current two-dimensional position coordinates of the vehicle to be controlled on the road and the world two-dimensional position coordinates of each fellow vehicle on the road are analyzed respectively to obtain a real-time vehicle distance value between the vehicle to be controlled and each fellow vehicle on the road.

3. The remote human-machine co-driving method according to claim 1, characterized in that: The specific steps for taking corresponding control measures based on the comparison and analysis results are as follows: If the driving safety index of the vehicle to be controlled on the road is higher than the preset upper limit of the driving safety range, the first control measure is executed; If the driving safety index of the vehicle to be controlled on the road is within a preset driving safety interval, the current vehicle speed value of the vehicle to be controlled on the road and the current distance warning vehicle speed value of the distance warning vehicle are obtained for comprehensive analysis to obtain an automatic control data set of the vehicle to be controlled on the road, and a second control measure is executed based on the automatic control data set of the vehicle to be controlled on the road, wherein the automatic control data set includes a current command acceleration value, a current command speed value, and a current command steering wheel deflection angle value; If the driving safety index of the vehicle to be controlled on the road is lower than the preset lower limit of the driving safety range, the third control measure is executed.

4. The remote human-machine co-driving method according to claim 3, characterized in that: The specific steps to obtain the automatic control data set of the vehicle to be controlled on the road are as follows: Performing a difference analysis on the current speed value of the controlled vehicle on the traveling road and the current approaching speed value of the distance warning vehicle to obtain the current relative speed difference between the controlled vehicle on the traveling road and the distance warning vehicle; Read the real-time vehicle distance value between the vehicle to be controlled and the distance warning vehicle on the road, and obtain the maximum safe distance value of the vehicle to be controlled on the road, and comprehensively analyze the maximum safe distance value of the vehicle to be controlled on the road, the real-time vehicle distance value of the vehicle to be controlled and the distance warning vehicle, and the current relative speed difference to obtain the current command acceleration value of the vehicle to be controlled on the road; Performing a difference analysis on the current controlled vehicle speed value and the current command acceleration value of the controlled vehicle on the road to obtain the current command speed value of the controlled vehicle on the road; Obtain the wheelbase value of the vehicle to be controlled, and make a comprehensive analysis with the maximum safe distance value on the road and the real-time vehicle distance value from the distance warning vehicle to obtain the current command steering wheel deflection angle value of the vehicle to be controlled on the road; The current command acceleration value, the current command speed value, and the current command steering wheel deflection angle value of the vehicle to be controlled on the road are marked as an automatic control data set.

5. A remote human-machine co-driving system, applying the remote human-machine co-driving method according to any one of claims 1 to 4, characterized in that: include: Data acquisition module, identification and analysis module, judgment and analysis module, first execution module, and second execution module; The data acquisition module is used to acquire the current two-dimensional position coordinates of the vehicle to be controlled on the road and the image data of the road; The identification and analysis module is used to identify and analyze the current two-dimensional position coordinates of the vehicle to be controlled on the road and the road image data, and obtain a number of fellow vehicles on the road and a real-time vehicle distance value to each fellow vehicle, wherein the road image data is specifically the pixel value and two-dimensional coordinates of each pixel point in the road image; The judgment and analysis module is used to judge and analyze the real-time vehicle distance value between the vehicle to be controlled and each of the other vehicles on the road and the preset vehicle distance warning range; The first execution module is used for, when the real-time vehicle distance value of each fellow vehicle of the to-be-controlled vehicle on the road is higher than the preset vehicle distance warning range, continuously acquiring the driving instruction data received several times by the to-be-controlled vehicle on the road, and the vehicle actual control driving data, and performing comprehensive analysis to obtain the vehicle actual driving index received each time by the to-be-controlled vehicle on the road, and performing trend analysis to obtain the comprehensive driving change index of the to-be-controlled vehicle on the road, wherein the instruction data is specifically the instruction value of each operation instruction, and the vehicle actual control driving data is specifically the actual operation value of the to-be-controlled vehicle for each operation instruction, and performing judgment and analysis on the comprehensive driving change index of the to-be-controlled vehicle on the road and the preset comprehensive driving change index threshold, and taking corresponding control measures based on the judgment and analysis results; The second execution module is used to mark the corresponding fellow vehicle as a distance warning vehicle when the real-time vehicle distance value between the vehicle to be controlled and the fellow vehicle on the road is within a preset vehicle distance warning range, and to continuously obtain the vehicle condition data of the vehicle to be controlled at several time points on the road and the adjacent vehicle condition data of the distance warning vehicle at several time points in real time, and to perform comprehensive analysis to obtain the driving safety index of the vehicle to be controlled on the road, the vehicle condition data specifically being the vehicle speed value and the vehicle acceleration value, and the adjacent vehicle condition data specifically being the distance warning vehicle speed value and the distance warning vehicle acceleration value, and to compare and analyze the driving safety index of the vehicle to be controlled on the road with the preset driving safety interval, and to take corresponding vehicle control measures based on the comparison and analysis results.

6. A remote human-machine co-driving device, using the remote human-machine co-driving method according to any one of claims 1 to 4, characterized in that: include: A data acquisition unit, a peer vehicle identification unit, a judgment unit, a first acquisition unit, a first analysis unit, a first measure execution unit, a vehicle marking unit, a second acquisition unit, a second analysis unit, and a second measure execution unit; The data acquisition unit is used to acquire the current two-dimensional position coordinates of the vehicle to be controlled on the road and the road image data, wherein the road image data is specifically the pixel value and two-dimensional coordinates of each pixel point in the road image; The peer vehicle identification unit is used to identify and analyze the acquired road image data of the vehicle to be controlled on the road, and obtain a number of peer vehicles on the road of the vehicle to be controlled, and a real-time vehicle distance value to each peer vehicle; The judgment unit is used to judge and analyze the real-time vehicle distance value between the vehicle to be controlled and each of the other vehicles on the road and the preset vehicle distance warning range; The first acquisition unit is used to continuously acquire the travel instruction data received several times by the vehicle to be controlled on the travel road, and the actual control driving data of the vehicle in real time when the real-time vehicle distance value of each fellow vehicle of the vehicle to be controlled on the travel road is higher than the preset vehicle distance warning range, wherein the instruction data is specifically the instruction value of each operation instruction, and the actual control driving data of the vehicle is specifically the actual operation value of the vehicle to be controlled for each operation instruction; The first analysis unit is used to comprehensively analyze the driving instruction data received several times on the road of the controlled vehicle and the actual control driving data of the vehicle, obtain the actual driving index of the vehicle received each time on the road of the controlled vehicle, and perform trend analysis to obtain the comprehensive driving change index of the vehicle on the road of the controlled vehicle; The first measure execution unit is used to judge and analyze the comprehensive driving change index of the vehicle to be controlled on the road and a preset comprehensive driving change index threshold, and take corresponding control measures based on the judgment and analysis results; The vehicle marking unit is used to mark the corresponding vehicle as a distance warning vehicle when the real-time vehicle distance value between the vehicle to be controlled and the vehicle on the road is within a preset vehicle distance warning range; The second acquisition unit is used to continuously acquire in real time the vehicle condition data of the to-be-controlled vehicle at several time points on the road and the adjacent vehicle condition data of the distance warning vehicle at several time points, wherein the vehicle condition data is specifically the vehicle speed value and the vehicle acceleration value, and the adjacent vehicle condition data is specifically the distance warning vehicle speed value and the distance warning vehicle acceleration value; The second analysis unit is used to comprehensively analyze the vehicle condition data of the vehicle to be controlled at several time points on the road and the vehicle condition data of the vehicle to be warned at several time points to obtain the driving safety index of the vehicle to be controlled on the road; The second measure execution unit compares and analyzes the driving safety index of the vehicle to be controlled on the road with a preset driving safety interval, and takes corresponding vehicle control measures based on the comparison and analysis results.

Images