Motor position estimation and control method and system for suppressing nonlinearity of inverter

Abstract

The invention discloses a motor position estimation and control method and system for suppressing nonlinearity of an inverter. A rotor position estimation error caused by a dead zone effect of the inverter is effectively suppressed through combination of positive and negative sequence alternating high-frequency voltage injection and a composite filtering phase-locked loop based on weighted fusion. For the problem of direct current bias error of an estimation position caused by fundamental wave disturbance caused by a dead zone effect, position information is extracted from high-frequency current by alternately injecting positive and negative sequence voltage signals into an A phase, a B phase and a C phase and a filter-free signal processing method, and the fundamental wave voltage disturbance is eliminated; secondly, in order to suppress sixth harmonics caused by the dead zone effect, a composite filtering phase-locked loop structure is designed, one phase-locked loop adopts an iteration method to construct a target function to filter out the sixth harmonics, the rotor position and the rotating speed of the double phase-locked loops are adjusted by reasonably setting a weight coefficient, optimal balance is achieved between the filtering capacity and the bandwidth of the phase-locked loops, and the filtering capacity of the double phase-locked loops is improved. According to the control method, the problems of direct current bias errors and sixth harmonic disturbance are solved at the same time.

Description

Technical Field

[0001] The present invention belongs to the field of motor control, and in particular relates to a motor position estimation and control method and system for suppressing inverter nonlinearity. Background Art

[0002] As motor control systems evolve toward higher efficiency and lower cost, sensorless control technology has attracted significant attention due to its cost and reliability advantages, especially due to its elimination of mechanical sensors. This method accurately estimates rotor position and speed by analyzing motor terminal voltage and current signals in real time, combining motor mathematical models with advanced estimation algorithms. It has been widely used in industrial drives, home appliances, and new energy vehicles. However, further optimization is needed in low-speed position estimation accuracy and parameter robustness.

[0003] Traditional high-frequency square-wave voltage injection methods inject voltage into a rotating dq coordinate system. Dead-zone effects introduce fundamental and sixth-order harmonic voltage disturbances, leading to DC offset and harmonic distortion in position estimation. However, existing dead-zone compensation methods (such as lookup tables and disturbance observers) are ineffective under light load or zero-clamp conditions. Furthermore, the limited bandwidth of traditional phase-locked loops (PLLs) makes it difficult to effectively filter out the sixth-order harmonic at low speeds. Consequently, existing technologies still fail to simultaneously address both DC offset errors and sixth-order harmonic disturbances. Summary of the Invention

[0004] This invention addresses the technical problem that existing technologies fail to simultaneously address both DC bias error and sixth harmonic disturbances, thereby providing a motor position estimation and control method and system that suppresses inverter nonlinearity. Differentiating from traditional control processes, the present invention explores alternative control approaches, sequentially injecting alternating positive and negative sequence high-frequency voltage signals into the three phases A, B, C, and C. This approach eliminates fundamental voltage disturbances through differential current processing, and employs a composite filtered phase-locked loop (PLL) to weightedly fuse the position and speed outputs, balancing filtering capability with PLL bandwidth to achieve closed-loop motor control while simultaneously addressing both DC bias error and sixth harmonic disturbances.

[0005] To this end, the present invention provides the following technical solutions:

[0006] In one aspect, the present invention provides a motor position estimation method for suppressing inverter nonlinearity, comprising:

[0007] The three phases A, B, and C of the motor are injected with positive and negative sequence alternating high-frequency voltage signals of equal amplitude in sequence. That is, the three corresponding injection signals of A, B, and C are recorded as stage one, stage two, and stage three respectively. In each stage, positive and negative voltage signals are alternately injected into one of the three phases; for example, stage one is injected into phase A, stage two is injected into phase B, and stage three is injected into phase C.

[0008] The difference equation between the current differences after the positive and negative voltage signals are injected in each stage is calculated to eliminate the fundamental frequency disturbance voltage. The difference equations of the current differences in stages one, two, and three are then combined to obtain the total current difference equation that carries the rotor position information.

[0009] A composite filter phase-locked loop based on weighted fusion is used to demodulate the total current difference equation to obtain the rotor position and speed of the motor;

[0010] The composite filter phase-locked loop is a filter phase-locked loop branch based on an iterative algorithm added to the existing phase-locked loop branch, and the motor position estimation results of the two phase-locked loop branches are weighted to obtain the rotor position and speed of the motor.

[0011] Preferably, the process of constructing the total current difference equation carrying the rotor position information is as follows:

[0012] After injecting positive and negative sequence alternating high-frequency voltage signals of equal amplitude into the motor, the current difference equations under the injection of positive and negative voltage signals in stages one, two, and three are constructed respectively;

[0013] Subtract the current differences under the positive voltage and negative voltage signals in the same phase to obtain the difference equation of the current difference in each phase;

[0014] Perform coordinate transformation on the difference equations of stage 2 and stage 3, respectively, with the coordinate transformation angles being 120° and 240°;

[0015] The difference equation of stage one is summed with the difference equations of stages two and three after coordinate transformation to obtain the total current difference equation carrying the rotor position information.

[0016] Preferably, if the motor is a permanent magnet synchronous motor, the total current difference equation carrying the rotor position information is:

[0017]

[0018] Where, I αΔ , I βΔ is the sum of the difference equation of stage 1 corresponding to the α, β directions and the difference equations of stage 2 and stage 3 after coordinate transformation; I α1Δ ,I β1Δ are the current differences in the α and β directions after only injecting positive voltage and negative voltage into phase A; I' α2Δ ,I' β2Δ The difference between the current difference in the α and β directions after the positive voltage is injected into the B phase and the negative voltage is injected into the B phase respectively. α2Δ , I β2Δ Result after coordinate transformation; I' α3Δ ,I' β3ΔThe current difference I in the α and β directions after the positive voltage is injected into the C phase and the negative voltage is injected into the C phase respectively. α3Δ , I β3Δ Result after coordinate transformation; θ r Indicates the rotor position, custom parameter K1 = K2V h , ΔT is the switching period, the average inductance ΣL=(L dh +L qh ) / 2, average inductance difference ΔL=(L dh -L qh ) / 2,L dh and L qh are the d and q axis inductances after the high frequency voltage signal is injected, a1 ,ζ b1 ,ζ a2 ,ζ b2 ,ζ a3 ,ζ b3 , ε1, ε2 are all custom functions; Δu ahx and Δu bhx is the high-frequency disturbance voltage error in the α and β directions, and x represents 1, 2, and 3, where "1" means that positive and negative voltages are injected only into phase A, "2" means that positive and negative voltages are injected only into phase B, and "3" means that positive and negative voltages are injected only into phase C.

[0019] Preferably, the process of constructing the current difference equation when positive voltage and negative voltage signals are injected in any stage is as follows:

[0020] After injecting positive and negative sequence alternating high-frequency voltage signals of equal amplitude into the motor, the three-phase voltage equation is converted into a high-frequency voltage equation in a two-phase stationary coordinate system;

[0021] The high-frequency voltage equation in the two-phase stationary coordinate system is transformed to obtain the current difference equation when positive voltage and negative voltage signals are injected respectively.

[0022] Preferably, the filtering phase-locked loop branch based on the iterative algorithm is to filter out harmonics by subtracting the estimated sixth harmonic from the position error, wherein the function f of the position error minus the estimated sixth harmonic is e Expressed as:

[0023]

[0024] exist: Among them, θ err2 is the position error, I αΔ , I βΔ It is the sum of the difference equation of stage 1 corresponding to the α and β directions and the difference equations of stage 2 and stage 3 after coordinate transformation; is the 6th harmonic θ h The estimated value of6a , e 6b is the actual amplitude, is the estimated amplitude, is the rotor position estimate of the filtered phase-locked loop branch based on the iterative algorithm;

[0025] Based on the function f e Construct the objective function E(x), that is, convert the harmonic suppression problem into an optimization problem to solve its minimum point. The objective function E(x) is expressed as:

[0026]

[0027] Custom variables T is the matrix transpose symbol.

[0028] Preferably, the objective function E(x) is minimized by using a gradient descent method, a Gauss-Newton method or a Levenberg-Marquardt method to obtain a motor position estimation result of a filtered phase-locked loop branch based on an iterative algorithm.

[0029] Preferably, the Levenberg-Marquardt method is used to solve the minimum value of the objective function E(x) to obtain the motor position estimation result of the filtered phase-locked loop branch based on the iterative algorithm;

[0030] The parameter update rules are as follows:

[0031] x(k+1)=x(k)-H k -1 J k f e (k);

[0032] In the formula, k represents the number of iterations, x(k) is the variable x corresponding to the kth iteration, and f e (k) is the function f corresponding to the kth iteration e , J k is the Jacobian matrix, H k is the Hessian matrix, and we can infer that:

[0033]

[0034] Where λ is the damping coefficient, and the functions p(k) and q(k) satisfy:

[0035]

[0036] Where, T s is the sampling period, k p2 is the proportional gain of the filter phase-locked loop transfer function, k i2is the integral gain of the filtered phase-locked loop transfer function. N0, N1, N2, D0, D1, and D2 are all user-defined variables / constants.

[0037] In another aspect, the present invention provides a motor control method based on the above-mentioned motor position estimation method, comprising:

[0038] Estimate the rotor position and speed of the motor using the above motor position estimation method;

[0039] Integrate the estimated rotor position and speed into the motor closed-loop control instead of the position encoder;

[0040] Among them, the speed outer loop outputs the q-axis current reference value through the PI regulator according to the deviation between the given speed and the estimated rotor speed, performs inverse Park transformation and SVPWM modulation based on the estimated rotor position, and generates the drive signal of the inverter.

[0041] In a third aspect, the present invention provides a control system based on the above method, comprising:

[0042] The voltage signal injection module is used to sequentially inject positive and negative sequence alternating high-frequency voltage signals of equal amplitude into the three phases A, B, and C of the motor. That is, the three injection signals are recorded as phase one, phase two, and phase three respectively. In each phase, positive and negative voltage signals are alternately injected into one of the three phases.

[0043] A signal processing module is used to calculate the difference equation between the current differences after the positive and negative voltage signals are injected in each stage to eliminate the fundamental frequency disturbance voltage, and then combine the difference equations of the current differences in stages one, two, and three to obtain the total current difference equation that carries the rotor position information; and to demodulate the total current difference equation using a composite filter phase-locked loop based on weighted fusion to obtain the rotor position and speed of the motor;

[0044] The composite filter phase-locked loop is used for demodulating the total current difference equation by using a composite filter phase-locked loop based on weighted fusion to obtain the rotor position and speed of the motor.

[0045] In a fourth aspect, the present invention provides a motor comprising a controller and a memory;

[0046] Wherein, the controller calls the control program stored in the memory to implement: the motor position estimation method or the motor control method.

[0047] The beneficial effects are:

[0048] 1. The existing technology is a method of estimating the motor position by injecting positive and negative high-frequency voltages into the dq rotating coordinate system, but it still cannot eliminate the fundamental frequency voltage disturbance caused by the nonlinearity of the inverter. Commonly used inverter nonlinearity suppression methods (such as lookup table method and disturbance observer) are not effective under light load or current zero clamping conditions; and the bandwidth of the traditional phase-locked loop is limited, making it difficult to effectively filter out the sixth harmonic in the low-speed area. The technical solution of the present invention proposes a unique concept. By alternately injecting positive and negative voltages into the ABC three phases, the difference equation between the current difference after the positive and negative voltage signals are injected in each stage is mathematically derived to directly eliminate the fundamental frequency voltage disturbance. In addition, a composite filter phase-locked loop based on weighted fusion is proposed, which can balance the bandwidth and filtering capability even under low-speed conditions, and suppress the sixth harmonic interference caused by the nonlinearity of the inverter on the position estimation. Therefore, the technical solution of the present invention effectively solves the problems of position estimation DC bias error and sixth harmonic interference caused by the nonlinear dead zone effect of the inverter in motor control through the injection of alternating high-frequency voltage signals with positive and negative sequences and the composite orthogonal filtering phase-locked loop method. At the same time, no additional dead zone compensation circuit is required, which significantly improves the system accuracy and response speed and reduces the hardware cost. It is particularly suitable for high-precision control scenarios such as electric vehicles and household appliance motors.

[0049] 2. The technical solution of the present invention preferably adopts the Levenberg-Marquardt method for calculation of the filtered phase-locked loop branch based on the iterative algorithm because the algorithm has the advantages of both the global convergence of the gradient descent method and the fast local convergence characteristics of the Gauss-Newton method, which is more conducive to quickly and accurately solving the minimum value. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is the timing diagram for the injection of positive and negative sequence alternating high frequency voltage signal;

[0051] Figure 2 This is a block diagram of a motor closed-loop control system based on positive-negative sequence alternating high-frequency voltage signal injection according to an embodiment of the present invention;

[0052] Figure 3 This is a comparison diagram of the position observation steady-state performance of a motor running at 200 r / min under a constant load condition using an embodiment of the present invention;

[0053] Figure 4 This is a comparison diagram of the position observation dynamic performance of a motor running at 150 r / min under load disturbance using an embodiment of the present invention. DETAILED DESCRIPTION

[0054] Figure 1This is a timing diagram for injecting alternating positive and negative high-frequency voltage signals. One injection cycle in this timing diagram consists of three phases: Phase 1, Phase 2, and Phase 3. In each phase, high-frequency positive and negative voltage signals are injected alternately into the motor's three phases, A, B, and C. Each voltage signal is injected within a single pulse-width modulation cycle. Frequencies significantly above the motor's operating frequency and below the PWM carrier frequency are generally referred to as high frequencies.

[0055] Figure 2 This is a block diagram of a motor parameter identification system based on positive and negative sequence alternating high-frequency voltage signal injection according to an embodiment of the present invention. The system primarily includes three components: positive and negative sequence alternating high-frequency voltage signal injection, signal processing, and a composite filtering phase-locked loop. Taking a permanent magnet synchronous motor as an example, when the motor is running at low speed, after the high-frequency voltage signal is injected, the resistance and speed-related terms account for a small proportion and can be ignored. Therefore, the high-frequency voltage equation in a two-phase stationary coordinate system is expressed as:

[0056]

[0057] Where, the subscript "h" represents a high-frequency signal, and are the voltage and current in the α and β coordinate systems after the high-frequency voltage signal is injected, respectively, and u αh ,u βh is the voltage in the α and β directions, i αh ,i βh is the current in the direction of α, β, the average inductance ΣL=(L dh +L qh ) / 2, average inductance difference ΔL=(L dh -L qh ) / 2,L dh and L qh After injecting high-frequency voltage signal, d,q axis inductance, θ r It should be understood that the above high-frequency voltage equation is universal for motors with salient polarity; for other feasible motors, the high-frequency voltage equation is constructed according to the same technical ideas.

[0058] Further transforming equation (1) to obtain the current difference equation, expressed as:

[0059]

[0060] Where, custom parameters Δi αh , Δi βh They are the current differences in the α and β directions after the high-frequency voltage signal is injected, and ΔT is the switching period.

[0061] According to existing literature, considering the influence of inverter dead zone, for any stage, the voltage disturbance error Δu of the three phases ABC caused by the high-frequency voltage injection into the motor is A , Δu B , Δu C It can be modeled as:

[0062]

[0063] Where, the subscript "f" represents the fundamental frequency signal, ΔU represents the terminal voltage error, and i xf ,R xh ,i xh They represent the fundamental frequency current, inverter equivalent high-frequency resistance and high-frequency current in the ABC three-phase shaft system respectively, and x is A, B, and C.

[0064] After injecting positive and negative sequence alternating high-frequency voltage pulses of equal amplitude into the motor, the three-phase voltage equation and voltage disturbance error are converted into high-frequency voltage equation and voltage disturbance error in a two-phase stationary coordinate system, respectively, and recorded as stage one, stage two, and stage three. The current difference equation expression in stage one is:

[0065] Phase 1:

[0066]

[0067] Where, the voltage disturbance error Δu in the α direction is ax =Δu afx +Δu ahx and the voltage disturbance error Δu in the β direction bx =Δu bfx +Δu bhx , x represents "+" and "-", "+" and "-" represent the injected voltages are positive and negative respectively, and "1" represents the injection of positive and negative voltages only into phase A; Δu afx and Δu bfx is the fundamental frequency disturbance voltage error in the α, β directions, Δu ahx and Δu bhx is the high-frequency disturbance voltage error in the α and β directions; Δu af1+ and Δu bf1+ They are the fundamental frequency disturbance voltage errors in the α and β directions after positive voltage is injected into phase A, respectively, and Δu ah1+ and Δu bh1+ is the high-frequency disturbance voltage error in the α and β directions after positive voltage is injected into phase A; Δu af1- and Δu bf1- They are the fundamental frequency disturbance voltage errors in the α and β directions after negative voltage is injected into phase A, respectively, and Δu ah1- and Δu bh1- is the high-frequency disturbance voltage error in the α and β directions after negative voltage is injected into phase A; V his the amplitude of the injected high-frequency voltage, Δi α1+ , Δi β1+ are the current differences in the α and β directions respectively; Δi α1- , Δi β1- They are the current differences in α and β directions after negative voltage is injected into phase A only.

[0068] From formula (3), it can be seen that during the period of superimposing positive and negative voltage injection on phase A, only the high-frequency current is affected, and the fundamental frequency current remains unchanged. Therefore, the fundamental frequency disturbance voltage is Δu under the positive and negative voltage injection. af1+ and Δu af1- , Δu bf1+ and Δu bf1- Therefore, the fundamental frequency disturbance voltage can be directly eliminated by subtracting the current difference equations in different positive and negative injection stages. The expression is:

[0069] Phase 1:

[0070]

[0071] Where, I α1Δ ,I β1Δ They are the differences in current in the α and β directions after only positive voltage is injected into phase A and negative voltage is injected into phase A, respectively.

[0072] Similarly, the expression after subtracting the current difference equations of stage two and stage three is:

[0073] Phase 2:

[0074]

[0075] Where, I α2Δ , I β2Δ are the differences in the current in the α and β directions after only positive voltage is injected into phase B and negative voltage is injected into phase B, respectively. “2” means that only positive and negative voltage are injected into phase B; Δi α2+ , Δi β2+ They are the current differences in the α and β directions after the positive voltage is injected into phase B; Δi α2- , Δi β2- They are the current differences in the α and β directions after negative voltage is injected into phase B; Δu ah2+ and Δu bh2+ is the high-frequency disturbance voltage error in the α and β directions after positive voltage is injected into phase B; Δu ah2- and Δu bh2- It is the high-frequency disturbance voltage error in the α and β directions after negative voltage is injected into phase B.

[0076] Phase 3:

[0077]

[0078] Where, I α3Δ , I β3Δ The current differences in the α and β directions are respectively the differences after only positive voltage is injected into phase C and negative voltage is injected into phase C. “3” means that only positive and negative voltages are injected into phase C. Δi α3+ , Δi β3+ They are the current differences in the α and β directions after positive voltage is injected into phase C; Δi α3- , Δi β3- They are the current differences in the α and β directions after negative voltage is injected into phase C; Δu a3+ , Δu b3+ They are the voltage differences in the α and β directions after positive voltage is injected into phase C; Δu ah3+ and Δu bh3+ It is the high-frequency disturbance voltage error in the α and β directions after positive voltage is injected into phase C.

[0079] Formulas (6) and (7) are respectively and Coordinate transformation;

[0080] Phase 2:

[0081]

[0082] Phase 3:

[0083]

[0084] Where Δu ah2 =Δu ah2+ -Δu ah2- ,Δu bh2 =Δu bh2+ -Δu bh2- ,Δu ah3 =Δu ah3+ -Δu ah3- ,Δu bh3 =Δu bh3+ -Δu bh3- .

[0085] Adding equations (5), (8) and (9) together, we can obtain:

[0086]

[0087] Where:

[0088]

[0089] In the above formula, ζ a1 ,ζ b1 ,ζ a2 ,ζ b2 ,ζ a3 ,ζb3 , ε1, ε2 are all custom functions.

[0090] From formula (10), we can see that the fundamental frequency disturbance voltage Δu afx and Δu bfx The problem is completely eliminated, but there is still a high-frequency disturbance voltage that affects the estimation accuracy of the rotor position. This disturbance voltage will cause the sixth harmonic in the estimated position. The present invention proposes a composite filter phase-locked loop structure based on weighted fusion to filter out the sixth harmonic. The composite filter phase-locked loop (PLL) contains two phase-locked loops, one of which is a traditional phase-locked loop, recorded as PLL1, and its transfer function expression is:

[0091]

[0092] Where, the proportional gain k p =2ζω n , integral gain ζ is the damping coefficient, ω n is the natural frequency, and s is the complex frequency domain variable.

[0093] Traditional PLLs have a low-pass filter characteristic that naturally attenuates high-order harmonics by reducing bandwidth. However, the PLL's low-frequency bandwidth is limited, making harmonic attenuation less effective when the motor is running at low speeds. Therefore, it is necessary to combine the PLL with harmonic filtering technology to achieve better performance without sacrificing bandwidth.

[0094] First, use Equation (11) to obtain the rotor position and speed, which are marked as and The second PLL is recorded as PLL2, and the sixth harmonic is filtered out by iterative method to obtain the rotor position and speed, which are marked as and

[0095] Then, the position and speed obtained by the two PLLs are weighted to obtain the final estimated rotor position and speed. The weighting factor is 0.5, and the output of the weighted fusion of the two PLLs is expressed as:

[0096]

[0097] Figure 2 The harmonic estimation structure is embedded in PLL2, which can filter out harmonics by subtracting the estimated 6th harmonic from the position error. h Available in the form of sine and cosine variation amplitudes, the actual value θ h and estimated values Expressed as:

[0098]

[0099] Where, e 6a , e 6b is the actual amplitude; is the estimated amplitude; is the estimated motor speed.

[0100] Position error minus the estimated 6th harmonic function f e Expressed as:

[0101]

[0102] Where, position error I αΔ , I βΔ It is the sum of the difference equation of stage one recorded in formula (10) and the difference equations of stage two and stage three after coordinate transformation.

[0103] By constructing the objective function E(x), the harmonic suppression problem is converted into an optimization problem of solving its minimum point, which can be expressed as:

[0104]

[0105] Where, custom variables

[0106] In PLL2, an iterative algorithm can be used to solve the minimum value of the objective function. Common algorithms include: gradient descent method, Gauss-Newton method and Levenberg-Marquardt method. Since the Levenberg-Marquardt method has both the global convergence of the gradient descent method and the fast local convergence characteristics of the Gauss-Newton method, the present invention uses the Levenberg-Marquardt method for solution. The parameter update rules are as follows:

[0107] x(k+1)=x(k)-H k -1 J k f e (k)(17)

[0108] Where k represents the number of iterations, function f e Jacobian matrix of (x) is the Hessian matrix, λ is the damping coefficient, and I is the identity matrix.

[0109] The key to solving formula (17) is to obtain J k , can be calculated according to the following equation:

[0110]

[0111] Introducing the custom functions D(z), N(z), we can further deduce:

[0112]

[0113] Where:

[0114]

[0115] Where, T s is the sampling period, k p2 is the proportional gain of the filter phase-locked loop transfer function, k i2 is the integral gain of the filtered phase-locked loop transfer function. N0, N1, N2, D0, D1, and D2 are all custom variables that meet the above requirements.

[0116] Define functions p(k), q(k), and rewrite (19) using the differential form to obtain:

[0117]

[0118] Where:

[0119]

[0120] By substituting formula (20) into formula (17) and solving them together, we can finally get:

[0121]

[0122] Where λ is the damping coefficient and I is the unit matrix.

[0123] The estimated rotor position can be obtained by equations (12), (14), (15) and (22) and speed Among them, the proportional gain k p2 =2ζω n2 , integral gain

[0124] Based on the above theoretical reasoning, the technical solution of the present invention provides a method for estimating the motor position by suppressing the nonlinearity of the inverter. The technical idea is:

[0125] Positive and negative sequence alternating high-frequency voltage signals of equal amplitude are injected into the three phases A, B, and C of the motor in sequence, that is, the three injected signals are recorded as stage one, stage two, and stage three respectively. In each stage, positive and negative voltage signals are alternately injected into one of the three phases. The difference equation between the current differences after the positive and negative voltage signals are injected in each stage is calculated to eliminate the fundamental frequency disturbance voltage. The difference equations of the current differences in stages one, two, and three are then combined to obtain the total current difference equation carrying the rotor position information. The total current difference equation is demodulated using a composite filtering phase-locked loop based on weighted fusion to obtain the rotor position and speed of the motor.

[0126] It should be noted that the core of the present invention is to eliminate the fundamental frequency disturbance voltage by injecting positive and negative sequence alternating high-frequency voltage signals of equal amplitude into the three phases A, B, and C of the motor in sequence, and then deriving the difference equation between the current difference after the positive and negative voltage signals are injected in each stage; and further combining with a composite filter phase-locked loop to effectively suppress the sixth harmonic caused by the dead zone effect. Therefore, under the premise of meeting the technical ideas of the present invention, the motor of the technical solution of the present invention is not limited to a permanent magnet synchronous motor. When applied to other types of motors, adaptive transformation and adjustment are performed according to the motor equation.

[0127] In some embodiments, the technical solution of the present invention further provides a motor control method, including:

[0128] Estimate the rotor position and speed of the motor using the above motor position estimation method;

[0129] Integrate the estimated rotor position and speed into the motor closed-loop control instead of the position encoder;

[0130] Among them, the speed outer loop outputs the q-axis current reference value through the PI regulator according to the deviation between the given speed and the estimated rotor speed, performs inverse Park transformation and SVPWM modulation based on the estimated rotor position, and generates the drive signal of the inverter.

[0131] like Figure 2 As shown in the figure, positive and negative voltage signals of equal amplitude are alternately injected into the three phases A, B, and C. Current sensors are used to collect the three-phase currents and transform them into a two-phase stationary coordinate system through coordinate transformation. A filterless signal demodulation method is then used to separate the high-frequency current response component and suppress the fundamental frequency disturbance caused by inverter nonlinearity. A weighted fusion-based phase-locked loop (PLL) is used to extract the rotor position error signal from the high-frequency current in the stationary coordinate system, suppressing the sixth harmonic caused by inverter nonlinearity. Accurate rotor position and speed estimates are obtained through closed-loop regulation. Furthermore, a PI regulator is used in the current inner loop to control the d-axis and q-axis currents, respectively. The speed outer loop outputs a PI regulator based on the deviation between the given speed and the estimated speed to generate a q-axis current reference value. Finally, an inverse Park transform and SVPWM modulation are performed based on the estimated rotor position to generate the inverter drive signal. The parameters of the composite filter PLL and the injected signal voltage amplitude are set to ensure control accuracy while balancing system stability and dynamic response performance.

[0132] In some embodiments, the present invention further provides a control system including: a voltage signal injection module, a signal processing module, and a composite filter phase-locked loop. It should be understood that in addition to the above three modules, other motor functional modules may also be included in the specific implementation.

[0133] The voltage signal injection module is used to inject positive and negative sequence alternating high-frequency voltage signals of equal amplitude into the three phases A, B, and C of the motor in sequence. That is, the three injected signals are recorded as stage one, stage two, and stage three respectively. In each stage, positive and negative voltage signals are alternately injected into one of the three phases.

[0134] The signal processing module is used to calculate the difference equation between the current differences after the positive and negative voltage signals are injected in each stage to eliminate the fundamental frequency disturbance voltage, and then combine the difference equations of the current differences in stages one, two and three to obtain the total current difference equation that carries the rotor position information; and is used to demodulate the total current difference equation using a composite filtering phase-locked loop based on weighted fusion to obtain the rotor position and speed of the motor.

[0135] The composite filter phase-locked loop is used to demodulate the total current difference equation using a composite filter phase-locked loop based on weighted fusion to obtain the rotor position and speed of the motor.

[0136] It should be understood that the specific implementation process of each module please refer to the above method content, the present invention will not go into details here, and the division of the above functional modules is only for example illustration. In some embodiments, some functional modules can be merged, and some functional modules can be split. Each functional module can be implemented in software or hardware or a combination of software and hardware. Among them, the software and hardware equipment includes but is not limited to general-purpose computer equipment, programmable gate arrays, digital signal processors, microprocessors and their corresponding programming or burning software.

[0137] In some embodiments, the present invention further provides a motor including a controller and a memory; wherein the controller calls a control program stored in the memory to implement: the above-mentioned motor position estimation method or the above-mentioned motor control method of the claim.

[0138] Please refer to the description of the aforementioned method for the specific implementation process of each step. The memory may include a read-only memory and a random access memory, and provide instructions and data to the processor. A portion of the memory may also include a non-volatile random access memory.

[0139] Figure 3 This is a comparison chart of the position observation steady-state performance of the motor running at 200r / min under constant load conditions using an embodiment of the present invention. Under constant load conditions (when the motor runs at 200r / min and the phase current is 1.6A), the position observation performance of the traditional high-frequency voltage injection method is compared with that of the proposed positive and negative sequence alternating high-frequency voltage signal injection. Figure 3 (a) is the traditional high-frequency injection method. It can be seen that the phase current shows a zero-crossing distortion effect. Since the influence of the fundamental wave disturbance voltage is weakened after the load, the rotor position deviation of the traditional high-frequency voltage injection method is reduced compared with the no-load condition. However, the high-frequency disturbance voltage still has a significant negative impact, and the position observation error reaches about -24°. In comparison, Figure 3 (b) shows that the proposed method can significantly suppress the harmonic components in the observed position, reducing the position error by approximately 66.7% and effectively eliminating the DC bias. Experiments have shown that this method can effectively reduce the impact of inverter nonlinearity and magnetic field spatial harmonics on position observation.

[0140] Figure 4 The comparison chart of the dynamic performance of position observation of the motor running at 150r / min under load disturbance of the embodiment of the present invention is shown. The position observation performance of the motor running at 150r / min under sudden load disturbance condition using the traditional high-frequency voltage injection method is compared with the proposed positive and negative sequence alternating high-frequency voltage signal injection method. The experimental setting is that the motor initially runs at no load, then suddenly adds load to increase the phase current to 3A, and finally removes the load. Figure 4 As shown in (a), the peak position observation error of the traditional high-frequency voltage injection method reaches 30° when a sudden load is applied; Figure 4 (b) shows that the proposed method can control the maximum error to within 24°. When the load is removed, the position error of the proposed method quickly converges from 20° to within 15°, while the error of the traditional high-frequency voltage injection method remains around 20°. Under steady-state conditions, the harmonic content of the proposed method is significantly lower than that of the traditional high-frequency voltage injection method. Experimental results demonstrate that the positive-negative sequence alternating high-frequency voltage signal injection method has superior position observation accuracy under load disturbance conditions.

[0141] It should be emphasized that the examples described in the present invention are illustrative rather than restrictive. Therefore, the present invention is not limited to the examples described in the specific embodiments. Any other embodiments derived by those skilled in the art based on the technical solution of the present invention that do not depart from the purpose and scope of the present invention, whether modified or replaced, also fall within the scope of protection of the present invention.

Claims

1. A motor position estimation method for suppressing inverter nonlinearity, characterized by: include: The three phases A, B, and C of the motor are injected with positive and negative sequence alternating high-frequency voltage signals of equal amplitude in sequence. That is, the three corresponding injection signals of A, B, and C are recorded as stage one, stage two, and stage three respectively. In each stage, positive and negative voltage signals are alternately injected into one of the three phases. The difference equation between the current differences after the positive and negative voltage signals are injected in each stage is calculated to eliminate the fundamental frequency disturbance voltage. The difference equations of the current differences in stages one, two, and three are then combined to obtain the total current difference equation that carries the rotor position information. A composite filter phase-locked loop based on weighted fusion is used to demodulate the total current difference equation to obtain the rotor position and speed of the motor; The composite filter phase-locked loop is a filter phase-locked loop branch based on an iterative algorithm added to the existing phase-locked loop branch, and the motor position estimation results of the two phase-locked loop branches are weighted to obtain the rotor position and speed of the motor.

2. The method according to claim 1, wherein: The construction process of the total current difference equation carrying the rotor position information is as follows: After injecting positive and negative sequence alternating high-frequency voltage signals of equal amplitude into the motor, the current difference equations under the injection of positive and negative voltage signals in stages one, two, and three are constructed respectively; Subtract the current differences under the positive voltage and negative voltage signals in the same phase to obtain the difference equation of the current difference in each phase; Perform coordinate transformation on the difference equations of stage 2 and stage 3, respectively, with the coordinate transformation angles being 120° and 240°; The difference equation of stage one is summed with the difference equations of stages two and three after coordinate transformation to obtain the total current difference equation carrying the rotor position information.

3. The method according to claim 2, wherein: If the motor is a permanent magnet synchronous motor, the total current difference equation carrying the rotor position information is: Where, I αΔ , I βΔ It is the sum of the difference equation of stage 1 corresponding to the α and β directions and the difference equations of stage 2 and stage 3 after coordinate transformation; I α1Δ ,I β1Δ are the differences in current in the α and β directions after only positive voltage is injected into phase A and negative voltage is injected into phase A; I' α2Δ ,I' β2Δ The difference between the current difference in the α and β directions after the positive voltage is injected into the B phase and the negative voltage is injected into the B phase respectively. α2Δ , I β2Δ Result after coordinate transformation; I' α3Δ ,I' β3Δ The current difference I in the α and β directions after the positive voltage is injected into the C phase and the negative voltage is injected into the C phase respectively. α3Δ , I β3Δ Result after coordinate transformation; θ r Indicates the rotor position, custom parameter K1 = K2V h , ΔT is the switching period, the average inductance ΣL=(L dh +L qh ) / 2, average inductance difference ΔL=(L dh -L qh ) / 2,L dh and L qh are the d and q axis inductances after the high frequency voltage signal is injected, a1 ,ζ b1 ,ζ a2 ,ζ b2 ,ζ a3 ,ζ b3 , ε1, ε2 are all custom functions; Δu ahx and Δu bhx is the high-frequency disturbance voltage error in the α and β directions, and x represents 1, 2, and 3, where "1" means that positive and negative voltages are injected only into phase A, "2" means that positive and negative voltages are injected only into phase B, and "3" means that positive and negative voltages are injected only into phase C. V h is the amplitude of the injected high-frequency voltage.

4. The method according to claim 2, wherein: The process of constructing the current difference equation when positive and negative voltage signals are injected in any stage is as follows: After injecting positive and negative sequence alternating high-frequency voltage signals of equal amplitude into the motor, the three-phase voltage equation is converted into a high-frequency voltage equation in a two-phase stationary coordinate system; The high-frequency voltage equation in the two-phase stationary coordinate system is transformed to obtain the current difference equation when positive voltage and negative voltage signals are injected respectively.

5. The method according to claim 1, wherein: The filter phase-locked loop branch based on the iterative algorithm filters out harmonics by subtracting the estimated sixth harmonic from the position error, where the function f e Expressed as: exist: Among them, θ err2 is the position error, I αΔ , I βΔ It is the sum of the difference equation of stage 1 corresponding to the α and β directions and the difference equations of stage 2 and stage 3 after coordinate transformation; is the 6th harmonic θ h The estimated value of 6a , e 6b is the actual amplitude, is the estimated amplitude, is the rotor position estimate of the filtered phase-locked loop branch based on the iterative algorithm; Based on the function f e Construct the objective function E(x), that is, convert the harmonic suppression problem into an optimization problem to solve its minimum point. The objective function E(x) is expressed as: Custom variables T is the matrix transpose symbol.

6. The method according to claim 5, characterized in that: The objective function E(x) is minimized by using the gradient descent method, Gauss-Newton method or Levenberg-Marquardt method to obtain the motor position estimation result of the filtered phase-locked loop branch based on the iterative algorithm.

7. The method according to claim 5 or 6, characterized in that: The objective function E(x) is minimized using the Levenberg-Marquardt method to obtain the motor position estimation result of the filtered phase-locked loop branch based on the iterative algorithm. The parameter update rules are as follows: x(k+1)=x(k)-H k -1 J k f e (k); In the formula, k represents the number of iterations, x(k) is the variable x corresponding to the kth iteration, and f e (k) is the function f corresponding to the kth iteration e , J k is the Jacobian matrix, H k is the Hessian matrix, and we can infer that: Where λ is the damping coefficient, and the functions p(k) and q(k) satisfy: Where, T s is the sampling period, k p2 is the proportional gain of the filter phase-locked loop transfer function, k i2 is the integral gain of the filtered phase-locked loop transfer function. N0, N1, N2, D0, D1, and D2 are all user-defined variables / constants.

8. A motor control method based on the motor position estimation method according to any one of claims 1 to 7, characterized in that: include: Using the method of claim 1 to estimate the rotor position and speed of the motor; Integrate the estimated rotor position and speed into the motor closed-loop control instead of the position encoder; Among them, the speed outer loop outputs the q-axis current reference value through the PI regulator according to the deviation between the given speed and the estimated rotor speed, performs inverse Park transformation and SVPWM modulation based on the estimated rotor position, and generates the drive signal of the inverter.

9. A control system based on the method according to any one of claims 1 to 8, characterized in that: include: The voltage signal injection module is used to sequentially inject positive and negative sequence alternating high-frequency voltage signals of equal amplitude into the three phases A, B, and C of the motor. That is, the three injection signals are recorded as phase one, phase two, and phase three respectively. In each phase, positive and negative voltage signals are alternately injected into one of the three phases. The signal processing module is used to calculate the difference equation between the current differences after the positive and negative voltage signals are injected in each stage to eliminate the fundamental frequency disturbance voltage. The difference equations of the current differences in stages one, two, and three are then combined to obtain the total current difference equation that carries the rotor position information. and for demodulating the total current difference equation using a composite filter phase-locked loop based on weighted fusion to obtain the rotor position and speed of the motor; The composite filter phase-locked loop is used for demodulating the total current difference equation by using a composite filter phase-locked loop based on weighted fusion to obtain the rotor position and speed of the motor.

10. A motor, characterized in that: including a controller and a memory; Wherein, the controller calls the control program stored in the memory to implement: the motor position estimation method described in any one of claims 1 to 7 or the motor control method described in claim 8.

Images