A thermal-electric coupling simulation interface and method for large transport unmanned aerial vehicles

Abstract

This invention discloses a thermal-electric coupling simulation interface and method for large transport drones. The simulation interface is located between the power electronics system simulation subsystem and the thermal management system simulation subsystem, and includes: a physical layer signal conditioning module for capturing transient waveforms of the power electronics system simulation subsystem and converting them into discrete power loss sequences; a cross-domain parameter mapping matrix for mapping the discrete power loss sequences to spatial heat source distribution terms identifiable by the thermal management system simulation subsystem; a multi-timescale synchronization coordinator including a dual-buffered storage architecture for data synchronization; and a real-time closed-loop feedback controller for real-time correction of the physical model parameters of semiconductor components in the power electronics system simulation subsystem based on temperature data fed back from the thermal management system simulation subsystem, thereby constructing a bidirectional closed-loop simulation circuit. This invention achieves real-time bidirectional closed-loop interaction of thermal-electric energy flow, significantly improving the simulation accuracy and numerical stability of large transport drones.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) simulation technology, and particularly relates to a thermal-electric coupling simulation interface and method for large transport UAVs. Background Technology

[0002] Against the backdrop of the current global air transport system's leapfrog development towards intelligence and unmanned operation, large transport drones, with their significant advantages such as strong payload capacity, long range, high mission flexibility, and relatively low operation and maintenance costs, have gradually become an indispensable core force in the fields of modern logistics distribution, emergency rescue and support, and long-distance material delivery.

[0003] With the continuous evolution of aviation electric propulsion technology, large transport drones are rapidly moving towards high voltage, high power, and high power density. Their internal integrated energy systems, power propulsion systems, and complex avionics systems are showing a high degree of electrification and integration.

[0004] Within the existing technological framework, simulation studies for large transport drones are typically conducted based on a functional decoupling approach. Power electronic system simulations primarily focus on current ripples, voltage fluctuations, power conversion efficiency, and the dynamic response of control laws, tending to provide detailed modeling of the power conversion process on microsecond or even nanosecond timescales. Thermal management system simulations, on the other hand, emphasize evaluating the temperature rise history, cooling medium flow field distribution, and radiator efficiency of motors, battery packs, and power conversion modules through computational fluid dynamics (CFD) or thermal circuit models, with response times typically on the order of seconds or minutes. This divide-and-conquer modeling strategy can indeed provide valuable engineering data at a relatively low computational cost in low power density or steady-state operation scenarios, effectively supporting the development of early drone systems at specific historical stages.

[0005] However, as large transport drone applications place increasingly stringent demands on power density and system reliability, the inherent characteristics of the aforementioned decoupled simulation scheme at the principle level have gradually revealed significant limitations in addressing the challenges of strong thermal-electric coupling under complex operating conditions.

[0006] Current technologies lack a simulation interface that is compatible with heterogeneous physical fields and supports cross-scale time coordination. In actual simulations, because electrical and thermal simulations belong to different software platforms or solver kernels, data exchange often relies on manual offline mapping or low-frequency hard synchronization. This not only leads to severe data lag but also fails to capture system-level transient anomalies caused by thermoelectric coupling effects under extreme flight envelopes or sudden load changes. Furthermore, the limited airframe space of large transport UAVs restricts the physical redundancy of the cooling system, significantly increasing the system's sensitivity to temperature rise. Any prediction deviations due to insufficient simulation interface accuracy could potentially evolve into thermal breakdown of power modules or thermal runaway risks in actual flight.

[0007] Based on this, how to establish a thermal-electric coupling simulation interface with high dynamic response capability, high parameter mapping accuracy and multi-timescale collaborative characteristics while maintaining the advantage of independent solution of each physical field has become a key technical bottleneck restricting the leap of large transport drones towards high performance and high safety.

[0008] Therefore, in response to the aforementioned profound technical contradictions, developing a simulation interface capable of achieving deep interaction and precise matching between electrical energy flow and thermal energy flow has become a key challenge and an urgent technical problem for those skilled in the art. Summary of the Invention

[0009] This invention aims to provide a thermal-electric coupling simulation interface and method for large transport drones. By constructing a complete architecture that includes physical layer signal conditioning, cross-domain mapping, and multi-scale coordination, it enables real-time bidirectional closed-loop interaction between power loss data and transient temperature distribution data, significantly improving the simulation accuracy and numerical stability of large transport drones.

[0010] According to one aspect of the present invention, a thermal-electric coupling simulation interface for large transport unmanned aerial vehicles is provided, which is set between a power electronics system simulation subsystem and a thermal management system simulation subsystem. The simulation interface includes a cross-domain coupling and synchronization execution module, which includes a physical layer signal conditioning module, a cross-domain parameter mapping matrix, a multi-timescale synchronization coordinator, and a real-time closed-loop feedback controller, wherein: The physical layer signal conditioning module is used to capture the transient waveforms of the power electronic system simulation subsystem and convert them into discrete power loss sequences. The cross-domain parameter mapping matrix is ​​used to map the power loss discrete sequence into a spatial heat source distribution item that can be identified by the thermal management system simulation subsystem. The multi-timescale synchronization coordinator includes a dual-buffered storage architecture for synchronizing data between the first simulation step of the power electronic system simulation subsystem and the second simulation step of the thermal management system simulation subsystem. The real-time closed-loop feedback controller is used to correct the physical model parameters of semiconductor components in the power electronic system simulation subsystem in real time based on the temperature data fed back by the thermal management system simulation subsystem, so as to construct a bidirectional closed-loop simulation circuit.

[0011] According to some embodiments, the multi-timescale synchronization coordinator includes a double-buffered storage architecture: The dual-buffered storage architecture includes a first buffer and a second buffer; The first buffer is used to accumulate and write the power loss discrete sequence at the first simulation step size, and to output the average thermal power at the trigger time of the second simulation step size, wherein the first simulation step size is shorter than the second simulation step size. The second buffer is used to receive low-frequency temperature data fed back by the thermal management system simulation subsystem, and to generate a pseudo-continuous temperature signal that matches the first simulation step size through an interpolation algorithm.

[0012] According to some embodiments, the real-time closed-loop feedback controller includes a parameter correction engine, which dynamically adjusts the simulation parameters of the carrier mobility, on-resistance, and turn-off delay time of the semiconductor elements in the power electronic system simulation subsystem based on the received temperature feedback signal, so as to simulate the thermally induced performance degradation process under real physical environment.

[0013] According to some embodiments, the physical layer signal conditioning module includes a high-bandwidth current sampling unit, a transient voltage monitoring unit, and a signal preprocessing unit: The high-bandwidth current sampling unit is used to collect the transient current on the output side of the inverter; The transient voltage monitoring unit is used to monitor the transient voltage of the power bus; The signal preprocessing unit is used to filter and perform analog-to-digital conversion on the acquired transient current and transient voltage to generate the input data of the power loss discrete sequence. The signal preprocessing unit also has a data compression function, which removes redundant stationary segments from the electromagnetic simulation waveform using a compression algorithm based on slope threshold determination.

[0014] According to some embodiments, the cross-domain parameter mapping matrix includes a loss calculation unit and a heat source mapping unit: The loss calculation unit includes a lookup table model for switching losses and an analytical model for conduction losses of semiconductor devices, which are used to calculate energy loss based on transient electrical waveforms and real-time junction temperature data. The heat source mapping unit distributes the calculated losses to the corresponding spatial grid nodes according to the geometric topology of the power conversion module, forming a non-uniformly distributed heat source input. The heat source mapping unit also identifies the differences in thermal diffusivity of different material layers by modeling the thermal characteristics of the heterogeneous packaging material, and proportionally distributes the power loss to the three-dimensional coordinate nodes of each material layer.

[0015] According to some embodiments, the cross-domain coupling and synchronous execution module further includes: An environmental envelope simulation unit is connected to the interface of the UAV flight control system to obtain ambient temperature, air pressure and convective heat transfer coefficient, and inject them as dynamic boundary conditions into the thermal management system simulation subsystem. The flow field state monitoring unit is used to acquire the flow rate and velocity of the cooling medium and the operating power consumption of the pump, and convert them into real-time correction instructions for the convective heat transfer boundary conditions in the thermal management system simulation subsystem.

[0016] According to some embodiments, the multi-timescale synchronization coordinator further includes a dynamic step size management strategy: When the power electronic system simulation subsystem is in steady-state operation, the data aggregation time window is increased to reduce the synchronization frequency; When a sudden load change or entry into a violent maneuver envelope is detected, the multi-timescale synchronization coordinator immediately switches to high-frequency synchronization mode.

[0017] According to some embodiments, the simulation interface further includes a system control and state assurance module, which includes a dual-core processor architecture, a self-diagnostic monitoring module, and a non-volatile memory module, wherein: The dual-core processor architecture includes a first processor core and a second processor core. The first processor core is responsible for running the real-time operating system and communication protocol stack, while the second processor core serves as a dedicated mathematical operation accelerator, responsible for performing cross-domain parameter mapping operations. The self-diagnostic monitoring module is used to continuously monitor the power supply voltage, operating temperature and logic clock frequency inside the interface, and to force the simulation system into a safe suspension state when there is a hardware abnormality, and drive the status indicator array to display the corresponding error code. The non-volatile storage module is used to record the data flow during the simulation process to support offline backtracking analysis after the simulation task is completed.

[0018] According to some embodiments, the system control and status protection module and the cross-domain coupling and synchronization execution module are encapsulated in an electromagnetic compatibility (EMC) protective enclosure, and the inner wall of the EMC protective enclosure is coated with a microwave absorbing material with a thickness of not less than 0.5 mm.

[0019] According to another aspect of the present invention, a method for thermal-electric coupling simulation of large transport unmanned aerial vehicles is provided, comprising: Load electrothermal characteristic parameters and establish a synchronous data link between the simulation interface and the power electronics system simulation subsystem; The electrical simulation waveform output by the power electronic system simulation subsystem is captured in real time by the physical layer signal conditioning module, converted into a discrete power loss sequence, and mapped to a spatial heat source distribution term through a cross-domain parameter mapping matrix. A multi-timescale synchronization coordinator is used to perform time-scale scaling and averaging on the electrical simulation data of the first time scale and the thermal simulation data of the second time scale, and the processed heat source data is transmitted to the thermal management system simulation subsystem for temperature distribution solution. The temperature distribution data from the thermal simulation is received by the real-time closed-loop feedback controller, which drives the parameter correction engine to dynamically adjust the physical model parameters of the components in the power electronic system simulation subsystem, forming a deep closed-loop coupling between thermal energy flow and electrical energy flow.

[0020] The technical solutions provided by the embodiments of the present invention have the following beneficial effects: This invention achieves high-precision mapping of power electronic signals to thermal field distribution, multi-timescale data synchronization, and electro-thermal bidirectional closed-loop feedback control by constructing a cross-domain coupling and synchronization execution module, a physical layer signal conditioning module, a cross-domain parameter mapping matrix, a multi-timescale synchronization coordinator, and a real-time closed-loop feedback controller. In particular, the dual-buffered storage architecture employed in the multi-timescale synchronization coordinator effectively solves the computational bottleneck caused by the mismatch between the electro-thermal simulation step sizes. While ensuring real-time transmission of thermal response data, it avoids numerical divergence caused by excessively large data spans, thus ensuring the numerical stability of the thermo-electric closed-loop simulation of large transport UAVs in complex flight missions.

[0021] The real-time closed-loop feedback controller enables dynamic real-time correction of physical characteristics. By directly injecting temperature effects into the underlying component model, the simulation process possesses adaptive dynamic physical processes, realistically reproducing the nonlinear characteristics of large transport drones under extreme environments or high power output, such as thermal aging and efficiency drops. It solves the problem of severe disconnect between electrical and thermal physical fields in traditional drone simulations, achieving a continuous physical description from microsecond-level switching transients to minute-level temperature rise evolution, significantly improving simulation accuracy and accurately capturing electrical performance instability caused by localized overheating.

[0022] The simulation interface's cross-domain coupling and synchronous execution module also integrates an environmental envelope simulation unit and a flow field monitoring unit, greatly enhancing the engineering applicability of the simulation scheme. It can not only simulate the internal heat flow interaction of the system but also couple the influence of external flight dynamics on heat dissipation characteristics, providing precise data support for the design of UAV cooling systems, the formulation of power limiting strategies, and the optimization of thermal protection schemes.

[0023] The system control and status assurance module in the simulation interface integrates a dual-core processor architecture, a self-diagnostic monitoring module, a non-volatile storage module, and an electromagnetic compatibility protection structure. It is used to schedule and control the overall operation of the interface, indicate its status, record data, and protect against environmental interference, ensuring the high reliability and stability of the simulation interface under complex working conditions.

[0024] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0025] Figure 1 This is a block diagram illustrating the structural composition of a thermal-electric coupling simulation interface for large transport drones, as shown in an example embodiment of the present invention.

[0026] Figure 2 This is a schematic diagram of the connection of the simulation interface in an example embodiment of the present invention.

[0027] Figure 3 This is a schematic diagram of the physical layer signal conditioning module in an example embodiment of the present invention.

[0028] Figure 4 This is a schematic diagram of the logical structure of the cross-domain parameter mapping matrix in an example embodiment of the present invention.

[0029] Figure 5 This is a flowchart of a thermal-electric coupling simulation method for large transport drones, as an example embodiment of the present invention.

[0030] Figure 6 This is a schematic diagram of the simulation process described in this invention. Detailed Implementation

[0031] To further understand the content of this invention, a detailed description of the invention is provided in conjunction with the accompanying drawings and embodiments. The invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the relevant invention and are not intended to limit the invention. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0032] This embodiment proposes a thermal-electric coupling simulation interface for large transport drones. This solution is not a simple signal pass-through, but rather aims to achieve real-time bidirectional closed-loop interaction between power loss data and transient temperature distribution data within the microsecond to millisecond range during the operation of the large transport drone's propulsion system.

[0033] like Figure 1 , Figure 2 This embodiment proposes a thermal-electric coupling simulation interface for large transport drones.

[0034] According to the example embodiment, the simulation interface, serving as the core data exchange and physical property adjustment hub, is located between the power electronics system simulation subsystem 100 and the thermal management system simulation subsystem 200. The simulation interface includes a cross-domain coupling and synchronization execution module 001 and a system control and state assurance module 002. The cross-domain coupling and synchronization execution module 001 includes a physical layer signal conditioning module 1, a cross-domain parameter mapping matrix 2, a multi-timescale synchronization coordinator 3, and a real-time closed-loop feedback controller 4.

[0035] Specifically, the physical layer signal conditioning module 1 is used to capture the transient waveform of the power electronic system simulation subsystem 100 and convert it into a discrete power loss sequence. As the signal acquisition front end of the interface, the physical layer signal conditioning module 1 determines the fidelity of the electrical signal acquisition.

[0036] Specifically, the cross-domain parameter mapping matrix 2 is used to map the power loss discrete sequence into a spatial heat source distribution item that can be identified by the thermal management system simulation subsystem 200. The cross-domain parameter mapping matrix 2 is deployed in the field programmable gate array processor inside the interface system and undertakes the key task of converting the loss information in the electric domain into the energy distribution information in the thermal domain.

[0037] The multi-timescale synchronization coordinator 3 includes a dual-buffered storage architecture for synchronizing data between the first simulation (electrical simulation) step size of the power electronic system simulation subsystem 100 and the second simulation (thermal simulation) step size of the thermal management system simulation subsystem 200. The dual-buffered storage architecture includes a first buffer and a second buffer. The first buffer is used to accumulate and write the discrete power loss sequence at the first simulation step size and output the average thermal power at the trigger time of the second simulation step size, wherein the first simulation step size is shorter than the second simulation step size. The second buffer is used to receive low-frequency temperature data fed back by the thermal management system simulation subsystem 200 and generate a pseudo-continuous temperature signal that matches the first simulation step size through an interpolation algorithm.

[0038] When the thermal management system simulation subsystem 200 triggers a solution cycle, the first buffer outputs the average thermal power within that time period, thus achieving a smooth transition from microsecond-level discrete data to the step-size load required for thermal simulation. The second buffer undertakes the reverse transmission task, receiving low-frequency spatial temperature distribution data generated by thermal simulation, and converting the second-level temperature signal into a pseudo-continuous signal with microsecond-level steps through built-in linear interpolation logic, thereby completely eliminating the numerical oscillation phenomenon in electrical simulation caused by the step jump of the temperature signal.

[0039] Furthermore, the multi-timescale synchronization coordinator 3 also includes a dynamic step size management strategy. When the power electronic system simulation subsystem 100 is running in steady state, the data aggregation time window is increased to reduce the synchronization frequency, thereby saving computing resources. When a sudden load change or entry into a severe maneuver envelope is detected, the multi-timescale synchronization coordinator immediately switches to a high-frequency synchronization mode to ensure that transient thermal shocks are handled in a timely manner. For example, when the real-time monitored inverter bus current change rate exceeds 200% / ms of the preset rated current, or the roll angular velocity returned by the flight control interface exceeds 30 deg / s, the step size controller is triggered to switch the synchronization frequency from 100Hz to 10kHz.

[0040] Furthermore, the multi-timescale synchronization coordinator 3 adopts a multi-stage pipelined processing structure, parallelizing various processes such as data acquisition, loss calculation, scale conversion, and message encapsulation. This structural design minimizes the physical delay of signals passing through the interface system, locking the loop delay time of electro-thermal coupling to the microsecond level, thus meeting the stringent requirements for high-fidelity simulation of the rapid transient processes of large transport UAVs.

[0041] Specifically, the real-time closed-loop feedback controller 4 is used to correct the physical model parameters of the semiconductor components in the power electronic system simulation subsystem 100 in real time based on the temperature data fed back by the thermal management system simulation subsystem 200, so as to construct a bidirectional closed-loop simulation circuit. The real-time closed-loop feedback controller 4 integrates a parameter correction engine, which dynamically adjusts the simulation parameters of the carrier mobility, on-resistance, and turn-off delay time of the semiconductor components in the power electronic system simulation subsystem 100 according to the received temperature feedback signal, so as to simulate the thermally induced performance degradation process under real physical environment.

[0042] According to some embodiments, the parameter correction engine operates by dynamically adjusting the controlled source parameters of semiconductor components in the power electronic system simulation subsystem 100. Specifically, when the received temperature feedback signal indicates an increase in junction temperature, the parameter correction engine automatically corrects the carrier mobility parameters in real time based on the semiconductor physical property file. This correction manifests as an increase in the on-resistance of the power devices and a drift in the gate drive impedance. Furthermore, the engine simultaneously reduces dynamic parameters such as turn-off delay time, thereby realistically reproducing the efficiency reduction or enhanced electromagnetic interference characteristics of the power system caused by thermal performance degradation in the simulation environment. This transforms the simulation process from an isolated, unidirectional calculation into a dynamic physical system model with adaptive evolution capabilities.

[0043] Specifically, the parameter correction engine incorporates a temperature-electrical parameter mapping relationship based on semiconductor physical properties. This mapping relationship defines a functional model of key electrical parameters as a function of junction temperature, thereby quantitatively and in real-time converting the spatial temperature distribution data fed back from the thermal management system simulation subsystem into correction instructions for the component models in the power electronics system simulation subsystem. Carrier mobility is a key parameter determining the conductivity of semiconductor devices and is closely related to temperature. The temperature-dependent correction parameter for carrier mobility (μ) is set in the correction engine with the following temperature-dependent function: μ(T) = μ0*(T0 / T) αμ Where μ(T) is the calculated carrier mobility at the current junction temperature T, μ0 is the reference carrier mobility value calibrated at the reference temperature T0 (usually 25°C or 300K), T is the current junction temperature (unit: K) received by the real-time closed-loop feedback controller from the thermal management system simulation subsystem, and αμ is the mobility temperature decay coefficient.

[0044] The mobility temperature decay coefficient is an empirical constant related to semiconductor materials (typically ranging from 1.5 to 2.5), which characterizes the decay rate of carrier mobility as temperature increases. The specific value is obtained by fitting characteristic curves or experimental data provided by the device manufacturer and is preset in the lookup table of the parameter correction engine.

[0045] As the junction temperature T increases, the μ(T) calculated from this function will be less than μ0. The decrease in mobility will directly lead to an increase in the channel resistance and a decrease in the transconductance of the device, thus realistically reproducing the weakening of the device's driving capability and the change in switching speed caused by high temperature in the simulation. The parameter correction engine injects this calculated μ(T) value into the semiconductor device physical model of the power electronic system simulation subsystem in real time.

[0046] Temperature-dependent correction of on-resistance (Rdson) is simulated using piecewise linear or specific function models. For example, the on-resistance of power MOSFETs or IGBTs has a significant positive temperature coefficient, and the following linear approximation is used for fast real-time correction: Rdson(T)= Rdson0* [1+kR* (T-T0)] Where Rdson(T) is the calculated on-resistance at the current junction temperature T, Rdson0 is the reference on-resistance calibrated at the reference temperature T0, and kR is the temperature coefficient of on-resistance.

[0047] According to some embodiments, the on-resistance is linearly correlated with temperature, such as "Rdson increases by 15% for every 10°C increase". The parameter correction engine calculates the current Rdson(T) based on the feedback temperature T according to the above relationship, and dynamically corrects the on-resistance parameter of the corresponding power device in the simulation circuit accordingly, so that the simulation can accurately calculate the increased on-state loss (I²R) caused by the increase in on-resistance at high junction temperatures.

[0048] The switching delay time is affected by factors such as carrier recombination rate and threshold voltage drift, and is also temperature-dependent. The temperature correction model for the turn-off delay time (td(off)) can be expressed as: td(off)(T)=td(off)0* exp[(Ea / k)*(1 / T0-1 / T)] Where td(off)(T) is the calculated turn-off delay time at the current junction temperature T, td(off)0 is the reference turn-off delay time at the reference temperature T0, Ea is the activation energy (unit: eV) related to the turn-off process, which is a parameter related to the device physical mechanism, and k is the Boltzmann constant.

[0049] By adjusting the turn-off delay time parameter, the parameter correction engine enables electrical simulations to model the effects of temperature on switching transients. For example, increased switching delay time at high temperatures may result in longer switching overlap times in the simulation, affecting switching losses and voltage and current stresses, which are crucial for predicting electromagnetic interference (EMI) and system efficiency.

[0050] The parameter correction engine's workflow is as follows: Upon receiving the interpolated junction temperature signal T corresponding to the current electrical simulation moment from the multi-timescale synchronization coordinator, it calculates or retrieves the target parameter values ​​such as μ(T), Rdson(T), and td(off)(T) at the current temperature T in parallel, based on pre-stored mathematical functions or lookup tables. The calculated new parameter set is then sent to the power electronics system simulation subsystem in real-time via the output interface of the real-time closed-loop feedback controller in the form of a "model parameter update command." At the start of the next simulation step, the power electronics system simulation subsystem uses the updated parameters to solve the electrical network equations.

[0051] Furthermore, the real-time closed-loop feedback controller 4 also has a fault injection function. During the simulation, the real-time closed-loop feedback controller 4 can be manually set to typical thermal management system fault modes such as cooling fan failure, coolant leakage, or aging of thermal interface materials, and observe in real time how these faults evolve into abnormal electrical parameters or even system collapse through the thermal-electric coupling path, providing a highly realistic virtual test environment for the development of fault diagnosis algorithms and the verification of fault-tolerant control strategies for large transport UAVs.

[0052] In a preferred embodiment of the present invention, the real-time closed-loop feedback controller 4 considers the temperature sensitivity of the packaged parasitic inductance when correcting electrical parameters. Based on the thermal expansion coefficients of each layer of the packaged material, it calculates the subtle geometric deformations at the wire bonding points and bus structures caused by temperature increases, thereby correcting the parasitic inductance parameters in the simulation model and accurately predicting the potential enhancement of turn-off voltage spikes under high-temperature operating conditions.

[0053] To address the complexities of large transport drones operating at different flight altitudes and in varying geographical environments, the simulation interface is additionally equipped with an environmental envelope simulation unit 5 and a flow field state monitoring unit 6. The environmental envelope simulation unit 5 is connected to the drone flight control system interface and is used to acquire ambient temperature, air pressure, and convective heat transfer coefficients, which are then injected as dynamic boundary conditions into the thermal management system simulation subsystem 200.

[0054] According to some embodiments, the environmental envelope simulation unit 5 interfaces with the simulation data link of the UAV flight control system through a dedicated physical interface to acquire key environmental parameters such as flight altitude, flight speed, and external air pressure in real time. The environmental envelope simulation unit 5 converts these parameters into convective heat transfer coefficients and boundary temperatures, and injects them into the boundary conditions of the thermal management system simulation subsystem 200 in real time. For example, when simulating a UAV cruising at an altitude of 30,000 feet, this unit automatically reduces the reference value of the convective heat transfer coefficient of the radiator based on the physical properties of the thin atmosphere at high altitudes. This allows the entire thermo-electric coupling simulation process to cover the entire mission profile of the UAV from takeoff, high-altitude cruising, to landing, greatly improving the accuracy of system-level thermal balance analysis.

[0055] The flow field state monitoring unit 6 is used to acquire the flow rate and velocity of the cooling medium and the operating power consumption of the pump, and convert them into real-time correction instructions for the convective heat transfer boundary conditions in the thermal management system simulation subsystem 200. Preferably, the cross-domain parameter mapping matrix 2 also introduces a reduction-order processing method based on a three-dimensional thermal sensitivity coefficient matrix when performing heat source allocation. When dealing with the complex liquid cooling system or forced air cooling system of large transport UAVs, the complex fluid dynamics mesh can be simplified to the algebraic coupling relationship between key physical nodes. This simplification process reduces the time cost of a single thermal simulation solution from several seconds to milliseconds while retaining the core heat conduction path and convection characteristics. By monitoring the flow rate and velocity of the cooling medium in real time through the flow field state monitoring unit 6, and dynamically correcting the heat dissipation boundary conditions in the thermal management system simulation subsystem 200 according to the pump operating status, a deep integration of fluid dynamic effects and electrothermal characteristics is achieved.

[0056] According to some embodiments, the simulation interface further includes a system control and status protection module 002, which includes a dual-core processor architecture, a self-diagnostic monitoring module 15, and a non-volatile storage module 14. The dual-core processor architecture includes a first processor core 7 and a second processor core 8. The first processor core 7 is responsible for running the real-time operating system and communication protocol stack, and the second processor core 8 serves as a dedicated mathematical operation accelerator, responsible for performing cross-domain parameter mapping operations.

[0057] Specifically, the self-diagnostic monitoring module 15 is used to continuously monitor the power supply voltage, operating temperature and logic clock frequency inside the interface, and to force the simulation system into a safe suspension state when there is a hardware abnormality, and drive the status indicator array 10 to display the corresponding error code. The non-volatile storage module 14 is used to record the data flow during the simulation process to support offline backtracking analysis after the simulation task ends.

[0058] In a preferred embodiment of the present invention, the status indicator array 10 displays the current data transmission load, synchronization error status, and connectivity of each simulation channel in real time using LEDs of different colors. Internally, the physical layer signal conditioning module 1 employs a differential signal transmission-based wiring scheme and is equipped with a common-mode choke, effectively suppressing conducted interference from complex power lines in an airborne environment and ensuring the integrity of weak simulation signals.

[0059] According to some embodiments, the simulation interface adopts a multi-core heterogeneous architecture in its hardware implementation. The first processor core 7 serves as the system's main control unit, responsible for running the real-time operating system (RTOS), handling complex communication protocol stacks, and managing the status indicator array 10. The second processor core 8 is configured as a dedicated mathematical operation accelerator, employing a single-instruction multiple-data (SIMD) architecture, focusing on performing high-frequency cross-domain parameter mapping operations and interpolation processing of multi-timescale data. This dual-core processor architecture design ensures that even when handling complex flow field coupling simulations with tens of thousands of mesh nodes, the end-to-end processing latency of the interface system can be kept within 500 microseconds.

[0060] Furthermore, the self-diagnostic monitoring module 15 operates independently of the main processor. Once it detects an abnormal hardware status, such as channel saturation or memory verification errors, the self-diagnostic monitoring module 15 immediately forces the simulation task into a safe suspension mode via a hardware interrupt signal and drives the status indicator array 10 to display the corresponding error code. This prevents erroneous simulation conclusions caused by simulation interface failures from misleading engineering designs. Simultaneously, the front end of the physical layer signal conditioning module 1 integrates a self-calibration unit, automatically injecting a standard calibration pulse before each simulation task starts to eliminate sampling deviations caused by device aging or environmental temperature drift.

[0061] Furthermore, the non-volatile storage module 14 is a non-volatile high-speed flash memory used to record the key state data stream throughout the simulation process. Specifically, the data stream includes the complete thermal-electric coupling evolution process, used to record the evolution process throughout the simulation and support offline backtracking analysis after the simulation task is completed. By analyzing historical data, technicians can identify insufficient thermal redundancy or electromagnetic compatibility weaknesses in the system, providing quantitative basis for optimizing the physical structure of the UAV.

[0062] To cope with the high-intensity electromagnetic interference environment inside large transport drones, the electromagnetic compatibility protection shell 9 of the simulation interface is made of ADC12 aluminum alloy through one-piece die casting. The inner wall is additionally coated with a 0.5mm thick carbon-based absorbing material, which can effectively suppress electromagnetic radiation from 10 MHz to 3 GHz, ensuring the signal-to-noise ratio of the internal sensitive sampling circuit.

[0063] Furthermore, the simulation interface employs a customized real-time Ethernet protocol at the data link layer. Each data frame of this protocol is forcibly inserted with a 64-bit absolute timestamp, which is uniformly distributed by a global clock synchronization bus distributed across all subsystems. This hierarchical, tree-structured clock architecture ensures picosecond-level phase consistency across electrical and thermal simulation tasks distributed across different computing nodes. Additionally, the protocol incorporates a priority scheduling mechanism, assigning the highest priority to critical fault states such as junction temperature threshold data and coolant leaks related to thermal runaway warnings, ensuring that this critical information has absolute priority in network congestion situations.

[0064] See Figure 3 The physical layer signal conditioning module 1 includes a high-bandwidth current sampling unit 11, a transient voltage monitoring unit 12, and a signal preprocessing unit 13.

[0065] According to some embodiments, the high-bandwidth current sampling unit 11 is used to collect transient current on the output side of the inverter. The high-bandwidth current sampling unit 11 is specially designed as a through-core current transformer architecture with an electromagnetic shielding structure. The core magnetic core is made of nanocrystalline alloy material with high initial permeability and low coercivity. The response bandwidth is strictly set above 10 MHz, which can accurately capture the microsecond-level current ripple generated by large transport drones under the high-frequency pulse width modulation control of the three-phase inverter.

[0066] The transient voltage monitoring unit 12 is used to monitor the transient voltage of the power bus. It is connected to the power bus through a voltage divider circuit composed of a low inductive impedance network and monitors the voltage drop and overshoot of the DC side voltage in real time when the load changes suddenly, such as when the UAV suddenly switches from cruise mode to climb mode.

[0067] The signal preprocessing unit 13 is used to filter and perform analog-to-digital conversion on the acquired transient current and transient voltage to generate the input data of the power loss discrete sequence. The signal preprocessing unit 13 integrates a multi-order Butterworth low-pass filter, the cutoff frequency of which is dynamically configured according to the sampling rate, aiming to eliminate non-physical high-frequency oscillation noise caused by switching actions. The processed analog signal is then converted into a digital code stream that can be processed by subsequent logic through a high-speed analog-to-digital converter with 14-bit resolution.

[0068] The signal preprocessing unit 13 also has a data compression function. Through a compression algorithm based on slope threshold determination, it removes redundant stationary segments from the electromagnetic simulation waveform, retaining transient overshoot information with thermal effect characteristics. Specifically, the compression algorithm based on slope threshold determination can use slope calculation to compare the voltage change rate between adjacent sampling points. The slope calculation formula is: k=(V n -Vn-1 ) / dt Among them, V n The instantaneous voltage value at the current sampling moment (unit: volts, V), V n-1 Δt represents the instantaneous voltage value at the previous sampling moment, dt is the fixed sampling time interval (in seconds) of the physical layer signal conditioning module, and k is the instantaneous slope (in V / s). The absolute value of Δt directly characterizes the drastic change in the voltage waveform. The fixed sampling time interval is determined by the sampling rate of the high-speed analog-to-digital converter; for example, for a sampling rate of 1 GHz, Δt = 1 ns.

[0069] In this embodiment, the slope threshold k is set to 0.05, meaning that a dynamic fluctuation range of less than 5% is considered a "redundant stable segment." This allows the threshold to adapt to different voltage levels and sampling rates while avoiding the inapplicability of a fixed numerical threshold in different application scenarios. The 5% ratio is an empirical value determined based on extensive simulation and measurement comparisons. It effectively filters out minor fluctuations caused by quantization noise or simulator numerical oscillations, while ensuring the capture of all true power transient edges.

[0070] Furthermore, the sampling front end of the physical layer signal conditioning module 1 is integrated with a self-calibration unit, which automatically injects standard voltage / current pulses into the acquisition circuit before each simulation task begins, detects the transmission gain and phase deviation of the signal link, and generates compensation coefficients stored in the register of the interface system to eliminate acquisition errors caused by hardware aging or ambient temperature drift.

[0071] In a preferred embodiment of the present invention, the input terminal of the physical layer signal conditioning module 1 is configured with an adjustable attenuation / gain network. For power supply architectures of different voltage levels in large transport drones (such as a high-voltage DC 270V system and a low-voltage 28V system), users can remotely switch the input range via software commands to ensure that the sampling accuracy is always within the optimal quantization range, avoiding simulation distortion caused by signal saturation or excessive quantization noise.

[0072] See Figure 4 The cross-domain parameter mapping matrix 2 includes a loss calculation unit 21 and a heat source mapping unit 22. The loss calculation unit 21 includes a semiconductor device switching loss lookup table model and a conduction loss analytical model, which are used to calculate the energy loss based on transient electrical waveforms and real-time junction temperature data.

[0073] According to some embodiments, the loss calculation unit 21 does not perform real-time physical equation solving, but instead employs a fast retrieval mechanism based on a three-dimensional characteristic table. It comprehensively considers transient voltage, current, and real-time junction temperature fed back by the real-time closed-loop feedback controller 4 by pre-setting loss characteristic models of third-generation semiconductor power devices, including silicon carbide and gallium nitride. When the physical layer signal conditioning module 1 detects a switching action, the loss calculation unit 21 immediately retrieves the corresponding energy loss value from the pre-stored three-dimensional characteristic table.

[0074] The heat source mapping unit 22 distributes the calculated losses to corresponding spatial grid nodes according to the geometric topology of the power conversion module, forming a non-uniformly distributed heat source input. The heat source mapping unit also identifies the differences in thermal diffusivity between different material layers by modeling the thermal characteristics of the heterogeneous encapsulation material, and proportionally distributes the power losses to the three-dimensional coordinate nodes of each material layer. Furthermore, the heat source mapping unit 22 maps these discrete loss points to spatial grid nodes recognizable by the thermal management system simulation subsystem 200, based on the actual physical geometric topology of the power module within the large transport UAV.

[0075] Specifically, a three-dimensional Gaussian distribution weight function is adopted, with the geometric center of the semiconductor chip as the origin. The distribution variance is determined based on the difference in thermal diffusivity, and the single-point loss is transformed into a volumetric heat generation rate distribution, which is then mapped to the discrete element mesh of the thermal management system simulation subsystem (such as CFD software).

[0076] The volumetric heat distribution is defined by a three-dimensional Gaussian function, and the formula is: Q(x,y,z)=P*G(x,y,z) Where Q(x,y,z) is the three-dimensional volumetric heat generation rate distribution (unit: W / m³), (x,y,z) is the local spatial coordinates with the heat source center as the origin (unit: m), P is the transient power loss calculated by the loss calculation unit (unit: W), and G(x,y,z) is the normalized three-dimensional Gaussian distribution kernel function.

[0077] According to some embodiments, the standard deviation of the Gaussian distribution physically characterizes the range of heat diffusion. The standard deviation of the Gaussian distribution is not a fixed value but is determined by the heterogeneous thermal properties of the packaging material. The heat source mapping unit 22, based on pre-stored device packaging geometry and material properties (such as the thermal diffusivity of each layer including the chip attachment material, substrate, and base plate), determines or adjusts the standard deviation of the Gaussian distribution differently for different directions (x, y, z) or different material layers by simplifying the heat conduction model or empirical formulas. For example, in directions with high thermal diffusivity (such as through the copper substrate), a larger standard deviation of the Gaussian distribution is set to simulate faster heat diffusion; in directions with low thermal diffusivity (such as through the ceramic insulating layer), a smaller standard deviation of the Gaussian distribution is set. Based on this, the thermal diffusivity of different material layers is identified by modeling the thermal properties of the heterogeneous packaging material. Finally, the continuous volumetric heat rate distribution function Q(x,y,z) is mapped onto the discrete volume grid (cell) or node (node) of the thermal management system simulation subsystem (such as CFD software), thus constituting the non-uniformly distributed heat source input.

[0078] By introducing a physical mapping model based on a three-dimensional Gaussian distribution, the heat source mapping unit of this invention can more realistically simulate the diffusion process of heat generated from the micrometer-level chip junction region within the millimeter-level package, avoiding the numerical singularity and distortion caused by directly imposing a point heat source onto a single grid node. By correlating the standard deviation parameter of the Gaussian distribution with the material's thermal diffusivity, it effectively accommodates heterogeneous materials and supports modeling of complex packaging structures.

[0079] According to some embodiments, the matrix is ​​internally subdivided into a loss calculation unit 21 and a heat source mapping unit 22. The loss calculation unit 21 does not perform real-time physical equation solving, but instead employs a fast retrieval mechanism based on a three-dimensional characteristic table. This unit pre-configures loss characteristic models for third-generation semiconductor power devices, including silicon carbide and gallium nitride, which comprehensively consider transient voltage, current, and real-time junction temperature fed back by the real-time closed-loop feedback controller 4. When the physical layer signal conditioning module 1 detects a switching action, the loss calculation unit 21 immediately retrieves the corresponding energy loss value from the pre-stored three-dimensional characteristic table. The heat source mapping unit 22 maps these discrete loss points to spatial grid nodes recognizable by the thermal management system simulation subsystem 200, based on the actual physical geometry of the power modules within the large transport UAV. In specific engineering implementations, the mapping matrix employs heterogeneous packaging material modeling technology, capable of identifying the differences in thermal diffusivity between different materials such as ceramic substrates, direct copper plating layers, and silicon gels, ensuring that the description of heat source distribution is accurate to the micrometer-level physical structure.

[0080] In a preferred embodiment of the present invention, the cross-domain parameter mapping matrix of the simulation interface integrates nonlinear regression correction logic. This logic, through learning from a large amount of experimental data, can empirically compensate for complex thermophysical phenomena (such as the cold-start characteristics of power devices in extremely low-temperature environments) that cannot be fully covered by theoretical physical models, thereby further converging the deviation between simulation results and measured data.

[0081] In a preferred embodiment of the present invention, the cross-domain parameter mapping matrix of the simulation interface employs a reduction-order processing method based on a three-dimensional thermal sensitivity coefficient matrix when performing heat source allocation. This reduction-order processing method simplifies the complex fluid dynamics mesh to the mutual coupling relationships of key nodes, reducing the single-run solution time of thermal simulation to the millisecond level while preserving the main heat conduction paths and convection characteristics, thereby meeting the real-time requirements of UAV mission-level simulation.

[0082] The simulation interface has multi-channel expansion capability, and multiple interface modules can be cascaded through the backplane bus, thereby realizing centralized, integrated thermo-electric coupling simulation of multiple high-power heat sources such as the multi-engine propulsion system, energy storage system and airborne radar of large transport UAVs.

[0083] Furthermore, the interface system follows a modular and open software architecture, providing standardized application programming interfaces. Users can define new physical field coupling logic (such as introducing electromagnetic compatibility and thermal field coupling) or add third-party simulation solver drivers according to the development needs of specific models, thus realizing the continuous evolution and functional expansion of simulation capabilities.

[0084] The simulation interface supports a device identification protocol based on physical addresses. When multiple modules operate in parallel, the system can automatically assign logical numbers and synchronize the internal counters of all modules, ensuring that the distributed heat source data collected by different interface modules have a completely consistent time reference. This is of great significance for analyzing the unbalanced heat distribution of the propulsion system on both sides of the wing of a large transport UAV.

[0085] Furthermore, the communication link layer of the simulation interface implements a data error correction mechanism based on redundancy check. In scenarios with strong radiation interference, such as airborne simulation environments, if a bit flip error is detected in the transmitted data frame, the interface can automatically initiate a retransmission request or use the built-in error correction algorithm to recover the original data, ensuring the continuity of the UAV full envelope simulation mission for tens of hours and the absolute reliability of the data.

[0086] In summary, the thermal-electric coupling simulation interface for large transport UAVs presented in this example embodiment constructs a highly integrated multiphysics simulation platform through deep integration of hardware links, cross-domain physical mapping, time-scale coordination, and closed-loop feedback mechanisms. The design fully considers the technical characteristics of large transport UAVs under high power density, complex flight dynamics, and stringent reliability requirements, solving the predictive failure problem caused by electrothermal decoupling simulation. This lays a solid technical foundation for the forward development of a new generation of intelligent unmanned transport platforms. It not only improves the transparency of UAV power systems during the design phase but also provides strong technical support for subsequent system integration, ground verification, and airworthiness certification. Its application prospects are broad, with significant social and economic benefits.

[0087] According to another embodiment, such as Figure 5 , Figure 6 This embodiment proposes a thermal-electric coupling simulation method for large transport drones.

[0088] In S101, load the electrothermal characteristic parameters and establish a synchronous data link between the simulation interface and the power electronic system simulation subsystem.

[0089] According to some embodiments, during the simulation initialization phase, the interface system loads the structured geometric model of the power system of the large transport UAV and the electro-thermal characteristic parameter files of the semiconductor devices. The environmental envelope simulation unit initializes the thermophysical parameters of the external atmosphere according to the preset flight mission trajectory. Subsequently, the physical layer signal conditioning module enters the real-time acquisition state and establishes a synchronous link with the power electronic system simulation subsystem.

[0090] In S103, the electrical simulation waveform output by the power electronic system simulation subsystem is captured in real time by the physical layer signal conditioning module, converted into a discrete power loss sequence, and mapped to a spatial heat source distribution term through a cross-domain parameter mapping matrix.

[0091] According to some embodiments, during the simulation operation phase, the power electronic system simulation subsystem solves the electrical network equations in nanosecond-level steps, generating high-frequency changing switching waveform data. The physical layer signal conditioning module captures the electrical simulation waveforms output by the power electronic system simulation subsystem in real time and passes them to the cross-domain parameter mapping matrix for loss decomposition. The cross-domain parameter mapping matrix converts the calculated transient power loss into the corresponding heat flow injection item.

[0092] In S105, a multi-timescale synchronization coordinator is used to perform time-scale scaling and averaging on the electrical simulation data of the first timescale and the thermal simulation data of the second timescale, and the processed heat source data is transmitted to the thermal management system simulation subsystem for temperature distribution solution.

[0093] According to some embodiments, when the solution time of the thermal management system simulation subsystem arrives, the multi-timescale synchronization coordinator pushes the accumulated heat flow data to the thermal simulation kernel. The multi-timescale synchronization coordinator is used to perform time-scale scaling and averaging processing on the electrical simulation data of the first time scale and the thermal simulation data of the second time scale. Based on finite element analysis or equivalent thermal network model, the thermal management system simulation subsystem solves for the temperature distribution map of the power module, battery pack and heat sink surface at that moment, and transmits the processed heat source data to the thermal management system simulation subsystem for temperature distribution solution. The thermal simulation results are fed back to the real-time closed-loop feedback controller through the second buffer.

[0094] In S107, the temperature distribution data from the thermal simulation feedback is received through a real-time closed-loop feedback controller, which drives the parameter correction engine to dynamically adjust the physical model parameters of the components in the power electronic system simulation subsystem, forming a deep closed-loop coupling between thermal energy flow and electrical energy flow.

[0095] According to some embodiments, the real-time closed-loop feedback controller utilizes received spatial temperature information to update the component physical models in the power electronic system simulation subsystem in real time through a parameter correction engine. The correction engine dynamically adjusts the controlled source parameters and impedance parameters in the simulation circuit based on the influence of temperature on carrier concentration, thereby achieving adaptive evolution of electrical performance with temperature changes. This cyclical process continues throughout the entire simulation task, achieving deep coupling between thermal and electrical energy flow.

[0096] In summary, this invention achieves significant breakthroughs in key technical dimensions such as accurate mapping of physical parameters, smooth coordination over time scales, and real-time closed-loop control feedback through system-level design innovation. It provides a scientific, rigorous, and highly engineering-valued thermo-electric coupling simulation technology for the development of large transport drones, a high-power-density system.

[0097] The present invention provides a thermal-electric coupling simulation interface and method for large transport drones, and its excellent engineering performance can be further confirmed by the following specific embodiments and comparative data comparison.

[0098] In this comparative test, a large transport drone main propulsion inverter with a rated power of 50 kilowatts was selected as the test object. This inverter uses silicon carbide (SiC) MOSFET modules, and the switching frequency is set to 20 kilohertz. The simulation environment simulates the typical operating condition of a drone performing maximum power climb at an altitude of 8,000 meters, where the ambient air pressure is approximately one-third of standard atmospheric pressure, resulting in extremely harsh heat dissipation conditions.

[0099] Example 1: Simulation using the thermal-electric coupling simulation interface described in this invention. In this embodiment, the power electronic system simulation subsystem 100 operates with a fixed step size of fifty nanoseconds, while the thermal management system simulation subsystem 200 operates with a step size of ten milliseconds. The simulation interface described in this invention establishes a real-time bidirectional closed loop between the two. The physical layer signal conditioning module 1 captures the voltage spikes generated by the switching transients, and the loss calculation unit 21 automatically increases the MOSFET on-state voltage drop simulation parameters based on the real-time feedback of the junction temperature (which has now reached 135 degrees Celsius) due to poor heat dissipation.

[0100] Comparative Example 1: Using the traditional thermal and electrical decoupling simulation method (static parameter configuration) In this comparative example, the parameters of the electrical simulation were set to standard parameters at room temperature (25 degrees Celsius) before the task began, while the thermal simulation was performed in an open-loop manner based only on the estimated rated losses. Although the thermal simulation also predicted the temperature rise, this temperature rise data was not transmitted back to the electrical simulation subsystem in real time, causing the resistance and loss parameters to remain in an ideal state throughout the electrical simulation process.

[0101] The key parameters for the two schemes after 30 minutes of simulation are shown in Table 1: Table 1 Comparison of Key Parameters

[0102] The quantitative data above shows that the present invention successfully captured the nonlinear behavior of power semiconductors at high temperatures by dynamically correcting the physical properties of components through a real-time closed-loop feedback controller.

[0103] In Comparative Example 1, due to the lack of thermo-electric feedback, the simulation system ignored the increase in on-resistance caused by temperature rise (positive temperature coefficient), resulting in an optimistic but incorrect efficiency result (98.1%) and causing the predicted junction temperature to be more than 32 degrees Celsius lower than the actual value. In practical engineering design, this could lead to unexpected overheating protection or even hardware burnout in the UAV's propulsion system during the climb phase.

[0104] In Embodiment 1 of this invention, the simulation interface not only accurately predicted the sharp rise in junction temperature, but also precisely simulated the enhanced turn-off voltage spike caused by the temperature increase (DC-side voltage ripple increased from 12.3V to 18.5V) by adjusting the gate drive impedance in real time. This high-fidelity simulation capability provides valuable quantitative support for the margin design of the cooling system and the thermal robustness of the motor control algorithm of large transport UAVs.

[0105] Furthermore, the simulation interface in this embodiment demonstrates exceptional fault injection simulation capabilities. In a simulation experiment targeting cooling loop flow attenuation, the flow field monitoring unit artificially set the coolant pump efficiency to decrease by 30%. The simulation interface immediately fed back the deterioration of heat exchange capacity to the simulation closed loop via a multi-timescale synchronization coordinator. The real-time closed-loop feedback controller rapidly detected the heat accumulation inside the power module and drove the power electronics system simulation subsystem to simulate the abnormal increase in carrier concentration due to overheating, ultimately successfully reproducing the complete physical evolution chain of the system from localized hot spots to thermal breakdown under extreme fault conditions. This functionality is crucial for verifying the derating strategy of the UAV power system.

[0106] Furthermore, in multi-module cascaded application scenarios, the simulation interface of this embodiment supports an automatic device identification protocol based on physical addresses. For ultra-large transport UAVs with four or six engines, multiple interface modules can be cascaded via a backplane bus to achieve integrated control of distributed heat sources. A unified reference frequency signal ensures seamless alignment of multiple independent propulsion branches on the simulation time axis. Each physical layer signal conditioning module can independently switch its front-end attenuation / gain network for different voltage levels (such as 270V high-voltage systems and 28V low-voltage systems), ensuring that sampling accuracy remains within the optimal quantization range in complex hybrid voltage architectures and avoiding the impact of quantization noise on the coupled simulation results.

[0107] Regarding the flexible design of the software architecture, the simulation interface in this embodiment provides a standardized application programming interface (API), allowing engineers to switch loss models based on specific material technology routes. For example, when verifying a UAV platform using domestically produced SiC power devices, engineers only need to import the temperature-loss characteristic curve provided by the corresponding manufacturer into the loss calculation unit, and the interface can automatically update the cross-domain mapping logic. This modular and scalable design approach ensures that the interface can adapt to future UAV avionics and energy systems with higher power densities and more complex integration.

[0108] In summary, the thermal-electric coupling simulation interface and method for large transport UAVs described in this invention completely breaks the deadlock of isolated electrical and thermal physical fields in traditional simulations through high-bandwidth capture at the physical layer, cross-domain mapping accuracy optimization, multi-scale time synchronization, and closed-loop feedback deeply embedded in the physical layer. It not only achieves aerospace-grade standards in the real-time and synchronization of data transmission but also represents a significant breakthrough in the depth of physical mechanism reconstruction. The widespread application of this simulation interface will provide a complete verification foundation for the forward development of large transport UAVs, from underlying component characteristics to system-level thermal balance, and has profound engineering significance for improving the operational safety and system reliability of UAVs in extreme flight environments.

[0109] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A thermal-electric coupling simulation interface for large transport drones, positioned between a power electronics system simulation subsystem and a thermal management system simulation subsystem, characterized in that... The simulation interface includes a cross-domain coupling and synchronization execution module, which comprises a physical layer signal conditioning module, a cross-domain parameter mapping matrix, a multi-timescale synchronization coordinator, and a real-time closed-loop feedback controller, wherein: The physical layer signal conditioning module is used to capture the transient waveforms of the power electronic system simulation subsystem and convert them into discrete power loss sequences. The cross-domain parameter mapping matrix is ​​used to map the power loss discrete sequence into a spatial heat source distribution item that can be identified by the thermal management system simulation subsystem. The multi-timescale synchronization coordinator includes a dual-buffered storage architecture for synchronizing data between the first simulation step of the power electronic system simulation subsystem and the second simulation step of the thermal management system simulation subsystem. The real-time closed-loop feedback controller is used to correct the physical model parameters of semiconductor components in the power electronic system simulation subsystem in real time based on the temperature data fed back by the thermal management system simulation subsystem, so as to construct a bidirectional closed-loop simulation circuit.

2. The simulation interface according to claim 1, characterized in that, The multi-timescale synchronization coordinator includes a dual-buffered storage architecture: The dual-buffered storage architecture includes a first buffer and a second buffer; The first buffer is used to accumulate and write the power loss discrete sequence at the first simulation step size, and to output the average thermal power at the trigger time of the second simulation step size, wherein the first simulation step size is shorter than the second simulation step size. The second buffer is used to receive low-frequency temperature data fed back by the thermal management system simulation subsystem, and to generate a pseudo-continuous temperature signal that matches the first simulation step size through an interpolation algorithm.

3. The simulation interface according to claim 1, characterized in that, The real-time closed-loop feedback controller includes a parameter correction engine. The parameter correction engine dynamically adjusts the simulation parameters of the carrier mobility, on-resistance, and turn-off delay time of the semiconductor elements in the power electronic system simulation subsystem according to the received temperature feedback signal, so as to simulate the thermally induced performance degradation process under real physical environment.

4. The simulation interface according to claim 1, characterized in that, The physical layer signal conditioning module includes a high-bandwidth current sampling unit, a transient voltage monitoring unit, and a signal preprocessing unit. The high-bandwidth current sampling unit is used to collect the transient current on the output side of the inverter; The transient voltage monitoring unit is used to monitor the transient voltage of the power bus; The signal preprocessing unit is used to filter and perform analog-to-digital conversion on the acquired transient current and transient voltage to generate the input data of the power loss discrete sequence. The signal preprocessing unit also has a data compression function, which removes redundant stationary segments from the electromagnetic simulation waveform using a compression algorithm based on slope threshold determination.

5. The simulation interface according to claim 1, characterized in that, The cross-domain parameter mapping matrix includes a loss calculation unit and a heat source mapping unit: The loss calculation unit includes a lookup table model for switching losses and an analytical model for conduction losses of semiconductor devices, which are used to calculate energy loss based on transient electrical waveforms and real-time junction temperature data. The heat source mapping unit distributes the calculated losses to the corresponding spatial grid nodes according to the geometric topology of the power conversion module, forming a non-uniformly distributed heat source input. The heat source mapping unit also identifies the differences in thermal diffusivity of different material layers by modeling the thermal characteristics of the heterogeneous packaging material, and proportionally distributes the power loss to the three-dimensional coordinate nodes of each material layer.

6. The simulation interface according to claim 1, characterized in that, The cross-domain coupling and synchronous execution module also includes: An environmental envelope simulation unit is connected to the interface of the UAV flight control system to obtain ambient temperature, air pressure and convective heat transfer coefficient, and inject them as dynamic boundary conditions into the thermal management system simulation subsystem. The flow field state monitoring unit is used to acquire the flow rate and velocity of the cooling medium and the operating power consumption of the pump, and convert them into real-time correction instructions for the convective heat transfer boundary conditions in the thermal management system simulation subsystem.

7. The simulation interface according to claim 1, characterized in that, The multi-timescale synchronization coordinator also includes a dynamic step size management strategy: When the power electronic system simulation subsystem is in steady-state operation, the data aggregation time window is increased to reduce the synchronization frequency; When a sudden load change or entry into a violent maneuver envelope is detected, the multi-timescale synchronization coordinator immediately switches to high-frequency synchronization mode.

8. The simulation interface according to claim 1, characterized in that, The simulation interface also includes a system control and status assurance module, which comprises a dual-core processor architecture, a self-diagnostic monitoring module, and a non-volatile memory module, wherein: The dual-core processor architecture includes a first processor core and a second processor core. The first processor core is responsible for running the real-time operating system and communication protocol stack, while the second processor core serves as a dedicated mathematical operation accelerator, responsible for performing cross-domain parameter mapping operations. The self-diagnostic monitoring module is used to continuously monitor the power supply voltage, operating temperature and logic clock frequency inside the interface, and to force the simulation system into a safe suspension state when there is a hardware abnormality, and drive the status indicator array to display the corresponding error code. The non-volatile storage module is used to record the data flow during the simulation process to support offline backtracking analysis after the simulation task is completed.

9. The simulation interface according to claim 8, characterized in that, The system control and status assurance module and the cross-domain coupling and synchronization execution module are encapsulated in an electromagnetic compatibility (EMC) protective enclosure, and the inner wall of the EMC protective enclosure is coated with a wave-absorbing material with a thickness of not less than 0.5 mm.

10. A thermal-electric coupling simulation method for large transport drones, characterized in that, include: Load electrothermal characteristic parameters and establish a synchronous data link between the simulation interface and the power electronics system simulation subsystem; The electrical simulation waveform output by the power electronic system simulation subsystem is captured in real time by the physical layer signal conditioning module, converted into a discrete power loss sequence, and mapped to a spatial heat source distribution term through a cross-domain parameter mapping matrix. A multi-timescale synchronization coordinator is used to perform time-scale scaling and averaging on the electrical simulation data of the first time scale and the thermal simulation data of the second time scale, and the processed heat source data is transmitted to the thermal management system simulation subsystem for temperature distribution solution. The temperature distribution data from the thermal simulation is received by the real-time closed-loop feedback controller, which drives the parameter correction engine to dynamically adjust the physical model parameters of the components in the power electronic system simulation subsystem, forming a deep closed-loop coupling between thermal energy flow and electrical energy flow.

Images