Efficiency optimization method of electric aircraft controller based on improved loss model

Abstract

The present invention discloses an efficiency optimization method for an electric aircraft controller based on an improved loss model, including: S1: establishing a permanent magnet synchronous motor loss model to obtain the optimal stator current when the motor loss is minimum; S2: constructing a multi-parameter online identification strategy based on a model reference adaptive system; S3: combining the multi-parameter online identification strategy with the permanent magnet synchronous motor loss model control to suppress the loss model parameter mismatch and optimize the efficiency of the electric aircraft controller. The optimization method of the present invention can suppress the influence of the loss model parameter mismatch on the motor loss. By building a mathematical model of the electric aircraft permanent magnet synchronous motor, the energy efficiency optimization of the electric aircraft electric controller is completed and applied to the electric aircraft. The system energy consumption calculation results are in good consistency with the test results, which effectively reduces the motor's operating loss and the heating of the controller and the motor, and increases the aircraft's flight time.

Description

Technical Field

[0001] The invention relates to the technical field of electric aircraft optimization control, and in particular to an electric aircraft controller efficiency optimization method based on an improved loss model. Background Art

[0002] Electric aircraft have a very broad application prospect. In the military field, it can be used for battlefield reconnaissance, plateau border patrol, plateau emergency material delivery and intelligence acquisition. In the civilian field, it can be used for short-distance transportation around general aviation airports, commuter flights in mountainous and hilly areas, training flights and sightseeing. With the rapid development of new energy technologies (solar energy, hydrogen energy, etc.), electric general aviation aircraft will become one of the main means to reduce aviation carbon emissions. Electric aircraft will surely become the main product in the aviation market and an important direction for the development of green aviation in the world. Carrying out research on key technologies for electric aircraft design and verification can improve the research and development capabilities and localization rate of key technologies of general electric aircraft in my country, and play an important role in promoting the development of my country's general aviation industry.

[0003] Unlike traditional aircraft, electric aircraft use electric propulsion systems instead of internal combustion engines. The electric propulsion system is mainly composed of batteries, controllers and permanent magnet synchronous motors (PMSM). The controller converts the DC power provided by the battery into AC power, and drives the motor to rotate the propeller to provide power for the aircraft. PMSM often uses a vector control method, which does not take into account the loss changes caused by the temperature rise of the motor winding resistance when the motor runs for a long time. The long-term operation of PMSM may cause the motor to overheat and the aircraft to "stop", endangering the safety of aircraft operation. Therefore, the quality of the PMSM control strategy directly affects the reliability of the electric propulsion system of the electric aircraft. Optimizing the PMSM control algorithm and reducing the operating loss of the PMSM are important means to ensure the safe flight of electric aircraft.

[0004] The loss of PMSM is mainly concentrated in the copper loss and iron loss of the stator. Loss model control (LMC) can be used to effectively reduce the copper loss and iron loss of the motor. The LMC strategy is based on the mathematical model of the motor and determines the motor stator current value with the minimum motor copper loss and iron loss as the optimization goal. However, as the motor runs for a long time, the motor temperature rises, causing the stator copper loss and iron loss to increase. If the parameters in the LMC strategy are not updated in time, it will cause LMC parameter mismatch, increase motor loss, and rapidly increase temperature rise, and even cause the aircraft to "shut down" due to motor overheating, endangering the safety of aircraft operation. Summary of the invention

[0005] In view of this, the present invention discloses an electric aircraft controller efficiency optimization method based on an improved loss model to solve the problem of increased operating losses of electric aircraft motors caused by traditional LMC parameter mismatch.

[0006] The technical solution of the present invention is: an efficiency optimization method for an electric aircraft controller based on an improved loss model, comprising:

[0007] S1: Establish a permanent magnet synchronous motor loss model, with the goal of minimizing the motor iron loss resistance and stator resistance loss, and obtain the optimal stator current when the motor loss is minimized;

[0008] S2: Based on the model reference adaptive system, construct a multi-parameter online identification strategy; wherein the multi-parameter online identification strategy includes an adaptive rate based on Popov's hyperstability theorem and an adaptive rate based on the LMS algorithm;

[0009] S3: Incorporate a multi-parameter online identification strategy into the permanent magnet synchronous motor loss model control to suppress loss model parameter mismatch and optimize the efficiency of the electric aircraft controller.

[0010] Specifically, a permanent magnet synchronous motor loss model considering iron loss is established in S1, and the optimal stator current with the minimum motor loss is obtained according to the extreme value theory:

[0011] By considering the permanent magnet synchronous motor loss model of iron loss, the motor voltage equation under the d and q coordinate axes is obtained as follows:

[0012]

[0013] Where: u d 、u q are the d-axis and q-axis stator voltage components respectively; i d 、i q are the current components of the stator current under the d and q axes respectively; i wd 、i wq are the active components of the stator current on the d and q axes respectively; i fd 、i fq are the reactive components of the stator current on the d and q axes respectively; R s is the stator resistance; R f is the iron loss resistance; L is the inductance component of the stator inductance under the d and q axes (L for surface-mounted permanent magnet synchronous motors d =L q =L);ω e is the motor electrical angular velocity; ψ f is the permanent magnet flux.

[0014] Kirchhoff's law gives the current equation:

[0015]

[0016] Electromagnetic torque equation T e It can be expressed as:

[0017] T e =n p i wq ψ f (3)

[0018] Where: n p is the number of motor pole pairs;

[0019] When the motor is in steady state, the iron loss current under the d and q axes can be expressed as:

[0020]

[0021] Motor copper loss P cu and iron loss P f It can be expressed as:

[0022]

[0023] The sum of copper loss and iron loss is the total motor loss P loss :

[0024]

[0025] The loss equation is obtained from equations (3), (4) and (6):

[0026]

[0027] The problem of finding the optimal current can be transformed into the problem of finding the extreme value of the function. The optimal current i can be obtained wd-opt :

[0028]

[0029] Finally, the optimal d-axis current i can be obtained from equations (2), (3), (4) and (8): d-opt :

[0030]

[0031] Specifically, the adaptive rate based on Popov's hyperstability theorem is obtained according to the following steps:

[0032] According to the theory of model reference adaptive system, an adjustable model and a reference model based on Popov's hyperstability theorem are established;

[0033] The reference model is:

[0034] From (1) and (2), we can get the reference model: (10)

[0035]

[0036] Rewrite formula (10) as:

[0037]

[0038] Among them: i=i wd i wq T ; I=[i d i q ] T ;

[0039] The adjustable model is obtained by expressing the iron loss resistance and flux linkage in equation (11) with estimated values:

[0040]

[0041] in:

[0042] is the estimated value of iron loss resistance; is the estimated value of magnetic linkage; for i wd An estimated value of for i wq The estimated value of; From equations (11) and (12), the error state equation can be obtained:

[0043]

[0044] in:

[0045] make From formula (13), we can get:

[0046]

[0047] Formula (14) is a nonlinear feedback system, which consists of two parts: a linear forward channel and a nonlinear feedback channel;

[0048] According to Popov's superstability theorem and its adjustable model and reference model, the two conditions for accurately identifying the motor parameters are:

[0049] (1) The transfer function matrix of the linear forward channel H(s) = D(sI-A) -1 Strictly positive:

[0050] (2) The nonlinear feedback channel satisfies the inequality:

[0051]

[0052] in: is any finite positive number;

[0053] For condition (1), the positive reality lemma is used to ensure that the transfer function is strictly positive definite. The general linear time-invariant system is:

[0054]

[0055] Among them: A1 is the system matrix; B1 is the input matrix; C1 is the output matrix; J is the direct transfer matrix; x is the state vector; u is the input vector; y is the output vector.

[0056] Assuming that system (16) is controllable and observable, its transfer function H(s) is:

[0057] H(s)=J+C1(sI-A1) -1 B1 (17)

[0058] If equation (17) is a strictly positive definite transfer matrix, the conditions that need to be met are that there are positive definite matrices P, Q, and non-singular matrices K, L that satisfy the following conditions:

[0059]

[0060] Among them: A1 T is the transposed matrix of A1; B1 T is the transposed matrix of B1; L T is the transposed matrix of L; J T is the transposed matrix of J; K T is the transposed matrix of K.

[0061] According to condition (1) and formula (18), we can get:

[0062]

[0063] Assumptions From formula (19), we can get:

[0064]

[0065] Where: k is a constant.

[0066] From equation (20), we can see that if we want to ensure that the matrices Q and P are positive definite, k>0 is required. Here, k=1 is taken, and the matrices P and D are unit matrices. Therefore, when D is the unit matrix, condition (1) is satisfied, and the transfer function matrix of the linear forward channel is H(s)=D(sI-A) -1 Strictly positive;

[0067] For condition (2), we need to design an adaptive rule first, substitute it into the inequality and solve the inequality in reverse. Expand the matrix W in condition (2) to get:

[0068]

[0069] For the convenience of proof, equation (20) is decomposed into:

[0070]

[0071] in: eta(0,t1)=eta1(0,t1)+eta2(0,t1);

[0072] When designing the adaptive rate according to Popov stability theorem, the proportional integral form is adopted, and the adaptive rate of iron loss resistance and permanent magnet flux is:

[0073]

[0074] Among them: F1(τ), F2(τ), G1(τ), G2(τ) are functions of the motor's current, voltage and electrical angular velocity.

[0075] First, the adaptive rate of the iron loss resistor is designed. From equations (22) and (23), we can get:

[0076]

[0077] Formula (24) is divided into two parts:

[0078]

[0079] in: η1(0,t1)=η 11 (0,t1)+η 12 (0,t1).

[0080] make:

[0081]

[0082] Where: k i and k p are proportional constant and integral constant greater than 0 respectively;

[0083] It can be proved from equations (25) and (26) that

[0084]

[0085] Similarly, it can be proved Therefore, from equation (27) and equation (23), the adaptive rate of permanent magnet flux linkage and iron loss resistance can be obtained as follows:

[0086]

[0087] Specifically, the adaptive rate based on the LMS algorithm is obtained according to the following steps:

[0088] According to the model reference adaptive system theory, an adjustable model and a reference model based on the LMS algorithm are established;

[0089] When the motor is running in steady state, the reference model is obtained by discretizing equation (1):

[0090]

[0091] Where: u d (k) and u q (k), i d (k), i q (k), i wd (k), i wq (k) are u d 、u q 、i d 、i q 、i wd 、i wq The sample value at time k.

[0092] The stator resistance and stator inductance in equation (29) are expressed as estimated values ​​to obtain an adjustable model:

[0093]

[0094] in: u d (k) and u q (k) is the estimated value.

[0095] The input-output function relationship of the LMS algorithm is:

[0096]

[0097] Where: X i is the input of the algorithm; W i is the weight coefficient of each input; O(W i , X i ) is the output of the algorithm.

[0098] LMS updates the weight W based on the gradient descent theory. i , and its weight adjustment rule is:

[0099]

[0100] Where: η is the convergence factor; d(k) is the actual system output.

[0101] In order to ensure good convergence of the identification results during the weight adjustment process, the convergence factor should satisfy:

[0102] 0<2ηX 2 <1 (33)

[0103] According to equations (29), (30) and (32), the adaptive rates of stator inductance and stator resistance are obtained as follows:

[0104] (1) Stator inductance adaptation rate

[0105]

[0106] (2) Stator resistance adaptation rate

[0107]

[0108] The electric aircraft controller efficiency optimization method based on the improved loss model provided by the present invention first builds a loss model of PMSM, and obtains the optimal stator current when the motor loss is the smallest through the loss model; secondly, a multi-parameter online identification strategy is designed based on the model reference adaptive system (MRAS) to ensure the stable identification of parameters; finally, the parameter identification is added to the LMC to effectively reduce the operating loss of the motor. In addition, in order to solve the problem of lack of rank in multi-parameter online identification, the present invention is based on MRAS, and designs two adaptive rates according to Popov's superstability theorem and the least mean square error (LMS) algorithm, which effectively solves the non-convergence of parameter identification results caused by the mutual influence between parameters.

[0109] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0110] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0111] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0112] Figure 1A schematic diagram of a PMSM mathematical model considering iron loss provided in an embodiment disclosed in the present invention;

[0113] Figure 2 A schematic diagram of the MRAS identification parameter principle provided in the disclosed embodiment of the present invention;

[0114] Figure 3 A schematic diagram of the principle of the identification system provided in the embodiment disclosed in the present invention;

[0115] Figure 4 A schematic flow chart of an electric aircraft controller efficiency optimization method based on an improved loss model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0116] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present invention. Instead, they are merely examples of systems consistent with some aspects of the present invention as detailed in the appended claims.

[0117] In order to solve the problem that the permanent magnet synchronous motor of electric aircraft is prone to parameter mismatch when using the traditional LMC strategy, which leads to increased motor operating losses, causes system overheating, and endangers the operating safety of the aircraft, this implementation scheme provides an electric aircraft controller efficiency optimization method based on an improved loss model. By combining the motor parameter identification strategy with the LMC strategy, the influence of the loss model parameter mismatch on the motor operating losses is reduced.

[0118] The specific steps include:

[0119] S1: Establish a permanent magnet synchronous motor loss model, with the goal of minimizing the motor iron loss resistance and stator resistance loss, and obtain the optimal stator current when the motor loss is minimized;

[0120] In the electric propulsion system of an aircraft, the output efficiency of the motor is particularly important. If the motor loss is too high, the aircraft may stop due to motor overheating, which will endanger the safety of the aircraft. In order to reduce the motor loss, the motor is first mathematically modeled and the loss composition is analyzed. The PMSM equivalent circuit considering the iron loss resistance is as follows: Figure 1 shown.

[0121] Figure 1 Medium: u d 、u q are the d-axis and q-axis stator voltage components respectively; i d 、i q are the current components of the stator current under the d and q axes respectively; i wd 、i wqare the active components of the stator current on the d and q axes respectively; i fd 、i fq are the reactive components of the stator current on the d and q axes respectively; R s is the stator resistance; R f is the iron loss resistance; L is the inductance component of the stator inductance under the d and q axes (L for surface-mounted permanent magnet synchronous motors d =L q =L);ω e is the motor electrical angular velocity; ψ f is the permanent magnet flux.

[0122] From the above figure, the motor voltage equation under the d and q coordinate axes is:

[0123]

[0124] Where: u d 、u q are the d-axis and q-axis stator voltage components respectively; i d 、i q are the current components of the stator current under the d and q axes respectively; i wd 、i wq are the active components of the stator current on the d and q axes respectively; i fd 、i fq are the reactive components of the stator current on the d and q axes respectively; R s is the stator resistance; R f is the iron loss resistance; L is the inductance component of the stator inductance under the d and q axes (L for surface-mounted permanent magnet synchronous motors d =L q =L);ω e is the motor electrical angular velocity; ψ f is the permanent magnet flux.

[0125] Kirchhoff's law gives the current equation:

[0126]

[0127] Electromagnetic torque equation T e It can be expressed as:

[0128] T e =n p i wq ψ f (3)

[0129] Where: n p is the number of motor pole pairs.

[0130] When the motor is in steady state, the iron loss current under the d and q axes can be expressed as:

[0131]

[0132] Depend on Figure 1 It can be seen that the motor copper loss P cu and iron loss P f It can be expressed as:

[0133]

[0134] The sum of copper loss and iron loss is the total motor loss P loss :

[0135]

[0136] The loss equation is obtained from equations (3), (4) and (6):

[0137]

[0138] The problem of finding the optimal current can be transformed into the problem of finding the extreme value of the function. The optimal current i can be obtained wd-opt :

[0139]

[0140] Finally, the optimal d-axis current i can be obtained from equations (2), (3), (4) and (8): d-opt :

[0141]

[0142] From formula (9), we can see that i d-opt It is related to the motor parameters resistance, inductance and flux. During the flight of the aircraft, as the flight time increases, the heating of the motor will cause changes in parameters such as resistance, inductance and flux. In order to ensure the acquisition of the real-time optimal current i d-opt , real-time parameter identification is required. When performing motor multi-parameter identification, the identification result is inaccurate due to too many parameters. In order to achieve accurate identification, the present invention is based on MRAS and solves the equation under-rank problem by designing two different adaptive algorithms, thereby achieving accurate identification of multiple motor parameters.

[0143] Specific setting S2: Based on the model reference adaptive system, a multi-parameter online identification strategy is constructed; wherein the multi-parameter online identification strategy includes an adaptive rate based on Popov's hyperstability theorem and an adaptive rate based on the LMS algorithm;

[0144] MRAS takes the motor body as the reference model, represents the parameters to be identified in the reference model with estimated values, establishes an adjustable model, gives the same input u to the two models, takes the difference e between the two model outputs as the input of the adaptive algorithm, and adjusts the parameters to be identified in the adjustable model through the adaptive algorithm, so that the difference between the two model outputs is 0, and the parameter identification effect can be achieved. The principle of MRAS parameter identification is as follows: Figure 2 shown.

[0145] First, we construct an adjustable model and a reference model based on Popov's superstability theorem:

[0146] From (1) and (2), we can get the reference model:

[0147]

[0148] Rewrite formula (10) as:

[0149]

[0150] Among them: i=[i wd i wq T ; I=i d i q T ;

[0151] The adjustable model is obtained by expressing the iron loss resistance and flux linkage in equation (11) with estimated values:

[0152]

[0153] in: is the estimated value of iron loss resistance; is the estimated value of magnetic linkage; for i wd An estimated value of for i wq The estimated value of .

[0154] From equations (11) and (12), we can get the error state equation:

[0155]

[0156] in:

[0157] make From formula (13), we can get:

[0158]

[0159] Formula (14) can be regarded as a nonlinear feedback system, which consists of two parts: a linear forward channel and a nonlinear feedback channel. Figure 3 As shown, where: D is the matrix to be determined, and the D matrix needs to be designed later to ensure identification stability; V is the system output, V=De.

[0160] Adaptive rate based on Popov's hyperstability theorem:

[0161] The two conditions for stable identification of system parameters using Popov's superstability theorem are:

[0162] (1) The transfer function matrix of the linear forward channel H(s) = D(sI-A) -1 Strictly positive.

[0163] (2) The nonlinear feedback channel satisfies the inequality:

[0164]

[0165] in: is any finite positive number.

[0166] For condition (1), the positive reality lemma is used to ensure that the transfer function is strictly positive definite. The general linear time-invariant system is:

[0167]

[0168] Among them: A1 is the system matrix; B1 is the input matrix; C1 is the output matrix; J is the direct transfer matrix; x is the state vector; u is the input vector; y is the output vector.

[0169] Assuming that system (16) is controllable and observable, its transfer function H(s) is:

[0170] H(s)=J+C1(sI-A1) -1 B1 (17)

[0171] If equation (17) is a strictly positive definite transfer matrix, the conditions that need to be met are that there are positive definite matrices P, Q, and non-singular matrices K, L that satisfy the following conditions:

[0172]

[0173] Among them: A1 T is the transposed matrix of A1; B1 T is the transposed matrix of B1; L T is the transposed matrix of L; J T is the transposed matrix of J; K T is the transposed matrix of K.

[0174] According to condition (1) and formula (18), we can get:

[0175]

[0176] Assumptions From formula (19), we can get:

[0177]

[0178] Where: k is a constant.

[0179] From equation (20), we can see that if we want to ensure that the matrices Q and P are positive definite, k>0 is required. Here, k=1 is taken, and the matrices P and D are unit matrices. Therefore, when D is the unit matrix, condition (1) is satisfied, and the transfer function matrix of the linear forward channel is H(s)=D(sI-A) -1 Strictly positive.

[0180] For condition (2), we need to design an adaptive rule first, substitute it into the inequality and solve the inequality inversely. Expand the matrix W in condition (2) to get:

[0181]

[0182] For the convenience of proof, equation (21) is decomposed into:

[0183]

[0184] in: eta(0,t1)=eta1(0,t1)+eta2(0,t1);

[0185] When designing the adaptive rate according to the Popov stability theorem, it is generally designed in the form of proportional integration. The adaptive rate of iron loss resistance and permanent magnet flux is:

[0186]

[0187] Among them: F1(τ), F2(τ), G1(τ), G2(τ) are functions of the motor's current, voltage and electrical angular velocity.

[0188] First, the adaptive rate of the iron loss resistor is designed. From equations (22) and (23), we can get:

[0189]

[0190] Formula (24) is divided into two parts:

[0191]

[0192] in: η1(0,t1)=η11 (0,t1)+η 12 (0,t1).

[0193] make:

[0194]

[0195] Where: k i and k p are proportional constant and integral constant greater than 0 respectively.

[0196] It can be proved from equations (25) and (26) that

[0197]

[0198] Similarly, it can be proved Therefore, from equations (23) and (27), the adaptive rate of permanent magnet flux and iron loss resistance can be obtained as follows:

[0199]

[0200] Constructing adjustable models and reference models based on LMS algorithm:

[0201] When the motor is running in steady state, the reference model is obtained by discretizing equation (1):

[0202]

[0203] Where: u d (k) and u q (k), i d (k), i q (k), i wd (k), i wq (k) are u d 、u q 、i d 、i q 、i wd 、i wq The sample value at time k.

[0204] The stator resistance and stator inductance in equation (29) are expressed as estimated values ​​to obtain an adjustable model:

[0205]

[0206] in: u d (k) and u q (k) is the estimated value.

[0207] S3: Incorporate a multi-parameter online identification strategy into the permanent magnet synchronous motor loss model control to suppress loss model parameter mismatch and optimize the efficiency of the electric aircraft controller.

[0208] LMS is a commonly used optimization algorithm that adjusts the weights of the input signal to make the estimated value close to the actual output value of the system, thereby solving the optimal parameters in the parameter estimation problem.

[0209] The input-output function relationship of the LMS algorithm is:

[0210]

[0211] Where: X i is the input of the algorithm; W i is the weight coefficient of each input; O(W i , X i ) is the output of the algorithm.

[0212] LMS updates the weight W based on the gradient descent theory. i , and its weight adjustment rule is:

[0213]

[0214] Where: η is the convergence factor, d(k) is the actual system output.

[0215] In order to ensure good convergence of the identification results during the weight adjustment process, the convergence factor should satisfy:

[0216] 0<2ηX 2 <1 (33)

[0217] According to equations (29), (30) and (32), the adaptive rates of stator inductance and stator resistance are obtained as follows:

[0218] (1) Stator inductance adaptation rate

[0219]

[0220] (2) Stator resistance adaptation rate

[0221]

[0222] In summary, this implementation plan firstly establishes the loss model of the motor, and aims to obtain the optimal stator current with the minimum motor iron loss resistance and stator resistance loss; secondly, in view of the parameter mismatch problem in the loss model, a multi-parameter online identification strategy is designed based on MRAS; finally, the parameter identification strategy is combined with LMC to suppress the influence of the loss model parameter mismatch on the motor loss. Figure 4 shown.

[0223] This implementation plan optimizes the energy efficiency of the electric aircraft's electric controller by building a mathematical model of the electric aircraft's permanent magnet synchronous motor and applying it to the electric aircraft. The system energy consumption calculation results are highly consistent with the test results, effectively reducing the motor's operating losses and the heating of the controller and motor, and increasing the aircraft's flight time.

[0224] Those skilled in the art will readily appreciate other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art that are not disclosed by the present invention. The specification and examples are to be considered exemplary only, and the true scope and spirit of the present invention are indicated by the claims.

Claims

1. An efficiency optimization method for electric aircraft controller based on an improved loss model, characterized in that: include: S1: Establish a permanent magnet synchronous motor loss model, with the goal of minimizing the motor iron loss resistance and stator resistance loss, and obtain the optimal stator current when the motor loss is minimized; S2: Building a multi-parameter online identification strategy based on the model reference adaptive system; The multi-parameter online identification strategy includes: Adaptive rate based on Popov's hyperstability theorem: In the formula, is the estimated value of iron loss resistance; is the estimated value of magnetic linkage; for i wd An estimated value of for i wq The estimated value of; F1(τ), F2(τ), G1(τ), G2(τ) are functions of the motor’s current, voltage and electrical angular velocity; k i and kp are proportional constant and integral constant respectively greater than 0; i d 、i q are the current components of the stator current under the d and q axes respectively; i wd 、i wq are the active components of the stator current on the d and q axes respectively; ω e is the motor electrical angular velocity; Adaptation rate based on LMS algorithm: (1) Stator inductance adaptation rate (2) Stator resistance adaptation rate In the formula, u d (k) and u q (k), i d (k), i q (k), i wd (k), i wq (k) are u d 、u q 、i d 、i q 、i wd 、i wq The sample value at time k; u d (k) and u q (k) an estimated value; η is the convergence factor; i wq is the active component of the stator current on the q axis; i q is the current component of the stator current under the q axis; S3: Incorporate a multi-parameter online identification strategy into the permanent magnet synchronous motor loss model control to suppress loss model parameter mismatch and optimize the efficiency of the electric aircraft controller.

2. The electric aircraft controller efficiency optimization method based on the improved loss model according to claim 1 is characterized in that: In S1, a permanent magnet synchronous motor loss model considering iron loss is established, and the optimal stator current with the minimum motor loss is obtained according to the extreme value theory: By considering the permanent magnet synchronous motor loss model of iron loss, the motor voltage equation under the d and q coordinate axes is obtained as follows: Where: u d 、u q are the d-axis and q-axis stator voltage components respectively; i d 、i q are the current components of the stator current under the d and q axes respectively; i wd 、i wq are the active components of the stator current on the d and q axes respectively; i fd 、i fq are the reactive components of the stator current on the d and q axes respectively; R s is the stator resistance; R f is the iron loss resistance; L is the inductance component of the stator inductance under the d and q axes. The surface-mounted permanent magnet synchronous motor L d =L q =L;ω e is the motor electrical angular velocity; ψ f is the permanent magnet flux; Kirchhoff's law gives the current equation: Electromagnetic torque equation T e It can be expressed as: T e =n p i wq ψ f (3) Where: n p is the number of motor pole pairs; When the motor is in steady state, the iron loss current under the d and q axes can be expressed as: Motor copper loss P cu and iron loss P f It can be expressed as: The sum of copper loss and iron loss is the total motor loss P loss : The loss equation is obtained from equations (3), (4) and (6): The problem of finding the optimal current can be transformed into the problem of finding the extreme value of the function. The optimal current i can be obtained wd-opt : Finally, the optimal d-axis current i can be obtained from equations (2), (3), (4) and (8): d-opt :

3. The electric aircraft controller efficiency optimization method based on the improved loss model according to claim 1 is characterized in that: The adaptive rate based on Popov's hyperstability theorem is obtained according to the following steps: According to the theory of model reference adaptive system, an adjustable model and a reference model based on Popov's hyperstability theorem are established; The reference model is: From (1) and (2), we can get the reference model: Rewrite formula (10) as: Where: i = [i wd i wq ] T ; I = [i d i q ] T ; The adjustable model is obtained by expressing the iron loss resistance and flux linkage in equation (11) with estimated values: in: R f is the estimated value of iron loss resistance; is the estimated value of magnetic linkage; for i wd An estimated value of for i wq An estimated value of From equations (11) and (12), we can get the error state equation: in: make From formula (13), we can get: Formula (14) is a nonlinear feedback system, which consists of two parts: a linear forward channel and a nonlinear feedback channel; According to Popov's superstability theorem and its adjustable model and reference model, the two conditions for accurately identifying motor parameters are: (1) The transfer function matrix of the linear forward channel H(s) = D(sI-A) -1 Strictly positive: (2) The nonlinear feedback channel satisfies the inequality: in: γ0 is any finite positive number; For condition (1), the positive reality lemma is used to ensure that the transfer function is strictly positive definite. The general linear time-invariant system is: Among them: A1 is the system matrix; B1 is the input matrix; C1 is the output matrix; J is the direct transfer matrix; x is the state vector; u is the input vector; y is the output vector; Assuming that system (16) is controllable and observable, its transfer function H(s) is: H(s)=J+C1(sI-A1) -1 B1 (17) If equation (17) is a strictly positive definite transfer matrix, the conditions that need to be met are that there are positive definite matrices P, Q, and non-singular matrices K, L that satisfy the following conditions: Among them: A1 T is the transposed matrix of A1; B1 T is the transposed matrix of B1; L T is the transposed matrix of L; J T is the transposed matrix of J; K T is the transposed matrix of K; According to condition (1) and formula (18), we can get: Assumptions From formula (19), we can get: Where: k is a constant; From equation (20), we can see that if we want to ensure that the matrices Q and P are positive definite, k>0 is required. Here, k=1 is taken, and the matrices P and D are unit matrices. Therefore, when D is the unit matrix, condition (1) is satisfied, and the transfer function matrix of the linear forward channel is H(s)=D(sI-A) -1 Strictly positive; For condition (2), we need to design an adaptive rule first, substitute it into the inequality and solve the inequality in reverse. Expand the matrix W in condition (2) to get: For the convenience of proof, equation (20) is decomposed into: in: eta(0,t1)=eta1(0,t1)+eta2(0,t1); The adaptive rate is designed according to Popov stability theorem, using proportional integral form. The adaptive rate of iron loss resistance and permanent magnet flux linkage is: Among them: F1(τ), F2(τ), G1(τ), G2(τ) are functions of the motor’s current, voltage and electrical angular velocity; First, the adaptive rate of the iron loss resistor is designed. From equations (22) and (23), we can get: Formula (24) is divided into two parts: Among them: η1(0,t1)=η 11 (0,t1)+η 12 (0,t1); make: Where: k i and k p are proportional constant and integral constant greater than 0 respectively; It can be proved from equations (25) and (26) that Similarly, it can be proved Therefore, from equation (27) and equation (23), the adaptive rate of permanent magnet flux linkage and iron loss resistance can be obtained as follows:

4. The electric aircraft controller efficiency optimization method based on the improved loss model according to claim 1 is characterized in that: The adaptive rate based on the LMS algorithm is obtained according to the following steps: According to the model reference adaptive system theory, an adjustable model and a reference model based on the LMS algorithm are established; When the motor is running in steady state, the reference model is obtained by discretizing equation (1): Where: u d (k) and u q (k), i d (k), i q (k), i wd (k), i wq (k) are u d 、u q 、i d 、i q 、i wd 、i wq The sample value at time k; The stator resistance and stator inductance in equation (29) are expressed as estimated values ​​to obtain an adjustable model: in: u d (k) and u q (k) an estimated value; The input-output function relationship of the LMS algorithm is: Where: X i is the input of the algorithm; W i is the weight coefficient of each input; O(W i , X i ) is the output of the algorithm; LMS updates the weight W based on the gradient descent theory. i , and its weight adjustment rule is: Where: η is the convergence factor; d(k) is the actual system output; In order to ensure good convergence of the identification results during the weight adjustment process, the convergence factor should satisfy: 0<2ηX 2 <1 (33) According to equations (29), (30) and (32), the adaptive rates of stator inductance and stator resistance are obtained as follows: (1) Stator inductance adaptation rate (2) Stator resistance adaptation rate

Images