Data transmission lag compensation and consistency evaluation method and system in virtual-real mapping test

Abstract

The application discloses a kind of data transmission lag compensation and consistency evaluation method and system in virtual-real mapping test.The method obtains real moving object pose as reference true value, is unified to simulation coordinate system by space coordinate conversion and extracts speed and other state information;Determine transmission delay after time synchronization, use current speed and delay to carry out linear compensation to position and obtain predicted position;Compensated data is injected into simulation platform to drive virtual object, compare real response curve to calculate amplitude error and phase error, and accordingly quantify the consistency of virtual and real.The system includes true value acquisition, coordinate mapping, time synchronization, lag compensation, data injection and reproduction, and consistency evaluation module.The application realizes space-time systematic alignment with low computational overhead, and provides a double-dimensional quantifiable evaluation system, effectively improving the reliability of simulation test.

Description

Technical Field

[0001] This invention relates to the field of virtual-real mapping testing and simulation technology, specifically to a method and system for data transmission lag compensation and consistency evaluation in virtual-real mapping testing. Background Technology

[0002] With the rapid development of technologies such as intelligent driving, robotics, and drones, virtual-real hybrid simulation testing has become an important means of product development and verification. The core objective of virtual-real mapping testing is to enable virtual objects in a virtual simulation environment to accurately reproduce the motion state of real moving objects in the physical world, thereby completing functional verification, performance evaluation, and other testing tasks in the simulation environment.

[0003] However, in practical engineering applications, virtual-real mapping testing faces several key technical challenges: First, there's the issue of inconsistent spatial coordinate systems. Real-world moving objects typically use absolute positioning methods like GPS to obtain pose data, while simulation platforms usually operate in a relative coordinate system. Accurately mapping the pose data of real objects to the simulation coordinate system is a fundamental prerequisite for virtual-real mapping.

[0004] Second, there is the issue of data transmission latency. The pose data of a real moving object needs to go through multiple stages, including acquisition, encoding, transmission, decoding, and injection, before reaching the simulation platform. This process inevitably introduces millisecond-level transmission latency. For real objects moving at high speeds, even a delay of a few milliseconds can cause a significant deviation between the virtual mapped position and the real position, severely affecting the reliability of the simulation test.

[0005] Third, there is a lack of systematic evaluation methods for virtual-real consistency. In existing technologies, the evaluation of virtual-real mapping accuracy often relies on manual observation of curve comparisons, lacking quantitative indicators and automated evaluation processes, making it difficult to form reproducible and comparable acceptance criteria.

[0006] Therefore, there is an urgent need for a technical solution that can systematically solve the above-mentioned problems of spatial alignment, time delay compensation, and consistency evaluation. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for data transmission lag compensation and consistency evaluation in virtual-real mapping tests, so as to achieve high-precision synchronization between real moving objects and virtual simulation environments, and to provide a quantifiable evaluation system to improve the credibility of simulation tests.

[0008] To achieve the above objectives, this invention provides a method for data transmission lag compensation and consistency evaluation in virtual-real mapping testing, comprising: S1. Obtain the pose data of the real moving object as the baseline truth value; S2. Perform spatial coordinate transformation on the reference true value to unify it to the simulation coordinate system, and extract motion state information, which includes at least velocity; S3. Synchronize the data acquisition terminal with the simulation platform and obtain the data transmission delay time; S4. Based on the motion state information at the current moment and the delay time, the position data at the current moment is compensated to obtain the predicted position; S5. Inject the motion state data containing the predicted position into the simulation platform to drive the virtual object to generate a simulation response curve; S6. Compare the simulated response curve with the real response curve formed by the reference true value, and calculate the amplitude error and phase error; S7. Evaluate the consistency between virtual and real data based on the amplitude error and phase error, and output the evaluation results.

[0009] Furthermore, the compensation of the current position data based on the current motion state information and the delay time includes: Using the current velocity V and the delay time Δt, the compensated position Pcompensated = Pcurrent + V·Δt is obtained.

[0010] Furthermore, the compensation also includes compensation using the acceleration 'a' at the current moment, with the formula P_compensation = P_current + V·Δt + ½·a·(Δt)².

[0011] Furthermore, the acquisition of data transmission delay time includes measuring the system's average transmission delay or measuring the transmission delay of each frame of data in real time.

[0012] Furthermore, the evaluation of virtual-real consistency based on amplitude error and phase error includes calculating a comprehensive score and determining whether the comprehensive score reaches a preset threshold.

[0013] Furthermore, the preset threshold is 90%.

[0014] Furthermore, the actual moving object is a vehicle, the pose data is provided by an inertial navigation system, and the motion state information also includes three-axis acceleration and attitude angles, including pitch angle, yaw angle and angular velocity.

[0015] Furthermore, the comparison between the simulated response and the actual response curve specifically includes: The dynamic data of the real moving object is injected into the simulation platform in the form of a data stream; The simulation platform generates realistic response curves based on the injected data; The simulated response curve and the real response curve are automatically overlaid and displayed on the same time axis.

[0016] The present invention also provides a virtual-real mapping test system, comprising: The truth acquisition module is used to obtain the pose data of the real moving object as the baseline truth value. The coordinate mapping module is used to perform spatial coordinate transformation on the reference true value to unify it to the simulation coordinate system, and to extract motion state information, which includes at least velocity. The time synchronization module is used to synchronize the data acquisition terminal and the simulation platform in time and to obtain the data transmission delay time. The lag compensation module is used to compensate the current position data based on the current motion state information and the delay time to obtain the predicted position; The data injection and reproduction module is used to inject motion state data containing the predicted position into the simulation platform and drive the virtual object to generate a simulation response. The consistency evaluation module is used to compare the simulated response with the real response curve formed by the reference true value, calculate the amplitude error and phase error, evaluate the consistency between the virtual and real responses based on the amplitude error and phase error, and output the evaluation result.

[0017] Furthermore, the lag compensation module is specifically used to obtain the compensated position P_compensated = P_current + V·Δt using the current velocity V and the delay time Δt, or to use the current acceleration a for compensation, with the formula P_compensated = P_current + V·Δt + ½·a·(Δt)². The consistency evaluation module is specifically used to calculate the comprehensive score of the amplitude error and the phase error, and to determine whether the comprehensive score reaches a preset threshold.

[0018] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention establishes a systematic "space-time-evaluation" virtual-real alignment framework: spatial dimension unification is achieved through spatial coordinate transformation, time dimension alignment is achieved through time synchronization and lag compensation, and consistency quantification is achieved through dual-index evaluation. The three work together to form a complete virtual-real mapping accuracy assurance system.

[0019] 2. The hysteresis compensation algorithm of this invention has extremely low computational overhead, requiring only two parameters: the current frame speed and the delay time, to complete position compensation. It is suitable for in-loop testing scenarios with high real-time requirements. At the same time, it supports the introduction of an acceleration term to improve the compensation accuracy under emergency acceleration / deceleration conditions, and has good scalability.

[0020] 3. This invention is the first to define a two-dimensional evaluation model for amplitude error and phase error in engineering. It quantifies the consistency between virtual and real from two dimensions: “accuracy of magnitude” and “speed of time response”. Combined with a comprehensive score threshold, it provides an executable acceptance standard for simulation credibility.

[0021] 4. This invention establishes a closed-loop automatic benchmarking and reproduction process based on truth value injection. The dynamic data of the real object is used as the benchmark truth value to directly drive the virtual object, automatically generate the simulation response curve and superimpose and compare it with the real response curve to form a repeatable and quantifiable verification closed loop. Attached Figure Description

[0022] Figure 1 This is the overall flowchart of the virtual-real mapping test method of the present invention; Figure 2 This is a module structure diagram of the virtual-real mapping test system of the present invention.

[0023] Attached reference numerals: 100, Truth value acquisition module; 200, Coordinate mapping module; 300, Time synchronization module; 400, Lag compensation module; 500, Data injection and reproduction module; 600, Consistency evaluation module. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] Example 1

[0026] This embodiment uses vehicle-in-the-loop (VIL) testing as an application scenario to describe the technical solution of the present invention in detail. In VIL testing, a real vehicle travels on a real road, and simultaneously, through a data acquisition and transmission system, the vehicle's motion state is mapped in real time to a virtual vehicle in the simulation platform, achieving virtual-real synchronization. In this embodiment, the real moving object is the vehicle, the virtual object running on the simulation platform is the virtual vehicle, and the baseline truth is provided by the RT3000 series inertial navigation system installed on the vehicle.

[0027] like Figure 1 As shown, the method in this embodiment includes the following steps: S1, obtain the pose data of the real moving object as the baseline truth value.

[0028] In this step, the RT3000 inertial navigation system installed on the actual vehicle outputs high-precision pose data at a frequency of 100Hz, including latitude and longitude, altitude, speed, three-axis acceleration, pitch angle, yaw angle, angular velocity, and other information. This data serves as the "baseline truth" for all subsequent processing, assuming that the RT3000's output can accurately reflect the vehicle's actual motion state.

[0029] S2 performs spatial coordinate transformation on the reference true value to unify it to the simulation coordinate system and extracts motion state information.

[0030] The simulation platform operates in a relative coordinate system, with its origin typically set at a fixed point on the test site. Therefore, the latitude, longitude, and altitude data output by the RT3000 need to be converted into X, Y, and Z coordinates in the simulation coordinate system. This embodiment uses a preset coordinate offset method for conversion. Before the test begins, the latitude and longitude position of the simulation coordinate system origin in the real world is measured, and the offset is calculated accordingly. During the test, the absolute position output by the RT3000 is subtracted from this offset to obtain the relative position in the simulation coordinate system.

[0031] Meanwhile, the motion state information of the vehicle is extracted from the data output by RT3000, including velocity V, three-axis acceleration (ax, ay, az) and attitude angles (pitch angle, yaw angle, angular velocity) to ensure that the motion trajectory of the virtual vehicle in three-dimensional space is highly consistent with that of the real vehicle.

[0032] S3 synchronizes the data acquisition terminal with the simulation platform in time and obtains the data transmission delay.

[0033] Time synchronization is crucial for achieving high-precision virtual-real mapping. In this embodiment, the data acquisition end (i.e., the onboard computing device) and the simulation platform (i.e., the computer running the simulation software) synchronize their clocks using the IEEE 1588 Precision Time Protocol or the Network Time Protocol to ensure that both ends use a unified time base. Each frame of data acquired from the RT3000 is marked with a transmission timestamp at the data acquisition end and a reception timestamp when received by the simulation platform.

[0034] Obtaining the latency Δt is the core of this step. In this embodiment, the average system transmission latency is determined by measuring the network transmission latency multiple times and taking the average. Specifically, before the test begins, the data acquisition terminal continuously sends 100 frames of test data to the simulation platform, records the sending and receiving timestamps of each frame, calculates the difference, and takes the arithmetic mean to obtain the average system transmission latency Δt. In another embodiment, the IEEE 1588 clock synchronization protocol can also be used to calculate the actual transmission latency of each frame of data in real time during the test to account for network latency jitter.

[0035] S4, based on the motion state information and delay time at the current moment, compensates the position data at the current moment to obtain the predicted position.

[0036] This is the core step of the invention. Due to the transmission delay Δt measured in S3, the position data received by the simulation platform is actually the position of the real vehicle before time Δt. If the virtual vehicle is driven directly using this "lagging" position, the movement of the virtual vehicle will always lag behind that of the real vehicle.

[0037] To solve this problem, this step uses a linear extrapolation method for position compensation. The specific formula is: P compensation = P current + V·Δt Where Pcurrent represents the position data contained in the current frame received by the simulation platform, V represents the velocity data contained in the current frame, and Δt represents the average transmission delay measured by the system. The physical meaning of this formula is: the predicted position that can be reached by a real vehicle moving at a constant speed within a delay time Δt at its current velocity.

[0038] This compensation method is based on a reasonable assumption: within a short time window of milliseconds, the vehicle's motion can be approximated as uniform motion. In this embodiment, the system's average transmission delay Δt is typically 10 to 20 milliseconds; at this time scale, the uniform motion assumption has sufficient accuracy for most operating conditions.

[0039] In another embodiment, to further improve the compensation accuracy under emergency acceleration / deceleration conditions, an acceleration term can be introduced into the compensation formula, namely: P compensation = P current + V·Δt + ½·a·(Δt)² Where 'a' represents the acceleration data contained in the current frame. This formula assumes that the vehicle undergoes uniform acceleration during the delay time Δt, which is better suited to scenarios involving sharp acceleration / deceleration.

[0040] S5 injects motion state data containing the predicted position into the simulation platform, driving the virtual object to generate simulation response curves.

[0041] After S4 hysteresis compensation processing, complete motion state data, including predicted position, velocity, acceleration, and attitude angles, is injected into the simulation platform as a data stream. Based on the injected data, the simulation platform drives the virtual vehicle to move within the virtual environment and generates simulation response curves. These curves may include the virtual vehicle's position trajectory, velocity curve, and acceleration curve, depending on the specific testing requirements.

[0042] S6 compares the simulated response curve with the real response curve formed by the reference true value, and calculates the amplitude error and phase error.

[0043] In this step, the simulation platform automatically overlays the simulated response curve generated by S5 with the real response curve formed by the baseline true value (i.e., the original output data of RT3000) on the same time axis. By comparing the two curves, the accuracy of the virtual-real mapping can be intuitively judged.

[0044] More importantly, this step defines two quantization error metrics: Amplitude error is used to determine the degree of deviation between the simulated and real objects in terms of quantitative values. For example, when a real vehicle performs emergency braking, if the peak deceleration measured by the RT3000 is 8 m / s², while the peak deceleration of the virtual vehicle in the simulation platform is 7 m / s², then the amplitude error will accurately reflect this 1 m / s² deviation. Amplitude error can be obtained by calculating the statistical measures of the numerical difference between the two curves at the same moment (such as root mean square error, maximum absolute error, etc.).

[0045] Phase error is used to determine the difference in "reaction time" between simulated and real objects, primarily to detect delays in sensor data links or dynamic model calculations. For example, if a virtual vehicle's steering response lags by 50 milliseconds compared to a real vehicle when the real vehicle begins to turn, the phase error will reflect this time lag. Phase error can be obtained by calculating the peak offset of the cross-correlation function of two curves.

[0046] S7 evaluates the consistency between virtual and real data based on amplitude error and phase error, and outputs the evaluation results.

[0047] In this step, a comprehensive score is generated based on the amplitude error and phase error calculated in S6. The comprehensive score can be calculated as a weighted average of the amplitude error score and the phase error score, and the weights can be adjusted according to specific testing requirements. For example, for tests that emphasize dynamic accuracy, the weight of the amplitude error can be appropriately increased; for tests that emphasize real-time verification, the weight of the phase error can be appropriately increased.

[0048] In this embodiment, the preset comprehensive score threshold is 90%. When the comprehensive score reaches or exceeds 90%, the current virtual-real mapping accuracy is deemed to meet the requirements, and the simulation reliability is good. When the comprehensive score is below 90%, the system issues an alarm, prompting technicians to check and optimize the data link, compensation parameters, or simulation model. The final evaluation result is output in report form, including amplitude error, phase error, comprehensive score, and a conclusion on whether the standard is met.

[0049] Example 2

[0050] This embodiment provides a virtual-real mapping test system that implements the above method. For example... Figure 2As shown, the system includes a truth value acquisition module 100, a coordinate mapping module 200, a time synchronization module 300, a lag compensation module 400, a data injection and reproduction module 500, and a consistency evaluation module 600.

[0051] The truth acquisition module 100 is used to acquire the pose data of a real moving object as a reference truth value. In this embodiment, the truth acquisition module 100 is communicatively connected to the RT3000 inertial navigation system installed on a real vehicle to receive high-precision pose data streams in real time.

[0052] The coordinate mapping module 200 is used to perform spatial coordinate transformation on the reference ground truth to unify it to the simulation coordinate system and extract motion state information. The coordinate mapping module 200 pre-stores the offset between the origin of the simulation coordinate system and the absolute position in the real world, and completes spatial alignment through coordinate translation operations. The extracted motion state information includes at least velocity V, and in practical applications, it also includes six degrees of freedom information such as acceleration a and attitude angles.

[0053] The time synchronization module 300 is used to synchronize the time between the data acquisition terminal and the simulation platform, and to obtain the data transmission delay. In this embodiment, the time synchronization module 300 achieves clock synchronization between the two ends based on the IEEE 1588 precise time protocol, and measures the average transmission delay Δt of the system by sending and receiving test data packets multiple times, and provides the Δt value to the hysteresis compensation module 400.

[0054] The hysteresis compensation module 400 is used to compensate the current position data based on the current motion state information and the delay time Δt to obtain the predicted position. The hysteresis compensation module 400 executes a linear extrapolation algorithm, and by default uses the velocity compensation formula P compensation = P current + V·Δt; in scenarios requiring higher accuracy, it can switch to the acceleration compensation formula P compensation = P current + V·Δt + ½·a·(Δt)².

[0055] The data injection and reproduction module 500 is used to inject motion state data, including the predicted position, into the simulation platform to drive the virtual object to generate a simulation response. The data injection and reproduction module 500 packages the complete motion state data processed by the hysteresis compensation module 400 into a data format recognizable by the simulation platform, injects it into the simulation engine through a high-speed data interface, drives the virtual vehicle to move, and generates a simulation response curve.

[0056] The consistency evaluation module 600 compares the simulated response with the real response curve formed by the benchmark true value, calculates the amplitude error and phase error, evaluates the consistency between the simulated and real responses based on the amplitude and phase errors, and outputs the evaluation results. The consistency evaluation module 600 includes an amplitude error calculation unit, a phase error calculation unit, and a comprehensive scoring unit. The comprehensive scoring unit weights and fuses the two types of errors according to preset weights, generates a comprehensive score, and compares it with a preset threshold (e.g., 90%), finally outputting an evaluation report containing various indicators and judgment conclusions.

[0057] In this embodiment, the hysteresis compensation module 400 is specifically used to obtain the compensated position Pcompensated = Pcurrent + V·Δt using the current velocity V and the delay time Δt, or it can also use the current acceleration a for compensation, with the formula Pcompensated = Pcurrent + V·Δt + ½·a·(Δt)². The consistency evaluation module 600 is specifically used to calculate the comprehensive score of amplitude error and phase error, and determine whether the comprehensive score reaches a preset threshold.

[0058] Through the collaborative work of the above modules, this system can achieve high-precision real-time synchronization between real vehicles and virtual simulation environments, and provide quantifiable and reproducible consistency evaluation results, effectively improving the simulation credibility of vehicle-in-the-loop testing.

[0059] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for data transmission lag compensation and consistency evaluation in virtual-real mapping testing, characterized in that, include: S1. Obtain the pose data of the real moving object as the baseline truth value; S2. Perform spatial coordinate transformation on the reference true value to unify it to the simulation coordinate system, and extract motion state information, which includes at least velocity; S3. Synchronize the data acquisition terminal with the simulation platform and obtain the data transmission delay time; S4. Based on the motion state information at the current moment and the delay time, the position data at the current moment is compensated to obtain the predicted position; S5. Inject the motion state data containing the predicted position into the simulation platform to drive the virtual object to generate a simulation response curve; S6. Compare the simulated response curve with the real response curve formed by the reference true value, and calculate the amplitude error and phase error; S7. Evaluate the consistency between virtual and real data based on the amplitude error and phase error, and output the evaluation results.

2. The method according to claim 1, characterized in that, The compensation of the current position data based on the current motion state information and the delay time includes: Using the current velocity V and the delay time Δt, the compensated position Pcompensated = Pcurrent + V·Δt is obtained.

3. The method according to claim 2, characterized in that, The compensation also includes compensation using the acceleration 'a' at the current moment, with the formula P_compensation = P_current + V·Δt + ½·a·(Δt)².

4. The method according to claim 1, characterized in that, The acquisition of data transmission delay includes measuring the system's average transmission delay or measuring the transmission delay of each frame of data in real time.

5. The method according to claim 1, characterized in that, The evaluation of virtual and real consistency based on amplitude error and phase error includes calculating a comprehensive score and determining whether the comprehensive score reaches a preset threshold.

6. The method according to claim 5, characterized in that, The preset threshold is 90%.

7. The method according to claim 1, characterized in that, The actual moving object is a vehicle, the pose data is provided by an inertial navigation system, and the motion state information also includes three-axis acceleration and attitude angles, including pitch angle, yaw angle and angular velocity.

8. The method according to claim 1, characterized in that, The simulation response curve is compared with the real response curve formed from the baseline true value, specifically including: The dynamic data of the real moving object is injected into the simulation platform in the form of a data stream; The simulation platform generates realistic response curves based on the injected data; The simulated response curve and the real response curve are automatically overlaid and displayed on the same time axis.

9. A virtual-real mapping test system, characterized in that, include: The truth acquisition module is used to obtain the pose data of the real moving object as the baseline truth value. The coordinate mapping module is used to perform spatial coordinate transformation on the reference true value to unify it to the simulation coordinate system, and to extract motion state information, which includes at least velocity. The time synchronization module is used to synchronize the data acquisition terminal and the simulation platform in time and to obtain the data transmission delay time. The lag compensation module is used to compensate the current position data based on the current motion state information and the delay time to obtain the predicted position; The data injection and reproduction module is used to inject motion state data containing the predicted position into the simulation platform and drive the virtual object to generate a simulation response. The consistency evaluation module is used to compare the simulated response with the real response curve formed by the reference true value, calculate the amplitude error and phase error, evaluate the consistency between the virtual and real responses based on the amplitude error and phase error, and output the evaluation result.

10. The system according to claim 9, characterized in that, The lag compensation module is specifically used to obtain the compensated position P_compensation = P_current + V·Δt by using the current velocity V and the delay time Δt, or to use the current acceleration a for compensation, with the formula P_compensation = P_current + V·Δt + ½·a·(Δt)². The consistency evaluation module is specifically used to calculate the comprehensive score of the amplitude error and the phase error, and to determine whether the comprehensive score reaches a preset threshold.

Images