Road stop line signal lamp countdown simulation method based on digital twin engine

Abstract

This invention relates to the field of intelligent traffic control and discloses a method for simulating road stop line traffic light countdowns based on a digital twin engine. The method includes: acquiring basic vehicle safety information and comparing it with physical observation coordinates; filtering forged messages using spatial coordinate residuals; instantiating a virtual state machine to replay historical actuation events; driving detector state updates in chronological order and calculating the states of the gap timer and the maximum green light timer; extracting virtual-real phase transition residuals and internal reference states; calculating physical braking limits based on instantaneous kinematic parameters; constructing an anchored deceleration curve and generating a reconstructed trajectory by combining historical countdowns; introducing the reconstructed trajectory, actuation information of other vehicles in the same phase, and competing phase request pulses to perform dual-threaded deduction, generating absolute deduction timestamp boundaries; and using a monotonically decreasing constraint function to process the original remaining time to generate the final countdown and send it out. This method can reduce the risk of countdown jumps and reverse increases, and improve the early warning transmission guarantee capability under abnormal operating conditions.

Description

Technical Field

[0001] This invention relates to the field of intelligent traffic control, specifically to a method for simulating the countdown of road stop line traffic lights based on a digital twin engine. Background Technology

[0002] With the development of vehicle-to-everything (V2X) technology, traffic signal countdown systems are typically deployed at sensor-controlled intersections to assist drivers or autonomous vehicles in decision-making. Existing countdown systems primarily receive basic safety messages broadcast by vehicles via roadside communication units, combining this with traffic flow data collected by physical sensing devices at the intersection to deduce the current signal phase timing scheme. The system dynamically adjusts the signal light duration based on vehicle arrivals and sends the calculated remaining countdown value to the onboard terminal, providing a time reference for vehicle longitudinal speed planning and stop / start operations.

[0003] In practical applications, because the timing scheme of inductively controlled traffic signals relies on real-time detected traffic flow conditions, the random arrival of vehicles can cause fluctuations in the phase transition time calculated by the system. These fluctuations can lead to jumps or even inverse increases in the countdown values ​​sent to vehicles, preventing the autonomous driving system from generating a smooth longitudinal control strategy. Furthermore, current countdown systems, when processing multi-source input data, typically use trajectory coordinates reported by the vehicle-to-everything (V2X) network directly, lacking a mechanism to verify the authenticity of these messages. When maliciously forged trajectory data is received, the system is prone to using false information as a basis for timing, leading to misjudgments in the virtual state machine's timing.

[0004] Existing traffic signal countdown systems lack a deep integration mechanism with the underlying communication layer when dealing with complex traffic safety scenarios. When a vehicle experiences brake failure or malicious lane-jumping as it approaches the stop line at an intersection, the system still packages and sends signal phases and timing messages according to standard business logic. This approach cannot readily allocate limited roadside communication bandwidth, resulting in high-priority warning messages for lane-crossing collision avoidance lacking a dedicated, low-latency transmission channel, making it difficult to guarantee timely communication at intersections under abnormal conditions. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a road stop line signal light countdown simulation method based on a digital twin engine. This method solves the problem that existing sensor-controlled intersection signal countdown systems, when facing random traffic flow, experience jumps or reverse increases in the countdown values ​​sent to vehicles due to dynamic vehicle arrivals, which in turn interferes with the longitudinal control algorithm of autonomous driving. It also addresses the technical issues of existing systems being susceptible to timing misjudgments due to falsified trajectory data when receiving vehicle network data, and the lack of underlying direct intervention and communication resource scheduling preemption mechanisms when vehicles approach intersections and experience braking failure or runaway events.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a road stop line traffic light countdown simulation method based on a digital twin engine, comprising the following steps:

[0007] S10, obtain the basic safety information and physical entity trajectory of the target vehicle, extract the absolute coordinate position in the basic safety information and the physical observation coordinate in the physical entity trajectory, compare the absolute coordinate position and the physical observation coordinate to obtain the spatial coordinate residual, and retain the basic safety information of the vehicle when the spatial coordinate residual does not exceed the anti-deception tolerance threshold.

[0008] S20, instantiate a virtual state machine and inject historical actuation events for state synchronization playback. Drive the corresponding detector state update one by one according to the absolute timestamp order of the historical actuation events. When the phase is in the green light state and a vehicle arrival event or detector occupation event is detected, reset the gap timer of the corresponding detector to a preset unit to extend the green light time. When no new vehicle arrival event is detected, decrease the gap timer according to the preset time step. When the phase enters the green light state, start the maximum green light timer to accumulate until the maximum green light threshold is reached. Compare the phase transition calculation timestamp output by the virtual state machine with the real physical timestamp issued by the traffic signal controller to extract the virtual and real phase transition residuals and extract the reference state of the internal register of the virtual state machine.

[0009] S30: Calculate the physical braking limit of the target vehicle based on the instantaneous kinematic parameters of the target vehicle and the remaining spatial distance to the stop line. When the physical braking limit is within the safety threshold, construct the anchored deceleration curve by combining the historical countdown value of the previous cycle, and generate the reconstructed trajectory based on the anchored deceleration curve.

[0010] S40 introduces the reconstructed trajectory with the internal register reference state as the initial condition to perform dual-thread extrapolation. The reconstructed trajectory of the target vehicle, the measured actuation events of other observed vehicles in the phase of the target vehicle or the supplementary arrival pulses extrapolated from the most recent actuation records, and the background traffic flow request pulses of the competing phase are input into the virtual state machine to establish the original time boundary. The virtual-real phase transition residual is used to correct the original time boundary and generate the absolute extrapolation timestamp boundary.

[0011] S50 extracts the original remaining time based on the absolute extrapolation timestamp boundary, uses a monotonically decreasing constraint function to process the original remaining time and the historical countdown value of the previous cycle to generate the final output countdown value, and sends the final output countdown value to the vehicle terminal.

[0012] Furthermore, in S10, when the spatial coordinate residual is greater than the anti-spoofing tolerance threshold, the vehicle basic safety message is determined to be a maliciously forged message. The maliciously forged message is isolated and removed from the input source of the virtual state machine, and the operation of the virtual state machine is maintained entirely by physical observation coordinates. The false trajectory is blocked from the data source, and the virtual state machine is prevented from being interfered with and causing timing jumps.

[0013] Furthermore, in S20, the process of extracting the virtual-real phase transition residual is as follows: search for the last valid phase switching event within the time interval covering the historical actuation events as the comparison anchor point, and extract the real physical timestamp of the valid phase switching event broadcast by the traffic signal controller; record the phase transition calculation timestamp of the virtual state machine triggering the corresponding phase transition logic when processing the same historical actuation event; subtract the real physical timestamp from the phase transition calculation timestamp to obtain the virtual-real phase transition residual, so as to eliminate the clock deviation between the physical entity and the twin model.

[0014] The historical actuation events are injected into the virtual state machine one by one in absolute timestamp order. When a phase is in the green light state, if a vehicle arrival event or a detector occupation event is detected, the gap timer of the corresponding detector is reset to a preset unit to extend the green light time. When no new vehicle arrival event is detected, the gap timer is decremented according to a preset time step. When the phase enters the green light state, the maximum green light timer is started to accumulate and continues to accumulate to the maximum green light threshold. Thus, the gap timer value and the maximum green light timer value at the current moment are calculated.

[0015] Furthermore, in S30, after calculating the physical braking limit of the target vehicle, it also includes: when the positive rate of change of instantaneous acceleration in the instantaneous kinematic parameters exceeds the preset safety tolerance, and the absolute value of the required deceleration calculated based on the instantaneous velocity and remaining spatial distance in the instantaneous kinematic parameters exceeds the preset maximum deceleration envelope, it is determined that a braking failure and vehicle slippage or malicious red light running and checkpoint breaching event has occurred, triggering the underlying hardware interrupt mechanism to abolish and reconstruct the trajectory generation logic, and pushing a channel preemption command to broadcast the highest priority stop line crossing collision avoidance warning message.

[0016] Furthermore, the steps for constructing the anchored deceleration curve and generating the reconstructed trajectory in S30 include: reading the constraint feedback flag of the previous cycle; when the constraint feedback flag triggers the forced hovering logic and it is determined that the target vehicle cannot safely pass through the intersection within the current green light cycle, using the remaining spatial distance as the spatial physical boundary and the historical countdown value of the previous cycle as the time dimension boundary, setting the starting point velocity to the instantaneous velocity in the instantaneous kinematic parameters, strictly setting the ending point velocity to zero, applying acceleration continuity constraints to perform polynomial curve fitting to generate the anchored deceleration curve, and using the anchored deceleration curve to cover the original kinematic prediction model to generate the reconstructed trajectory within the digital twin engine.

[0017] Furthermore, the steps for establishing the original time boundary in S40 include: configuring the interrupted traffic flow state in the first calculation thread; after the target vehicle and the currently observed vehicles in the phase where the target vehicle is located have completed the release, cutting off the subsequent supplementary analog pulse injection of all phase-related detectors; and recording the first original extrapolation timestamp when the gap timer is reset to zero and the gap timing end logic is triggered.

[0018] In the second calculation thread, a continuous saturated traffic flow state is configured, and simulated vehicle arrival pulses are continuously injected into all relevant detectors in the phase where the target vehicle is located. When the green light duration reaches the maximum green light limit and the maximum green light timeout logic is triggered, the second original simulation timestamp is recorded.

[0019] The first original projection timestamp and the second original projection timestamp are combined as the original time boundary. The steps in S4 to generate the absolute projection timestamp boundary by correcting the original time boundary using the virtual-real phase transition residual include: executing the linear translation addition correction logic, adding the virtual-real phase transition residual to the first original projection timestamp and the second original projection timestamp respectively, and generating the absolute lower boundary timestamp and the absolute upper boundary timestamp accordingly. The two are combined to form the absolute projection timestamp boundary.

[0020] Furthermore, the steps in S50 for extracting the original remaining time based on the absolute extrapolation timestamp boundary include: selecting the absolute lower boundary timestamp as the reference calculation timestamp in the green light countdown scenario; selecting the absolute upper boundary timestamp as the reference calculation timestamp in the red light countdown scenario; and subtracting the current physical timestamp from the reference calculation timestamp to calculate the original remaining time.

[0021] Furthermore, the steps in S50 for processing the original remaining time and the historical countdown values ​​of the previous cycle using the monotonically decreasing constraint function include:

[0022] Calculate the difference between the historical countdown value of the previous period and the preset time step to obtain the maximum decreasing countdown;

[0023] The original remaining time is compared with the maximum countdown value, and the smaller value between the original remaining time and the maximum countdown value is selected as the final output countdown value.

[0024] When the final output countdown value is the maximum decreasing countdown, it is determined that a forced smooth intervention has been applied to the target vehicle, and the constraint feedback flag corresponding to the target vehicle is set to the triggered state; otherwise, the constraint feedback flag corresponding to the target vehicle is set to the non-triggered state.

[0025] Furthermore, the step in S50 that sends the final countdown value to the vehicle terminal also includes:

[0026] The final output countdown value is encapsulated into signal phase and timing message;

[0027] The system monitors the channel resource status of the underlying communication interface. When it detects that the roadside communication unit is in a channel preemption state, it prioritizes blocking the transmission link of the final output countdown value at the application layer, stops the encapsulation and push process of signal phase and timing messages, and relinquishes communication bandwidth to ensure the exclusive transmission channel of the stop line crossing anti-collision warning message.

[0028] This invention provides a method for simulating countdown times for road stop line traffic lights based on a digital twin engine. It offers the following advantages:

[0029] 1. This invention compares the absolute coordinates in the vehicle's basic safety messages with the physical observation coordinates in the physical entity's trajectory to calculate the spatial coordinate residual, thereby filtering out malicious forged messages that exceed the anti-spoofing tolerance threshold. This mechanism, based on cross-validation of physical perception and communication data, can block the interference of false trajectories on the virtual state machine from the data input end, reduce the risk of timing misjudgment and timing jumps induced by false data in the virtual state machine, and improve the reliability of the basic data on which countdown simulations rely.

[0030] 2. This invention injects historical actuation events into a virtual state machine in chronological order to calculate the states of the gap timer and the maximum green light timer. It then combines this with the target vehicle's reconstructed trajectory, measured actuation or supplementary arrival pulses of other vehicles within the target vehicle's phase, and competing phase request pulses to perform parallel simulations, establishing absolute simulation timestamp boundaries. Finally, a monotonically decreasing constraint function is used to constrain the original remaining time and the maximum decreasing countdown. This combination of features improves the ability of the countdown simulation at inductively controlled intersections to represent implicit control states, suppresses countdown fluctuations caused by random vehicle arrivals, and reduces the probability of sudden jumps or reverse increases in traditional countdown values, thus providing a smoother and more stable time reference for vehicle longitudinal speed planning.

[0031] 3. This invention calculates the physical braking limit of the target vehicle and monitors changes in instantaneous kinematic parameters. When a vehicle is determined to have experienced braking failure or a collision, it triggers a low-level hardware interrupt mechanism and actively pushes a channel preemption command. This low-level intervention mechanism can prioritize blocking the transmission link of regular countdown messages under abnormal operating conditions, prompting roadside communication units to prioritize the relinquishment of communication bandwidth. This provides low-latency transmission conditions for stop-line collision warning messages, improving communication response capabilities and operational safety in abnormal intersection scenarios. Attached Figure Description

[0032] Figure 1 This is a flowchart of the method of the present invention;

[0033] Figure 2This is a flowchart of the multi-source sensing data acquisition and historical state backtracking mechanism of the present invention;

[0034] Figure 3 This is a flowchart of the implicit state pre-playback and virtual-real phase transition residual extraction mechanism of the present invention;

[0035] Figure 4 This is a flowchart of the microscopic trajectory intention anchoring and reconstruction mechanism of the present invention;

[0036] Figure 5 This is a flowchart of the digital twin parallel advance extrapolation and feedforward residual compensation mechanism of the present invention;

[0037] Figure 6 This is a flowchart of the monotonically decreasing constraint mapping and closed-loop feedback generation mechanism of the present invention;

[0038] Figure 7 This is a comparison curve of the countdown smoothing effect of the present invention. Detailed Implementation

[0039] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] See Figure 1 , Figure 1 This is a flowchart of a road stop line traffic light countdown simulation method based on a digital twin engine according to an embodiment of the present invention. The present invention provides a road stop line traffic light countdown simulation method based on a digital twin engine. This method is executed cyclically by edge computing nodes within a discrete computing cycle at a preset time step, and includes the following steps:

[0041] S10: Perform multi-source perception data acquisition and historical state backtracking. Obtain historical actuation events of the intersection through roadside perception devices and build a sliding window of fixed time length in memory. Receive basic vehicle safety messages sent by the vehicle terminal of the target vehicle through the roadside communication unit to extract the instantaneous kinematic parameters of the vehicle. Read the countdown value and constraint feedback flag issued to the target vehicle in the previous cycle. Perform spatiotemporal fitting cross-verification between the instantaneous kinematic parameters of the target vehicle and the physical entity trajectory captured by the roadside perception devices. If the spatial coordinate residual exceeds the preset anti-deception tolerance threshold, determine that the basic safety message is a maliciously forged message and isolate and remove it from the input source of the digital twin engine to prevent the virtual state machine from being poisoned by false trajectories and causing abnormal extension of green lights or timing jumps due to human manipulation.

[0042] S20, perform implicit state pre-playback and virtual-real phase transition residual extraction, instantiate a virtual state machine in the internal digital twin engine, inject the actuation event sequence in the sliding window into the virtual state machine for state synchronization playback, and drive the corresponding detector state update one by one according to the absolute timestamp order of the actuation event sequence, and discretly calculate the gap timer and maximum green light timer of each phase based on the vehicle arrival event, detector occupation event and detector release event.

[0043] Specifically, when the target phase enters the green light phase, the maximum green light timer is started to accumulate. When a vehicle arrival or detector occupancy event is detected, the gap timer is reset to a preset unit to extend the green light time. If no new vehicle arrival event is detected within the subsequent distance step, the gap timer is decremented by a preset time step until the gap timer returns to zero or the maximum green light timer reaches the maximum green light threshold. This obtains the gap timer value and the maximum green light timer value of the virtual state machine at the current moment. The phase transition calculation timestamp of the virtual state machine is compared with the real physical timestamp output by the traffic signal controller to calculate the virtual-real phase transition residual and extract the internal register reference state of the virtual state machine to advance to the current moment.

[0044] S30 executes micro-trajectory intention anchoring and reconstruction, extracts the instantaneous acceleration of the target vehicle and the remaining spatial distance to calculate the real-time physical braking limit. If the positive change in instantaneous acceleration exceeds the maximum deceleration envelope allowed by the current remaining spatial distance, it is determined that the target vehicle has experienced brake failure and rollback or malicious red light running and checkpoint breaching. The underlying hardware interrupt mechanism is immediately triggered to abolish the smooth reconstruction logic of the target vehicle, and a channel preemption command is pushed to the roadside communication unit to broadcast the highest priority stop line crossing collision warning message to the intersection area. If the physical braking limit is within the safety threshold, the intervention branch continues to be executed according to the state value of the constraint feedback flag. When the constraint feedback flag triggers the forced hovering logic, the anchoring deceleration curve of the vehicle is constructed by combining the remaining spatial distance and the countdown value of the previous cycle. The reconstructed trajectory is generated by covering the original kinematic prediction model with the anchoring deceleration curve in the virtual space of the digital twin engine.

[0045] S40 performs parallel advance simulation and feedforward residual compensation using digital twins. Taking the internal register reference state as the initial condition, it injects the reconstructed trajectory of the target vehicle, the measured trajectory of other observed vehicles in the phase where the target vehicle is located, or the arrival pulse extrapolated based on the most recent actuation record, and the background traffic flow request pulse of the competing phase into the virtual state machine to form the simulation input for the phase-level release demand of the entire intersection. It starts ultra-real-time dual-thread simulation towards the future time axis, establishes the original time boundaries for triggering the virtual state machine gap timing end condition and the maximum green light timing end condition, and uses the virtual-real phase transition residual to perform linear translation correction on the original time boundaries to generate the absolute simulation timestamp boundary.

[0046] S50 executes monotonically decreasing constraint mapping and closed-loop feedback generation, extracts the safety margin remaining time based on the absolute extrapolation timestamp boundary, inputs the safety margin remaining time into the monotonically decreasing constraint function, combines it with the countdown value issued in the previous cycle to generate the final output countdown value for the current cycle, updates the constraint feedback flag bit used for trajectory reconstruction judgment in the next cycle, and sends the final output countdown value to the target vehicle's on-board terminal after being encapsulated by the communication protocol. When the roadside communication unit is in a channel preemption state, it prioritizes blocking the transmission link of the final output countdown value to ensure low-latency transmission of collision avoidance warning messages.

[0047] See Figure 2 , Figure 2 This is a flowchart of a multi-source sensing data acquisition and historical state backtracking mechanism according to an embodiment of the present invention.

[0048] In this embodiment, the edge computing node advances the discrete computing cycle with a preset time step, for example, setting the time step to 100 milliseconds. Within the current computing cycle, the node receives raw sensing data sent by the roadside sensing device through the communication interface. This data includes physical actuation records of vehicles crossing the detection area inside the intersection. To manage this time-series data, the edge computing node obtains the current physical timestamp of the system and allocates storage space in the memory area to construct a data sliding window. As a preferred approach, the time length of the data sliding window is set to a fixed time constant, for example, the duration covering at least one complete traffic signal cycle. As the system clock advances, the system dynamically updates the time interval covered by the data sliding window and stores the extracted physical actuation records in chronological order, while simultaneously clearing historical data exceeding the time interval to maintain a dynamic balance in the memory space.

[0049] In addition, the edge computing node also receives basic vehicle safety messages broadcast by the on-board terminals of target vehicles within the communication coverage area via the roadside communication unit. Specifically, the system decodes the basic safety messages to extract the instantaneous kinematic parameters of the target vehicle at the current physical timestamp, including the target vehicle's absolute coordinate position, instantaneous velocity, and instantaneous acceleration. The message decoding process of the basic safety messages and the conversion calculation from the vehicle's latitude and longitude coordinates to the plane projection coordinates can be implemented by those skilled in the art using existing vehicle network communication protocol parsing stacks and conventional coordinate system transformation algorithms. The specific process is well-known in the field and will not be elaborated here.

[0050] Subsequently, the system accesses the local cache space and, based on the target vehicle's unique vehicle identifier, such as a temporary pseudo-random code or underlying physical MAC address defined by the vehicle networking standard, backtracks to retrieve the historical status data generated by the system for the target vehicle in the previous calculation cycle. It then reads and obtains the historical countdown values ​​and corresponding constraint feedback flags, which are used as input conditions for subsequent trajectory reconstruction judgment.

[0051] Considering the potential for cyberattacks in open communication environments, and to prevent false trajectory data from inducing abnormally prolonged green lights or timing jumps in the virtual state machine, this embodiment introduces a spatiotemporal fitting cross-validation mechanism. The basic principle of this mechanism is to use sensor observations, which are unforgeable in the physical world, to verify the authenticity of radio messages in the digital space.

[0052] In practice, edge computing nodes extract physical observation coordinates aligned with the current physical timestamp from the physical entity trajectories captured by roadside sensing devices, such as millimeter-wave radar or video cameras. It should be noted that if the corresponding physical observation coordinates are missing at the current moment due to physical factors such as blind spots, the system will suspend the current cross-validation and retain the vehicle's trust assessment status from the previous cycle. After successfully acquiring data from both sides, the system compares the absolute coordinates parsed from the vehicle's basic safety message with these physical observation coordinates, calculating the spatial coordinate residual between the two. The calculation formula is as follows:

[0053] ;

[0054] In the formula, Represents spatial coordinate residuals. and This represents the horizontal and vertical coordinates of the absolute coordinate position extracted from the vehicle's basic safety messages. and The horizontal and vertical coordinates represent the physical observation coordinates captured by the roadside sensing device.

[0055] Based on this, the system will calculate the spatial coordinate residuals. The value is compared with a preset anti-spoofing tolerance threshold. This anti-spoofing tolerance threshold is usually calibrated based on the inherent positioning error of the roadside sensor and the physical width of the standard lane, for example, it is set as a fixed value in the range of 1.5 meters to 3.0 meters. When the spatial coordinate residual is greater than the anti-spoofing tolerance threshold, the edge computing node determines that the basic security message is a malicious forged message, and then isolates and removes the malicious forged message from the input source of the digital twin engine, interrupts the twin data injection process of the target vehicle in the current cycle, and relies entirely on the physical perception data collected by the roadside perception device to maintain the basic operation of the virtual state machine.

[0056] Conversely, when the spatial coordinate residual is less than or equal to the anti-spoofing tolerance threshold, the edge computing node determines that the basic security message is valid and inputs it along with historical state data into the subsequent data processing stage, thereby ensuring the completeness and security of the entire data acquisition and backtracking logic.

[0057] See Figure 3 , Figure 3 This is a flowchart of the implicit state pre-playback and virtual-real phase transition residual extraction mechanism according to an embodiment of the present invention.

[0058] In this embodiment, the edge computing node initiates the construction of the virtual environment by allocating independent computing threads within its internal digital twin engine. In practical engineering, since physical traffic signal controllers typically operate as black-box systems, only broadcasting the current light color status and unable to directly output internal interval timers or maximum green light timer values, it is necessary to construct a digital twin copy to invert these implicit states.

[0059] To achieve digital replication of the physical control logic, edge computing nodes instantiate a standardized inductive control state machine within this computing thread. This virtual state machine essentially serves as a copy of the control logic operation of the physical traffic signal controller at the intersection. Internally, it is configured with a phase loop structure, green light interval time, and extended green light parameters for each detector that are completely identical to those of the real intersection. For the basic phase transition logic operation rules within the virtual state machine, those skilled in the art can refer to existing standard inductive control architectures for implementation; its basic operating mechanism is well-known in the field and will not be elaborated upon here.

[0060] In addition, during the system cold start phase, due to the lack of historical benchmarks, the system will set the initial virtual-real phase transition residual to zero by default, and then dynamically update it after capturing the first valid phase switching event.

[0061] Based on this, to reconstruct the aforementioned implicit working state, the edge computing node reads the standard event stream sequence within the previously constructed data sliding window. This standard event stream sequence typically includes the occupancy and departure pulse signals generated when a physical vehicle passes over the buried detection coil or radar virtual coil. Because the state machine's state transition logic has strict time-series dependencies, the system does not directly jump to the current time for computation. Instead, it gradually injects the standard event stream sequence into the standardized sensing control state machine according to the absolute timestamps of each actuation event.

[0062] Because the state transition logic of the state machine has strict time-series dependencies, the system does not directly jump to the current time for calculation. Instead, it gradually injects the standard event stream sequence into the standardized sensor control state machine according to the absolute timestamp of each actuation event.

[0063] To clarify the process of deriving the gap timer and the maximum green light timer from the actuation event, let the distance from the walk be numbered as follows: The corresponding physical time is ; Indicates the first Individual walk-in detector Was a vehicle arrival or possession event detected? Indicates the current phase at the th Is the area under a green light for taking a walk? This indicates the remaining value of the gap timer. This represents the maximum accumulated value of the green light timer. This indicates that the unit has extended the green light time. This represents the maximum green light threshold. Therefore, the timer update rule for the virtual state machine during event replay is:

[0064] ;

[0065] ;

[0066] ;

[0067] when When the timer for the time interval ends, the timer for the time interval is determined. or When the maximum green light timer ends, the edge computing node continuously executes the above discrete updates according to the time sequence of the actuation events within the sliding window until the virtual time advances to the current physical timestamp, thereby obtaining the gap timer value corresponding to each detector at the current moment and the cumulative value or remaining value of the maximum green light timer for the current phase.

[0068] As a preferred approach, a virtual state machine executes rapid replay and update of its internal state at a computational speed higher than that of real time, based on the original time interval of the input sequence. For example, it can mobilize the highest priority idle computing power of edge computing nodes to execute at a speed of hundreds of times, thereby reproducing the current accumulated state of the physical device in the digital space.

[0069] To eliminate the unavoidable communication delays and computational scheduling latency between physical devices and the digital twin system, and to obtain a high-precision synchronization reference, the system needs to find a spatiotemporally aligned anchor point. During the playback process, edge computing nodes synchronously parse the actual signal phases and timing message sequences received from the physical traffic signal controller via the communication interface.

[0070] Specifically, the system searches for the last valid phase transition event within the time interval covered by the sliding window, sets it as the comparison anchor point, and extracts the actual physical timestamp of the event broadcast by the physical traffic signal controller. Simultaneously, the system records the computation timestamps of the corresponding phase transition logic triggered by the virtual state machine when processing the same standard event stream sequence. Edge computing nodes perform residual calculation operations, subtracting the computation timestamp from the actual physical timestamp to obtain the virtual-real phase transition residual, calculated using the following formula:

[0071] ;

[0072] In the formula, This represents the residual of the phase transition between virtual and real. This represents the actual physical timestamp output by the physical traffic signal controller. This represents the timestamp of the computation triggered by the virtual state machine to execute the corresponding phase transition logic. For cases where the virtual state machine remains in the same phase for an extended period within the sliding window, when the system does not detect any valid phase transition events, the edge computing node will directly read the historical virtual-real phase transition residuals generated in the previous computation cycle stored in its local register and assign them to the current cycle's value. This is to prevent the algorithm from getting stuck in a logic dead zone where no reference is available during the phase maintenance phase.

[0073] After residual extraction is completed, the next core step is to accurately align the virtual state with the current real world. The edge computing node continues to drive the timeline evolution of the virtual state machine until its internal virtual time reaches the current physical timestamp, at which point the system immediately suspends the state replay task.

[0074] As the final output of this stage, the edge computing node reads the internal register variables of the virtual state machine at the moment of suspension. These variables include details such as the current operating phase identifier, the remaining values ​​of the inter-detector timers obtained by progressively discretizing and updating according to the actuation event sequence, and the remaining time of the maximum green light timer. The edge computing node losslessly packages all the read internal register variables to generate a synchronization reference state and writes this reference state into a low-latency, high-speed memory space, thereby providing high-precision initial physical boundary conditions for the timeline advance extrapolation in subsequent steps.

[0075] See Figure 4 , Figure 4 This is a flowchart of a microscopic trajectory intention anchoring and reconstruction mechanism according to an embodiment of the present invention.

[0076] In this embodiment, after completing the implicit state pre-playback, the edge computing node enters the stage of analyzing the microscopic kinematic intent of the target vehicle. To prevent extreme dangerous events that may occur in the actual road environment, the system extracts the instantaneous acceleration of the target vehicle from the vehicle's basic safety information, and combines it with the high-precision roadside map to obtain the remaining spatial distance from the vehicle's current position to the intersection stop line, and then calculates the real-time physical braking limit of the target vehicle.

[0077] As a preferred method, the system combines the target vehicle's current instantaneous speed with the aforementioned remaining spatial distance to calculate the required deceleration for the vehicle to safely stop before the stop line. To avoid mathematical overflow due to the remaining spatial distance approaching zero when the vehicle approaches the stop line, the system presets a minimum effective distance threshold, for example, 0.5 meters. When the remaining spatial distance is greater than this minimum effective distance threshold, the formula for calculating the required deceleration is as follows:

[0078] ;

[0079] In the formula, This indicates the deceleration required for a safe stop. This represents the instantaneous speed of the target vehicle at the current physical timestamp. This represents the absolute remaining space distance for the target vehicle to reach the stop line at the intersection. If the remaining space distance is less than or equal to this minimum effective distance threshold, the system will pause the calculation of this formula and directly use the required deceleration of the target vehicle obtained in the previous calculation cycle.

[0080] After acquiring the required deceleration, the system compares it with the maximum deceleration envelope allowed by the vehicle's physical kinematics. This maximum deceleration envelope is typically determined by the target vehicle's dynamic mechanical limits and the friction coefficient of the current road surface. In practical engineering implementation, the system can set the maximum deceleration threshold of this envelope to a fixed value within the range of -5 m / s² to -8 m / s² based on preset weather conditions. If the target vehicle experiences a sudden positive acceleration change, i.e., the rate of change of acceleration in the acceleration direction exceeds the preset safety tolerance, and the absolute value of the required deceleration calculated by the system exceeds the upper limit of the threshold allowed by the maximum deceleration envelope for the current remaining spatial distance, then the edge computing node determines that the target vehicle is unable to complete safe braking before the stop line, i.e., a brake failure and rollback or malicious red light violation has occurred.

[0081] Faced with the aforementioned emergency scenario, the edge computing node immediately triggers a hardware interrupt mechanism by calling the underlying driver interface, forcibly abolishing the smooth reconstruction logic for the target vehicle and abandoning the subsequent countdown fitting. Simultaneously, the system generates a channel preemption command and pushes it to the roadside communication unit. Upon receiving this command, the roadside communication unit directly interrupts the currently transmitting low-priority regular service data stream at the media access control layer, allocates the highest-priority radio resource block, and broadcasts a stop-line collision warning message to all connected vehicles within the intersection area. This preemption mechanism, which directly schedules underlying communication resources across the application layer, minimizes alarm latency, thereby maximizing the safety redundancy for traffic participants at the intersection.

[0082] It should be noted that if the comparison results show that the physical braking limit is within the safety threshold, it indicates that the target vehicle is within the normal driving and braking capacity range. The edge computing node continues to execute the conventional intent anchoring and intervention branches based on the constraint feedback flag value read back from the previous cycle. When the constraint feedback flag triggers the forced hovering logic, it indicates that the target vehicle has been assigned the expected action of stopping and waiting during the red light in the previous calculation cycle or has experienced the anti-jump smoothing constraint of the green light countdown.

[0083] At this point, the system needs to further analyze the current actual traffic light color status. If it is determined that the vehicle cannot safely pass through the intersection within the current green light cycle, the edge computing node needs to combine the remaining spatial distance at the current moment with the countdown value issued for the target vehicle in the previous cycle to plan a vehicle anchoring deceleration curve that conforms to physical kinematic constraints.

[0084] Specifically, in the process of generating the anchored deceleration curve, the system uses the remaining spatial distance as the physical boundary and the countdown value issued in the previous cycle as the time dimension boundary, and performs polynomial curve fitting or quadratic integral programming. By setting the starting point velocity to the current instantaneous velocity, strictly setting the ending point velocity to zero, and applying acceleration continuity constraints, the system ensures that the generated deceleration curve transitions smoothly in space and that the ending velocity precisely decreases to zero when reaching the stop line.

[0085] Conversely, if the constraint feedback flag does not trigger the forced hovering logic, or if the system confirms through secondary analysis that the target vehicle can pass through the intersection normally before the current green light countdown ends, it means that the system determines that the vehicle is currently in a green light permitting state or a free driving state without intervention. In this case, the system retains the target vehicle's current original kinematic prediction state and does not make any forced intervention, thus forming a complete logical judgment closed loop.

[0086] Within the virtual space of the digital twin engine, for the target vehicle requiring intervention, edge computing nodes use a calculated anchored deceleration curve to directly overwrite the target vehicle's default kinematic prediction model, such as constant speed or constant acceleration. The system then performs forward recursive calculations along this anchored deceleration curve on the time axis to generate a reconstructed trajectory of the target vehicle over a continuous period in the future.

[0087] This reconstructed trajectory directly replaces the original prediction model and is mapped into the actual motion coordinate sequence of the virtual vehicle in the digital twin engine. It is then used as the core boundary condition input into the subsequent advanced simulation process to ensure that the triggering timing of the detection coils in the virtual state machine is strictly consistent with the theoretical motion state of the vehicle after it is constrained.

[0088] See Figure 5 , Figure 5 This is a flowchart of a digital twin parallel advance extrapolation and feedforward residual compensation mechanism according to an embodiment of the present invention.

[0089] In this embodiment, after acquiring the synchronization baseline state and the reconstructed trajectory, the edge computing node initiates the dual-threaded ultra-real-time simulation environment initialization program of the digital twin engine. Specifically, the system uses the aforementioned lossless packaged internal register variables as the physical starting point of the virtual state machine. Based on this, the system maps the reconstructed trajectory generated during the micro-trajectory intention reconstruction stage to the virtual space. It should be noted that the green light countdown corresponds to the overall release boundary of the signal phase where the target vehicle is located, rather than being determined solely by the target vehicle.

[0090] Therefore, while introducing the target vehicle to reconstruct the trajectory, the system also reads the actuation events of other vehicles related to the target phase within the sliding window before the current physical time, the positions of currently observed queued vehicles, and the occupancy status of each lane detector, and generates the measured trajectories or arrival pulse sequences of other vehicles in the same phase, and injects them into the virtual state machine together with the background traffic flow request pulses of the competing phase.

[0091] The reconstructed trajectory of the target vehicle is used to constrain the future triggering sequence of the detector corresponding to the target vehicle. The measured trajectory or arrival pulse of other vehicles in the same phase is used to characterize the remaining release demand in that phase. The background traffic flow request pulse of the competing phase is used to characterize the request intensity of other phases, thus forming a parallel inference input for the phase-level demand of the entire intersection.

[0092] As a preferred approach, to address uninterrupted free-moving vehicles or background traffic flow beyond sight, the system synchronously incorporates a default historical traffic flow distribution model as supplementary input. This distribution model is specifically generated based on the average traffic volume statistics of the intersection over a fixed time period in the past, and is directed into the virtual state machine to other intersection phases that conflict with the target vehicle, i.e., competing phases, simulating the request pulses of competing traffic flows. This avoids the simulation environment falling into a logical dead zone of phase transition prediction stagnation due to a lack of pulse data from surrounding detectors.

[0093] To accurately determine the timing of signal phase transitions under complex traffic flow conditions, the system employs an extreme value boundary approximation strategy. The basic principle is that, due to the randomness of future vehicle arrivals, the system defines the earliest and latest times of phase transition by simulating two extreme scenarios: the fewest and the most vehicles arriving. Specifically, the system explores the time interval of the signal phase transition in parallel across two independent computation threads.

[0094] The edge computing node is configured with an interrupted traffic flow state in the first computing thread. In this state, the system assumes that after the target vehicle and the currently confirmed queued vehicles have completed their release, no new vehicles will arrive at the detection coils in each detection area of ​​the target phase. Therefore, the system cuts off the subsequent supplementary simulated pulse injection of all relevant detectors of the target phase in the digital twin engine to obtain the earliest time boundary where the target phase may undergo a phase transition.

[0095] The virtual state machine rapidly extrapolates along the future timeline according to a set extrapolation rate factor. This extrapolation rate factor is dynamically allocated based on the current idle computing power of the edge computing nodes and is typically set to an integer between one hundred and five hundred. During the extrapolation process, its internal gap timer continuously decrements until it reaches zero due to the lack of new extended green light pulses. At this point, the virtual state machine forcibly triggers the gap timer termination logic, and the system records the original extrapolation timestamp when this phase transition occurs, using it as the original lower boundary timestamp.

[0096] Meanwhile, the edge computing node is configured with a continuous saturated traffic flow state in the second computing thread. In this state, the system continuously injects high-frequency simulated vehicle arrival pulses into all relevant detectors of the target phase in the virtual state machine, so that the target phase is in a continuous saturated arrival state at the phase level, rather than injecting pulses only for a single detector corresponding to the target vehicle, thereby obtaining the latest time boundary at which the target phase may undergo a phase transition.

[0097] To ensure the gap timer is continuously reset and never reaches zero, the injection interval of the aforementioned high-frequency pulses must be strictly less than the unit extended green light time set in the standard inductive control architecture. As the simulation progresses, the green light duration of the current phase of the virtual state machine eventually reaches the maximum green light limit set internally by the system. At this point, the virtual state machine forcibly breaks through the gap control, triggering the maximum green light timeout termination logic. The system records the original simulation timestamp when this forced phase transition occurs, using it as the original upper boundary timestamp.

[0098] After obtaining the original projection timestamp in the virtual space, to eliminate the clock skew between the physical state machine and the virtual state machine in the early stages, the edge computing nodes need to utilize the virtual-to-real phase transition residual extracted during the implicit state pre-playback stage to execute a heterogeneous clock feedforward residual compensation mechanism. The system uses linear translation and addition correction logic to map the original projection timestamp to the absolute physical time axis, generating an absolute projection timestamp boundary. The specific compensation calculation formula is as follows:

[0099] ;

[0100] In the formula, This represents the absolute extrapolated timestamp boundary after compensation and correction, encompassing both the absolute lower boundary timestamp and the absolute upper boundary timestamp. This represents the original derivation timestamp of the virtual state machine output, i.e., the original lower boundary timestamp or the original upper boundary timestamp. This represents the virtual-to-real phase transition residual extracted from historical state backtracking. Through the aforementioned feedforward residual compensation, the system successfully maps the abstract phase transition logic derived from the digital space into time nodes with real physical meaning, thus providing reliable data support for the final closed-loop feedback.

[0101] For any detector within the target phase , in the The total kinetic input for a walk is denoted as The total actuation input is the target vehicle actuation corresponding to the reconstructed trajectory of the target vehicle. Measured actuation of other observed vehicles in the same phase And supplementary actuation based on extrapolation of the most recent actuation record Together they constitute, that is:

[0102] ;

[0103] The virtual state machine drives the detector state update of the target phase with the aforementioned total actuation input, and drives the phase competition logic with the background traffic flow request pulse of the competing phase. Therefore, the green light end time boundary obtained subsequently is the release state for the entire signal phase, rather than just the movement trajectory of a single target vehicle.

[0104] See Figure 6 , Figure 6 This is a flowchart of a monotonically decreasing constraint mapping and closed-loop feedback generation mechanism according to an embodiment of the present invention.

[0105] In this embodiment, after completing the parallel advance simulation of the digital twin, the edge computing node obtains the absolute simulation timestamp boundary, which includes the absolute lower boundary timestamp and the absolute upper boundary timestamp. In order to provide safe and reliable intersection passage guidance to target vehicles within the communication coverage area, the system needs to convert this simulation boundary into a countdown value that is intuitively readable by the vehicles.

[0106] For different traffic light colors, the system employs a differentiated extreme value boundary selection mechanism. Specifically, in the green light countdown scenario, to ensure vehicles have sufficient safe braking distance, the edge computing nodes adhere to a conservative safety strategy, selecting the aforementioned absolute lower boundary timestamp as the baseline calculation basis. Conversely, in the red light countdown scenario, to prevent vehicles from starting prematurely and causing intersection conflicts, the system instead selects the absolute upper boundary timestamp as the baseline calculation basis. The system subtracts the current physical timestamp from the selected baseline timestamp to calculate the original remaining time for the current calculation cycle.

[0107] In real-world sensor-controlled intersections, the phase transition time derived by the virtual state machine within consecutive calculation cycles often fluctuates due to the continuous influx of new vehicles randomly entering the detection area. Directly sending the original remaining time to the vehicle can easily cause the countdown displayed on the onboard terminal to jump or even increase in the opposite direction. This not only violates the normal perception of human drivers but also interferes with the longitudinal control algorithm of autonomous vehicles. To address this issue, edge computing nodes introduce a monotonically decreasing constraint mapping mechanism.

[0108] In implementing this mechanism, the system needs to determine in advance whether the current calculation period is the starting calculation period of the signal phase. If so, the system directly sets the original remaining time as the countdown value of the current period and updates the historical countdown values ​​simultaneously to establish the initial calculation benchmark.

[0109] If the current calculation cycle is in the phase maintenance phase, the system accesses the local cache to read the historical countdown value from the previous calculation cycle and the system's preset time step. By calculating the difference between the historical countdown value and the time step, the theoretically maximum allowed countdown reduction under the current physical timestamp is obtained. Based on this, the system compares the original remaining time with the maximum countdown reduction value and selects the smaller value as the final countdown value for the current cycle. The specific calculation formula for the above constraint logic is as follows:

[0110] ;

[0111] In the formula, This represents the countdown value for the current period generated after the monotonically decreasing constraint mapping. This indicates the base timestamp selected by the system based on the current light color status. Indicates the current physical timestamp. This represents the historical countdown value from the previous calculation period. This represents the preset time step for edge computing nodes to advance discrete computing cycles. In specific engineering configurations, it is usually set to a fixed, tiny time constant such as 0.1 seconds. This represents the function that takes the minimum value.

[0112] Effectively correlating the macroscopic signal deduction state with the microscopic vehicle control intent is crucial for achieving a closed loop in the entire core control logic. To this end, based on the aforementioned calculation results, edge computing nodes synchronously generate constraint feedback flags to guide the microscopic trajectory intent reconstruction phase. In practice, the system determines the data source of the actual values ​​in the above formula. When the final output countdown value is the maximum decreasing countdown, the edge computing node determines that the system has applied forced smoothing intervention to the target vehicle and then sets the constraint feedback flag corresponding to that target vehicle to the triggered state.

[0113] Conversely, when the system selects the original remaining time as the final output, it means that the current deduction result conforms to the natural decreasing law, and the system sets the constraint feedback flag to the untriggered state. Through the dynamic assignment of this flag, the system successfully decouples the macroscopic signal deduction state from the microscopic vehicle control intention and establishes a state association, completing the closed loop of the entire core control logic.

[0114] As the final output of this embodiment, after generating the final countdown value for the current period and related phase information, the edge computing node executes a closed-loop feedback broadcast task. As a preferred approach, the system encapsulates the above information into signal phase and timing messages according to the vehicle-to-everything (V2X) standard data dictionary. The edge computing node transmits this message to the roadside communication unit via the underlying communication bus, and the roadside communication unit periodically broadcasts it to all connected vehicles within the intersection area via a radio channel. It is particularly noteworthy that while performing the regular broadcast task, the edge computing node simultaneously monitors the channel resource status of the underlying communication interface in real time.

[0115] When the system detects that the roadside communication unit is in a channel preemption state due to abnormal vehicle slippage or checkpoint breaching events detected by the pre-processing steps, the edge computing node immediately blocks the transmission link of the final output countdown value at the application layer, ceasing the encapsulation and push of signal phase and timing messages. By actively relinquishing computing power and communication bandwidth, the system ensures that collision avoidance warning messages have an absolutely exclusive transmission channel, thereby maximizing the low-latency delivery of alarm information under extreme emergency conditions. The specific encoding and encapsulation rules for signal phase and timing messages, as well as the radio modulation and transmission process of the roadside communication unit, can be implemented by those skilled in the art using existing vehicle-to-everything (V2X) communication standard protocol stacks. The specific process is well-known in the field and will not be elaborated upon here.

[0116] To further aid in understanding the present invention, a specific application embodiment is described below. Imagine a physical vehicle equipped with vehicle-to-everything (V2X) communication capabilities... A vehicle with an instantaneous speed of 15 m / s approaches an intersection under sensor control. The current traffic light is green. The absolute remaining distance between the vehicle and the stop line at the intersection is... =80m. Edge computing nodes operate at a preset time step. =0.1s to execute the current discrete calculation cycle. During the multi-source sensing data acquisition and historical state backtracking phase, the edge computing node receives the vehicle's basic safety information, extracts its absolute coordinate position, and performs spatiotemporal fitting and cross-validation with the physical observation coordinates captured by the roadside millimeter-wave radar. The calculated spatial coordinate residuals... =0.8, which is lower than the system's preset anti-spoofing tolerance threshold of 1.5. The system determines that the basic security message is valid and allows the data to be injected into subsequent processes.

[0117] Subsequently, during the implicit state pre-playback phase, the system injects a 120-second sequence of actuation events from the previous sliding window into the virtual state machine. By comparing the last valid phase transition event within the window, the virtual-to-real phase transition residual is calculated. =0.2s, and extract the reference state of the internal register after the virtual state machine has completed its advancement at the current moment. During the micro-trajectory intention anchoring and reconstruction phase, the system calculates the required deceleration for the vehicle to safely stop. =152 / (2×80)≈1.41m / s 2 The calculated value is within the maximum deceleration envelope threshold allowed by the current road surface, and there is no positive abrupt change in instantaneous acceleration. The system determines that the physical braking limit is within the safe threshold. Since the retrospective display shows that the constraint feedback flag of the previous cycle was not triggered, the system retains the vehicle's original kinematic prediction state and maps it to the virtual space of the digital twin engine.

[0118] During the parallel advance simulation phase of the digital twin, the system initiates dual-thread simulation to determine the original time boundary. After the first calculation thread separately cuts off the subsequent simulated pulse injection of the phase where the target vehicle is located, it triggers the gap timing end logic and obtains the original lower boundary timestamp of 15.0s;

[0119] The second calculation thread, under the state of continuous high-frequency injection of simulated pulses, triggers the maximum green light timer termination logic and obtains the original upper boundary timestamp of 25.0s. This is combined with the previously obtained virtual-real phase transition residuals. =0.2s, the system performs linear translation correction, generating an absolute lower boundary timestamp of 15.2s and an absolute upper boundary timestamp of 25.2s. During the monotonically decreasing constraint mapping and closed-loop feedback generation stage, since the current actual traffic light color is green, the system adopts a conservative safety strategy and selects the absolute lower boundary timestamp of 15.2s as the reference timestamp.

[0120] Assuming the current physical timestamp is 0.0s and the historical countdown value of the previous cycle is 15.3s, the system calculates that the theoretically allowed maximum countdown reduction is 15.3 - 0.1 = 15.2s. The system compares the original remaining time of 15.2s corresponding to the selected baseline timestamp with the maximum countdown reduction of 15.2s and takes the minimum value. The system then outputs the current cycle countdown value as 15.2s and updates the constraint feedback flag to the untriggered state. This final value is encapsulated by the underlying communication bus and periodically broadcast to the target vehicle by the roadside communication unit, completing one full closed-loop calculation.

[0121] To verify the actual technical effectiveness of the above solution, this embodiment conducted experimental verification and effect comparison in a hardware-in-the-loop intersection micro-traffic simulation platform. The experiment selected 100 consecutive cycles of inductive control signals, inputting the same high-frequency random traffic flow pulses. The smoothness of the countdown output and vehicle operation safety indicators using the traditional vehicle-to-everything (V2X) direct estimation method and the method based on the digital twin engine of this invention were statistically analyzed and compared.

[0122] See Figure 7 , Figure 7 This is a comparison curve of the countdown smoothing effect according to an embodiment of the present invention. Traditional vehicle networking direct estimation methods cannot invert the internal implicit state of the physical sensing controller and lack extreme value derivation and monotonically decreasing constraint mechanisms. As a result, the countdown curves output by these methods exhibit obvious step jumps and non-monotonic increases under random traffic pulse interference. The maximum single count jump amplitude recorded in the experiment reached 4.5s.

[0123] After applying the solution of this invention, the countdown curve output by the system maintains a monotonically decreasing trend throughout the entire countdown period, completely eliminating the phenomenon of numerical jumps. Statistical data shows that the solution of this invention reduces the countdown jump frequency from an average of 6.2 times per cycle in the traditional method to 0 times, reducing the incidence of abnormal emergency braking events caused by sudden changes in countdown values ​​by 98.5%. The above experimental results demonstrate that the solution of this invention can provide high-precision and high-stability signal timing prediction in complex random sensing traffic flow environments, effectively ensuring the longitudinal control stability and vehicle-road cooperative safety margin of intelligent connected vehicles in intersection areas.

[0124] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for simulating countdown times for road stop line traffic lights based on a digital twin engine, characterized in that: Includes the following steps: S10: Obtain the target vehicle's basic safety information, physical observation coordinates, and historical actuation events; compare the absolute coordinate position with the physical observation coordinates to obtain the spatial coordinate residual; retain the basic safety information when the residual does not exceed the threshold. S20 injects historical actuation events into the virtual state machine in chronological order, resets the gap timer when a vehicle arrives or occupies the spacer, decrements the gap timer when no new vehicle arrives, and accumulates the maximum green light timer when the phase enters the green light, extracting the virtual-real phase transition residual and the baseline state. S30 calculates the physical braking limit and generates a reconstructed trajectory based on instantaneous kinematic parameters, remaining spatial distance, and historical countdown values; S40, using the reference state as the initial condition, input the reconstructed trajectory, the measured actuation or extrapolated arrival pulses of other vehicles in the same phase, and the competing phase request pulses into the virtual state machine, and deduce the original time boundary in parallel, and obtain the absolute deduced timestamp boundary through the residual correction; S50 extracts the original remaining time based on the absolute extrapolation timestamp boundary, and generates the output countdown value by combining it with the historical countdown value through a monotonically decreasing constraint function and sends it to the vehicle terminal.

2. The method for simulating a road stop line traffic light countdown based on a digital twin engine according to claim 1, characterized in that, S10 further includes: when the spatial coordinate residual is greater than the anti-spoofing tolerance threshold, determining that the vehicle basic safety message is a maliciously forged message, isolating and removing it from the input source of the virtual state machine, and maintaining the operation of the virtual state machine based on the physical observation coordinates.

3. The method for simulating a road stop line traffic light countdown based on a digital twin engine according to claim 1, characterized in that, The extraction of the virtual-real phase transition residual in S20 includes: searching for the last valid phase switching event within the time interval covering the historical actuation events as a comparison anchor point, extracting the real physical timestamp of the valid phase switching event broadcast by the traffic signal controller; recording the phase transition calculation timestamp when the virtual state machine triggers the corresponding phase transition logic when processing the same historical actuation event; and subtracting the real physical timestamp from the phase transition calculation timestamp to obtain the virtual-real phase transition residual.

4. The method for simulating a road stop line traffic light countdown based on a digital twin engine according to claim 1, characterized in that, S30 further includes: when the instantaneous acceleration positive change rate of the target vehicle exceeds the preset safety tolerance, and the absolute value of the required deceleration calculated based on the instantaneous speed and the remaining spatial distance exceeds the preset maximum deceleration envelope, it is determined that a braking failure and vehicle slippage or malicious red light running and checkpoint breaching event has occurred, triggering the underlying hardware interrupt mechanism to abolish and reconstruct the trajectory generation logic, and pushing a channel preemption instruction to the roadside communication unit to broadcast a stop line crossing anti-collision warning message.

5. The method for simulating a road stop line traffic light countdown based on a digital twin engine according to claim 1, characterized in that, The process of constructing the anchored deceleration curve and generating the reconstructed trajectory in S30 includes: reading the constraint feedback flag from the previous cycle; when the constraint feedback flag triggers the forced hovering logic and it is determined that the target vehicle cannot safely pass through the intersection within the current green light cycle, a polynomial curve is fitted using the remaining spatial distance as the spatial boundary, the historical countdown value from the previous cycle as the time boundary, the instantaneous speed as the starting speed, and zero as the ending speed, and an acceleration continuity constraint is applied to generate the anchored deceleration curve; and the reconstructed trajectory is generated by covering the original kinematic prediction model with the anchored deceleration curve in the digital twin engine.

6. The method for simulating a road stop line traffic light countdown based on a digital twin engine according to claim 1, characterized in that, The original time boundary obtained in S40 includes: configuring an interrupted traffic flow state in the first calculation thread, cutting off the subsequent supplementary simulated pulse injection of all relevant detectors in the phase after the target vehicle and the currently observed vehicles in the phase where the target vehicle is located have completed the release, and recording the first original extrapolation timestamp when the gap timer returns to zero and the gap timing end logic is triggered; configuring a continuous saturated traffic flow state in the second calculation thread, continuously injecting simulated vehicle arrival pulses into all relevant detectors in the phase where the target vehicle is located, and recording the second original extrapolation timestamp when the green light duration reaches the maximum green light limit and the maximum green light timing end logic is triggered; and combining the first original extrapolation timestamp and the second original extrapolation timestamp as the original time boundary.

7. The road stop line traffic light countdown simulation method based on a digital twin engine according to claim 6, characterized in that, The step S40 of correcting the original time boundary using the virtual-real phase transition residual includes: adding the first original projection timestamp to the virtual-real phase transition residual to obtain the absolute lower boundary timestamp; adding the second original projection timestamp to the virtual-real phase transition residual to obtain the absolute upper boundary timestamp; and combining the absolute lower boundary timestamp and the absolute upper boundary timestamp as the absolute projection timestamp boundary.

8. The method for simulating a road stop line traffic light countdown based on a digital twin engine according to claim 7, characterized in that, Extracting the original remaining time in S50 includes: in the green light countdown scenario, selecting the absolute lower boundary timestamp as the base calculation timestamp; in the red light countdown scenario, selecting the absolute upper boundary timestamp as the base calculation timestamp; and subtracting the current physical timestamp from the base calculation timestamp to obtain the original remaining time.

9. The method for simulating a road stop line traffic light countdown based on a digital twin engine according to claim 8, characterized in that, The step S50, which generates the final output countdown value through a monotonically decreasing constraint function, includes: calculating the difference between the historical countdown value of the previous cycle and the preset time step to obtain the maximum decreasing countdown; comparing the original remaining time with the maximum decreasing countdown and taking the smaller value as the final output countdown value; when the final output countdown value is equal to the maximum decreasing countdown, setting the constraint feedback flag corresponding to the target vehicle to the triggered state, and otherwise setting the constraint feedback flag to the untriggered state.

10. The method for simulating a road stop line traffic light countdown based on a digital twin engine according to claim 1, characterized in that, The S50 process for sending data to the vehicle terminal also includes: encapsulating the final output countdown value into a signal phase and timing message; monitoring the channel resource status of the underlying communication interface; and when it is detected that the roadside communication unit is in a channel preemption state, prioritizing blocking the transmission link of the final output countdown value at the application layer, stopping the encapsulation and push process of the signal phase and timing message, and relinquishing communication bandwidth to ensure the exclusive transmission channel of the stop line crossing collision warning message.

Images