Double-vehicle linkage cross coupling synchronous control method and system

Abstract

The invention relates to the technical field of vehicle control, in particular to a double-vehicle linkage cross coupling synchronous control method and system. Comprising the following steps: collecting a quality index of a wireless communication link of a partner vehicle; calculating a weight factor of communication quality based on the quality index; obtaining a report state of the partner vehicle, and generating a reckoning state of the partner vehicle; obtaining an effective partner vehicle state according to the report state and the reckoning state of the partner vehicle based on the weight factor; calculating a synchronization error vector based on the real-time state of the vehicle and the effective partner vehicle state; and generating a control instruction according to the synchronization error vector so as to drive a power and / or steering execution mechanism to implement cross coupling synchronization control. The problems that due to the fact that a communication system has problems, a control system receives inaccurate or delayed partner vehicle state information, a synchronization error is misjudged, an improper correction instruction is sent out, and the stability and safety of a carrying task are affected are solved.

Description

Technical Field

[0001] This application relates to the field of vehicle control technology, and more specifically, to a method and system for synchronous control of dual-vehicle linkage cross coupling. Background Technology

[0002] In modern industrial production and logistics, multiple independent vehicles are often used in coordinated transport to efficiently and safely transport large, extra-long, or irregularly shaped heavy objects. The vehicles exchange motion status in real time via wireless communication and maintain consistent relative position, speed, and attitude through cross-coupling synchronous control. This avoids structural damage caused by torsion, shearing, or compressive stress, ensuring safe and stable handling.

[0003] However, in real-world environments, communication links harbor hidden risks. For example, slight degradation of the antenna's physical connection can cause intermittent attenuation, weakening the reliability of data exchange between vehicles. When the controller receives inaccurate or delayed status information from partner vehicles, it may misjudge synchronization errors and issue inappropriate correction commands. This not only fails to eliminate the deviation but also amplifies the inconsistency in the movement of the two vehicles, potentially triggering the underlying drive motor protection and affecting mission stability and safety.

[0004] Specifically, when vehicles operate on uneven roads for extended periods, the impact and vibration cause unstable contact at the antenna connection points and fluctuations in contact impedance, resulting in inconsistent wireless signal strength. As a result, critical data packets (position, speed, attitude, etc.) suffer from compromised integrity or reduced timeliness, with retransmissions and losses occurring simultaneously. Critical synchronization commands and feedback experience unacceptable delays or omissions.

[0005] When the two vehicle controllers perform cross-coupling calculations based on this, the estimation of the partner's state deviates, and the resulting synchronization error deviates from the actual relative motion, forming a "pseudo-synchronization error." The speed loop and steering commands generated based on this error will incorrectly correct the deviation, further widening the difference in relative position and speed, causing the shared load to be subjected to uneven shear, torsional, or compressive stress.

[0006] Under continuous communication disturbances and error correction, the control algorithm misinterprets the phenomenon as a serious physical deviation, frequently issuing large-amplitude commands that require the motor to output excessive torque for short periods. This repeatedly triggers the drive overload protection, causing the vehicle to intermittently become sluggish and sluggish in response. In conditions requiring high instantaneous response, such as narrow curves, the system struggles to provide sufficient torque for precise steering and speed adjustments, posing a safety hazard. Furthermore, remote monitoring often only displays "excessive synchronization error," failing to reveal the vicious cycle of "intermittent communication—false error misjudgment—hardware protection," making fault location and handling even more difficult.

[0007] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0008] This application discloses a dual-vehicle linkage cross-coupling synchronization control method and system, which aims to solve the problem in the prior art where communication system problems lead to the control system receiving inaccurate or delayed status information of partner vehicles, resulting in incorrect judgment of synchronization error and issuance of inappropriate correction commands, thus affecting the stability and safety of the handling task.

[0009] The technical solution of this application is as follows:

[0010] In a first aspect, this application discloses a dual-vehicle linkage cross-coupling synchronous control method, including:

[0011] Collect quality indicators of the wireless communication link of the partner vehicle. The quality indicators include at least the received signal strength, data packet delay fluctuation, data packet loss rate and data freshness.

[0012] Calculate the weighting factors for communication quality based on quality indicators;

[0013] Obtain the reported status of the partner vehicle, and generate the estimated status of the partner vehicle based on the latest reliable reported status of the partner vehicle, the vehicle kinematic model, and odometer and / or inertial measurement data. The reported status includes at least one of position, speed, and / or attitude.

[0014] The reported status and the estimated status of the partner vehicle are linearly weighted and fused according to the weight factor to obtain the effective partner vehicle status. When the weight factor decreases, the weight of the reported status is reduced and the weight of the estimated status of the partner vehicle is increased.

[0015] Based on the real-time status of this vehicle and the status of effective partner vehicles, calculate the synchronization error vector, which includes at least one of position, speed and / or attitude error.

[0016] Control commands are generated based on the synchronization error vector to drive the power and / or steering actuators to implement cross-coupled synchronization control.

[0017] Furthermore, the reported status and the estimated status of the partner vehicles are linearly weighted and fused according to weighting factors to obtain the effective partner vehicle status, including:

[0018] Monitor the vehicle's motion status and preset path information, and assess the situational risk level based on the motion status and preset path information;

[0019] Calculate the communication confidence level of the report status based on the situational risk level and in conjunction with quality indicators;

[0020] Based on the situational risk level and the vibration characteristics of the vehicle obtained from the detection, the sensor confidence level for the estimated state of the partner vehicle is calculated.

[0021] When the communication confidence level is lower than the preset first threshold and the sensor confidence level is lower than the preset second threshold, the weighting factor is adjusted according to the situational risk level, and the reported status and the estimated status of the partner vehicle are linearly weighted and fused according to the adjusted weighting factor to obtain the effective partner vehicle status.

[0022] Furthermore, after monitoring the vehicle's motion status and preset path information, and assessing the situational risk level based on the motion status and preset path information, the process also includes:

[0023] Obtain local map information of the environment surrounding the vehicle;

[0024] The local map information is compared with the preset path information to obtain the comparison results, so as to identify the abnormal parts of the preset path information that are blocked by temporary obstacles, occupied, or deviate from the actual passable area.

[0025] Based on the comparison results, update the preset path information and adjust the situation risk level according to the updated preset path information;

[0026] The thresholds and weighting factors used for calculating communication confidence and sensor confidence are scheduled based on the adjusted situational risk level.

[0027] Furthermore, obtain local map information of the vehicle's surrounding environment, including:

[0028] Based on the size of the load being transported and the vehicle's motion state, the field of view obstruction area of ​​the on-board environmental perception sensor is calculated.

[0029] Based on preset path information and the known distribution of fixed obstacles, identify fixed occlusion areas;

[0030] Obtain supplementary local map information in areas with obstructed view and / or fixed obstruction areas;

[0031] The local map information is merged with the supplementary local map information to generate blind spot-free local map information, which is then compared with the preset path information.

[0032] Furthermore, the reported status and the estimated status of the partner vehicles are linearly weighted and fused according to weighting factors to obtain the effective partner vehicle status, including:

[0033] Perform timestamp freshness verification on the reported status of partner vehicles. When the data freshness exceeds a preset time threshold, the reported status is judged as expired.

[0034] When the report status is determined to be expired or a preset number of consecutive packet losses are detected within a preset time window, a short-term retransmission is triggered.

[0035] If the updated report status is not obtained within the maximum waiting time, the weight factor will be reduced to no less than the lower limit value according to the preset weight reduction rule.

[0036] The reported status and the estimated status of partner vehicles are linearly weighted and fused according to the reduced weighting factors to obtain the effective partner vehicle status.

[0037] Furthermore, based on the dimensions of the transported load and the vehicle's motion state, the field-of-view obstruction area of ​​the onboard environmental perception sensor is calculated, including:

[0038] Obtain real-time deformation information of the outer contour of the heavy object;

[0039] The three-dimensional geometric model of the object is dynamically updated based on its dimensions and real-time deformation information.

[0040] The system acquires the installation location and field-of-view parameters of the vehicle's environmental perception sensors, and calculates the vehicle's real-time pose information based on the vehicle's motion state.

[0041] Based on the dynamically updated 3D geometric model, installation location and field of view parameters, as well as the vehicle's real-time pose information, the occlusion area of ​​the field of view is obtained through 3D geometric calculation.

[0042] Furthermore, real-time deformation information of the external contour of the heavy object is obtained, including:

[0043] Acquire deformation data of the hinged joints and / or key flexible areas of the heavy object;

[0044] Based on deformation data and the structural connection relationship and material flexibility properties of the heavy object, a multi-segment deformation topology structure of the heavy object is constructed.

[0045] Based on a multi-segment deformation topology, the real-time deformation information of the outer contour of the heavy object is calculated.

[0046] Furthermore, before calculating the real-time deformation information of the object's external contour, the following steps are also included:

[0047] Acquire minute relative displacement and / or angle change data output by micro-motion sensors pre-placed at the connection points of the heavy structure;

[0048] By performing correlation analysis between small relative displacement and / or angle change data and deformation data, the identification results of connection points where implicit changes in connection stiffness or flexibility characteristics are obtained;

[0049] Based on the identification results, adjust the local stiffness or flexibility parameters in the structural connection relationships;

[0050] Based on the adjusted structural connection relationship and material flexibility properties, the multi-segment deformation topology is updated, and the updated multi-segment deformation topology is used to calculate the real-time deformation information of the outer contour of the heavy object.

[0051] Furthermore, based on the identification results, the local stiffness or flexibility parameters in the structural connections are adjusted, including:

[0052] The ambient temperature at the connection point is obtained from the temperature sensor at the hinge of the heavy object.

[0053] Based on the ambient temperature and the coefficient of thermal expansion and contraction of the material, temperature compensation is performed on the data of minute relative displacement and / or angle changes to obtain the compensated micro-motion data.

[0054] The vibration frequency and impact load amplitude of the hinge joint of the heavy object are monitored and the duration of action is accumulated. The degree of fatigue damage is assessed based on the duration of action and the fatigue characteristic curve of the heavy object material.

[0055] Based on the compensated micromotion data and the degree of fatigue damage, the local stiffness or flexibility parameters in the structural connection relationship are corrected;

[0056] Based on the corrected parameters and material flexibility properties, the multi-segment deformation topology is updated, and the updated topology is used to calculate the real-time deformation information of the outer contour of the heavy object.

[0057] Secondly, this application also discloses a dual-vehicle linkage cross-coupling synchronous control system, comprising:

[0058] The monitoring module is used to collect quality indicators of the wireless communication link of the partner vehicle. The quality indicators include at least the received signal strength, data packet delay fluctuation, data packet loss rate and data freshness.

[0059] The weighting factor calculation module is used to calculate the weighting factor of communication quality based on quality indicators.

[0060] The estimated state calculation module is used to obtain the reported state of the partner vehicle and generate the estimated state of the partner vehicle based on the latest reliable reported state of the partner vehicle, the vehicle kinematic model, and odometer and / or inertial measurement data. The reported state includes at least one of position, speed, and / or attitude.

[0061] The fusion module is used to linearly weight and fuse the reported status and the estimated status of the partner vehicle according to the weight factor to obtain the effective partner vehicle status. When the weight factor decreases, the weight of the reported status is reduced and the weight of the estimated status of the partner vehicle is increased.

[0062] The error calculation module is used to calculate the synchronization error vector based on the real-time status of the vehicle and the status of the effective partner vehicles. The synchronization error vector includes at least one of position, speed and / or attitude error.

[0063] The instruction generation module is used to generate control instructions based on the synchronization error vector to drive the power and / or steering actuators to implement cross-coupled synchronous control.

[0064] Beneficial effects

[0065] The dual-vehicle linkage cross-coupling synchronization control method disclosed in this application first collects wireless link quality indicators (RSSI, latency fluctuation, packet loss rate, data freshness) of the partner vehicle and calculates the communication quality weighting factor; then, it obtains the partner vehicle's reported status and infers the partner vehicle's status by combining kinematic models and odometer / IMU; the reported and inferred statuses are linearly weighted according to weights, with the proportion of reported status decreasing and the proportion of inferred status increasing when the weight decreases, forming an effective partner vehicle status; based on the real-time status of the vehicle and this effective status, a synchronization error vector (including position / vehicle speed / attitude) is calculated, and drive and steering commands are generated accordingly to achieve robust synchronization through cross-coupling. This mechanism addresses hidden communication problems such as antenna degradation and intermittent attenuation, suppresses "pseudo-synchronization errors," avoids frequent triggering of error correction and motor overload protection, and maintains drive continuity and trajectory stability. Advantages: ① Improved robustness, adapting to link fluctuations while still accurately estimating partner status; ② More precise and safer control, reducing uneven stress; ③ Reduced fault risk, mitigating overload protection triggering; ④ More accurate diagnosis, with alarms better reflecting actual physical deviations. In summary, the goal is to improve adaptability, accuracy, and safety under complex working conditions. Attached Figure Description

[0066] Figure 1 This is a flowchart illustrating a dual-vehicle linkage cross-coupling synchronous control method provided in this application.

[0067] Figure 2 A flowchart of a dual-vehicle linkage cross-coupling synchronous control system provided in this application.

[0068] In the diagram: 1. Monitoring module; 2. Weight factor calculation module; 3. Estimated status calculation module; 4. Fusion module; 5. Error calculation module; 6. Instruction generation module. Detailed Implementation

[0069] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0070] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0071] Traditional multi-vehicle collaborative material handling systems, when facing complex industrial environments, may experience intermittent signal attenuation due to hidden issues such as physical connection degradation in their communication systems. This can affect the reliability of data exchange between vehicles. When the control system receives inaccurate or delayed status information from partner vehicles, it may incorrectly determine synchronization errors and issue inappropriate corrective commands. This not only fails to solve the problem but may also exacerbate motion deviations between vehicles, or even trigger the protection mechanisms of the underlying drive motors, ultimately affecting the stability and safety of the material handling task.

[0072] Reference Figure 1 In response, this application proposes a dual-vehicle linkage cross-coupling synchronous control method, including:

[0073] S1000: Collects quality indicators of the wireless communication link of the partner vehicle. The quality indicators include at least received signal strength, data packet delay fluctuation, data packet loss rate, and data freshness.

[0074] S2000: Weighting factors for calculating communication quality based on quality indicators;

[0075] S3000: Obtain the reported status of the partner vehicle, and generate the estimated status of the partner vehicle based on the latest reliable reported status of the partner vehicle, the vehicle kinematic model, and odometer and / or inertial measurement data. The reported status includes at least one of position, speed, and / or attitude.

[0076] S4000: The reported status and the estimated status of the partner vehicle are linearly weighted and fused according to the weight factor to obtain the effective partner vehicle status. When the weight factor decreases, the weight of the reported status is reduced and the weight of the estimated status of the partner vehicle is increased.

[0077] S5000: Based on the real-time status of this vehicle and the status of effective partner vehicles, calculate the synchronization error vector, which includes at least one of position, speed and / or attitude error;

[0078] S6000: Generates control commands based on the synchronization error vector to drive the power and / or steering actuators to implement cross-coupled synchronization control.

[0079] Specifically, a partner vehicle refers to another vehicle that works in conjunction with the main vehicle to complete the task of moving heavy objects.

[0080] The quality indicators of a wireless communication link are parameters that measure the reliability of inter-vehicle communication. These include Received Signal Strength (RSSI), packet delay fluctuation, packet loss rate, and data freshness. Received signal strength reflects the strength of the signal; packet delay fluctuation indicates the degree of variation in data transmission delay; packet loss rate refers to the proportion of packets lost within a certain period; and data freshness measures the time difference between the received data and the actual occurrence of the data.

[0081] Reporting status refers to motion information such as position, speed, and / or attitude that a partner vehicle sends to its own vehicle in real time via wireless communication.

[0082] Estimated state refers to the prediction or estimation of the current motion state of the partner vehicle by the vehicle itself based on the partner vehicle's historical reported state, its own kinematic model, and odometer and / or inertial measurement unit (IMU) data.

[0083] Effective partner vehicle status is the result of a linearly weighted fusion of reported status and estimated status. It combines the advantages of both and provides a more accurate and reliable estimate of partner vehicle status.

[0084] The synchronization error vector is the difference between the real-time state of the vehicle and the state of the effective partner vehicle, used to quantify the degree of inconsistency between the two vehicles in position, speed and / or attitude.

[0085] Cross-coupling synchronization control is a control strategy that uses the synchronization error between two vehicles as a control input to generate corresponding control commands to drive the power and / or steering actuators, thereby keeping the two vehicles moving in sync.

[0086] Specifically, various methods can be used to collect quality indicators of the wireless communication link of partner vehicles. For example, the received signal strength (RSSI) can be monitored in real time through the vehicle's wireless communication module, while simultaneously recording the timestamp of each received data packet and comparing it with the sending timestamp to calculate data packet latency fluctuations and data freshness. The data packet loss rate can be obtained by counting the number of expected data packets that are not received within a certain period of time. The collection of these indicators is continuous to ensure real-time monitoring of the communication link quality.

[0087] When calculating the weighting factor for communication quality based on the collected quality indicators, a pre-defined mapping function or lookup table can be used. For example, when the received signal strength is high, the packet delay fluctuation is small, the packet loss rate is low, and the data freshness is high, a higher weighting factor can be assigned to the communication quality. Conversely, when the communication quality indicators deteriorate, the weighting factor will decrease accordingly. This calculation of the weighting factor can be linear or non-linear, depending on the required sensitivity to communication quality.

[0088] In obtaining the reported state of the partner vehicle and generating its estimated state, the partner vehicle periodically transmits its position, velocity, and / or attitude information to the vehicle via wireless communication; this information constitutes the reported state. Simultaneously, the vehicle uses the partner vehicle's latest reliable reported state as an initial value, combined with a pre-established vehicle kinematic model (e.g., based on an Ackerman steering model or differential drive model) and its own collected odometer and / or inertial measurement data (such as accelerometer and gyroscope data), to predict the partner vehicle's future motion state using state estimation algorithms such as Kalman filtering or extended Kalman filtering, thereby generating the estimated state of the partner vehicle.

[0089] Subsequently, the reported status and the estimated status of the partner vehicle are linearly weighted and fused according to a weighting factor to obtain the effective partner vehicle status. Specifically, the effective partner vehicle status can be expressed as: Effective Partner Vehicle Status = Weighting Factor × Reported Status + (1 - Weighting Factor) × Estimated Status. When the communication quality is good and the weighting factor is high, the reported status dominates the fusion result; when the communication quality deteriorates and the weighting factor decreases, the weight of the estimated status will increase accordingly to compensate for the potential unreliability of the reported status. This dynamically adjusted fusion mechanism ensures that a relatively accurate partner vehicle status can be obtained under different communication conditions.

[0090] Based on the real-time status of the vehicle and the status of its effective partner vehicles, a synchronization error vector is calculated. The real-time status of the vehicle can be accurately sensed through its own sensors (such as GPS, IMU, wheel speed encoders, etc.). The calculation of the synchronization error vector involves comparing the real-time status of the vehicle with the status of its effective partner vehicles one by one in dimensions such as position, velocity, and / or attitude to obtain the corresponding error values. For example, the position error can be the difference between the coordinates of the center of mass of the two vehicles, the velocity error can be the difference between the linear velocity or angular velocity of the two vehicles, and the attitude error can be the difference between the yaw angle, pitch angle, or roll angle of the two vehicles.

[0091] Finally, control commands are generated based on the synchronization error vector to drive the power and / or steering actuators to implement cross-coupled synchronization control. The generation of control commands can employ proportional-integral-derivative (PID) controllers, model predictive control (MPC), or other advanced control algorithms. These algorithms calculate the adjustments required for the vehicle's power system (e.g., motor torque) and steering system (e.g., steering angle) based on the magnitude and trend of the synchronization error vector, thereby driving the vehicle's actuators to make precise motion adjustments to reduce synchronization errors and ensure coordinated movement between the two vehicles.

[0092] The dual-vehicle linkage cross-coupling synchronization control method proposed in this application effectively solves the pseudo-synchronization error problem caused by communication instability in traditional methods by introducing real-time monitoring and evaluation of the wireless communication link quality and using it as a weighting factor to dynamically adjust the status of partner vehicles. In traditional schemes, when communication quality deteriorates, the controller may directly use inaccurate reported status, leading to incorrect synchronization error calculation, which in turn triggers incorrect correction actions and may even trigger underlying hardware protection.

[0093] The method in this application reduces the weight of the reported state and increases the weight of the inferred state generated based on the vehicle kinematic model and odometer / inertial measurement data when communication quality deteriorates. This intelligent weighted fusion mechanism allows the system to rely more heavily on its own prediction of the partner vehicle's motion trend when the communication link experiences intermittent attenuation or data loss, thereby obtaining a more stable and reliable effective partner vehicle state. As a result, the calculated synchronization error vector can more accurately reflect the actual relative motion between the two vehicles, avoiding the generation of pseudo-synchronization errors.

[0094] In another embodiment of this application, step S4000 is further proposed to include:

[0095] S4100: Monitors the vehicle's motion status and preset path information, and assesses the situational risk level based on the motion status and preset path information;

[0096] S4200: Calculate the communication confidence level of the report status based on the situational risk level and in conjunction with quality indicators;

[0097] S4300: Calculate the sensor confidence level for the estimated state of the partner vehicle based on the situational risk level and the vibration characteristics of the vehicle obtained from the detection.

[0098] S4400: When the communication confidence level is lower than the preset first threshold and the sensor confidence level is lower than the preset second threshold, the weighting factor is adjusted according to the situation risk level, and the reported status and the estimated status of the partner vehicle are linearly weighted and fused according to the adjusted weighting factor to obtain the effective partner vehicle status.

[0099] Specifically, monitoring the vehicle's motion status and preset path information, and assessing the situational risk level accordingly, can be understood as follows: acquiring the vehicle's real-time position, speed, acceleration, angular velocity, and other motion parameters through onboard sensors (such as GPS, inertial measurement unit, wheel speed sensors, etc.); combining this with pre-planned or dynamically generated path information (such as path curvature, width and height restrictions, temporary traffic control, and hazardous sections). Based on these factors, the system comprehensively judges the risk level of the current operating environment and classifies the situational risk level into discrete levels such as low risk, medium risk, and high risk, which are used as higher-level constraints for subsequent confidence calculations and weight adjustments.

[0100] The communication confidence level, calculated based on the situational risk level and communication link quality indicators, refers to the quantifiable result of the reliability of the partner vehicle's reported status by comprehensively considering indicators such as Received Signal Strength Indicator (RSSI), packet delay fluctuation, packet loss rate, and data freshness under a defined risk context. The communication confidence level can take a continuous value from 0 to 1, with higher values ​​indicating more reliable links. In high-risk scenarios, to improve safety margins, even if link indicators are at a "fair" level, the communication confidence level can be lowered according to the situation to reflect a cautious approach to accepting external data.

[0101] In practical applications, the sensor confidence level used to estimate the state of partner vehicles is calculated based on the situational risk level and the vibration characteristics of the vehicle itself. This refers to quantifying the reliability of the outputs from internal sensors such as the vehicle's odometer / IMU based on the vibration frequency, amplitude, and duration detected by sensors such as accelerometers and gyroscopes. The sensor confidence level also takes a continuous value from 0 to 1. When the vehicle is on a bumpy road or experiencing strong vibrations, even if the sensor itself is not faulty, the measurement noise and drift increase, so the confidence level in its own estimation results should be reduced. In high-risk situations, this weighting is even more stringent.

[0102] Furthermore, when the communication confidence level is lower than a preset first threshold and the sensor confidence level is lower than a preset second threshold, the weighting factor is adjusted according to the situational risk level. Specifically, when both external reports and internal calculations have high uncertainty, the system no longer maintains a fixed fusion ratio, but triggers a closed-loop adaptive mechanism of "risk-confidence-weight". Under low risk, a small adjustment can be made to maintain efficiency; under medium / high risk, a more conservative linear weighting is adopted (e.g., tightening the weight of the reporting status, or prioritizing internal calculation for a short time when the communication is extremely poor) to reduce the "pseudo-synchronization error" caused by the distortion of a single information source, avoid error correction and control jitter, and suppress the frequent triggering of the drive system overload protection.

[0103] In summary, this solution is based on the following main framework: "Risk assessment provides context—communication / sensor confidence quantification reliability—fusion weights are dynamically scheduled according to risk and confidence level."

[0104] (1) The situational risk level is responsible for calibrating the environmental complexity and safety margin;

[0105] (2) Communication confidence and sensor confidence measure the “acceptability” of external and internal data, respectively;

[0106] (3) The fusion weight factor is reallocated in real time so that the effective partner vehicle status remains robust under uncertain conditions, thereby ensuring the stability and safety of cross-coupling synchronization control.

[0107] In some preferred embodiments, the following scenario is taken as an example:

[0108] Two heavy trucks are traveling in platooning on a highway. This vehicle is traveling in a straight line at a speed of 80 km / h along a main highway route, assessed as low risk; link indicators are excellent, communication confidence is high; vehicle vibration is stable, and sensor confidence is high. At this point, the fusion weights remain balanced, and the reported state and the inferred state are successfully fused.

[0109] Upon entering a construction area, the preset path indicates upcoming traffic control and temporary obstacles. The vehicle begins to decelerate and change lanes, raising the scenario risk to medium risk. Construction interference causes a decrease in RSSI and an increase in packet loss, reducing communication confidence. Uneven road surfaces exacerbate vehicle vibration, leading to a simultaneous decrease in sensor confidence. If the communication confidence falls below 0.6 and the sensor confidence falls below 0.7 at this point, the system redistributes weighting factors according to a "medium risk" strategy: while ensuring safety, it appropriately reduces the weight of reported states and increases the weight of inferred states (or adopts a finer-grained ratio and time window strategy) to ensure sufficient robustness of the fused output. Through this dynamic scheduling, the system can maintain a reliable estimate of the partner vehicle's status even under conditions of poor communication and sensor noise, ensuring the smooth execution of dual-vehicle cross-coupling synchronous control.

[0110] In another embodiment of this application, it is further proposed that after S4100, the following is also included:

[0111] S4110: Obtain local map information of the environment surrounding the vehicle;

[0112] S4120: Compare the local map information with the preset path information to obtain the comparison result, so as to identify the abnormal parts of the preset path information that are blocked by temporary obstacles, occupied, or deviate from the actual passable area.

[0113] S4130: Based on the comparison results, update the preset path information and adjust the situational risk level according to the updated preset path information;

[0114] S4140: Schedule the thresholds and weighting factors used for calculating communication confidence and sensor confidence based on the adjusted situational risk level.

[0115] Specifically, obtaining local map information of the vehicle's surrounding environment refers to using onboard environmental perception sensors (such as lidar, millimeter-wave radar, and cameras) to collect data on roads, obstacles, and traffic participants in real time, and constructing a local environmental map near the vehicle's current location; this map is more detailed and updated more promptly than the preset path information.

[0116] The process involves comparing local map information with preset path information to identify differences and output the comparison results. Key detection areas include: temporary obstacles (construction barriers, temporarily parked vehicles, etc.) on the preset path, path occupancy / narrowing, and deviations between the preset geometry and the actual passable area. This allows for the accurate location of abnormal sections within the preset path.

[0117] In practical applications, the system dynamically updates the preset path information based on the comparison results and reassesses the risk level of the situation based on the updated path: if there is a detour, narrowing of the passage, or a decrease in safety margin, the risk level is increased accordingly; if the obstacle is removed or the passage conditions are improved, the risk level is decreased.

[0118] Furthermore, based on the adjusted situational risk level, the system schedules the thresholds for calculating communication confidence and sensor confidence, as well as the fusion weight factors: in high-risk situations, the confidence threshold is increased, the acceptance criteria are tightened, and the fusion weights are adjusted, with a greater preference for using more reliable data sources (increasing the proportion of the vehicle's estimated state when necessary); in low-risk situations, the thresholds are relaxed to maintain efficiency and responsiveness.

[0119] Through the aforementioned mechanism, the solution uses a closed loop of "local map—path comparison—path update—risk reassessment—threshold / weight scheduling" to solve the problem that relying solely on preset paths may not match reality. The introduction of local maps can quickly expose temporary obstacles and traffic deviations, allowing path information to be updated in a timely manner, and risk assessments to converge rapidly with environmental changes. On this basis, dynamically scheduling confidence thresholds and fusion weights makes the fusion of partner vehicle reported status and estimated status more intelligent and robust, improving the accuracy and availability of effective partner vehicle status.

[0120] In some preferred embodiments:

[0121] Two heavy trucks operate in a synchronized, cross-coupled manner along a pre-defined logistics route. The truck's lidar and camera detect a construction area approximately 50 meters ahead: some lanes are blocked by barriers, the passage width is narrowed, and temporary obstacles exist. The system compares the local map with the pre-defined route, identifying this section as an abnormal section that is "occupied / narrowed and deviates from the actual passable area." The system then updates the pre-defined route (e.g., planning a temporary detour, or marking this section of the original route as high-risk) and adjusts the situational risk level to medium-high risk. Based on this risk level, the system raises the communication and sensor confidence thresholds, tightening data acceptance standards; simultaneously, it adjusts the fusion weights, appropriately reducing the weight of the partner vehicle's reported status and increasing the weight of the truck's estimated status in situations where communication quality may fluctuate. Therefore, even in complex environments with temporary obstacles and narrowed passages, the system can robustly fuse the partner vehicle's status, ensuring the accuracy and safety of synchronized control.

[0122] In another embodiment of this application, S4110 further includes:

[0123] S4111: Calculate the field-of-view occlusion area of ​​the on-board environmental perception sensor based on the size of the load being transported and the motion state of the vehicle.

[0124] S4112: Identify fixed occlusion areas based on preset path information and known fixed obstacle distribution;

[0125] S4113: Obtain supplementary local map information in areas with obstructed view and / or fixed obstruction areas;

[0126] S4114: Merge local map information with supplementary local map information to generate blind spot-free local map information, and use the blind spot-free local map information to compare with preset path information.

[0127] Specifically, calculating the obstructed area of ​​the vehicle's environmental perception sensors involves using 3D geometric modeling and ray tracing to accurately determine the spatial range invisible to the sensors. This is based on the geometric dimensions and shape of the load carried by the vehicle, the vehicle's current motion (pitch, roll, yaw, etc.), and the installation location and field-of-view parameters of the environmental perception sensors (LiDAR, millimeter-wave radar, cameras, etc.). For large or irregularly shaped heavy objects, turning, acceleration, and deceleration can cause the obstruction range to change dynamically, thus creating time-varying blind spots.

[0128] Identifying fixed occlusion areas refers to locating occlusion areas that are unrelated to vehicle movement, exist for a long time, and continuously affect the sensor's line of sight by combining a preset path with known environmental information (buildings, bridges, tunnels, large fixed equipment, etc. marked in high-precision maps).

[0129] Acquiring supplementary local map information refers to using compensation perception strategies or external information sources to fill blind spots in identified and / or fixed occlusion areas. For example, this can be achieved by using perception results shared by partner vehicles, calling historical / high-precision map fragments, adjusting sensor attitude or scanning mode, or enabling auxiliary / temporary sensors to obtain environmental data that was not originally directly observable.

[0130] The process of fusing and generating blind-spot-free local map information involves consistently fusing the local map directly perceived by the vehicle with blind-spot information obtained through supplementary methods. Methods such as Kalman filtering, particle filtering, probabilistic grids, or learning-based fusion can be employed to eliminate redundancy, fill gaps, and weight confidence levels, resulting in a more comprehensive, accurate, and spatiotemporally consistent "blind-spot-free" local map. This map is then compared with a preset path to more accurately identify abnormal sections blocked by temporary obstacles, occupied, or deviating from the actual passable area. This provides reliable input for situational risk assessment and avoids misjudgments due to incomplete maps.

[0131] Through the above process, this solution forms a closed loop of "occlusion solution → fixed occlusion identification → blind spot supplementation → multi-source fusion," systematically solving the environmental cognition gap caused by the limited field of view of sensors: it considers both dynamic occlusion caused by heavy objects and attitudes, as well as static occlusion that has existed in the environment for a long time; on this basis, blind spots are supplemented and multi-source robust fusion is performed to make local map information complete and accurate, thereby improving the reliability of path comparison and situational risk assessment, ensuring that the scheduling of communication / sensor confidence thresholds and fusion weights is more reasonable, and enhancing the robustness and security of linkage and synchronization control.

[0132] In a preferred embodiment of this example:

[0133] A heavy-duty transport vehicle, carrying a large cargo on its roof, is operating in a densely populated urban area with temporary construction sites. Based on a 3D model of the cargo and the vehicle's real-time attitude, speed, and acceleration, the system calculates the dynamic occlusion areas of the lidar and cameras. When the vehicle turns, the lateral occlusion of the cargo increases. Combining a preset path with a high-precision urban map, the system identifies fixed occlusions (high-rise buildings, streetlights, construction barriers, etc.) and areas with reduced satellite visibility. Within dynamic / fixed occlusion areas, the system accesses shared perception data from partner vehicles, supplements high-precision map details, and can briefly adjust sensor scanning modes or activate wide-angle cameras to capture blind spot information. The system then performs probabilistic raster fusion of the vehicle's direct perception data and supplementary data to obtain a blind-spot-free local map. This map is compared with the preset path to accurately identify abnormal sections blocked by temporary barriers, occupied by other vehicles, or deviating from the actual passable area, providing a solid data foundation for subsequent risk assessment and control command generation.

[0134] In another embodiment of this application, S4000 further includes:

[0135] S4500: Performs timestamp freshness verification on the reported status of partner vehicles. When the data freshness exceeds a preset time threshold, the reported status is judged as expired.

[0136] S4600: When the report status is determined to be expired or a preset number of consecutive packet losses are detected within a preset time window, a short-term retransmission is triggered;

[0137] S4700: If the updated report status is not obtained within the maximum waiting time, the weight factor will be reduced to no less than the lower limit value according to the preset weight reduction rule.

[0138] S4800: The reported status and the estimated status of the partner vehicle are linearly weighted and fused according to the weighted factors after weight reduction to obtain the effective partner vehicle status.

[0139] Specifically, performing timestamp freshness verification on the reported status of partner vehicles means comparing the timestamp of the received data with the current system time to calculate the data freshness. If the freshness exceeds a preset time threshold (such as 50 / 100 / 200ms), it is judged as expired to prevent outdated information from participating in the fusion.

[0140] Furthermore, the conditions for triggering a short-term retransmission are: ① the report status is judged to be expired; or ② within a preset time window (e.g., the most recent 1–2 seconds), a preset number of packet losses are achieved (e.g., 3 or 5 times). Short-term retransmission is used to request partner vehicles to immediately send back the latest report status in order to quickly restore effective communication in the event of a momentary interruption.

[0141] If no updated data is obtained after retransmission within the maximum waiting time (e.g., 500ms or 1s), it is determined that the communication problem persists. The fusion weight of the report status is dynamically reduced according to the preset weight reduction rules (linear / exponential / segmented reduction, etc.), and a constraint of not less than the lower limit is set to retain the minimum reference degree. At the same time, the weight of the partner vehicle status calculated based on the vehicle model and the odometer / IMU is increased accordingly.

[0142] Therefore, the system performs linear weighted fusion of the reported status and the estimated status according to the weighted factors after weight reduction. When the reliability or real-time performance of the reported data decreases, its contribution to the fusion result automatically decreases, while the contribution of the estimated status increases accordingly, thus obtaining a more robust and continuous effective partner vehicle status.

[0143] This solution significantly improves robustness and reliability under wireless link fluctuations, delays, or packet loss through a closed loop of "freshness verification → short-term retransmission → timeout weight reduction → robust integration": expired data is promptly removed, momentary interruptions are quickly replenished through retransmission, and continuous anomalies are adaptively migrated to vehicle-to-vehicle estimation through weight adaptation, maintaining the accuracy and continuity of state estimation and avoiding synchronization error expansion and control failure caused by poor communication.

[0144] In a preferred embodiment of this example:

[0145] Two heavy trucks are traveling in convoy in the mining area. This vehicle continuously receives status reports (position / speed / attitude) from the partner vehicle. The system is configured with: a time threshold of 100ms, a time window of 1s, a consecutive packet loss threshold of 3 times, and a maximum waiting time of 500ms.

[0146] If the difference between the timestamp of a received data and the current time is greater than 100ms, the data is considered expired and a retransmission is initiated immediately. If no data packets are received for three consecutive times within the last second, a short-term retransmission is also triggered.

[0147] If there is no update within 500ms after retransmission, then weight reduction is initiated: for example, the weight is reduced from the initial 0.7 (report) / 0.3 (calculated), and the report weight is reduced by 0.1 every 100ms according to a linear rule until the lower limit of 0.1 is reached; the weight sequence can be 0.7 / 0.3→0.6 / 0.4→0.5 / 0.5→…→0.1 / 0.9.

[0148] The system linearly fuses the "last frame's available reported status" and the "estimated status based on kinematic model + odometer / IMU" using downweighted coefficients to obtain the effective partner vehicle status. Even with severe link impairment, the estimation-driven fusion can still maintain a relatively accurate estimate of the partner vehicle status, ensuring platoon synchronization and driving safety.

[0149] In another embodiment of this application, S4111 further includes:

[0150] S4111-1: Obtain real-time deformation information of the outer contour of the heavy object;

[0151] S4111-2: Dynamically update the three-dimensional geometric model of the object based on its dimensions and real-time deformation information;

[0152] S4111-3: Obtain the installation position and field of view parameters of the vehicle's environmental perception sensor, and calculate the vehicle's real-time pose information based on the vehicle's motion state;

[0153] S4111-4: Based on the dynamically updated 3D geometric model, installation position and field of view parameters, as well as the real-time pose information of the vehicle, the field of view occlusion area is obtained through 3D geometric calculation.

[0154] Specifically, acquiring real-time deformation information of the external contour of a heavy object refers to online collection of data on the shape changes of the heavy object under its current carrying condition, based on multi-source sensing and / or model prediction. Strain / displacement / vision sensors can be deployed at hinge points or critical stress areas to continuously monitor deformations such as bending, torsion, and expansion. If necessary, an online deformation estimation model can be constructed by combining material properties and boundary conditions to compensate for noise and missing measurements, outputting deformation information that reflects the true geometric contour "at this very moment."

[0155] Dynamically updating the 3D geometric model of a heavy object based on its dimensions and real-time deformation information involves fusing nominal dimensions (design dimensions / loading list) with online deformation observations to form a continuously updated 3D digital model (mesh, voxels, or parametric surface) over time. This model adaptively corrects itself with deformation, ensuring consistency between the geometric representation and the physical form, and avoiding occlusion estimation biases caused by static dimension assumptions.

[0156] Acquiring the installation location and field of view parameters of the vehicle's environmental perception sensors, and calculating the vehicle's real-time pose information based on the vehicle's motion state, means, on the one hand, calling the calibrated external and internal parameters / field of view parameters (installation coordinates and attitude, field of view angle, detection distance, etc.), and on the other hand, calculating the vehicle's position and attitude in the global coordinate system in real time through the fusion of GNSS, IMU and odometer, thereby determining the precise spatial position and orientation of each sensor in the world coordinate system.

[0157] Based on the dynamically updated 3D geometric model, installation position and field of view parameters, and the vehicle's real-time pose information, the field of view occlusion area is obtained through 3D geometric calculation. This means using the updated heavy object model as the occlusion body, combined with the sensor's current pose and field of view constraints, and employing visibility determination methods such as ray tracing, frustum clipping, and point cloud occlusion removal / projection to solve the spatial distribution of the visible and occluded domains, and outputting the field of view occlusion area (which can be represented as a angular mask, voxel set, or probability grid) as the basis for subsequent blind spot re-measurement and risk assessment.

[0158] This solution employs a closed loop of "deformation perception—model update—pose / extrinsic parameter constraints—visibility calculation" to accurately characterize the impact of loading deformation on the sensor's field of view, resolving the distortion of occlusion range caused by traditional static size assumptions. The occlusion results change in real time with vehicle posture and load deformation, making blind spot recognition more closely resemble actual working conditions and providing reliable input for local map construction and situational risk assessment. Based on this, targeted blind spot re-measurement, multi-vehicle perception sharing, or map prior filling can be triggered, improving the completeness and accuracy of local maps and reducing the risks caused by blind spot misjudgments.

[0159] In some preferred embodiments:

[0160] Assuming two heavy trucks are engaged in joint transport, with one truck towing a large flexible structural component (such as a bundle of long steel pipes or a large roll of canvas). The system acquires the external contour deformation in real time through micro-motion / strain / displacement sensors or visual tracking deployed at key connection points and easily deformable areas; it fuses the deformation data with the original dimensions to dynamically update the 3D model of the flexible structural component; and it performs ray tracing calculations to determine sensor field-of-view occlusion masks by combining the calibration extrinsic parameters of the roof-mounted LiDAR and camera with the vehicle's pose calculated by GNSS / IMU / odometer. The results show that the lateral slight bending of the bundle of steel pipes during vehicle turns reduces the lateral visible sector. Based on this, the system triggers blind spot compensation within the occlusion angle domain: on the one hand, it utilizes the shared point cloud of the partner vehicle to cover the blind spot of this vehicle; on the other hand, it calls upon a high-precision map static prior for supplementation. After integrating direct sensing and supplementary information, a more complete local map is formed and accurately compared with the preset path. It can promptly identify the narrowing of the passage width caused by construction barriers, improve the situational risk level of the road section, and tighten the communication / sensor confidence threshold and state fusion weight in conjunction to ensure that robust synchronous control and safe passage can still be maintained under the conditions of dynamic occlusion and sudden environmental changes.

[0161] In another embodiment of this application, S4111-1 specifically includes:

[0162] S4111-11: Obtain deformation data of the hinged parts and / or key flexible areas of the load;

[0163] S4111-12: Based on deformation data and the structural connection relationship and material flexibility characteristics of the heavy object, construct a multi-segment deformation topology structure of the heavy object;

[0164] S4111-13: Calculate the real-time deformation information of the outer contour of a heavy object based on a multi-segment deformation topology.

[0165] Specifically, the hinged parts of a heavy object refer to the internal or external connections of the object, areas where relative movement or angular changes are permitted (such as robotic arm joints and trailer connection points); critical flexible areas refer to areas prone to elastic or plastic deformation during stress or movement (such as the sidewalls of large containers and flexible crane booms). Deformation data refers to the changes in position, angle, length, or curvature of the aforementioned parts in real time, which can be collected by displacement sensors, angle sensors, strain gauges, vision sensors, etc., to capture the dynamic changes of the non-rigid parts of the heavy object.

[0166] Among them, structural connection relationships describe the physical connection methods and geometric constraints (fixed connection, hinge connection, sliding connection, etc.) of the various components of the heavy object; material flexibility characteristics reflect the material's ability to deform under stress (such as elastic modulus, Poisson's ratio). Multi-segment deformation topology divides the heavy object into several segments with specific deformation characteristics, clarifies the connection relationships and deformation models between segments, and forms a structured representation that can reflect the overall deformation characteristics, which facilitates the decomposition and modeling analysis of complex deformation problems.

[0167] In practical applications, calculating the real-time deformation information of a heavy object's external contour based on a multi-segment deformable topology involves using a pre-constructed topology and real-time deformation data to derive the complete external contour at the current moment through geometric calculations, finite element analysis, or physics-based simulations. For example, updating the relative attitude of adjacent segments based on the angle changes at hinge points, updating the geometry of the segment based on the deformation of flexible regions, and then integrating the results of each segment to obtain the overall real-time external contour of the heavy object. This result serves as precise input for subsequent calculations of the field-of-view occlusion area.

[0168] The proposed solution involves: first, acquiring deformation data of hinged sections and / or key flexible areas to accurately perceive the dynamic changes of non-rigid parts; then, constructing a multi-segment deformation topology by combining structural connection relationships and material flexibility properties; and finally, calculating the real-time deformation information of the object's external contour. This refined modeling avoids oversimplifying complex deformations, fully incorporating dynamic deformations into the occlusion evaluation process, and improving the realism of the geometric description from the source.

[0169] This technical solution overcomes the problem of inaccurate occlusion estimation caused by the lack of real-time deformation in traditional methods for complex or flexible heavy-load scenarios. By measuring hinged joints and key flexible areas, and combining structural connections with material flexibility properties to construct a multi-segment deformation topology, the system can accurately calculate the real-time deformation information of the heavy object's external contour. This allows for more accurate dynamic updates to the 3D geometric model, significantly improving the accuracy and reliability of occlusion calculations by the vehicle's environmental perception sensors. This provides a more reliable environmental input for synchronized control of dual-vehicle linkage and cross-coupling, reduces the risk of collisions caused by occlusion misjudgments, and enhances system safety and control robustness.

[0170] In some preferred embodiments:

[0171] Consider a tractor unit towing a large semi-trailer: the semi-trailer's cargo box may slightly bend or twist during operation, and the tractor unit and semi-trailer are articulated. To obtain real-time deformation information of the semi-trailer's external contour, strain sensors or laser displacement sensors are deployed in key flexible areas of the cargo box (such as the middle and rear) to collect bending / torsional deformation data; angle sensors are installed at the articulation points between the tractor unit and semi-trailer to obtain relative rotation angles. Combining the material flexibility properties of the cargo box material (such as the elastic modulus of steel) and the structural connection relationship between the cargo box panel and the frame, a multi-segment deformation topology is constructed: the cargo box is modeled as multiple rigid segments connected by flexible or articulated segments, with the flexible segments driven by strain data and the articulated segments driven by angle data. Based on this topology, geometric transformations and interpolation are used to reconstruct the current complete external contour (including bending, torsion, and articulation morphology) in real time, and the 3D geometric model is dynamically updated accordingly, thereby more accurately calculating the area of ​​visual obstruction to the onboard environmental perception sensors.

[0172] In another embodiment of this application, it is further proposed that, before calculating the real-time deformation information of the outer contour of the heavy object, the following steps are also included:

[0173] S4111-130: Acquire small relative displacement and / or angle change data output by micro-motion sensors pre-positioned at the connection of the heavy structure;

[0174] S4111-131: Correlation analysis of small relative displacement and / or angle change data with deformation data to obtain identification results of connection points where implicit changes in connection stiffness or flexibility characteristics occur;

[0175] S4111-132: Adjust the local stiffness or flexibility parameters in the structural connection relationship based on the identification results;

[0176] S4111-133: Based on the adjusted structural connection relationship and material flexibility characteristics, update the multi-segment deformation topology and use the updated multi-segment deformation topology to calculate the real-time deformation information of the outer contour of the heavy object.

[0177] Specifically, acquiring minute relative displacement and / or angle change data output by micro-motion sensors pre-positioned at the structural connections of heavy objects refers to pre-installing micro-motion sensors (strain gauges, miniature displacement sensors, angle sensors, MEMS sensors, etc.) at the structural connections (such as hinge points, weld seams, bolted connections) of heavy objects (such as large robotic arms, crane booms, trailer vehicle connectors, etc.) to collect minute relative displacement and angle changes of the connection points under force or vibration in real time or near real time, in order to characterize the subtle deformation and mechanical state of the connection points.

[0178] Among them, the correlation analysis of small relative displacement and / or angle change data with deformation data to obtain the identification results of connection points where the connection stiffness or flexibility characteristics have hidden changes refers to the data fusion and pattern recognition (data fusion, pattern recognition or machine learning methods can be used) of micro-motion sensor data and deformation data (displacement, angle, strain, curvature, etc. from the hinge part and / or key flexible area of ​​the heavy object) and compared with historical benchmarks or theoretical models to locate suspicious connection points with hidden changes (microscopic changes that are difficult to observe directly on a macroscopic scale but will change the overall mechanical behavior).

[0179] Adjusting the local stiffness or flexibility parameters in the structural connection relationship based on the identification results means that once a change in the characteristics of the connection point is identified, the local stiffness or flexibility parameters at that point are corrected accordingly (if the stiffness decreases, the stiffness parameter is reduced; if the flexibility increases, the flexibility parameter is increased) so that the model is consistent with the actual mechanical state.

[0180] Therefore, based on the adjusted structural connection relationship and material flexibility characteristics, the multi-segment deformation topology is updated, and the updated multi-segment deformation topology is used to calculate the real-time deformation information of the outer contour of the heavy object. This means that the modified connection parameters and material flexibility characteristics (such as elastic modulus and Poisson's ratio) are applied to the multi-segment deformation topology to dynamically update the topology, and then the real-time deformation information of the outer contour of the heavy object is calculated accordingly, providing a more accurate input for subsequent field-of-view occlusion and synchronization control.

[0181] This application's solution, by placing micro-motion sensors at the connection points and performing correlation analysis with macroscopic deformation data, can promptly identify and quantify implicit changes in structural connection relationships and material flexibility properties. Based on this, local parameters are corrected online, and multi-segment deformation topologies are updated, making the real-time deformation calculation of the outer contour closer to the actual state, avoiding the accumulation of model-to-object deviations, and improving the accuracy of subsequent calculations of occluded areas by the onboard environmental perception sensors. Therefore, even under long-term use of heavy objects or complex operating conditions, the system can still maintain a high-fidelity representation of deformation, enhancing the accuracy, robustness, and safety of the dual-vehicle linkage cross-coupling synchronous control.

[0182] In some preferred embodiments:

[0183] In a dual-vehicle coordinated transport scenario, a tractor unit tows a large, heavy load connected to the tractor unit via multiple articulation points. Long-term operation and vibrations cause slight wear on the bearing at one of these articulation points, resulting in a slight decrease in local stiffness, though this is not macroscopically noticeable. Pre-positioned micro-sensors (micro-strain gauges and angle sensors) at this point continuously output minute relative displacement and angle changes; simultaneously, a vision sensor / LiDAR acquires overall deformation data. When the system detects that the minute angle change at this articulation point under the same load is approximately 5% larger than the baseline and exhibits a persistent deviation, it performs correlation analysis with the overall deformation data, identifying a hidden decrease in connection stiffness at this point. Based on this, the local stiffness coefficient of this connection point in the structural connection relationship is reduced by 5%, and the multi-segment deformation topology is updated in real-time, taking into account the material's flexibility. Based on the updated topology, the real-time deformation information of the load's external contour is recalculated to more accurately assess the obstruction areas of the onboard environmental perception sensors, thereby improving the reliability of path perception and obstacle avoidance, and ensuring the accuracy and safety of coordinated synchronous control under deformation conditions.

[0184] In another embodiment of this application, S4111-132 specifically includes:

[0185] S4111-1321: Acquire the ambient temperature at the connection point from the temperature sensor outputting the hinge joint of the heavy object;

[0186] S4111-1322: Based on ambient temperature and the coefficient of thermal expansion and contraction of materials, temperature compensation is performed on small relative displacement and / or angle change data to obtain compensated micro-motion data;

[0187] S4111-1323: Monitor the vibration frequency and impact load amplitude of the hinge joint of the heavy object and accumulate the action time, and evaluate the degree of fatigue damage based on the action time and the fatigue characteristic curve of the heavy object material;

[0188] S4111-1324: Based on the compensated micromotion data and the degree of fatigue damage, the local stiffness or flexibility parameters in the structural connection relationship are corrected;

[0189] S4111-1325: Based on the corrected parameters and material flexibility properties, update the multi-segment deformation topology and use the updated multi-segment deformation topology for real-time deformation information calculation of the outer contour of the heavy object.

[0190] Specifically, acquiring the ambient temperature of the connection point output by the temperature sensor at the hinge of the heavy object means placing temperature sensors (thermostats, thermocouples, or infrared temperature sensors, etc.) at the key hinge points or structural connections of the heavy object (such as hinge points, weld seams, bolt connections) to collect the ambient temperature at the connection point in real time, in order to characterize the thermal expansion and contraction effect of materials caused by temperature changes.

[0191] Specifically, based on the aforementioned ambient temperature and the material's coefficient of thermal expansion and contraction, temperature compensation is performed on the aforementioned minute relative displacement and / or angle change data to obtain compensated micro-motion data. This means that, based on the material's linear (or volumetric) expansion coefficient, the temperature-induced dimensional change is subtracted from / added to the original micro-motion data (minute relative displacement and / or angle change data), eliminating the systematic influence of temperature on sensor readings and making the micro-motion data more accurately reflect the intrinsic deformation of the structure.

[0192] In practical applications, monitoring the vibration frequency and impact load amplitude of the hinge joint of a heavy object and accumulating the duration of the impact, and assessing the degree of fatigue damage based on the aforementioned duration and the fatigue characteristic curve of the heavy object material, involves configuring vibration sensors / strain sensors at the hinge joint to continuously record the vibration frequency, impact load amplitude, and their accumulated time; then combining this with the material's SN curve or ε-N curve to calculate the current degree of fatigue damage and quantify the impact of long-term load history on the degradation of mechanical properties.

[0193] Furthermore, based on the compensated micro-motion data and the aforementioned fatigue damage degree, the local stiffness or flexibility parameters in the above structural connection relationship are corrected. This is done by combining the actual deformation after temperature compensation with the fatigue damage index as the basis, and adjusting the local stiffness or flexibility parameters of the corresponding connection points in the structural connection relationship (e.g., stiffness decreases, flexibility increases) so that the parameters reflect the combined effect of instantaneous deformation and cumulative damage.

[0194] Therefore, based on the corrected parameters and material flexibility properties, the multi-segment deformation topology is updated, and the updated multi-segment deformation topology is used to calculate the real-time deformation information of the outer contour of the heavy object: the corrected local parameters are written back to the model, and together with the material flexibility properties, the online update of the multi-segment deformation topology is driven, and the real-time deformation information of the outer contour of the heavy object is calculated accordingly, providing a more reliable input for subsequent calculation and control decisions of the field of view occlusion area.

[0195] The aforementioned mechanism addresses the insufficient accuracy of parameter adjustments relying solely on raw micro-motion data through two links: temperature compensation and fatigue damage assessment. The former eliminates measurement drift caused by temperature, while the latter explicitly characterizes performance degradation caused by long-term loads. The resulting parameter correction and topology update enable the model to more comprehensively represent the actual mechanical state and more accurately calculate the real-time deformation of the external contour, thereby improving the accuracy, robustness, and safety of the estimation of the field of view obstruction area and the synchronous control of the dual-vehicle linkage cross-coupling.

[0196] In some preferred embodiments:

[0197] Taking a crawler crane as an example, its boom operates for a long time under different ambient temperatures and operating loads. Micro motion sensors and temperature sensors are pre-arranged at each hinge point to obtain data on minute relative displacement / angle changes and ambient temperature online. When the original micro motion readings are偏大 due to material shrinkage under cold conditions, the system uses temperature and material expansion coefficients for temperature compensation to obtain more accurate micro motion data. At the same time, during the frequent lifting and slewing of the boom, the system records the vibration frequency, impact load amplitude, and cumulative action time, and evaluates the fatigue damage degree in combination with the material S-N / ε-N curve. If a hinge point is evaluated as moderately fatigued and the micro motion data after compensation shows a slight incremental deformation, the system will lower / raise the local stiffness or flexibility parameters of this point, and then update the multi-segment deformation topology. The real-time deformation information of the external contour of the heavy object calculated on this basis can be more accurately used for the evaluation of the field of view occlusion area, supporting path perception and safety control in the scenario of two-vehicle coordinated operation.

[0198] Referring to Figure 2 , the specific implementation manner of the present application also discloses a two-vehicle coordinated cross-coupled synchronization control system, including:

[0199] A monitoring module 1 for collecting quality indicators of the wireless communication link of the partner vehicle, where the quality indicators at least include received signal strength, packet delay fluctuation, packet loss rate, and data freshness;

[0200] A weight factor calculation module 2 for calculating the weight factor of communication quality based on the quality indicators;

[0201] An estimated state calculation module 3 for obtaining the reported state of the partner vehicle, generating the estimated state of the partner vehicle according to the latest credible reported state of the partner vehicle, the vehicle kinematic model, and odometer and / or inertial measurement data, and the reported state includes at least one of position, speed, and / or attitude;

[0202] A fusion module 4 for linearly weighted fusion of the reported state and the estimated state of the partner vehicle according to the weight factor to obtain the effective partner vehicle state, where when the weight factor decreases, the weight of the reported state is reduced and the weight of the estimated state of the partner vehicle is increased;

[0203] An error calculation module 5 for calculating the synchronization error vector based on the real-time state of the vehicle itself and the effective partner vehicle state, and the synchronization error vector includes at least one of position, speed, and / or attitude error;

[0204] An instruction generation module 6 for generating control instructions according to the synchronization error vector to drive the power and / or steering actuator to implement cross-coupled synchronization control.

[0205] This application's system, through a modular design, effectively integrates functions such as communication quality monitoring, status estimation and fusion, synchronization error calculation, and control command generation. The system can assess the wireless communication link quality of partner vehicles in real time and dynamically adjust the fusion weights of reported and estimated statuses based on the assessment results, thus ensuring accurate and reliable partner vehicle status even when communication is unstable. Consequently, the system can accurately calculate the synchronization error vector and generate appropriate control commands to drive the vehicle actuators to implement refined cross-coupling synchronization control. This effectively avoids pseudo-synchronization errors and error corrections caused by communication problems in traditional systems, significantly improving the synchronization accuracy, system robustness, and operational safety of dual-vehicle linked transport.

[0206] The above are merely embodiments of this application and are not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for synchronous control of two vehicles in a cross-coupling manner, characterized in that, include: The quality indicators of the wireless communication link of the partner vehicle are collected, and the quality indicators include at least the received signal strength, data packet delay fluctuation, data packet loss rate and data freshness; Calculate the weighting factor for communication quality based on the aforementioned quality indicators; Obtain the reported status of the partner vehicle, and generate the estimated status of the partner vehicle based on the latest reliable reported status of the partner vehicle, the vehicle kinematic model, and odometer and / or inertial measurement data. The reported status includes at least one of position, speed, and / or attitude. The reported status and the estimated status of the partner vehicle are linearly weighted and fused according to the weight factor to obtain the effective partner vehicle status. When the weight factor decreases, the weight of the reported status is reduced and the weight of the estimated status of the partner vehicle is increased. Based on the real-time status of this vehicle and the status of the effective partner vehicle, a synchronization error vector is calculated, wherein the synchronization error vector includes at least one of position, speed and / or attitude error. Control commands are generated based on the synchronization error vector to drive the power and / or steering actuators to implement cross-coupled synchronization control.

2. The dual-vehicle linkage cross-coupling synchronous control method according to claim 1, characterized in that, The reported status and the estimated status of the partner vehicle are linearly weighted and fused according to the weighting factor to obtain the effective partner vehicle status, including: Monitor the vehicle's motion status and preset path information, and assess the situational risk level based on the motion status and preset path information; Based on the situational risk level and the quality indicators, calculate the communication confidence level of the reporting status; Based on the risk level of the scenario and the vibration characteristics of the vehicle obtained from the detection, the sensor confidence level for the estimated state of the partner vehicle is calculated. When the communication confidence level is lower than a preset first threshold and the sensor confidence level is lower than a preset second threshold, the weighting factor is adjusted according to the situational risk level, and the reported status and the estimated status of the partner vehicle are linearly weighted and fused according to the adjusted weighting factor to obtain the effective partner vehicle status.

3. The dual-vehicle linkage cross-coupling synchronous control method according to claim 2, characterized in that, After monitoring the vehicle's motion status and preset path information, and assessing the situational risk level based on the motion status and preset path information, the method further includes: Obtain local map information of the environment surrounding the vehicle; The local map information is compared with the preset path information to obtain a comparison result, so as to identify abnormal parts of the preset path information that are blocked by temporary obstacles, occupied, or deviate from the actual passable area. Based on the comparison results, the preset path information is updated, and the situational risk level is adjusted according to the updated preset path information; The thresholds and weighting factors used for calculating communication confidence and sensor confidence are scheduled based on the adjusted situational risk level.

4. The dual-vehicle linkage cross-coupling synchronous control method according to claim 3, characterized in that, Obtain local map information of the environment surrounding the vehicle, including: Based on the size of the load being transported and the vehicle's motion state, the field of view obstruction area of ​​the on-board environmental perception sensor is calculated. Based on preset path information and the known distribution of fixed obstacles, identify fixed occlusion areas; In the obstructed field of view area and / or the fixed obstructed area, supplementary local map information is obtained; The local map information is fused with the supplementary local map information to generate blind spot-free local map information, which is then compared with the preset path information.

5. The dual-vehicle linkage cross-coupling synchronous control method according to claim 1, characterized in that, The reported status and the estimated status of the partner vehicle are linearly weighted and fused according to a weighting factor to obtain the effective partner vehicle status, including: Perform timestamp freshness verification on the reported status of partner vehicles. When the data freshness exceeds a preset time threshold, the reported status is judged as expired. When the report status is determined to be expired or a preset number of consecutive packet losses are detected within a preset time window, a short-term retransmission is triggered. If the updated report status is not obtained within the maximum waiting time, the weight factor will be reduced to a value no less than the lower limit according to the preset weight reduction rule. The reported status and the estimated status of the partner vehicle are linearly weighted and fused according to the reduced weighting factor to obtain the effective partner vehicle status.

6. The dual-vehicle linkage cross-coupling synchronous control method according to claim 4, characterized in that, Based on the dimensions of the transported load and the vehicle's motion state, the field-of-view obstruction area of ​​the onboard environmental perception sensor is calculated, including: Obtain real-time deformation information of the outer contour of the heavy object; The three-dimensional geometric model of the object is dynamically updated based on the object's dimensions and the real-time deformation information. The system acquires the installation location and field-of-view parameters of the vehicle's environmental perception sensors, and calculates the vehicle's real-time pose information based on the vehicle's motion state. Based on the dynamically updated 3D geometric model, the installation location and field of view parameters, and the real-time pose information of the vehicle, the field of view occlusion area is obtained through 3D geometric calculation.

7. The dual-vehicle linkage cross-coupling synchronous control method according to claim 6, characterized in that, Obtain real-time deformation information of the outer contour of the heavy object, including: Acquire deformation data of the hinged joints and / or key flexible areas of the heavy object; Based on the deformation data, combined with the structural connection relationship and material flexibility characteristics of the heavy object, a multi-segment deformation topology structure of the heavy object is constructed. Based on the multi-segment deformation topology, the real-time deformation information of the outer contour of the heavy object is calculated.

8. The dual-vehicle linkage cross-coupling synchronous control method according to claim 7, characterized in that, Before calculating the real-time deformation information of the object's outer contour, the following steps are also included: Acquire minute relative displacement and / or angle change data output by micro-motion sensors pre-placed at the connection points of the heavy structure; The correlation analysis between the tiny relative displacement and / or angle change data and the deformation data is performed to obtain the identification results of connection points where the connection stiffness or flexibility characteristics have undergone implicit changes. Based on the identification results, adjust the local stiffness or flexibility parameters in the structural connection relationship; Based on the adjusted structural connection relationship and material flexibility properties, the multi-segment deformation topology is updated, and the updated multi-segment deformation topology is used to calculate the real-time deformation information of the outer contour of the heavy object.

9. The dual-vehicle linkage cross-coupling synchronous control method according to claim 8, characterized in that, Based on the identification results, the local stiffness or flexibility parameters in the structural connection relationship are adjusted, including: The ambient temperature at the connection point is obtained from the temperature sensor at the hinge of the heavy object. Based on the ambient temperature and the coefficient of thermal expansion and contraction of the material, temperature compensation is performed on the small relative displacement and / or angle change data to obtain the compensated micro-motion data. The vibration frequency and impact load amplitude of the hinge joint of the heavy object are monitored and the duration of action is accumulated. The degree of fatigue damage is assessed based on the duration of action and the fatigue characteristic curve of the heavy object material. Based on the compensated micro-motion data and the degree of fatigue damage, the local stiffness or flexibility parameters in the structural connection relationship are corrected; Based on the corrected parameters and material flexibility properties, the multi-segment deformation topology is updated, and the updated topology is used to calculate the real-time deformation information of the outer contour of the heavy object.

10. A dual-vehicle linkage cross-coupling synchronous control system, characterized in that, include: The monitoring module is used to collect quality indicators of the wireless communication link of the partner vehicle. The quality indicators include at least received signal strength, data packet delay fluctuation, data packet loss rate and data freshness. The weighting factor calculation module is used to calculate the weighting factor of communication quality based on the quality index. The estimated state calculation module is used to obtain the reported state of the partner vehicle and generate the estimated state of the partner vehicle based on the latest reliable reported state of the partner vehicle, the vehicle kinematic model, and odometer and / or inertial measurement data. The reported state includes at least one of position, speed, and / or attitude. The fusion module is used to perform linear weighted fusion of the reported state and the estimated state of the partner vehicle according to the weight factor to obtain the effective partner vehicle state, wherein when the weight factor decreases, the weight of the reported state is reduced and the weight of the estimated state of the partner vehicle is increased. The error calculation module is used to calculate a synchronization error vector based on the real-time status of the vehicle and the status of the effective partner vehicle. The synchronization error vector includes at least one of position, speed and / or attitude error. The instruction generation module is used to generate control instructions based on the synchronization error vector to drive the power and / or steering actuators to implement cross-coupled synchronous control.

Images