Loss determination method, apparatus, and medium for electric aviation power motor controller

Abstract

The present application relates to the technical field of computer technology application, particularly related to a kind of electric aviation power motor controller loss determination method, equipment and medium, method is executed by computing device, comprising: generating time-varying operating condition sequence based on expected flight profile;Switching loss model is established by time slicing algorithm;Establish on-state loss model and diode switching loss model;Integrate all model calculations total loss power.The present application realizes the high-precision prediction of electric aviation power motor controller loss by multi-time scale analysis and parameter fitting, provides important basis for system optimization design.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer technology application, and particularly relates to a loss determination method, device and medium of an electric aviation power motor controller. BACKGROUND

[0002] With the rapid development of electric aircraft, extreme harsh requirements are put forward for the performance, efficiency and reliability of the core component power motor controller. The main power topology circuit in the controller usually adopts silicon carbide (SiC) MOSFET as a power switching device to realize high power density and high efficiency energy conversion. However, in the aviation application scenario, the controller faces challenges such as high switching frequency, long time large load operation, severe flight profile change and severe heat dissipation environment, which makes the accurate calculation of the switching loss and on-state loss of the power device, especially the SiC MOSFET, the core and difficulty of the controller topology design, device selection, thermal management and system reliability evaluation.

[0003] The mainstream method of power device loss modeling mainly includes physical model, behavior model and analytical model. The physical model is based on semiconductor physical principles, although it has high accuracy, but the model is complex, the parameter extraction is difficult, the calculation amount is large, and it is difficult to apply to system-level simulation and real-time control. The behavior model relies on external characteristics (such as current source, capacitor network) to fit the behavior of the device, although the simulation speed is fast, but its accuracy depends heavily on the completeness of the test data, and the physical meaning is not clear, and the generalization ability is poor. The analytical model seeks a balance between complexity and accuracy, and describes the voltage and current changes in the switching process through mathematical formulas, thereby calculating the energy loss, which is the ideal method in current engineering applications.

[0004] However, the existing analytical loss model still has significant deficiencies when applied to this special field of electric aviation:

[0005] Multi-scale dynamic response coupling problem: The flight profile of the aircraft is composed of a series of rapidly changing operation control points, and the time-varying sequence of operating parameters such as on-state current, junction temperature and switching frequency generated thereby has different response time scales. Most of the existing models are modeled for steady state or a single time scale, and fail to effectively integrate this multi-scale dynamic characteristic, resulting in large deviation in loss prediction under transient operating conditions.

[0006] Lack of interface with top-level system requirements: Most of the existing models are independent device-level models, and fail to establish a clear modeling path from the top-level task requirements of the aircraft (such as a specific flight profile) to the controller operating conditions (such as current and frequency instructions), and then to the device loss. The model input and system-level external characteristic interface are not clear, and it is difficult to accurately evaluate and optimize the system-level performance (such as flight range and thermal management requirements) at the design stage.

[0007] Multi-factor coupling modeling precision is insufficient: the existing analytical model is not competent when dealing with the strong coupling effect of multiple factors (such as device parameters dependent on junction temperature, nonlinear parasitic parameters, driving conditions, and load characteristics) under high frequency and high temperature conditions. In particular, switch loss dynamic modeling requires the decomposition of the detailed physical process of switch transient and the quantification of the influence of each parasitic parameter (such as driving resistance, parasitic inductance, and nonlinear capacitance). The existing method often simplifies the consideration, which is difficult to solve, resulting in difficulty in ensuring the precision of the model in a wide operating range.

[0008] Therefore, there is an urgent need for a high-precision loss modeling method that closely integrates the characteristics of electric aviation applications, can handle multi-scale dynamic operating conditions, establish a clear mapping from top-level requirements to bottom-level losses, and carefully consider the coupling of multiple physical factors, to support the design and optimization of high-reliability and high-efficiency electric aviation power motor controllers. SUMMARY

[0009] To solve the above technical problems, the technical scheme adopted by the present application is as follows:

[0010] According to the first aspect of the present application, a loss determination method for an electric aviation power motor controller is provided, which is executed by a computing device and includes the following steps:

[0011] S100, based on the expected flight profile of an electric aircraft, a time-varying operating condition sequence is extracted through multi-time scale time series analysis, and the sequence is stored in a memory, the time-varying operating condition sequence includes time series data of output phase current, switching frequency, modulation frequency and device junction temperature.

[0012] S200, the time-varying operating condition sequence is read by the processor, the energy loss of SiC power device in the switching transient process is solved based on numerical calculation method, and the switching loss calculation model is established; wherein the switching transient process is divided into multiple sub-stages by time slicing algorithm, the calculation of the switching loss calculation model depends on the time-varying operating condition sequence and the key device parameters of the power device, and the key device parameters are expressed as a function of the junction temperature in the time-varying operating condition sequence by the parameter fitting module.

[0013] S300, the processor executes the on-state loss calculation module, calculates the on-state loss of SiC power device and the on-state loss of freewheeling diode based on the time-varying operating condition sequence, and establishes the on-state loss calculation model of SiC power device and the on-state loss calculation model of freewheeling diode.

[0014] S400, the processor executes the freewheeling diode switching loss calculation module, calculates the switching loss of the freewheeling diode based on the time-varying operating condition sequence, and establishes the freewheeling diode switching loss calculation model.

[0015] S500, integrating each loss calculation model established by the model integration module, inputting the time-varying operation condition sequence, calculating and outputting the total loss power of the power motor controller in the entire expected flight profile.

[0016] According to the second aspect of the present application, an electronic device is provided, comprising a processor and a memory; the processor is configured to execute the steps of the method according to the first aspect of the present application by invoking programs or instructions stored in the memory.

[0017] According to the third aspect of the present application, a computer readable storage medium is provided, which stores programs or instructions for enabling a computer to execute the steps of the method according to the first aspect of the present application.

[0018] The loss determination method of the electric aviation power motor controller provided by the embodiments of the present application comprises: generating a time-varying operation condition sequence based on an expected flight profile; establishing a switching loss model by a time slicing algorithm; establishing an on-state loss model and a diode switching loss model; and integrating all the models to calculate the total loss power. Through multi-time scale analysis and parameter fitting, the present application realizes high-precision prediction of the loss of the electric aviation power motor controller, and provides an important basis for system optimization design.

[0019] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent through the following description. BRIEF DESCRIPTION OF DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative labor.

[0021] Figure 1 The flowchart of the loss determination method of the electric aviation power motor controller provided by the embodiments of the present application is shown in the figure;

[0022] Figure 2 The loss model schematic diagram is shown in the figure;

[0023] Figure 3 The half-bridge test circuit is shown in the figure;

[0024] Figure 4 The switch-dominated physical mechanism is shown in the figure;

[0025] Figure 5 The start transient process is shown in the figure;

[0026] Figure 6 To shut down the transient process. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. The terminology used in the description herein is for describing the specific embodiments only and is not intended to be limiting of the present application. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items.

[0029] It should be noted that some of the example embodiments are described as processes that are depicted as flow diagrams or methods. Although each can describe the operations as a sequential process, many of the operations can be performed in parallel, concurrently or simultaneously, and / or in a different order than that described. In addition, the processes can be terminated when their operations are completed, but can also have additional steps not included in the figure, which can also be performed after operations of the process are completed. The processes can correspond in whole, or in part, to a method, function, routine, subroutine, or the like.

[0030] The present application aims to provide a loss determination method of an electric aviation power motor controller, taking a single tube loss model of the controller as the core, through considering the aircraft operating conditions of the power flow path, multi-scale parameter extraction is performed to provide the required single tube level time-varying operating condition sequence data set (including time-varying parameters such as single tube on-state current, switching frequency, modulation frequency, junction temperature, etc.) for the single tube loss model, and finally realize output of single tube switching loss power and single tube on-state loss power corresponding to each short-time working point in the flight profile of the electric aircraft, and provide fine data support for thermal design, efficiency optimization and reliability evaluation of single power device of the controller.

[0031] Further, the embodiments of the present application provide a loss determination method of an electric aviation power motor controller, which is executed by a computing device, and the core task of the computing device is to adapt the computing requirement of the single tube loss model, complete parameter processing, model operation and result output related to the single tube loss.

[0032] In the embodiment of the present application, the computing device for performing the above loss determination method refers to a special or general electronic computing device with data processing, model operation, parameter interaction and result output capabilities of single tube loss model adaptation, and meets the high reliability and real-time requirements of electric aviation scenarios. The core function is to complete the automatic execution of the whole process of power motor controller single tube loss calculation (including single tube time-varying parameter analysis, single tube on-state and off-state loss calculation, loss result output, etc.) through the cooperation of hardware architecture (such as a special unit supporting transient integral operation) and software system (such as a single tube loss calculation module).

[0033] In the embodiment of the present application, the single tube specifically refers to a single power semiconductor device constituting a power conversion circuit of a power motor controller, which is a basic switching unit for realizing electric energy conversion. SiC MOSFET (Silicon Carbide Metal-Oxide-Semiconductor Field Effect Transistor) is mainly used, and the SiC Schottky barrier diode (freewheeling diode) in anti-parallel connection is regarded as an inherent component of the single tube. Because the two are integrated in physical structure or inseparable in function, they jointly participate in the switching process.

[0034] In the power topology of the controller, such as half-bridge and full-bridge, a single single tube exists as an independent switching unit, such as an upper tube or a lower tube in a half-bridge topology, which realizes the chopping and current conversion of electric energy through the switching of on / off states. The loss analysis of the single tube focuses on the energy dissipation of the device itself, including the on-state loss and switching loss of the SiC MOSFET, as well as the on-state loss and switching loss of the anti-parallel diode. It is the smallest unit of total loss calculation of the controller, and the sum of the losses of multiple single tubes constitutes the overall loss of the controller.

[0035] As shown in Figure 1 The loss determination method of the electric aviation power motor controller provided by the embodiment of the present application can include the following steps:

[0036] S100, based on the expected flight profile of the electric aircraft, a time-varying operating condition sequence is extracted through multi-time scale time sequence analysis, and the sequence is stored in a memory, wherein the time-varying operating condition sequence includes time sequence data of output phase current, switching frequency, modulation frequency and device junction temperature.

[0037] Specifically, S100 can include the following steps:

[0038] S101, flight profile analysis and task segmentation

[0039] Firstly, the expected flight profile of the electric aircraft is obtained. The flight profile defines the relationship between the speed, height, acceleration and time of the aircraft in a complete task (such as take-off, climb, cruise, descent, landing and ground taxiing). Then, the flight profile is divided according to the task stage, and each stage is assigned a unique operation control point identifier.

[0040] S102, mapping of multi-time scale operating conditions

[0041] For each task phase and its operation control point divided in S101, according to the performance characteristics and control strategy of the power motor and controller, the operating condition mapping is carried out, including:

[0042] (1) Output phase current sequence generation: according to the required propulsion power and motor speed of flight profile, through the motor map or torque-speed-current relationship, the time-varying instruction value sequence of motor controller output side phase current I m (t) is calculated. m (t) represents the target phase current value that the controller should output at time point t. For example, at t=10s (take-off acceleration phase), the current instruction may be large; at t=1800s (steady cruising phase), the current instruction will be significantly reduced.

[0043] (2) Switching frequency sequence generation: based on the pre-set switching loss-frequency optimization curve or thermal management strategy, the switching frequency instruction value sequence f sw m (t) associated with the current flight phase, output power and estimated thermal state is generated. sw m (t) represents the target switching frequency value given by the system decision at time point t. For example, in the low-speed high-torque phase, in order to reduce the loss, a lower switching frequency instruction may be selected; when high dynamic response is needed, a higher switching frequency instruction may be switched to.

[0044] (3) Modulation frequency / modulation ratio sequence generation: according to the motor speed control demand and the selected PWM modulation strategy (such as SPWM, SVPWM), the modulation wave frequency or modulation ratio instruction value sequence m a (t) or f mod (t) is calculated.

[0045] (4) Device junction temperature initial sequence generation: taking the ambient temperature or initial thermal state as the starting point, combining the power change course of the flight profile, through the pre-constructed electro-thermal coupling model, the time-varying sequence of power device junction temperature T jpre (t) is preliminarily estimated.

[0046] S103, time sequence data synthesis and storage:

[0047] The generated each operating condition instruction value or estimated sequence in S102 is aligned and synthesized according to a unified time axis to form the multi-time scale time-varying operating condition sequence. The sequence completely describes the dynamic change process of all key external conditions for the operation of the power motor controller in the entire expected flight profile. Finally, the sequence is written in the form of an array or a table in the memory of the computing device for subsequent modeling steps.

[0048] Optionally, the data points in the time-varying operating condition sequence are discretely stored at fixed sampling intervals, and the sampling intervals are dynamically adjusted according to the fastest changing stage in the flight profile to ensure that the sequence can accurately capture the transient characteristics of the operating conditions.

[0049] In S200, the time-varying operating condition sequence is read by a processor, the energy loss of the SiC power device in the switching transient process is solved based on a numerical calculation method, and a switching loss calculation model is established.

[0050] The switching transient process is divided into multiple sub-stages by a time slicing algorithm, and the calculation of the switching loss calculation model depends on the time-varying operating condition sequence and key device parameters of the power device, and the key device parameters are expressed as a function of the junction temperature in the time-varying operating condition sequence by a parameter fitting module.

[0051] In the embodiment of the application, to consider the multi-scale loss modeling accuracy, the change of the device junction temperature sequence is approximately modeled as a piecewise first-order dynamic differential equation, specifically: dT Jt / dt=(T Jobj -T Jt ) / τ Jt . Wherein, T Jobj is the steady-state value of the junction temperature expected to be reached after the operation of the control point, and the unit is °C or K; τ Jt is the thermal time constant corresponding to the rise or decay of the junction temperature curve, and the unit is second. T Jt is the real-time junction temperature of the power device at time t.

[0052] In the embodiment of the application, the loss model qualitatively considers the influence of parasitic parameters on the switching loss, and needs to introduce the driving resistance, parasitic inductance and parasitic capacitance. With the increase of the gate driving resistance R g , the switching speed decreases, and the device switching loss increases; with the increase of the source parasitic inductance L s , the di d / dt of the switching process decreases, and the switching loss increases. Especially, the turn-on loss has a greater impact; with the increase of the drain parasitic inductance L d , the di d / dt of the switching process decreases, and the di d / dt remains essentially unchanged. Turn-on losses decrease, while turn-off losses increase. d This is the transient drain current.

[0053] With the gate-source nonlinear parasitic capacitance C gs The increase of dv during the switching process gs / dt、dv ds / dt、d id As / dt decreases, turn-on losses increase and turn-off losses decrease. With the decrease of drain-source nonlinear parasitic capacitance C... ds The increase of dv during the switching process gs / dt、dv ds / dt、d id The switching losses remain essentially unchanged, and the parasitic parameters mentioned above are all inherent parameters of the initial circuit, which are already quantitatively expressed in the model.

[0054] Based on fixed DC bus voltage and drive source voltage, it is necessary to focus on the impact mechanism of four operating conditions on power device losses: output current, device junction temperature, switching frequency, and load variable frequency speed regulation frequency. The on-state resistance R of SiC power devices... ds(ON) Internal gate resistance R g(int) Transfer characteristics g fs Threshold voltage V th The forward on-state resistance RD of a Schottky barrier diode (freewheeling diode) (ON) Threshold voltage VD (th) All parameters are related to the junction temperature. The fitting interval is [25℃, 150℃], and R0... ds(ON) -T J It can be fitted using a quadratic polynomial function, specifically: .

[0055] Among them, TC1 and TC2 are both R ds(ON) -T J The quadratic fitting coefficient. T 25 =25℃.

[0056] R g(int) -T J The same principle applies to R in principle. ds(ON) -T J The quadratic function fitting often ignores the influence of temperature control, and it is modeled as a constant value. fs -T J V th -T J The curve can be fitted by a linear or higher-order function, specifically: ;

[0057] .

[0058] T 150 =150℃.

[0059] R D(ON) -T J , VD (th) -T J The curve can be fitted with a linear function as follows:

[0060] ;

[0061] .

[0062] Wherein, TC3, TC4 are linear fitting coefficients of R D(ON) -T J , VD (th) -T J .

[0063] In the embodiment of the application, a single tube loss model of the controller is constructed taking a half-bridge circuit as an example. Figure 2 As shown in the following figure:

[0064] The left side of the figure is a multi-scale loss modeling abstraction of a three-phase main power topology, wherein, corresponding to levels G0 and G1: the top layer demand is associated with mechanical-electrical system simulation, power device model and other links, four groups of time-varying sequences, i.e., output side phase current, switching frequency, device junction temperature and speed regulation frequency, are included in the discussion category of operating conditions (G1 is an abstract expression of operating conditions), so as to clearly anchor the path logic of multi-scale loss modeling.

[0065] The right side of the figure is a state and mode expression of a half-bridge circuit, wherein, corresponding to levels G3, G4 and G5: G3 shows three core working modes of the half-bridge circuit: upper tube conduction + lower tube off mode, upper tube off + lower tube conduction mode and upper and lower tube off mode; the mode switching process is accompanied by discrete switching events, and the multi-transient stage is divided by time slicing technology; G4 and G5 further expand the state sequence evolution under different transient states, and provide support for fine analysis of the loss characteristics of the switching process.

[0066] Figure 3 For a half-bridge test circuit, Figure 3 wherein, V DC is a DC bus voltage, R p is a parasitic resistance in the DC bus, L p is a parasitic inductance in the DC bus, L d(ext) is an external drain parasitic inductance, R g(ext) is an external gate parasitic resistance, L g(ext) is an external gate parasitic inductance, S1 is a switching tube 1, V drive is a gate drive voltage, Ls(ext) S2 is the external source parasitic inductance, and S2 is the switching transistor 2. L R is the load current. L L is the load resistance. L For the load-side inductance, C L C is the parasitic capacitance of the load inductor. jD For the nonlinear parasitic capacitance of the freewheeling diode, D f L is the ideal symbol for a freewheeling diode in forward orientation. d(int) For the internal drain parasitic inductance, C gd L is the nonlinear parasitic capacitance on the gate-drain side. g(int) R is the internal gate parasitic inductance. g(int) C is the internal gate parasitic resistance. gs L is the nonlinear parasitic capacitance on the gate-source side. s(int) For internal source parasitic inductance, C ds D is the nonlinear parasitic capacitance on the drain-source side. body This is the ideal symbol for a body diode.

[0067] The resistance parameters in the single-tube loss model include: DC bus parasitic resistance R. p SiC MOSFET (S i The external gate resistance R g（ext） With internal gate resistance R g（int） The on-state resistance R of a SiC MOSFET ds(ON) And SiC Schottky barrier diode D f On-state resistance RD (ON) .

[0068] The parasitic inductance parameters of the single-tube loss model include: DC bus parasitic inductance L p SiC MOSFET (S i External gate parasitic inductance L g（ext） Internal gate parasitic inductance L g（int） Drain parasitic inductance L d（int） Parasitic inductance L inside the source electrode s（int） And SiC Schottky barrier diode D f parasitic inductance L sD ; among which, L g（int） and L sD Often, because its impact on characteristics is relatively low, it is ignored; therefore, L sD Not reflected in Figure 3 middle.

[0069] Parasitic capacitance parameters in the single-transistor loss model include: SiC MOSFET (S i The gate-source nonlinear parasitic capacitance C gs, drain-source nonlinear parasitic capacitance C ds , nonlinear parasitic capacitance C of SiC Schottky barrier diode jD , and load capacitance C L . Among them, C gs and C L are modeled as constant capacitances, C gd , C ds are modeled as nonlinear functions of drain-source voltage v ds , and C jD is modeled as a nonlinear function of reverse voltage v r . In addition, C gs and C ds will exhibit characteristics related to the working region (such as ohmic region, saturation region, cutoff region) of the device under different working conditions of gate-source voltage v gd , drain-source voltage v ds ; the characteristics of C jD will also change with the reverse voltage of the diode, corresponding to different working states.

[0070] Assuming that the drain-source voltage v ds and the drain current i d change linearly during the switching transient process, and ignoring the change rates of dv ds / dt and di d / dt, the calculation is simplified. The model is based on the law of conservation of energy, expressed as: E loss =E source -E load -△E c . Among them, E loss is the loss, E source is the energy of the DC bus E sbus and E sdrive ; E load is the energy consumed in the load; and ΔE c is the change in the energy stored in the circuit. , E sdrive =V CC ×Q g . Among them, t p2 is the starting time node during the driving source pulse opening process, t p3 is the ending time node during the driving source pulse opening process, and Q g is the total gate charge.

[0071] , where the circuit parasitic inductance L stray is the sum of L p , L d , and L s .

[0072] The change in the energy stored in the circuit ΔEc With constant value capacitor C gs C L Nonlinear capacitor C gd C ds C jD The expression is related to different operating conditions. The change in energy stored in the circuit during the start-up process, ΔE, is related to this. c(ON) For E c(tp3) With E c(tp2) The difference, i.e., ΔE c（ON） =E c (t p3 )-E c (t p2 ), where E c (t p3 ) for t p3 The total energy stored in the circuit at any given moment, E c (t p2 ) for t p2 The total energy stored in the circuit at any given moment, E c (t p3 ) = (1 / 2) × (L) p I o 2 +C issH V CC 2 ) + (1 / 2)(C jDL +C L )×V DC 2 E c (t p2 )=[(1 / 2)×(C gs V EE 2 +C OssL V DC 2 ) + (1 / 2)(C jDH +C L )×V FD 2 ], where L p For the parasitic inductance of the DC bus, C issH C represents the gate-drain parasitic capacitance of a SiC MOSFET in the ohmic region. jDL C is the parasitic capacitance of the freewheeling diode in the cutoff region. L C is the parasitic capacitance of the load inductor. gs C is the gate-source parasitic capacitance of the SiC MOSFET. OssL C is the output parasitic capacitance of the SiC MOSFET cutoff region. jDH V is the junction capacitance of the conduction region of a SiC Schottky diode. FD This is the voltage of the freewheeling diode.

[0073] Circuit energy storage energy variation amount DE during turn-off process c(OFF) E c(tp1) E c(tp0) E c（OFF） E c E p1 E c E p0 E p0 E p1 E c E p1 E p1 E c E p0 E p0 E c E p1 E p E o E 2 E issH E CC E 2 E jDL E L E DC E 2 E c E p0 E gs E EE E 2 E OssL E DC E 2 E jDH E L E FD E 2 E p E issH E jDL E L E gs E ossL E jDH E FD E

[0074] In the embodiment of the present application, the switching loss calculation model of the SiC power device includes an opening loss calculation model and a turn-off loss calculation model, wherein the opening loss calculation model satisfies the following condition: P sw（ON） =E loss（ON） ×f sw ; P sw（ON） is the opening power loss of the SiC power device, f sw is the switching frequency, E loss（ON） is the opening energy loss, E loss（ON） =E loss1（ON） +E loss2（ON） , wherein E loss1（ON） is the main energy loss in the opening process, E loss2（ON） is the additional energy loss in the opening process, , V DC is the DC bus voltage, i d is the transient drain current, I o is the load current, v ds is the drain-source voltage, t p2 is the starting time node in the driving source pulse opening process, t p3 is the ending time node in the driving source pulse opening process; E loss2（ON） =L stray I o 2 +V CC Q g -△E c（ON） , L stray is the line parasitic inductance, V CC is the gate drive voltage, Q g is the total gate charge, and △E c（ON） is the change amount of the circuit storage energy in the opening process.

[0075] Figure 4 is an opening transient circuit topology evolution schematic diagram, Figure 5 is an opening transient waveform timing schematic diagram. Figure 4 In the opening process, the line parasitic inductance L stray originates from the parasitic inductance effect of the DC bus and the power device connection line, and affects the oscillation characteristics of the transient process. V drive is the gate drive voltage, which provides the opening / closing driving signal for the MOSFET gate, and D is a freewheeling diode (such as a SiC Schottky diode), which provides a freewheeling path for the current in the switching transient. i L is the load current, which is the current flowing through the load. Figure 5 In the opening process, V gs0 is the voltage level of the gate drive circuit. V Miller is the Miller platform voltage, which is caused by the Miller effect of C gd v gsThe change is slow at this stage. peak The peak value of drain current, the maximum value of drain current in the turn-on transient.

[0076] As Figure 4 and Figure 5 shown, according to the dominant physical mechanism of the turn-on transient, the MOSFET turn-on process can be divided into three core stages: the gate loop charging stage (the gate capacitor is charged, v gs rises), the drain current rises and the drain-source voltage drops stage (Miller effect dominates, the drain current gradually replaces the diode freewheeling, v ds rapidly drops), the drain current oscillation and steady state transition stage (line parasitic inductance and parasitic capacitance resonance, current / voltage oscillation and tends to be steady).

[0077] Through time slicing technology, the entire turn-on transient is divided into t0-t1, t1-t2, t2-t3, t3-t4, t4-t5, t5-t6, t6-t7, a total of 7 sub-stages (as shown in Figure 5 waveform timing):

[0078] t0-t1: the gate loop charges C gs , v gs rises but does not reach the threshold voltage V th , the drain current (i d =0), the drain-source voltage v ds ≈V DC ;

[0079] t1-t2: v gs reaches V th , i d begins to rise, the freewheeling diode D is still on, v ds maintains a high level;

[0080] t2-t3: i d continues to rise and gradually replaces the diode freewheeling, v ds begins to drop, v gs enters the "Miller plateau" (slow change) due to the Miller capacitor C gd ;

[0081] t3-t4: i d rises to the peak value I peak , v ds rapidly drops to the on-state voltage drop;

[0082] t4-t5: the line parasitic inductance L par and parasitic capacitance (C ds , C jD ) and other) resonates, i d , v ds oscillates;

[0083] t5-t6: The oscillation gradually decays, i d The tendency of load current I O v gs It disengages from the Miller plateau and rises to the drive voltage level;

[0084] t6-t7: Entering the steady-state conduction phase, i d Stable at I O Nearby, v ds For the on-state voltage drop, v gs It is stable at the driving voltage.

[0085] After analyzing the voltage and current dynamic characteristics of each sub-stage, the turn-on loss E is then analyzed. loss1(ON) By performing integration, the analytical loss expression for the start-up process can be derived.

[0086] The turn-off loss calculation model satisfies the following conditions:

[0087] P sw（OFF） =E loss（OFF） ×f sw ;P sw（OFF） For the turn-off power loss of SiC power devices, f sw E is the switching frequency. loss（OFF） To shut off energy loss, E loss（OFF） =E loss1（OFF） +E loss2（OFF） E loss1（OFF） E represents the main energy loss during the turn-off process. loss2（OFF） This refers to the additional energy loss during the shutdown process. , t p0 t is the starting time point in the drive source pulse turn-off process. p1 This is the end time point in the drive source pulse turn-off process; E loss2（OFF） =-L stray I o 2 -V EE Q g -△E c（OFF） V EE The lower limit of the driving voltage, ΔE c（OFF） This stores the change in energy during the circuit's shutdown process.

[0088] like Figure 6 As shown, based on the dominant physical mechanism of the turn-off transient, the turn-off process of a SiC MOSFET can be divided into three core stages: the gate circuit discharge stage (energy storage and release of the gate capacitor, gate-source voltage v) and the gate-source voltage discharge stage. gs (decline), drain-source voltage rise phase (Miller capacitance C) gd Dominant, drain-source voltage vds From on-state voltage drop to DC bus voltage V DC Rise, drain current i d Step by step to the freewheeling diode, drain current and drain-source voltage oscillation stage (line parasitic inductance and parasitic capacitance resonance, current, voltage presents oscillation and tends to be stable).

[0089] Through time slicing technology, the whole off transient is subdivided into t0-t1, t1-t2, t2-t3, t3-t4, t4-t5, t5-t6 a total of 6 sub-stages (combined with the current, voltage waveform timing):

[0090] t0-t1: Gate loop start discharge, v gs From the drive high level drop but still higher than the threshold voltage V th , drain current i d Maintain load current I O , drain-source voltage v ds It is the on-state voltage drop;

[0091] t1-t2: v gs Drop to the Miller platform voltage V Miller , Miller capacitor C gd Significant involvement in charging and discharging, v ds From the on-state voltage drop rises quickly, i d Still close to I O ;

[0092] t2-t3: v ds Continue to rise to V DC , i d Step by step to the freewheeling diode and start to decline, v gs C gd Effect on the Miller platform;

[0093] t3-t4: v ds Approach V DC , i d Quickly drop to a lower value, v gs Away from the Miller platform continues to decline to the drive low;

[0094] t4-t5: Line parasitic inductance and parasitic capacitance (such as drain-source parasitic capacitance C ds , diode junction capacitance C jD ) and other) resonance, i d , v ds Oscillation;

[0095] t5-t6: Oscillation gradually attenuates, i d Tends to 0 (freewheeling diode stable freewheeling), v ds Stable near V DC , vgs fall to the drive low level (e.g., V EE After analyzing the voltage and current dynamic characteristics of each sub-stage, the analytical loss expression of the turn-off process can be derived by integrating the turn-off loss.

[0096] S300, by the processor executing a on-state loss calculation module, based on the time-varying operating condition sequence, the on-state loss of the SiC power device and the on-state loss of the freewheeling diode are calculated respectively, and the on-state loss calculation model of the SiC power device and the on-state loss calculation model of the freewheeling diode are established.

[0097] In the embodiment of the application, the on-state loss calculation module is a functional module integrated in the processor, and its core function is: based on the input time-varying operating condition sequence (including but not limited to output side phase current, device junction temperature, switching frequency and other time-varying parameters), the energy loss (on-state loss) of SiC power device (such as SiC MOSFET) in the on-state and the energy loss (on-state loss) of freewheeling diode (such as SiC Schottky barrier diode) in the forward on-state are calculated respectively, and the calculation logic is fixed as reusable on-state loss calculation model (including SiC power device on-state loss model and freewheeling diode on-state loss model) through mathematical modeling. The core calculation logic of the module includes:

[0098] The on-state loss of SiC power device is calculated in combination with the on-state resistance (whose value changes nonlinearly with junction temperature and on-state current) and on-time of the device;

[0099] The on-state loss of the freewheeling diode is calculated in combination with the forward on-state resistance and forward voltage drop (which change nonlinearly with junction temperature and forward current) and on-time of the device;

[0100] Based on the dynamic characteristics of the time-varying operating condition sequence, the time-sequential calculation of the loss is realized (i.e. the loss value is updated in real time as the operating condition changes).

[0101] Through the execution of the module, the on-state loss can be converted from an abstract physical process to a quantitative mathematical expression, providing basic data support for power device total loss analysis and thermal design.

[0102] If the three-phase voltage type main power topology structure is considered, the on-state loss calculation model of the SiC power device meets the following conditions:

[0103] P C，M =R ds（ON） ×I Drms 2 , P C，M is the on-state loss of the SiC power device, R ds（ON） is the on-state resistance of the SiC power device, I Drms is the root mean square current of the SiC power device, IDrms =I oc ×(1 / 8+(m a ×cosφ) / 3π)I oc The peak value of the output current; m a φ is the modulation index, which is the ratio of the peak voltage of the modulating wave to the peak voltage of the carrier wave. φ is the power factor angle.

[0104] The on-state loss calculation model of the freewheeling diode satisfies the following conditions:

[0105] P C，D =u D0 ×I Fav +RD （ON） ×I Frms 2 P C，D For the conduction loss of the freewheeling diode, u D0 I is the forward voltage drop of the freewheeling diode. Fav I is the average current of the freewheeling diode. Fav =I oc ×(1 / 2π-(m a ×cosφ) / 8)I Frms I is the root mean square current of the freewheeling diode. Frms =I oc ×(1 / 8-(m a (×cosφ) / 3π).

[0106] S400, the processor executes the freewheeling diode switching loss calculation module to calculate the switching loss of the freewheeling diode based on the time-varying operating condition sequence, and establishes a freewheeling diode switching loss calculation model.

[0107] In this embodiment of the invention, the processor is a core hardware unit with data processing, module scheduling, and logic control capabilities. Its core function is to serve as the "central control and processing carrier" for the loss calculation process. Specifically:

[0108] Functional positioning: Receive and parse the time-varying operating condition sequence (such as dynamic parameters such as output phase current, switching frequency, device junction temperature, speed regulation frequency, etc.) output by the top-level requirements, and schedule the execution of functional modules such as "freewheeling diode switching loss calculation module" and "conduction loss calculation module" in sequence according to the preset logic;

[0109] Core capabilities: Possesses high-speed data processing capabilities, which can support complex computational tasks involved in switching loss calculations, such as transient physical process modeling, nonlinear parameter fitting (e.g., the effect of junction temperature on device characteristics), and integral operations (e.g., time integration of energy loss).

[0110] Output function: The calculation results of each module (such as the loss quantification data of the freewheeling diode switch and the loss calculation model parameters) are integrated to provide data support for subsequent total loss analysis, device thermal design, and controller optimization.

[0111] The freewheeling diode switch loss calculation module is a special function module integrated in the processor, and its core function is to quantify the energy loss of the freewheeling diode (such as a SiC Schottky barrier diode) in the switching process (turn-on transient and turn-off transient), and to establish a reusable switch loss calculation model. The specific definition is as follows:

[0112] Module positioning: Aiming at the switching physical characteristics of the freewheeling diode (such as the forward recovery process during turn-on, the reverse recovery process during turn-off, transient voltage / current oscillation, etc.), the loss calculation function unit is specially developed, and is the core component of the "switching loss branch" in the total loss calculation system.

[0113] Input parameters: The time-varying operating condition sequence is the core input, which specifically includes:

[0114] Electrical parameters: forward conduction current, reverse blocking voltage, switching frequency (determines the number of switches), gate drive parameters (such as gate voltage, gate resistance, affects switching speed) of the freewheeling diode;

[0115] Thermal parameters: device junction temperature (junction temperature will significantly affect the reverse recovery charge, forward voltage drop, etc. of the diode, and then change the switching loss);

[0116] Device inherent parameters: parasitic parameters (such as parasitic inductance, parasitic capacitance, reverse recovery time, reverse recovery charge) and other nonlinear characteristic parameters of the freewheeling diode;

[0117] Calculation logic:

[0118] Based on the switching transient physical mechanism of the freewheeling diode (such as the three-stage process of "forward current drop → reverse current peak → reverse current decay" during turn-off), the time slicing method is used to decompose the transient process into multiple sub-stages; for each sub-stage, a dynamic relationship model of voltage (such as the forward / reverse voltage across the diode), current (such as the forward conduction current, reverse recovery current) and time is established; by integrating the "voltage-current-time" relationship of each sub-stage, the quantification value of the turn-on loss and turn-off loss in a single switching period is obtained; combined with the dynamic changes of the time-varying operating condition sequence (such as the increase of Q rrThe calculation parameters of each sub-stage are corrected to realize the time-sequential dynamic calculation of the loss. The output results are: quantitative data: real-time switching loss value (including turn-on loss, turn-off loss and total switching loss) of the freewheeling diode changing with the operating conditions; and calculation model: the above calculation logic (including transient stage division, parameter correction rule, integral formula, etc.) is solidified into a "freewheeling diode switching loss calculation model", which can be directly used for subsequent integration analysis of total loss (on-state loss + switching loss), and pre-judgment and optimization of the controller on the device loss.

[0119] Further, the turn-on loss of the freewheeling diode is mainly caused by the reverse recovery process, and the freewheeling diode switching loss calculation model satisfies the following conditions:

[0120] P sw，D =(1 / 4)×Q rr ×U Drr ×f sw , wherein P sw，D is the turn-on power loss of the freewheeling diode, Q rr is the reverse recovery charge, U Drr is the voltage across the diode during reverse recovery, and f sw is the switching frequency.

[0121] The derivation process of P sw，D is as follows:

[0122] The turn-on loss energy E loss（ON），D of the freewheeling diode is obtained.

[0123] Wherein, ; wherein tri is the time for the drain current to rise from 0 to the on-state current Io of the SiC device; tfu is the time for the drain-source voltage to drop from UDD to R ds(on) •I o . E loss（ON），Drr is the reverse recovery loss in the turn-on stage of the freewheeling diode.

[0124] Since the turn-off loss energy E loss(OFF),D of the freewheeling diode is generally ignored, P sw，D =(E loss（ON），D +E loss（OFF），D )×f sw ≈E loss（ON），D ×f sw =(1 / 4)×Q rr ×U Drr ×f sw .

[0125] The influence of inductive load on the working mode of main power topology circuit is divided into two cases: power factor angle 0<Φ<π / 3 and Φ>π / 3, and the loss calculation needs to consider the role of freewheeling diode D. When 0<Φ<π / 3 is considered, the working mode has two kinds of 3S+0D and 2S+1D. The commutation process under inductive load cannot be completed instantaneously, and the load current needs to be freewheeled through the freewheeling diode of the target commutation switch. When Φ>π / 3 is considered, the working mode has two kinds of 2S+1D and 1S+2D.

[0126] S500, integrating each loss calculation model established by the model integration module, inputting the time-varying operation condition sequence, calculating and outputting the total loss power of the power motor controller in the entire expected flight profile.

[0127] The model integration module is the core functional module responsible for the loss calculation logic closed loop in the application, and its core role is: organically integrating each independent loss calculation model (including the on-state loss model of SiC power devices, the on-state loss model of freewheeling diodes, the switching loss model of SiC power devices, and the switching loss model of freewheeling diodes) established in the early stage, realizing the integrated calculation of the total loss of the power motor controller through time sequence matching, dynamic association and energy superposition. As the "final integration unit" of the loss calculation system, the model integration module connects each subdivided loss model and the total loss output, solves the problems of time sequence asynchronization and weak parameter correlation between independent models, and ensures the accuracy and dynamics of total loss calculation.

[0128] The input parameters of the model integration module include:

[0129] The output results of each subdivided loss model include the turn-on switching loss power and the turn-off switching loss power of SiC power devices, the on-state loss power of SiC power devices, the on-state loss power of freewheeling diodes, and the switching loss power of freewheeling diodes;

[0130] The original time-varying operation condition sequence is used to verify the time sequence consistency of each loss model, and ensure that the integration result matches the actual running state (such as the switching frequency and the change of phase current in different flight stages, which need to be reflected in the total loss);

[0131] The model association parameters, such as the working duty ratio and the switching period overlap coefficient of each device, are used to correct the loss superposition logic when multiple devices work at the same time.

[0132] The core function of the model integration module is:

[0133] Timing alignment: synchronize the loss data output by each sub-model on the time axis, eliminate the timing deviation caused by the difference in calculation granularity (such as switch loss calculated in transient state, on-state loss calculated in steady state interval, corresponding to the same timing through time slicing);

[0134] Dynamic superposition: based on the total loss power formula P=P sw（ON） +P sw（OFF） +P C，M +P C，D +P sw，D , dynamically accumulate each loss component according to real-time operating conditions to generate a timing curve of total loss power;

[0135] Boundary correction: for special working conditions (such as overlapping intervals of simultaneous conduction of devices, nonlinear variation of loss at extreme junction temperature), introduce correction coefficients to adjust the superposition result to ensure consistency with physical reality; profile adaptation: map the total loss calculation result to the full cycle of the "expected flight profile" (such as take-off, cruise, hovering, landing, etc.), output the total loss characteristic values (average value, peak value, integrated energy, etc.) of each stage.

[0136] The output of the model integration module is the total loss power timing data and characteristic parameters of the motor controller in the entire expected flight profile, providing core basis for controller thermal design, efficiency optimization and energy management.

[0137] Further, S500 specifically includes:

[0138] S501, receiving loss components output by each sub-model, combining input time-varying operating condition sequence (such as phase current, switching frequency, junction temperature data of each stage of flight profile), aligning all loss data on the time axis (for example, establishing a corresponding relationship between the transient value of switch loss and the steady state interval value of on-state loss through time slicing);

[0139] S502, based on the total loss power formula, dynamically superimposing each loss component to obtain the total loss power;

[0140] S503, for special working conditions in the flight profile (such as large current impact in the take-off stage, high frequency switching in the cruise stage), the module introduces boundary correction coefficients (such as a correction factor for R ds(ON) increase caused by junction temperature rise), adjusts the superposition result to match the actual physical characteristics;

[0141] S504, finally output the total loss power timing curve of the motor controller in the entire expected flight profile, and the total loss peak value, average value and cumulative energy loss of each flight stage (take-off, cruise, landing, etc.), providing complete loss data support for the heat dissipation design, efficiency optimization and endurance capability evaluation of the controller.

[0142] Further, the method provided by the embodiment of the present application further comprises the following steps:

[0143] S600, collecting, by a temperature sensor and a current and voltage sensor integrated in the motor controller, a junction temperature T of the SiC power device in real time j , a load current I o , and a drain-source voltage V ds as real-time monitoring data.

[0144] The present application realizes real-time collection of key electrical-thermal parameters of SiC power devices by integrating a high-reliability sensor array near the main power topology of an electric aviation power motor controller, and specifically comprises:

[0145] 1. Sensor selection and installation design

[0146] Temperature sensor: an NTC thermistor (accuracy ±1℃, measurement range -40℃~175℃) conforming to the aviation electronic environment standard (such as DO-160G) is selected, and is tightly attached to the packaging shell (such as the bottom heat dissipation surface of the TO-247-4 packaging) of the SiC power device, and a heat-conducting silicone grease is filled in the gap to reduce the thermal resistance error; one NTC sensor is independently configured for each SiC power device to ensure the spatial resolution of junction temperature monitoring (to avoid the averaging error caused by multiple devices sharing a sensor).

[0147] Current sensor: a Hall effect current sensor (bandwidth ≥1MHz, accuracy ±0.5%FS, range 0~500A) is used, which is connected in series in the drain loop of the SiC power device, and the parasitic inductance of the sensor and the main loop is controlled to be ≤5nH (to avoid measurement noise caused by high-frequency switching current); for a three-phase topology, one auxiliary current sensor is additionally configured on the output side of each phase to verify the consistency of the load current.

[0148] Voltage sensor: a differential voltage sensor (input impedance ≥100MΩ, common-mode rejection ratio ≥80dB, measurement range 0~1000V) is used, and the two collection ends of the sensor are connected to the drain and source of the SiC power device, respectively, and the length of the lead wire is controlled to be ≤5cm (to reduce the high-frequency voltage measurement deviation caused by the parasitic capacitance of the lead wire); the sensor output signal is transmitted to a signal conditioning module through a shielded twisted pair to avoid the influence of electromagnetic interference (EMI) in the aviation environment on the data.

[0149] 2. Data collection timing and synchronization control

[0150] The main clock (frequency 100 MHz) based on the controller generates a synchronous trigger signal to ensure that the collection time of temperature, current and voltage data deviates by less than 1 μs (to avoid matching errors of "junction temperature-voltage-current" caused by asynchronous timing, such as using the junction temperature at time t to correspond to the voltage data at t+5 μs); the collection frequency is dynamically adjusted according to the operating conditions: the collection frequency is set to 1 MHz in the switching transient stage (such as takeoff and landing), and is set to 100 kHz in the steady-state running stage (such as cruising), to balance the data accuracy and storage overhead.

[0151] The collected raw data is preprocessed:

[0152] Current / voltage data: high-frequency noise is removed by sliding average filtering (window size 10 sampling points), and pulse interference generated by switching transients is removed by a spike removal algorithm (threshold set to 1.5 times the rated value);

[0153] Temperature data: the slow changes of junction temperature are smoothed by first-order lag filtering (time constant 0.1 s) to avoid transient fluctuations caused by the thermal inertia of the sensor itself.

[0154] 3. Data storage and abnormality marking

[0155] The preprocessed real-time monitoring data (including collection time stamp, parameter type identification and sensor ID) are stored in the non-volatile memory (NVM) of the controller, and a circular coverage strategy is adopted (storage capacity ≥ 1 GB, capable of storing data for the last 24 hours); at the same time, the data are marked as abnormal: when the collection value exceeds the preset range (such as junction temperature > 175℃, current > 500A, voltage > 1000V), it is automatically marked as "abnormal data" and an alarm signal is triggered, and the subsequent parameter identification step will skip such data to avoid incorrect input.

[0156] S700, the key device parameters are dynamically updated based on the real-time monitoring data by a parameter identification algorithm.

[0157] Based on the real-time monitoring data preprocessed by S600, the key parameters of the SiC power device are dynamically updated by a phased parameter identification algorithm, and the core focuses on the functional relationship between on-state resistance and junction temperature. Specifically, in the steady-state stage of device conduction, the real-time on-state resistance is estimated according to a preset estimation method, and the estimated real-time on-state resistance is bound with the current junction temperature to update the functional relationship between on-state resistance and junction temperature.

[0158] The conduction steady-state period of the SiC power device can be screened out by multi-condition joint determination, and the determination conditions include: the load current I o : the fluctuation range of the continuous 100 sampling points is ≤±2% (to avoid nonlinear changes of V ds due to current transients); drain-source voltage V ds: the fluctuation range of 100 continuous sampling points is less than or equal to ±3% (to ensure that the device works in the ohmic conduction region, rather than in the saturation region); and the switch state: the PWM driving signal output by the controller is a continuous high level (the time length is greater than or equal to 100 microseconds, covering at least two current / voltage acquisition periods).

[0159] The period satisfying all the above conditions is marked as an "on-state steady stage", and the average values of the drain-source voltage Avg(V ds ), the load current Avg(I o ), and the junction temperature Avg(T j ) in the on-state steady stage are extracted as the basic data for parameter identification.

[0160] In the embodiment of the present application, the real-time on-state resistance R ds(ON) new = (Avg(V ds )-V th (T j )) / I o , wherein V th (T j ) is the threshold voltage of the SiC power device at the current junction temperature (calculated based on the initial fitting V th -T j function, such as V th (T j )=V th25 +TC Vth *(Avg(T j )-25)), V th25 is the threshold voltage at 25 DEG C, and TC Vth is the threshold voltage temperature coefficient); the V th (T j ) correction term is introduced to eliminate the influence of the threshold voltage on the drain-source voltage, so that the estimation error of R ds(ON) new is reduced from ±8% to within ±3%.

[0161] In the embodiment of the present application, the estimated real-time on-state resistance and the current junction temperature form a new sample pair, and the sliding window least square method is used to update the function relationship between R ds(on) -T j , which can specifically include:

[0162] Sample pair construction: the estimated R ds(ON) new and the corresponding Avg(T j ) form a new "resistance-junction temperature" sample pair, and are added to the sample pool; the sample pool adopts a sliding window mechanism, and the window size is set to 100 samples (considering data timeliness and statistical reliability), and when a new sample is added, the earliest 1 group of old samples are removed.

[0163] Least square fitting: Perform weighted least square fitting on the sample pairs within the sliding window (assign higher weight to the recent 30 groups of samples, weight coefficient linearly decreases over time), fit R ds(ON) -T j with a quadratic polynomial function: R ds(ON) (T j )=a·T j 2 +b·T j +c, where a, b, c are fitting coefficients; calculate the residual (deviation of actual sample value from fitted value) after fitting, if residual≤±5%, update the function relationship; if residual>±5%, trigger abnormality diagnosis (e.g. check if the sensor is offset, if the device is aged).

[0164] Fitting interval adaptation: When Avg(T j ) exceeds the initial fitting interval \([25℃,150℃]\) (e.g. Avg(T j )=160℃ under high temperature working condition), automatically extend the fitting model to a cubic polynomial: R ds(ON) (T j )=a·T j 3 +b·T j 2 +c·T j +d, ensure the validity of the function relationship in a wide junction temperature range (-40℃~175℃), d is a fitting coefficient.

[0165] S800, feedback the updated key device parameters into the loss calculation model established in S200, S300, S400, realize online correction and adaptive prediction of model parameters, to correct the deviation caused by device aging and junction temperature drift, after calibration, the accuracy is improved to within ±5%.

[0166] In S800, the updated key device parameters (the core is R ds(ON) (T j )) of S700 are fed back to the loss calculation model, and adaptive optimization of loss prediction is realized through hierarchical correction and dynamic verification, the specific steps are as follows:

[0167] 1. Model adaptation and substitution of parameters

[0168] Switching loss model correction: the updated R ds(ON) (T j ) is used to correct the "additional energy loss" term of switching loss: in the turn-on loss E loss2(ON) , the steady-state value of drain-source voltage after device conduction is corrected from R ds(ON) old ·Io Updated to R ds(ON) new ·I o , and then adjust the calculation of the energy change of the circuit (because △E c(ON) is related to the steady-state value of drain-source voltage); at the same time, in the off-state loss, the initial steady-state value of drain-source voltage is corrected for the same reason, to ensure that the calculation of switching loss matches the current device state.

[0169] On-state loss model correction: directly substitute the updated R ds(ON) (T j ) into the on-state loss formula of SiC power devices, and based on the change trend of R ds(ON) (T j ), correct the on-state resistance of the freewheeling diode, because the aging trends of the two types of devices are related, and R D(ON) new =R D(ON) old ·(R ds(ON) new / R ds(ON) old ), to ensure the consistency of the on-state loss calculation of the two types of devices.

[0170] Diode switching loss model correction: based on the device aging degree reflected by R ds(ON) (T j ), dynamically adjust the correction coefficient of reverse recovery charge Q rr : if R ds(ON) new / R ds(ON) old ≥1.2, indicating that the device is aging seriously, then increase Q rr by 5% (because aging will lead to an increase in the carrier storage effect of the diode), and the corrected diode switching loss formula is: P sw，D = (1 / 4) × Q rr × (1+k) U Drr ×f sw , k is the aging correction coefficient, 0≤k≤0.1.

[0171] 2. Dynamic verification of prediction accuracy

[0172] Real-time loss comparison verification: compare the "predicted total loss power Ppred" output by the corrected loss model with the "measured total loss power Pmeas" of the controller; wherein Pmeas is calculated by the difference between "input power" and "output power" (input power is the product of DC bus voltage and current, output power is the product of motor three-phase voltage and current, both are calculated based on sensor data of S600). Precision adjustment mechanism: if |Ppred-Pmeas|≤5% (satisfies the aviation level precision requirement), keep the current model parameters; if 5%<|Ppred-Pmeas|≤10%, shorten the parameter update period of S700 (from original 10s / time to 5s / time); if |Ppred-Pmeas|>10%, trigger "emergency correction", empty the current sample pool, re-collect 20 groups of on-state steady-state samples for fitting, and perform fault self-checking on the sensor (such as checking the zero drift of the voltage sensor).

[0173] 3. Engineering implementation of adaptive prediction

[0174] Integrate the corrected loss model into the real-time operating system (RTOS) of the controller, and adopt the "piecewise prediction + rolling update" strategy: for each flight phase of the aircraft (such as take-off, cruise), based on the time-varying operating condition sequence of the phase (S100 output), predict the loss power curve of the phase 10s in advance; every 1s, based on the latest sensor data (S600 output) and updated parameters (S700 output), roll the loss prediction curve of the remaining phase to ensure that the prediction result is always synchronized with the actual working condition; the prediction data is transmitted to the "energy management module" of the aviation power system in real time, providing dynamic basis for range estimation, cooling fan speed regulation, and power distribution strategy optimization (such as predicting the increase of loss in the cruise phase, reducing the power distribution of non-critical systems in advance to ensure endurance).

[0175] S600 to S800 realizes online self-calibration and adaptive prediction of the loss model throughout its life cycle, significantly improving the long-term accuracy and reliability of the model.

[0176] Once the parameters of the traditional loss model are established, they are fixed values and cannot reflect the parameter drift and aging conditions of the device during use. The present application can dynamically capture and update the relationship between key parameters such as on-state resistance and threshold voltage and junction temperature and aging degree by introducing real-time data monitoring and online parameter identification algorithm. This makes the loss prediction model change from a "static snapshot" to a "dynamic video", not only maintaining high accuracy at the beginning of the controller's use, but also continuously providing reliable loss prediction throughout its life cycle, laying a solid foundation for predictive health management (PHM) and fault warning.

[0177] (Embodiment)

[0178] Firstly, the core conditions of the benchmark test are defined, including the device junction temperature T J , the DC bus voltage V DS , the gate drive source voltage range V GS , the load current I D , the external gate drive resistance R gext) , the internal gate drive resistance R g(int) , and other key variables, the physical meaning of each variable and the corresponding circuit / device position are shown as follows Figure 3 .

[0179] Secondly, the characteristic parameters of the external circuit and the device are obtained through actual testing, numerical simulation combined with device manual data, including:

[0180] External circuit characteristic parameters: the DC bus equivalent parasitic inductance L stray , the load parallel capacitance C L , the load inductance L load , the bus equivalent resistance R p , etc.

[0181] SiC MOSFET device characteristic parameters: gate-source parasitic capacitance C gs , cutoff region gate-drain parasitic capacitance C gdL , ohmic region gate-drain parasitic capacitance C gdH , cutoff region drain-source parasitic capacitance C dsL , ohmic region drain-source parasitic capacitance C dsH , on-state resistance R ds(on) , transfer characteristic k fs , threshold voltage V th0 , total gate charge Q g , etc.

[0182] SiC Schottky barrier diode characteristic parameters: cutoff region diode parasitic capacitance C jDH , ohmic region diode parasitic capacitance C jDL , diode on-state resistance R D(on) , diode forward voltage constant quantity V D(on) , etc.

[0183] Finally, the calculation and extended analysis of switching loss are carried out:

[0184] (1) According to the time slice stage division of Figure 5 (turn-on transient state), Figure 6 (turn-off transient state), the voltage and current dynamics of each sub-stage are analyzed, and the turn-on energy loss E loss1(ON) and turn-off energy loss E loss2(ON) of the main switch process are calculated by integrating the relationship of “voltage-current-time”;

[0185] (2) Perform E loss1(ON) , E loss2(OFF) Supplementary calculation: Supplementary calculation is for "additional loss outside the main switching process", typical scenarios include:

[0186] Oscillation loss caused by resonance of parasitic inductance and parasitic capacitance (additional energy dissipation generated by interaction of line parasitic parameters in the late stage of switching transient);

[0187] Charge and discharge loss of gate drive circuit (power loss generated by charging and discharging of gate resistance to gate capacitance when driving signal is switched);

[0188] Edge region loss caused by non-ideal characteristics of devices (such as near threshold voltage, on / off switching transition zone, additional loss generated by deviation of device characteristics from ideal model);

[0189] (3) Combine the switching frequency f sw , and convert the sum of "main switching loss energy + supplementary loss energy" into average switching loss power (total switching energy loss x switching frequency);

[0190] (4) If the influence of different operating conditions needs to be analyzed, the above steps can be repeated, and key variables (such as SiC device on-state resistance R ds(ON) , internal gate resistance R g(int) , transfer characteristic g fs , threshold voltage V th , forward on-state resistance R D(ON) of Schottky barrier diode, threshold voltage V D(th) , etc.) are adjusted, and the loss characteristics are recalculated.

[0191] In summary, the present application has at least the following beneficial effects:

[0192] (1) Precise mapping and dynamic evaluation from the top-level task requirements of the aircraft to the bottom-level device loss are realized.

[0193] Traditional loss models are mostly isolated device-level models, which are disconnected from system-level operating requirements. The present application first establishes a clear and quantitative modeling path from aircraft flight tasks (such as climbing, cruising, and landing) to controller specific operating points (current, frequency, junction temperature), and then to device-level loss based on the expected flight profile to extract multi-time scale time-varying operating condition sequences. This makes it possible to accurately predict the total energy consumption and thermal load within the entire flight profile during the design phase, providing unprecedented data support for system-level optimization (such as range estimation and thermal management system design).

[0194] (2) The accuracy and reliability of loss prediction under complex dynamic conditions are significantly improved.

[0195] Aiming at the extreme conditions of high frequency, high temperature and transient working conditions in aviation applications, the application characterizes key device parameters (such as Rds (on), Vth) as a dynamic function of junction temperature, and uses a time slicing algorithm to finely model the switching transient, thereby fundamentally solving the problem of insufficient accuracy caused by ignoring parameter temperature characteristics and transient details in traditional analytical models. This method can accurately capture the coupling effect of junction temperature change and parasitic parameters on the switching process, thereby maintaining high prediction accuracy under a wide temperature range and various load conditions, effectively avoiding device over-temperature damage or system reliability decline caused by loss estimation deviation.

[0196] (3) A high-fidelity modeling solution is provided, which takes into account the computational efficiency and engineering practicality.

[0197] Compared with the physical model with large calculation and difficult parameter acquisition, and the behavior model lacking physical meaning, the analytical model of the application significantly reduces the computational complexity while ensuring high accuracy. The model established by the method is easy to embed into the system-level simulation tool chain, realizing fast and repeated simulation calculation, greatly shortening the design verification period. It provides an efficient and reliable theoretical tool and design basis for device selection, heat dissipation design, efficiency optimization and life prediction of motor controllers, and has important significance for promoting the development of high-power-density and high-reliability electric aviation power systems.

[0198] The embodiment of the application also provides an electronic device, comprising: at least one processor; and a memory in communication connection with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are configured to execute the method of the embodiment of the application.

[0199] The embodiment of the application also provides a computer-readable storage medium storing computer executable instructions, and the computer executable instructions are used to execute the method of the embodiment of the application.

[0200] It should be understood that the steps shown above can be reordered, added or deleted using various forms of flow. For example, the steps described in the application can be executed in parallel, sequentially or in different order, as long as the desired results of the technical solutions disclosed in the application can be achieved, which is not limited herein.

[0201] The above specific embodiments do not constitute a limitation on the scope of protection of the application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the application shall be included in the scope of protection of the application.

Claims

1. A method for determining the losses of an electric aircraft power motor controller, characterized in that, The method is executed by a computing device and includes the following steps: S100, based on the expected flight profile of the electric aircraft, extracts the time-varying operating condition sequence through multi-timescale time series analysis and stores the sequence in memory. The time-varying operating condition sequence includes time series data of output phase current, switching frequency, modulation frequency and device junction temperature. S200: The processor reads the time-varying operating condition sequence and calculates the energy loss of the SiC power device during the switching transient process based on numerical calculation methods, establishing a switching loss calculation model. The switching transient process is divided into multiple sub-stages using a time-slicing algorithm. The calculation of the switching loss calculation model depends on the time-varying operating condition sequence and the key device parameters of the power device. The key device parameters are represented as a function of the junction temperature in the time-varying operating condition sequence through a parameter fitting module. S300, the processor executes the conduction loss calculation module to calculate the conduction loss of SiC power devices and the conduction loss of freewheeling diodes based on the time-varying operating condition sequence, and establishes the conduction loss calculation model of SiC power devices and the conduction loss calculation model of freewheeling diodes. S400, the processor executes the freewheeling diode switching loss calculation module to calculate the switching loss of the freewheeling diode based on the time-varying operating condition sequence and establishes a freewheeling diode switching loss calculation model; S500 integrates the established loss calculation models through the model integration module, inputs the time-varying operating condition sequence, calculates and outputs the total loss power of the power motor controller within the entire expected flight profile.

2. The method according to claim 1, characterized in that, The key device parameters include the on-state resistance, threshold voltage, and transfer characteristics of SiC power devices, as well as the on-state resistance and threshold voltage of freewheeling diodes.

3. The method according to claim 2, characterized in that, The switching loss calculation model for SiC power devices includes a turn-on loss calculation model and a turn-off loss calculation model. The turn-on loss calculation model satisfies the following conditions: P sw（ON） =E loss（ON） ×f sw ;P sw（ON） f represents the turn-on power loss of SiC power devices. sw E is the switching frequency. loss（ON） To enable energy dissipation, E loss（ON） =E loss1（ON） +E loss2（ON） , of which E loss1（ON） E represents the main energy loss during the start-up process. loss2（ON） Additional energy loss during the startup process. V DC i is the DC bus voltage. d For transient drain current, I o For load current, v ds t is the drain-source voltage. p2 t is the starting time point in the process of driving the source pulse to turn on. p3 E represents the end time point in the process of driving the source pulse to start; loss2（ON） =L stray I o 2 +V CC Q g -△E c（ON） L stray For the parasitic inductance of the line, V CC Q is the gate drive voltage. g For the total gate charge, ΔE c（ON） To store the energy change of the circuit during the start-up process; The turn-off loss calculation model satisfies the following conditions: P sw（OFF） =E loss（OFF） ×f sw ;P sw（OFF） f represents the turn-off power loss of SiC power devices. sw E is the switching frequency. loss（OFF） To shut off energy loss, E loss（OFF） =E loss1（OFF） +E loss2（OFF） , of which E loss1（OFF） E represents the main energy loss during the turn-off process. loss2（OFF） This refers to the additional energy loss during the shutdown process. , t p0 t is the starting time point in the drive source pulse turn-off process. p1 This is the end time point in the drive source pulse turn-off process; E loss2（OFF） =-L stray I o 2 -V EE Q g -△E c（OFF） V EE The lower limit of the driving voltage, ΔE c（OFF） This stores the change in energy during the circuit's shutdown process.

4. The method according to claim 1, characterized in that, The method further includes the following steps: S600, through the temperature sensor and current / voltage sensor integrated in the motor controller, monitors and collects the junction temperature T of the SiC power device in real time. j Load current I o and drain-source voltage V ds As real-time monitoring data; S700 dynamically updates the parameters of the key components based on the real-time monitoring data through a parameter recognition algorithm; S800 feeds back the updated key component parameters into the loss calculation model established by S200, S300, and S400, enabling online correction and adaptive prediction of model parameters.

5. The method according to claim 4, characterized in that, Specifically, S700 includes: during the steady-state conduction phase of power devices, estimating the real-time on-state resistance according to a preset estimation formula, binding the estimated real-time on-state resistance to the current junction temperature, and updating the functional relationship between on-state resistance and junction temperature.

6. The method according to claim 2, characterized in that, The on-state loss calculation model for SiC power devices satisfies the following conditions: P C，M =R ds（ON） ×I Drms 2 P C，M R represents the on-state loss of a SiC power device. ds（ON） I is the on-state resistance of the SiC power device. Drms This represents the root mean square current of the SiC power device.

7. The method according to claim 2, characterized in that, The on-state loss calculation model of the freewheeling diode satisfies the following conditions: P C，D =u D0 ×I Fav +RD （ON） ×I Frms 2 P C，D For the conduction loss of the freewheeling diode, u D0 I is the forward voltage drop of the freewheeling diode. Fav I is the average current of the freewheeling diode. Frms RD is the root mean square current of the freewheeling diode. （ON） This is the forward on-state resistance of the freewheeling diode.

8. The method according to claim 2, characterized in that, The calculation model for the switching loss of the freewheeling diode satisfies the following conditions: P sw，D = (1 / 4) × Q rr ×U Drr ×f sw , where P sw，D Q is the turn-on power loss of the freewheeling diode. rr To recover the charge in the reverse direction, U Drr f is the voltage across the diode during reverse recovery. sw This refers to the switching frequency.

9. An electronic device, characterized in that, Including processor and memory; The processor executes the steps of the method as described in any one of claims 1 to 8 by invoking programs or instructions stored in the memory.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a program or instructions that cause a computer to perform the steps of the method as described in any one of claims 1 to 8.

Images