A parameter identification method for permanent magnet synchronous motor based on improved model reference adaptive system
By introducing first-order ADRC adaptive law in the MRAS system, the problem of poor anti-interference performance of traditional PI adaptive law is solved, effective filtering of high-frequency noise is achieved, and the accuracy of parameter identification of permanent magnet synchronous motor is significantly improved. and stability.
Patent Information
- Application Number
- CN202310975922.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-08-04
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2043-08-04
AI Technical Summary
The traditional model reference adaptive system (MRAS) based on PI adaptive law has poor immunity performance in the high frequency band and cannot effectively filter out the high frequency noise caused by the sensor, resulting in the identification results of permanent magnet synchronous motors being susceptible to noise disturbance.
Using first-order autoimmune controller (ADRC) as the adaptive law of MRAS, a reasonable ADRC adaptive law is designed to suppress high-frequency noise. The ADRC is inputted through the difference between the reference model and the adjustable model, and the output is feedbacked to the Adjust the model to form an improved MRAS system.
The jitter amplitude of the online parameter identification result is significantly reduced, the noise anti-noise performance of the MRAS system is improved, and the accuracy of parameter identification is significantly improved.
Smart Images

Figure CN117200621B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of motor control, and in particular relates to a permanent magnet synchronous motor parameter identification method based on an improved model reference adaptive system. Background Art
[0002] Due to the influence of various factors such as its own structure, power density limitation and various complex motor operating environments, the electromagnetic parameters of permanent magnet synchronous motors will change with the operation of the equipment, which is very likely to lead to a decrease in the motor vector control performance. Therefore, online parameter identification of permanent magnet synchronous motors has important scientific research and practical application value in high-performance motor control and fault warning.
[0003] Among many online parameter identification algorithms, the model reference adaptive method (MRAS) has the advantages of simple design process, small amount of calculation, high accuracy, and easy practical implementation. It is widely used in motor parameter identification and sensorless control. MRAS uses the reference model and the adjustable model to make a difference, inputs the error into a suitable adaptive mechanism, and feeds the output of the adaptive mechanism back to the parameter to be identified in the adjustable model, thereby realizing online parameter identification. However, in the application process of MRAS, since the amplitude of the Bode diagram of the classic PI adaptive law is zero in the high frequency band, MRAS cannot filter out the high-frequency noise that may be brought by the sensor, which makes the parameter identification results of MRAS susceptible to noise disturbance. Summary of the invention
[0004] In view of the problem that the traditional MRAS based on the PI adaptive law has poor anti-disturbance performance, the present invention proposes a permanent magnet synchronous motor parameter identification method based on an improved model reference adaptive system, and uses an active disturbance rejection controller (ADRC) as the adaptive law of MARS to suppress high-frequency noise. The design is reasonable, solves the shortcomings of the existing technology, and has good results.
[0005] A method for identifying parameters of a permanent magnet synchronous motor based on an improved model reference adaptive system comprises the following steps:
[0006] Step 1: Establish a state space equation model of the surface mounted permanent magnet synchronous motor as a reference model of MRAS;
[0007] Step 2: Establish an MRAS adjustable model;
[0008] Step 3: Establish the MRAS error state space equation;
[0009] Step 4: Design a first-order ADRC adaptive law.
[0010] Furthermore, in step 1, the state space equation mathematical model of the surface mounted permanent magnet synchronous motor with current as the state variable is:
[0011]
[0012] in, u=[u d u q ] T ; i = [i d i q ] T ;u d 、u q are the d-axis and q-axis components of the stator voltage respectively; i d 、i q are the d-axis and q-axis components of the stator current respectively; R s is the stator resistance; L is the stator inductance; w e is the rotor electrical angular velocity; ψ f is the flux linkage of the rotor permanent magnet; let equation (1) be the reference model of MRAS.
[0013] Furthermore, in step 2, based on formula (1), an adjustable model of MRAS is established:
[0014]
[0015] in, To identify inductance; To identify magnetic links.
[0016] Furthermore, in step 3, the reference model formula (2) is subtracted from the adjustable model formula (3) to obtain the error state space equation, which is expressed as:
[0017]
[0018] in,
[0019] According to Popov's superstability theory, in order to make the system of formula (3) stable, it is necessary to satisfy the Popov integral inequality, which can be expressed as:
[0020]
[0021] Among them, e T , w is the output and feedback loop of the error state equation; t1 is any time, r0 is any finite constant that does not depend on time t; Substituting equation (3) into equation (4) yields:
[0022]
[0023] in, r1 and r2 are arbitrary finite constants independent of time t.
[0024] Furthermore, in step 4, the first-order ADRC is introduced into equation (5), and the first-order ADRC is expressed as:
[0025]
[0026] Among them, β1 and β2 are the gains of ESO, β1 and β2 are twice and square of bandwidth w0 respectively; b0 is control gain; A is given quantity; B is feedback quantity; C is output quantity; the fal function in formula (6) is a nonlinear function, and its formula is:
[0027]
[0028] Among them, δ is the linear interval value; a is the power coefficient; e a is the observation error; z1 is the tracking of the feedback signal by ESO;
[0029] Let the given value A be equal to 0, and the feedback value B be:
[0030]
[0031] Among them, B1 is the component of the inductance adaptation law; B2 is the component of the flux adaptation rate;
[0032] Combining equations (5), (6), (7) and (8), we can obtain the new Popov integral inequality:
[0033]
[0034] in, are the estimated values of b and c respectively;
[0035] Formula (9) is the Popov integral inequality for online identification of inductance and flux linkage under the ADRC adaptive law. When the parameter values satisfy formula (9), the MRAS system is stable.
[0036] Beneficial technical effects brought by the present invention:
[0037] The present invention provides a permanent magnet synchronous motor parameter identification method based on an improved model reference adaptive system. In view of the problem that MRAS is easily affected by high-frequency noise, the traditional PI adaptive law is improved to the ADRC adaptive law. First, a suitable reference model is selected, and then an adjustable model is designed according to the reference model. Finally, the difference between the reference model and the adjustable model is sent to the first-order ADRC adaptive law and the output is fed back to the adjustable model to form a complete MRAS system. Compared with the traditional PI adaptive law, the ADRC adaptive law can effectively filter out noise and significantly reduce the jitter amplitude of the online parameter identification results. BRIEF DESCRIPTION OF THE DRAWINGS
[0038] Figure 1 This is a block diagram of the model reference adaptive system in the present invention;
[0039] Figure 2 It is the structure diagram of the error state space equation system in the present invention;
[0040] Figure 3 It is a first-order ADRC structure diagram in the present invention;
[0041] Figure 4 Bode plots of first-order ADRC and PI comparison for flux linkage identification;
[0042] Figure 5 This is the simulation result diagram of PI adaptive law flux identification;
[0043] Figure 6 This is the simulation result diagram of ADRC adaptive law flux identification;
[0044] Figure 7 This is the simulation result diagram of PI adaptive law inductance identification;
[0045] Figure 8 This is the simulation result diagram of ADRC adaptive law inductance identification; DETAILED DESCRIPTION
[0046] The specific implementation of the present invention is further described below in conjunction with specific embodiments:
[0047] A method for identifying parameters of a permanent magnet synchronous motor based on an improved model reference adaptive system comprises the following steps:
[0048] Step 1: Establish a state space equation model of the surface mounted permanent magnet synchronous motor as a reference model of MRAS;
[0049] The stator voltage equation of the surface-mounted permanent magnet synchronous motor under the dq axis is:
[0050]
[0051] Among them, u d 、u q are the d-axis and q-axis components of the stator voltage respectively; i d 、i q are the d-axis and q-axis components of the stator current respectively; R s is the stator resistance; L is the stator inductance; w e is the rotor electrical angular velocity; ψ f is the magnetic flux of the rotor permanent magnet;
[0052] Rewrite equation (1) into a state space equation with current as the state variable:
[0053]
[0054] in, u=[u d u q ] T ; i = [i d i q ] T ; Formula (2) is a state space equation set with voltage as input and current as output. Formula (2) is regarded as the reference model of MRAS;
[0055] Step 2: Establish an MRAS adjustable model;
[0056] The overall structure of MRAS is as follows Figure 1 As shown in the figure, based on the overall structure of MRAS and equation (2), the adjustable model of MRAS is established by taking online identification of inductance and flux linkage as an example:
[0057]
[0058] in, To identify inductance; To identify magnetic links;
[0059] Step 3: Establish the MRAS error state space equation;
[0060] Subtracting the reference model (2) from the adjustable model (3) yields the error state space equation, expressed as:
[0061]
[0062] in,
[0063] Since the values of the parameters to be identified in ΔA, ΔB, and ΔC are functions of the stator voltage and current, and the stator voltage and current are functions of time, this makes w have nonlinear and time-varying characteristics. The overall structure of formula (4) is as follows: Figure 2 As shown. Figure 2 In order to make the system stable and make e converge to zero, a suitable parameter adaptation law must be designed. According to Popov's superstability theory, Figure 2 For the system to be stable, it must satisfy the Popov integral inequality, which can be expressed as:
[0064]
[0065] Among them, e T , w is the output and feedback loop of the error state equation; t1 is any time, r0 is any finite constant that does not depend on time t; Substituting equation (4) into equation (5) yields:
[0066]
[0067] in, r1 and r2 are arbitrary finite constants that are independent of time t. Equation (5) can be used to identify inductance and flux linkage online.
[0068] Step 4: Design a first-order ADRC adaptive law;
[0069] The first-order ADRC is used as the adaptive law of MRAS, and its structure is as follows: Figure 3 As shown, the first-order ADRC is introduced into equation (6), and the first-order ADRC is expressed as:
[0070]
[0071] Among them, β1 and β2 are the gains of ESO, β1 and β2 are twice and square of bandwidth w0 respectively; b0 is control gain; A is given quantity; B is feedback quantity; C is output quantity; the larger the bandwidth w0, the faster the response speed of first-order ADRC, but the weaker the ability to resist high-frequency noise;
[0072] The fal function in formula (7) is a nonlinear function, and its formula is:
[0073]
[0074] Among them, δ is the linear interval value; a is the power coefficient; e a is the observation error; z1 is the tracking of the feedback signal by ESO; the Fal function uses a piecewise function to divide the controller working range into two parts: a large error range and a small error range. The large error range is responsible for adjusting the speed, and the small error range is responsible for steady-state anti-disturbance, which effectively solves the contradiction between the adjustment speed and the anti-disturbance performance.
[0075] Let the given value A be equal to 0, and the feedback value B be:
[0076]
[0077] Among them, B1 is the component of the inductance adaptive law; B2 is the component of the flux linkage adaptive law;
[0078] Combining equations (6), (7), (8) and (9), we can obtain the new Popov integral inequality:
[0079]
[0080] in, are the estimated values of b and c respectively;
[0081] Formula (10) is the Popov integral inequality for online identification of inductance and flux linkage under the ADRC adaptive law.
[0082] Taking the inductor online parameter identification as an example, the stability analysis of the ADRC adaptive law is carried out. a >δ, taking the derivative of C in (7) and simplifying it, we finally get:
[0083]
[0084] Since the feedback amount B may have a negative value, which may cause the inequality to be not satisfied, we can select larger β1 and β2 to make is greater than b0, which can make e a >δ monotonically increasing. Similarly, in e a <-δ, taking the derivative of C and simplifying it finally gives:
[0085]
[0086] make Much larger than b0, which can make e a <-δ Monotonically decreases. So ADRC is made by |e a |>δ interval to |e a |≤δ interval converges.
[0087] In the interval |e a When |≤δ, the ADRC formula is:
[0088]
[0089] Where b1 is b2 is The transfer function between the feedback quantity B and the controller output C is calculated from (13):
[0090]
[0091] Then, by performing the inverse Lapland transformation on (14), we can obtain:
[0092]
[0093] Substituting (15) back into (10) and splitting it into two equations, we can obtain:
[0094]
[0095] Among them, r 11 、r 12 is an arbitrary finite constant independent of time. It can be seen that is greater than 0 and less than 1, so when formula (16) When it is greater than zero, the inequality holds. Therefore, it can be proved that when the first-order ADRC converges to the linear interval |e a |≤δ, the designed adaptive law is stable. The derivation of the adaptive law for flux linkage identification and the proof of its stability are similar to those for inductance and are also omitted here.
[0096] The frequency domain performance of the first-order ADRC adaptive law is analyzed, and its linear region Bode diagram is as follows: Figure 4 As shown. The solid line is the first-order ADRC frequency domain response; the dotted line is the PI frequency domain response. In the figure, the amplitude and phase of PI in the high frequency band are both zero, indicating that PI cannot filter out high-frequency noise, while the amplitude of ADRC in the high frequency band is always less than zero and becomes smaller and smaller. It can be seen that the ADRC adaptive law has a low-pass characteristic in the linear range and can filter out high-frequency noise, that is, it has stronger anti-interference performance than PI.
[0097] Figure 5 , Figure 6 These are the simulation diagrams comparing the online identification effects of the classic PI adaptive law and the first-order ADRC adaptive law in a noisy environment. Figure 7 , Figure 8 The following are simulation diagrams of the online identification comparison of inductance under noise environment using the classic PI adaptive law and the first-order ADRC adaptive law. It can be seen that when the sensor has high-frequency noise, the jitter amplitude of the identification result of the classic PI adaptive law is larger due to the disturbance, while the jitter amplitude of the identification result of the first-order ADRC adaptive law is significantly smaller. In summary, the first-order ADRC adaptive law designed by the present invention has better anti-noise and anti-interference performance than the traditional PI adaptive law.
[0098] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the protection scope of the present invention.
Claims
1. A permanent magnet synchronous motor parameter identification method based on an improved model reference adaptive system, characterized in that: The following steps are involved: Step 1: Establish a state space equation model of the surface mounted permanent magnet synchronous motor as a reference model of MRAS; In the step 1, the state space equation mathematical model of the surface mounted permanent magnet synchronous motor with current as the state variable is: in, u=[u d u q ] T ; i = [i d i q ] T ;u d 、u q are the d-axis and q-axis components of the stator voltage respectively; i d 、i q are the d-axis and q-axis components of the stator current respectively; R s is the stator resistance; L is the stator inductance; w e is the rotor electrical angular velocity; ψ f is the flux linkage of the rotor permanent magnet; let equation (1) be the reference model of MRAS; Step 2: Establish an MRAS adjustable model; In the step 2, based on formula (1), an adjustable model of MRAS is established: in, To identify inductance; To identify magnetic links; Step 3: Establish the MRAS error state space equation; In step 3, the reference model is subtracted from the adjustable model to obtain the error state space equation, which is expressed as: in, According to Popov's superstability theory, in order to make the system of formula (3) stable, it is necessary to satisfy the Popov integral inequality, which can be expressed as: Among them, e T , w is the output and feedback loop of the error state equation; t1 is any time, r0 is any finite constant that does not depend on time t; Substituting equation (3) into equation (4) yields: in, r1 and r2 are arbitrary finite constants that do not depend on time t; Step 4: Design a first-order ADRC adaptive law; In step 4, the first-order ADRC is introduced into equation (5), and the first-order ADRC is expressed as: Among them, β1 and β2 are the gains of ESO, β1 and β2 are twice and square of bandwidth w0 respectively; b0 is control gain; A is given quantity; B is feedback quantity; C is output quantity; the fal function in formula (6) is a nonlinear function, and its formula is: Among them, δ is the linear interval value; a is the power coefficient; e a is the observation error; z1 is the tracking of the feedback signal by ESO; Let the given value A be equal to 0, and the feedback value B be: Among them, B1 is the component of the inductance adaptive law; B2 is the component of the flux linkage adaptive law; Combining equations (5), (6), (7) and (8), we can obtain the new Popov integral inequality: in, are the estimated values of b and c respectively; Formula (9) is the Popov integral inequality for online identification of inductance and flux linkage under the ADRC adaptive law. When the parameter values satisfy formula (9), the MRAS system is stable.
Citation Information
Patent Citations
Permanent magnet synchronous motor speed loop control parameter setting method based on fractional order sliding mode controller
CN111711396A
Asynchronous motor two-degree-of-freedom speed regulation method based on novel active-disturbance-rejection algorithm
CN116094390A
Cited By
Permanent magnet synchronous motor resistance on-line identification method and system
CN119496427A
