A system and method for dynamic verification and validation of aircraft platform digital simulations
The aircraft platform digital simulation dynamic verification system collects and verifies simulation data in real time. By using multi-dimensional capability weights and adaptive sliding windows to identify logic drift, it solves the problem that traditional methods cannot identify logic drift, and achieves accurate evaluation of digital simulation performance and scientific support for combat simulation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- AERONAUTICS RES INST OF CHINA
- Filing Date
- 2026-05-18
- Publication Date
- 2026-06-19
AI Technical Summary
Traditional digital simulation verification methods for aircraft platforms cannot effectively identify and quantify temporal logical drift, cannot evaluate the overall performance of digital simulation, and static simulation verification cannot identify systematic errors during long-term, highly dynamic adversarial processes.
A dynamic verification system for digital simulation of an aircraft platform is provided, including a data acquisition module, a pre-stored index database, and a data verification module. Through real-time data acquisition, dynamic environmental compensation, and dynamic verification units, and by utilizing multi-dimensional capability weights and adaptive sliding windows to identify logical drift, the system achieves real-time verification of digital simulation.
It achieves accurate capture and quantification of the overall performance of digital simulation, can identify logic drift, ensures that the simulation results are consistent with the combat capability of the actual aircraft platform, and prevents dimensional reduction attacks caused by inflated parameters.
Smart Images
Figure CN122242065A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of aircraft platform digital simulation, specifically relating to a dynamic verification system and method for aircraft platform digital simulation. Background Technology
[0002] Aircraft platform digital simulation refers to aircraft simulation controlled by artificial intelligence algorithms in a virtual environment. It can simulate the tactical behaviors, decision-making processes, and physical movements of human pilots. Aircraft platform digital simulation is an important tool for wargaming, effectiveness assessment, and personnel training. The combat effectiveness of the "AI-controlled aircraft platform" in digital simulation should be basically consistent with that of the "real aircraft operated by pilots" in the real world. This requires verification and validation. Verification ensures the model is accurate and free of derivation errors or logical flaws, while validation ensures that within the intended application scope, the digital simulation can accurately reproduce the system behavior of the real world, focusing on the degree of fit between the model and reality.
[0003] Digital simulation consists of behavioral and physical models. It is affected by both the intelligence level of the behavioral model and the refinement and parameters of the aircraft platform's physical model. Traditional simulation verification methods evaluate the behavioral and physical models separately, failing to assess the overall effectiveness of digital simulation. Furthermore, traditional static simulation verification methods cannot identify the systematic errors accumulated over time during long-term, highly dynamic adversarial processes, i.e., logic drift problems. Summary of the Invention
[0004] The purpose of this invention is to provide a dynamic verification system and method for digital simulation of aircraft platforms, which can accurately capture and quantify the timing-based logical drift, and achieve effective evaluation of the overall performance of digital simulation.
[0005] To achieve the above objectives, one aspect of the present invention provides a dynamic verification system for digital simulation of an aircraft platform, comprising a data acquisition module, a pre-stored index database, and a data verification module. The data acquisition module is used to receive raw data from digital simulation in real time. The raw data includes the spatiotemporal location and environmental information of the simulated entity. The spatiotemporal location includes altitude, and the environment includes meteorological and electromagnetic environments. The pre-stored index database contains the performance indicators of the aircraft platform under ideal conditions and performance correction mapping tables based on the effects of altitude, meteorological environment and electromagnetic environment. The data verification module includes an indicator calculation unit, an environmental dynamic compensation unit, and a dynamic verification unit. The index calculation unit is used to calculate the real-time observation value of the digital simulation based on the received raw data of the digital simulation. The environmental dynamic compensation unit is used to convert the performance indicators of the aircraft platform in the ideal environment into the dynamic expected capability value in the simulation environment by retrieving the performance correction mapping table and performing coupled calculations based on the spatiotemporal location and environmental information of the simulated entity. The dynamic verification unit is used to pre-set multi-dimensional capability weight vectors, identify the simulation stage and allocate multi-dimensional capability weights according to the stage. Based on real-time observations, dynamic expected capability values and multi-dimensional capability weights, it calculates individual capability residuals and multi-dimensional coupling comprehensive deviations. By using an adaptive sliding window to accumulate individual capability residuals and multi-dimensional coupling comprehensive deviations in the time domain, it determines whether logic drift has occurred.
[0006] Another aspect of the present invention provides a method for dynamic verification of digital simulation of an aircraft platform, which utilizes the above-described system to perform dynamic verification of digital simulation of an aircraft platform, including: Load the pre-stored index database to complete frame synchronization between the digital simulation system clock and the verification system clock; It receives the spatiotemporal location and environmental information of the simulated entity in real time and calculates the real-time observations of the digital simulation. Based on the current altitude, weather, and electromagnetic environment, the performance indicators of the aircraft platform under ideal conditions are converted into dynamic expected capability values under simulated conditions. Identify the current simulation stage, allocate multi-dimensional capability weights according to the stage, and calculate the residual of individual capabilities and the comprehensive deviation of multi-dimensional coupling. An adaptive sliding window is used to accumulate the residuals of individual capabilities and the deviation of multidimensional coupling in the time domain to determine whether logical drift has occurred.
[0007] According to the above-mentioned aircraft platform digital simulation dynamic verification system and method of the present invention, the timing-based logical drift can be accurately captured and quantified, thereby achieving effective evaluation of the overall performance of digital simulation. Attached Figure Description
[0008] To more clearly illustrate the technical solutions of the present invention, the accompanying drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort: Figure 1 This is a schematic diagram of the working scenario of an aircraft platform digital simulation dynamic verification system according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the operating logic of an aircraft platform digital simulation dynamic verification system according to an embodiment of the present invention; Figure 3 This is a schematic diagram of logic drift determination according to an embodiment of the present invention. Detailed Implementation
[0009] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.
[0010] One embodiment of the present invention provides a dynamic verification system for digital simulation of an aircraft platform. Figure 1 In the scenario shown, the dynamic verification system of this embodiment connects to the aircraft platform digital simulation system via middleware to perform dynamic verification of the aircraft platform digital simulation system. The dynamic verification system of this embodiment includes a data acquisition module, a pre-stored index database, a data verification module, and an evaluation display module. The following description, in conjunction with... Figure 2 The schematic diagram of the operating logic of the dynamic verification system of this invention provides a detailed description of each module.
[0011] The data acquisition module receives and analyzes raw data from digital simulations in the simulation environment in real time, including the spatiotemporal location of the simulated entity (latitude, longitude, altitude, velocity, attitude angle), sensor detection data (radar on / off status, radar target lock, alarm information), target characteristics (RCS (Radar Cross Section), infrared radiation), engagement information (weapon firing, weapon hit, jamming information), and environmental information (meteorological environment, electromagnetic environment). The pre-stored index database contains the static theoretical indices of the aircraft platform under standard atmospheric conditions and ideal conditions (performance indices under ideal conditions). It also includes a pre-set performance correction mapping table based on the effects of altitude, weather, and electromagnetic environment.
[0012] The data verification module is the core module of the dynamic verification system, which includes an indicator calculation unit, an environmental dynamic compensation unit, and a dynamic verification unit.
[0013] The indicator calculation unit is used to calculate real-time observations of the same dimension as the verification indicators based on the original data from the digital simulation. The calculation logic of the index calculation unit includes geometric feature calculation and kinematic feature calculation. Geometric feature calculation is based on the real-time coordinate vector difference between the simulated entity and the target, calculating the relative distance, azimuth angle, and approach angle. Kinematic feature calculation performs time differentiation on the attitude angles and velocity vectors of the simulated entity, calculating the instantaneous angular velocity and overload parameters (referring to the ratio of the aerodynamic force, mainly lift, experienced by the aircraft platform during maneuvering flight, to the total weight of the aircraft, usually quantified as a multiple of standard gravitational acceleration (G)).
[0014] The environmental dynamic compensation unit is used to extract real-time environmental parameters, retrieve the performance correction mapping table, and perform coupled calculations to adjust the aircraft platform performance indicators under ideal conditions. Mapped to the current digital simulation environment, output the dynamic expected capability value under the current environment. .
[0015] The calculation logic of the environmental dynamic compensation unit is as follows: based on the digital simulation dynamic data stream in the data acquisition module, including the location of the simulated entity, meteorological environment, and electromagnetic environment, it calls the static theoretical indicators in the pre-stored indicator database. Introducing a real-time environmental correction factor Including high correction factor Meteorological attenuation coefficient and electromagnetic interference correction factor The altitude correction coefficient is calculated based on interpolation from standard atmospheric tables, the meteorological attenuation coefficient is based on meteorological data models (such as clouds, rain, clear skies, snow, etc.), and the electromagnetic interference coefficient is based on a matching model of interference source power, distance, and frequency, to obtain the dynamic expected capability value under the current mission situation. :
[0016] Used to eliminate the influence of external environmental factors on the capabilities of digital simulation systems.
[0017] The dynamic verification unit is used to identify the simulation stage and assign weights according to the stage. and Multidimensional deviation calculations are performed, and consistency is determined based on a sliding window. In this embodiment, the simulation phase is, for example, the combat phase of an aircraft platform, which is divided into a search phase, an interception phase, and a dogfight phase (beyond visual range air combat only includes the search and interception phases). The search phase is defined as the simulated entity entering the mission airspace (radar silence) until the airborne sensors continuously and stably track the target and decide to launch an attack; the interception phase is defined as the simulated entity confirming the target and maneuvering to take position towards the target until the distance between the two sides is drastically reduced to within visual range (close-range dogfight) or beyond visual range missiles are launched and evasive maneuvers are taken (beyond visual range air combat); the dogfight phase is defined as the simulated entity entering within visual range and generating high maneuver overload until it hits the target and disengages from the battle or is shot down.
[0018] The dynamic verification unit has a dynamic weight allocation function, which can automatically identify the simulation stage based on situational data and perform weight allocation. In this embodiment, a multi-dimensional capability weight vector is preset. , These correspond to three capability dimensions: perception, decision-making, and execution. Each capability dimension includes multiple indicators. Specifically, the perception dimension indicators in digital simulation include radar acquisition range, target tracking stability, and multi-target locking capability; the decision-making dimension indicators include tactical switching timing and weapon launch position; and the execution dimension indicators include instantaneous angular velocity, energy loss, and available overload. The weighting ratio is automatically switched at different stages: the perception weight is increased during the search phase. Increase decision weight during the interception phase Increase execution weight during the combat phase .
[0019] The dynamic verification unit has a deviation calculation function, which can calculate the multi-dimensional coupled comprehensive deviation. The formula is:
[0020] in, It is the weight of the capability dimension (such as the search stage). M represents the number of capability dimensions, and N represents the number of capability dimensions. Number of indicators, dimensions The specific set of indicators can be represented as (such as perceptual dimensions) ); It is the allocation coefficient of a specific indicator within its respective dimension. This is used to illustrate the percentage contribution of radar acquisition range, target tracking stability, and multi-target lock-on capability to "sensing capability." Radar target acquisition is an event-driven message, while target tracking and target lock-on are periodic messages. When a radar target acquisition event occurs, When no radar target acquisition event occurs, ; It is the normalized deviation of a specific indicator for a single action (single deviation). .
[0021] In this embodiment, a single action specifically refers to a "single tactical action," which is a specific behavior with clearly defined boundaries performed by a digital simulation entity within a specific combat phase to achieve a certain tactical objective. A single tactical action is directly associated with the raw data acquired by the data acquisition module and is specifically divided into two mapping mechanisms: ① Event-triggered actions (such as the radar target acquisition time related to radar acquisition distance in perception capabilities, and the distance between the detecting and detected aircraft): determined by the state of "event-type messages" captured by the data acquisition module. ② State-persistent actions (such as the position of the detected aircraft and the position of the detecting radar target related to multi-target tracking stability in perception capabilities): determined based on "event-type messages" and combined with "periodic messages" continuously acquired by the data acquisition module.
[0022] The dynamic verification unit has an RMSE (Root Mean Square Error) logic drift determination function. It uses an adaptive sliding window T (e.g., adjusting the window length according to the high dynamic characteristics of the current adversarial situation; the greater the adversarial intensity, the shorter the time window T) to perform time-domain cumulative evaluation of the deviation data. Its core calculation formula is as follows:
[0023] Where n is the number of sampling frames contained in the sliding window T.
[0024] Specifically, for microscopic diagnosis, That is, the residual of a single ability. and These represent the real-time observation value and dynamic expected capability value of the k-th frame, respectively, and the micro root mean square error. The calculation formula is:
[0025] For macro-level assessment, That is, the comprehensive deviation sequence based on zero deviation. The multidimensional coupling comprehensive deviation of the k-th frame is represented by the macroscopic root mean square error. The calculation formula is:
[0026] Failure determination criteria: Logical drift: If the macro RMSE shows a significant upward trend over time t (i.e.) If the error exceeds a preset threshold, it is considered an accumulated error of the algorithm.
[0027] Performance parameter failure: If the microscopic RMSE deviates from zero and remains stable at a high level, it is determined that the underlying physical model parameters do not match the actual implementation.
[0028] The evaluation display module is used to provide multi-dimensional visualization of the verification results. In this embodiment, the evaluation display module can display the single deviation, overall deviation, and RMSE of the digital simulation. Specifically, the evaluation display module generates a bar chart based on the single deviation data to represent the capability deviation of a single tactical action; maps the overall deviation data to three-dimensional spatial coordinates to generate a three-dimensional spatial performance heatmap to represent the performance blind spot of the digital simulation in a specific spatial domain; and generates a line chart based on the time-series root mean square error (RMSE) data, and superimposes a dynamic judgment threshold to represent the drift state of the internal logic of the digital simulation.
[0029] Embodiments of the present invention also provide a method for dynamic verification of digital simulation of an aircraft platform, which uses the dynamic verification system of the above embodiments to perform dynamic verification of digital simulation of an aircraft platform, including the following steps.
[0030] Step 1: Initialization and Synchronization In this embodiment, a pre-stored index database is loaded, and static theoretical indices of a certain type of aircraft platform under standard atmospheric environment and ideal conditions are extracted. For example, setting its radar reference intercept range. Maximum overload available .
[0031] Complete frame synchronization between the digital simulation system clock and the verification system clock, and set the sampling frequency to 50Hz.
[0032] Step 2: Real-time data sampling It receives real-time spatiotemporal location information of simulated entities, periodic messages such as radar target lock, and event messages such as radar power-on, target acquisition, and missile launch. In this embodiment, for example... Figure 3 As shown, when the simulation reaches time t=15s (the period before t=15s is the steady state stage), the data acquisition module receives the spatiotemporal location message of the simulated entity (current height H=2000m), environmental information (moderate rain weather, encountering moderate electromagnetic suppression), and event-related messages (the radar target lock flag changes from 0 to 1).
[0033] Step 3: Real-time observations Dynamic solution The index calculation unit extracts the three-dimensional coordinate vector difference between the simulated entity and the target at the instant the "radar interception" event occurs, and calculates the actual radar interception distance observation value. At this point, an evasive maneuver is performed, and the velocity and attitude angles are differentiated to calculate the actual maximum usable overload observation value. .
[0034] Step 4: Dynamic Benchmark Mapping and Dynamic Expected Capability Value calculate The environmental dynamic compensation unit retrieves the corresponding correction coefficients from the performance correction mapping table based on the current altitude (2000m), weather conditions (moderate rain), and electromagnetic environment (moderate suppression). Assuming the table lookup yields: Altitude Correction Coefficient... Meteorological attenuation coefficient Electromagnetic interference coefficient .
[0035] Calculate the dynamic expected radar intercept range under the current environment using the following formula:
[0036] Thus, static theoretical indicators Converted into dynamic expected capability value .
[0037] Step 5: Dynamic allocation of identification and evaluation weights for the combat phase The system identifies the current combat phase and matches the corresponding weighting coefficients. In this embodiment, the system detects that fire control lock-on has been completed and accompanied by high-G maneuvers, determining that the air combat has transitioned from the "interception phase" to the "close-quarters combat phase." Based on preset rules, the system dynamically adjusts the multi-dimensional capability weight vector. Reduce the weight of the perception dimension to The decision-making dimension is set as The execution dimension weight has been increased to .
[0038] Step 6: Multidimensional Deviation Calculation and Coupling Analysis Calculate the normalized deviation of specific indicators (such as radar detection) for a single tactical maneuver:
[0039] Similarly, suppose the expected value of available overload is calculated based on the modified formula:
[0040] Calculate the multidimensional coupling comprehensive deviation by combining stage weights. Assuming that during the combat phase, the perception dimension only considers radar interception range ( The execution dimension only considers available overload ( Then the overall deviation of this frame is:
[0041] Step 7: Adaptive sliding window handling and logical drift determination The system detects that it is currently in a highly dynamic combat phase and adaptively shortens the sliding window length to a corresponding 2-second time window, including 100 sampled frames. The system uses the adaptive sliding window to perform a time-domain cumulative evaluation of the deviation data and calculates the root mean square error (RMSE). Within the window, the system simultaneously performs RMSE calculations at both the micro and macro levels: 1) Microscopic Diagnostic Level: This level assesses residuals for individual capabilities. The system extracts the absolute residuals between the actual radar interception distance and the expected value for 100 consecutive frames within a window. After substituting into the formula and calculating, we assume that the current window's micro-radar detection is obtained. If the microscopic RMSE is detected to deviate from zero and remain stable at a high level, it is determined that the underlying physical model parameters (such as radar basic gain settings) are mismatched with the actual implementation and have failed.
[0042] 2) Macro-level evaluation: This assesses the overall performance deviation. The system extracts a multi-dimensional coupled overall deviation sequence of 100 consecutive frames, i.e., ... After substituting into the formula and calculating, let's assume we obtain the macroscopic view of the current window. Furthermore, during the subsequent time t, within several consecutive sliding windows... It exhibits a significant upward trend of 8.2%-9.5%-12.1% (i.e., the derivative). And exceeding a preset threshold). At this point, the system determines that the overall decision-making or execution model of the digital simulation in the combat phase has accumulated algorithmic errors, resulting in a systematic logical drift, such as... Figure 3 As shown.
[0043] Step 8: Multidimensional Assessment and Visualization The evaluation and display module receives the processed deviation data and performs multi-dimensional visualization rendering and display: 1) Microscopic display of bar chart: The system extracts the deviation of a single tactical action in real time (such as the radar interception distance deviation of 10.55% and the overload deviation of 6.25% calculated in step 6), and represents the deviation of the digital simulation capability of a single action in real time through a dynamic bar chart.
[0044] 2) Macroscopic blind spot mining of heat map: This involves continuously calculating the overall deviation. The simulation is mapped to the corresponding three-dimensional spatial coordinates (latitude, longitude, and altitude) at the time of its generation for clustering and rendering. Since air combat takes place in three-dimensional space, the confidence level of the digital simulation varies in different airspaces. For example, when the digital simulation performs a cruise mission at an altitude of 10,000 meters, the model matching degree is high, with a deviation of only 2%, and the heatmap shows a green safe zone. However, when it descends to an altitude of 500 meters to perform low-altitude penetration, because the underlying physical model fails to fully calculate ground clutter or near-ground airflow, the deviation surges to 30%, and the heatmap shows a red high-risk warning color in the low-altitude region. This allows for the accurate identification of the digital simulation's performance blind spots.
[0045] 3) RMSE Line Chart Trend Determination: The system plots the accumulated RMSE data over time using a line chart in real time, marking dynamic thresholds with red step lines. Areas where RMSE exceeds the threshold are shaded, visually indicating to evaluators that logical drift (such as accumulated error) has occurred in the digital simulation. This effectively distinguishes between short-term "accidental errors" and long-term "systemic failures," assisting commanders in determining whether the simulation has failed and in implementing verification and intervention.
[0046] In summary, the aircraft platform digital simulation dynamic verification system and method of this invention, by periodically receiving simulation data in real time, realizes real-time comparison in dynamic processes such as aircraft maneuvering, sensor detection, and weapon strikes, and can capture transient logical conflicts that cannot be detected by static testing; it establishes a dynamic benchmark mapping mechanism based on physical characteristics, and calculates the "expected performance to be achieved under the current environment" in real time using a correction function according to the actual altitude, weather, and electromagnetic interference intensity of the current digital simulation, ensuring the scientific nature of the verification results; it proposes a mission-oriented weight allocation logic, which automatically identifies the air combat phase (search, interception, dogfight) and dynamically adjusts the evaluation weights of the three dimensions of perception, decision-making, and execution, and evaluates the contribution of various capabilities of digital simulation to the overall effectiveness according to different combat phases, making the verification results more consistent with the evaluation logic in actual combat.
[0047] The aircraft platform digital simulation dynamic verification system and method of this invention have the following beneficial effects: This invention ensures the objectivity and fairness of the evaluation benchmark under complex situations through an environmental dynamic benchmark mapping mechanism; it achieves precise benchmarking throughout the entire combat process using a task-oriented weight follow-up mechanism; and it effectively compensates for the shortcomings of traditional static verification in failing to capture dynamic failures by identifying deep logical drift and accumulated errors through an adaptive sliding window, providing scientific data support for improving the confidence of digital simulation in actual combat. In virtual-real combined combat, this invention can serve as a verification tool to determine whether digital simulation possesses the ability to confront actual aircraft platforms. By comparing the performance of digital simulation and actual equipment in real time, it ensures that both sides are confronting each other on the same tactical level, preventing "dimensional reduction attacks caused by inflated parameters in digital simulation."
[0048] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.
Claims
1. A digital simulation dynamic verification system for an aircraft platform, characterized in that, It includes a data acquisition module, a pre-stored indicator database, and a data verification module; The data acquisition module is used to receive raw data from digital simulation in real time. The raw data includes the spatiotemporal location and environmental information of the simulated entity. The spatiotemporal location includes altitude, and the environment includes meteorological and electromagnetic environments. The pre-stored index database contains the performance indicators of the aircraft platform under ideal conditions and performance correction mapping tables based on the effects of altitude, meteorological environment and electromagnetic environment. The data verification module includes an indicator calculation unit, an environmental dynamic compensation unit, and a dynamic verification unit. The index calculation unit is used to calculate the real-time observation value of the digital simulation based on the received raw data of the digital simulation. The environmental dynamic compensation unit is used to convert the performance indicators of the aircraft platform in the ideal environment into the dynamic expected capability value in the simulation environment by retrieving the performance correction mapping table and performing coupled calculations based on the spatiotemporal location and environmental information of the simulated entity. The dynamic verification unit is used to pre-set multi-dimensional capability weight vectors, identify the simulation stage and allocate multi-dimensional capability weights according to the stage. Based on real-time observations, dynamic expected capability values and multi-dimensional capability weights, it calculates individual capability residuals and multi-dimensional coupling comprehensive deviations. By using an adaptive sliding window to accumulate individual capability residuals and multi-dimensional coupling comprehensive deviations in the time domain, it determines whether logic drift has occurred.
2. The system as described in claim 1, characterized in that, The dynamic expected capacity value of the environmental dynamic compensation unit is calculated as follows. : in, Performance indicators under ideal conditions For high correction factor, This is the meteorological attenuation coefficient. This is the electromagnetic interference correction factor.
3. The system as described in claim 2, characterized in that, The dynamic verification unit calculates the multidimensional coupling comprehensive deviation as follows. : in, It is a capability dimension The weights are M, where M is the number of capability dimensions, and N is the number of capability dimensions. The number of indicators in the data. It is the allocation coefficient of indicator i within its respective capability dimension. It is the single deviation of indicator i. , For the real-time observed value of index i, Let i be the dynamic expected capability value of indicator i.
4. The system as described in claim 3, characterized in that, The dynamic verification unit utilizes an adaptive sliding window to accumulate the multidimensional coupled comprehensive deviation in the time domain and calculate the macroscopic root mean square error. : Where n is the number of sampling frames contained within the sliding window. The multidimensional coupling comprehensive deviation of the k-th frame; When the derivative of the macroscopic root mean square error is greater than 0 and exceeds the preset threshold, the dynamic verification unit determines that the algorithm has accumulated error and a systematic logical drift has occurred.
5. The system as described in claim 4, characterized in that, The dynamic verification unit utilizes an adaptive sliding window to accumulate individual capability residuals over time and calculate the micro root mean square error. : in, and These are the real-time observation value and the dynamic expected capability value of the k-th frame, respectively. Indicates the residual of a single ability; When the micro-root mean square error of the dynamic verification unit deviates from zero and remains stable at a high level, it is determined that the parameters of the underlying physical model do not match the actual installation.
6. The system as described in claim 5, characterized in that, It also includes an evaluation and display module, which is used to generate a bar chart based on single deviation data to characterize the capability deviation of a single action; map the comprehensive deviation data to three-dimensional spatial coordinates to generate a three-dimensional spatial performance heat map to characterize the performance blind zone of digital simulation in a specific spatial domain; and generate a line chart based on macroscopic and microscopic root mean square error data and superimpose a dynamic judgment threshold to characterize the drift state of the internal logic of digital simulation.
7. The system as described in any one of claims 1-6, characterized in that, The calculation of the index calculation unit includes geometric feature calculation and kinematic feature calculation. Geometric feature calculation is based on the real-time coordinate vector difference between the simulated entity and the target to calculate the relative distance, azimuth angle and approach angle. Kinematic feature calculation is to perform time differentiation processing on the attitude angle and velocity vector of the simulated entity to calculate the instantaneous angular velocity and overload parameters.
8. The system as described in any one of claims 3-6, characterized in that, The capability dimensions include perception, decision-making, and execution. Perception dimension indicators include radar acquisition range, target tracking stability, and multi-target locking capability. Decision-making dimension indicators include tactical switching timing and weapon launch position. Execution dimension indicators include instantaneous angular velocity, energy loss, and available overload.
9. A method for dynamic verification and validation of digital simulation of an aircraft platform, characterized in that, Using the system described in any one of claims 1-8 to perform dynamic verification of aircraft platform digital simulation includes: Load the pre-stored index database to complete frame synchronization between the digital simulation system clock and the verification system clock; It receives the spatiotemporal location and environmental information of the simulated entity in real time and calculates the real-time observations of the digital simulation. Based on the current altitude, weather, and electromagnetic environment, the performance indicators of the aircraft platform under ideal conditions are converted into dynamic expected capability values under simulated conditions. Identify the current simulation stage, allocate multi-dimensional capability weights according to the stage, and calculate the residual of individual capabilities and the comprehensive deviation of multi-dimensional coupling. An adaptive sliding window is used to accumulate the residuals of individual capabilities and the deviation of multidimensional coupling in the time domain to determine whether logical drift has occurred.