Methods, systems, and vehicles for optimizing the full-speed-range efficiency of electrically excited synchronous motors
By determining the operating region in an electrically excited synchronous motor, establishing a quartic equation and solving it analytically using the Ferrari method, the total copper loss is optimized, solving the problems of computational complexity and local optima in EESM control, and realizing the coordinated optimization of stator and rotor copper losses and high-efficiency control in the full speed domain.
Patent Information
- Application Number
- CN202611509990.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-24
- Publication Date
- 2026-11-17
AI Technical Summary
The existing field weakening optimization algorithm in the control of electrically excited synchronous motors (EESM) is computationally complex, sensitive to initial values, and may converge to a local optimum. It fails to achieve global coordinated optimization of stator and rotor copper losses, and the control strategies in the constant torque region and field weakening region are separated, resulting in discontinuous efficiency optimization across the entire speed domain.
By determining the motor's operating region, a quartic equation is established and solved to obtain the reference values of the d-axis and q-axis currents. The Ferrari method is used for analytical solution to optimize the total copper loss and realize current loop and PWM control. The speed loop PI controller, excitation current optimization module, current loop PI controller and SVPWM module are combined to perform unified current control.
It achieves global collaborative optimization of stator and rotor copper losses, avoids initial value sensitivity and local optima, ensures high efficiency and reliability in the full speed domain, and improves the real-time performance and full speed domain efficiency optimization effect of electrically excited synchronous motors.
Smart Images

Figure CN123122369A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of motor control, and more particularly to a method, system, and vehicle for optimizing the full-speed-range efficiency of an electrically excited synchronous motor. Background Technology
[0002] Electrically excited synchronous motors (EESMs) use excitation windings instead of permanent magnets and achieve flexible control of the air gap magnetic field by adjusting the excitation current, attracting widespread attention in the field of electric vehicle drives. Currently, EESM control strategies mainly operate under a vector control framework, coordinating the control of the stator quadrature axis current and the excitation current.
[0003] In existing technologies, maximum torque-to-current ratio control is typically used in the constant torque region to minimize the stator current amplitude under a given torque, thereby reducing stator copper losses. In the field weakening region, to extend the high-speed operating range, a negative direct-axis current is usually injected to weaken the air gap magnetic field, and numerical solution methods such as lookup table methods based on voltage limit ellipses or Newton's iteration method are often used.
[0004] The existing technology has the following shortcomings: the accuracy of the weak field region lookup table method depends on the previous calibration data and cannot adapt to the dynamic changes of motor parameters; numerical solutions such as Newton iteration are sensitive to initial values and have high computational complexity, and may not converge or converge to a local optimum, affecting real-time performance and reliability. Summary of the Invention
[0005] Based on the above problems, this invention proposes a method, system, and vehicle for optimizing the efficiency of an electrically excited synchronous motor across the entire speed domain. This invention solves the technical problems of complex calculations, sensitivity to initial values, and potential convergence to local optima in the field weakening region optimization algorithm of the existing electrically excited synchronous motor control; it also solves the technical problem that the existing control strategy only optimizes stator copper losses and does not include rotor excitation copper losses in a unified optimization objective; and it also solves the technical problems of the separation of control strategies in the constant torque region and the field weakening region, and the discontinuity of efficiency optimization across the entire speed domain.
[0006] This invention proposes a method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor, comprising: The motor operating region is determined based on the actual speed and DC bus voltage obtained; the motor operating region includes at least the constant torque region and the field weakening region. In the constant torque region, the d-axis current reference value and q-axis current reference value are obtained based on the acquired electromagnetic torque reference value and excitation current reference value. In the weak magnetic region, each excitation current is scanned with the goal of minimizing total copper loss. A fourth-order equation is established based on the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters, and the corresponding d-axis current and q-axis current are obtained by solving the equation. The total copper loss is calculated for each excitation current, and the excitation current that minimizes the total copper loss is taken as the excitation current reference value. The corresponding d-axis current reference value and q-axis current reference value are also obtained. Current loop control and PWM control are performed based on the excitation current reference value, d-axis current reference value, and q-axis current reference value.
[0007] In addition, determining the motor operating range includes: calculating the required stator voltage vector amplitude of the motor based on the electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters; If the required stator voltage vector amplitude is less than or equal to the voltage limit value, the motor is determined to be operating in the constant torque region; otherwise, the motor is determined to be operating in the weak field region.
[0008] In addition, in the constant torque region, the rated excitation current is read as the reference value of the excitation current; Based on the electromagnetic torque reference value and the excitation current reference value, the d-axis current reference value and the q-axis current reference value are calculated using the maximum torque-current ratio formula.
[0009] In addition, in the constant torque region, the rated excitation current is read; By querying the pre-stored two-dimensional table based on the electromagnetic torque reference value and the rated excitation current, and performing interpolation, the d-axis current reference value and the q-axis current reference value are obtained.
[0010] In addition, establishing the quartic equation includes: Substitute the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters into the voltage constraint equation and torque equation; By expressing the q-axis current as a function of the d-axis current using the torque equation, and substituting this into the voltage constraint equation, we obtain a fourth-order equation concerning the d-axis current.
[0011] Furthermore, the corresponding d-axis current and q-axis current obtained by solving include: Solve the quartic equation to obtain the real roots; Based on the stator current limit equation and voltage limit equation, the root that satisfies the current limit and voltage limit is selected from the real roots as the physically feasible d-axis current. If there are multiple physically feasible d-axis currents, select the d-axis current that minimizes the total copper loss. Then calculate the q-axis current based on the torque equation.
[0012] In addition, the Ferrari method was used to solve the quartic equation.
[0013] In addition, regarding the d-axis current The quartic equation is: , Among them, A, B, C, and D are determined by the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, d-axis inductance, q-axis inductance, mutual inductance between the excitation winding and the stator d-axis, and the number of pole pairs.
[0014] In addition, the total copper loss is calculated for each excitation current, including: Total copper loss The calculation formula is: , in, For stator resistance, For d-axis current, For q-axis current, For excitation resistor, This represents the excitation current detected during scanning.
[0015] In addition, the excitation current that minimizes total copper loss is used as the excitation current reference value, and the corresponding d-axis current reference values and q-axis current reference values are obtained, including: Select three scan points near the scan point corresponding to the minimum total copper loss; A quadratic function of total copper loss and excitation current is fitted based on three scanning points: Where a, b, and c are coefficients; Take the derivative of the quadratic function and set the derivative to zero:
[0016] Obtain the optimal excitation current : , The optimal excitation current is used as the excitation current reference value, and substituted into the voltage constraint equation and torque equation to obtain the corresponding d-axis current reference value and q-axis current reference value.
[0017] In addition, the stator voltage command is output based on the d-axis current reference value and the q-axis current reference value, and the excitation voltage command is output based on the excitation current reference value; Generate PWM signals to drive the three-phase inverter based on stator voltage commands; The actual excitation current in the excitation winding is controlled according to the excitation voltage command.
[0018] The present invention also proposes a system employing the full-speed-domain efficiency optimization method for electrically excited synchronous motors as described in any of the preceding claims, comprising: Speed loop PI controller, excitation current optimization module, current loop PI controller, SVPWM module and excitation power converter; A speed loop PI controller is used to output an electromagnetic torque reference value based on the error between the reference speed and the actual speed. The excitation current optimization module is used to receive electromagnetic torque reference value, actual speed and DC bus voltage, determine the motor operating area, and calculate and output a set of reference values, including: excitation current reference value, d-axis current reference value and q-axis current reference value; A current loop PI controller is used to output stator voltage commands and excitation voltage commands based on a set of reference values. The SVPWM module is used to generate PWM signals to drive the three-phase inverter based on stator voltage commands. The excitation power converter is used to adjust the actual excitation current in the excitation winding according to the excitation voltage command.
[0019] The present invention also proposes a vehicle that employs the full-speed-range efficiency optimization method for an electrically excited synchronous motor as described in any of the preceding claims.
[0020] The full-speed-domain efficiency optimization method for electrically excited synchronous motors provided by this invention determines the operating region based on the actual speed and DC bus voltage. In the constant torque region, the reference values for the d-axis and q-axis currents are obtained based on the reference values for electromagnetic torque and excitation current. In the field weakening region, each excitation current is scanned. A fourth-order equation is established and solved with the goal of minimizing total copper loss. The total copper loss is calculated, and the excitation current corresponding to the minimum value is taken as the reference value for excitation current. The corresponding reference values for the d-axis and q-axis currents are obtained. Finally, current loop control and PWM control are performed uniformly, thereby avoiding initial value sensitivity and local optima, and realizing global collaborative optimization of stator and rotor copper losses and full-speed-domain efficiency optimization. Attached Figure Description
[0021] Figure 1 A flowchart illustrating a method for optimizing the full-speed-domain efficiency of an electrically excited synchronous motor according to an embodiment of the present invention; Figure 2 An overall software architecture diagram provided for one embodiment of the present invention; Figure 3 A diagram illustrating the architecture of a motor control system according to an embodiment of the present invention; Figure 4 A schematic diagram of typical torque-speed characteristic curves of an electrically excited synchronous motor across the full speed range, provided in an embodiment of the present invention; Figure 5 An electrically excited synchronous motor provided in one embodiment of the present invention is The relationship between the current-limiting circle and the voltage-limiting ellipse in a coordinate system; Figure 6 This is a schematic diagram of the constant torque curves of a motor with different salient pole characteristics provided in one embodiment of the present invention; Figure 7 This is a comparison diagram of the constant torque curve and the MTPA curve provided in one embodiment of the present invention; Figure 8 A comparison diagram of the current limiting circle, voltage limiting ellipse, constant torque curve and MTPA curve provided in one embodiment of the present invention; Figure 9A schematic diagram of the topology of a two-level three-phase voltage source inverter provided in one embodiment of the present invention; Figure 10 A voltage space vector diagram and its sector division diagram are provided as an embodiment of the present invention; Figure 11 A detailed flowchart of a weak magnetic region optimization algorithm provided in one embodiment of the present invention; Figure 12 The flowchart illustrates the Ferrari method iterative fitting optimization process provided in one embodiment of the present invention. Detailed Implementation
[0022] The present invention will be further described in detail below with reference to specific embodiments and accompanying drawings. This description is intended only to illustrate specific embodiments of the invention and does not constitute any limitation on the invention. The scope of protection of the invention is defined by the claims.
[0023] Reference Figure 1 This invention proposes a method for optimizing the full-speed-domain efficiency of an electrically excited synchronous motor, comprising: Step S001: Determine the motor operating region based on the actual speed and DC bus voltage obtained; the motor operating region includes at least the constant torque region and the field weakening region; Step S002: In the constant torque region, the d-axis current reference value and q-axis current reference value are obtained based on the acquired electromagnetic torque reference value and excitation current reference value. Step S003: In the field weakening region, scan each excitation current with the goal of minimizing total copper loss. Establish a quartic equation based on the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters, and solve it to obtain the corresponding d-axis current and q-axis current. Calculate the total copper loss for each excitation current, and use the excitation current with the minimum total copper loss as the excitation current reference value, and obtain the corresponding d-axis current reference value and q-axis current reference value. Step S004: Perform current loop control and PWM control based on the excitation current reference value, d-axis current reference value, and q-axis current reference value.
[0024] Electrically Excited Synchronous Machines (EESMs), as a type of motor solution without rare-earth permanent magnets, have received widespread attention in the field of electric vehicle drives in recent years. Their rotors use excitation windings instead of permanent magnets, and the excitation current (…) is adjusted… This technology enables flexible control of the air gap magnetic field, fundamentally solving the dependence of permanent magnet synchronous motors (PMSMs) on rare earth materials and the resulting supply chain risks and costs. Currently, research on control strategies for EESMs mainly focuses on vector control frameworks, with the core being the control of the stator's right-angle axis current (…). , ) and excitation current ( Collaborative optimization of ).
[0025] In the constant torque region, the closest existing technology is the use of a maximum torque-to-current ratio (MTPA) control strategy. This strategy aims to minimize the stator current amplitude for a given torque, thereby reducing stator copper losses.
[0026] For example, existing technologies have proposed an MTPA control method for permanent magnet synchronous motors, which can determine the optimal current angle through table lookup or online calculation to achieve high-efficiency operation. Similarly, in electrically excited synchronous motors, research has also used the MTPA strategy to optimize stator-side efficiency. In the field weakening control region, to extend the high-speed operating range of the motor, a negative direct-axis current is usually injected to weaken the air gap magnetic field. Traditional field weakening control often employs table lookup methods or iterative calculations based on voltage limit ellipses. Related literature proposes a field weakening strategy based on Newton's iteration method, which adjusts the current command in real time by solving the voltage constraint equation online. In addition, some studies have also focused on the impact of inductor saturation on field weakening performance.
[0027] The overall situation in this technical field indicates that EESM control is a complex problem involving multiple variables, strong coupling, and nonlinearity. Existing technologies tend to treat the constant torque region and field weakening region separately, or to use numerical iterative methods with high computational complexity and convergence dependent on initial values in the field weakening region.
[0028] Although existing technologies have achieved the basic control functions of EESM to a certain extent, they still have the following significant defects and shortcomings, which are the core problems that the technical solution of this application aims to solve: 1. The optimization algorithm for the weak magnetic field region has bottlenecks in efficiency and reliability.
[0029] Existing field weakening control strategies, such as lookup table methods, heavily rely on the quantity and accuracy of pre-calibrated data for accuracy. They cannot adapt to dynamic changes in motor parameters due to temperature and magnetic saturation, leading to decreased control precision. Numerical methods like Newton's iteration suffer from sensitivity to initial iteration values, high computational complexity, heavy processor load, and potential convergence issues (e.g., non-convergence or convergence to a local optimum rather than the global optimum) under certain operating conditions. This directly impacts the real-time performance and reliability of the control system under complex vehicle operating conditions. The fundamental reason is that the optimization problem of EESM in the field weakening region is essentially a multivariable programming problem that requires simultaneous satisfaction of the torque equation and voltage constraint equations. Existing methods fail to provide an analytical approach that guarantees both computational efficiency and the attainment of a global optimum.
[0030] 2. Global collaborative optimization of stator and rotor copper losses has not been achieved.
[0031] Existing technologies primarily focus on optimizing stator copper losses, failing to incorporate excitation circuit copper losses into a unified optimization objective. For EESM, total copper losses... In the constant torque region, using only the MTPA strategy can only optimize the stator current, neglecting the effect of adjusting... This raises the possibility of further reducing total copper loss. In weak magnetic regions, existing methods typically aim to maintain maximum output capacity rather than minimize total copper loss. This isolated, localized optimization strategy fails to fully exploit the efficiency potential of the EESM, resulting in unnecessary energy waste and shortening the driving range of electric vehicles.
[0032] 3. The inherent defects of PMSM have not been fundamentally resolved.
[0033] Although EESM itself contains no rare-earth materials, if its control strategy fails to fully leverage its unique advantage of tunable magnetic fields, its efficiency and output capability in the high-speed region may even be inferior to PMSM. Existing technologies have failed to systematically utilize EESM by reducing... The direct magnetic weakening characteristic avoids the need for a large negative value injection, as is required for PMSM. The surge in stator copper losses and the risk of permanent magnet demagnetization caused by shaft current. In other words, existing control strategies have failed to translate the hardware advantages of being "rare-earth free" into systemic performance and reliability advantages.
[0034] Based on the deficiencies of existing technologies analyzed in the background section, this invention aims to solve the following key technical problems existing in the control of electrically excited synchronous motors (EESMs): 1. To address the problems of computational complexity, sensitivity to initial conditions, and potential convergence to local optima in traditional numerical optimization algorithms for weak magnetic regions, this paper proposes a computationally efficient and globally convergent analytical solution scheme to improve the real-time performance and reliability of the control system.
[0035] 2. To address the issue of incomplete efficiency optimization in existing control strategies, and to achieve a leap from local optimization (stator copper loss only) to global optimization (total stator and rotor copper loss), fully tapping the energy efficiency potential of EESM across the entire operating range, especially in the high-speed range.
[0036] 3. Systematically solve the inherent problems of rare earth dependence, demagnetization risk and sharp drop in efficiency in high-speed range of permanent magnet synchronous motors (PMSM). By giving full play to the unique advantage of adjustable magnetic field of EESM, it provides a rare earth-free, highly reliable, wide-speed-range and efficient electric drive solution.
[0037] In step S001 of this invention, the motor operating region is determined based on the actual rotational speed and DC bus voltage obtained; the motor operating region includes at least a constant torque region and a field weakening region; Optionally, the actual rotational speed can be acquired by a rotational speed sensor or obtained by a rotational speed estimation module.
[0038] Optionally, the DC bus voltage can be acquired through a bus voltage sampling circuit.
[0039] Optionally, determining the motor operating range includes: calculating the required stator voltage vector amplitude of the motor based on the electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters; If the required stator voltage vector amplitude is less than or equal to the voltage limit value, the motor is determined to be operating in the constant torque region; otherwise, the motor is determined to be operating in the weak field region.
[0040] In step S002, in the constant torque region, the d-axis current reference value and q-axis current reference value are obtained based on the acquired electromagnetic torque reference value and excitation current reference value. Optionally, in the constant torque region, the rated excitation current is read as the excitation current reference value; Based on the electromagnetic torque reference value and the excitation current reference value, the d-axis current reference value and the q-axis current reference value are calculated using the maximum torque-current ratio formula.
[0041] Alternatively, in the constant torque region, read the rated excitation current; based on the electromagnetic torque reference value and the rated excitation current, query the pre-stored two-dimensional table and perform interpolation to obtain the d-axis current reference value and the q-axis current reference value.
[0042] The pre-stored two-dimensional table is: pre-defined in multiple electromagnetic torques With multiple excitation current combinations, the optimal value for each combination is calculated offline, aiming at the maximum torque-to-current ratio. and And store it in a two-dimensional table.
[0043] First dimension: Electromagnetic torque From zero to maximum torque, the torque is divided into multiple torque sampling points according to the torque step size; Second dimension: Excitation current From the lower limit of the excitation current safety to the rated excitation current, the excitation current is divided into multiple excitation current sampling points according to the excitation current step size; Table content: Each grid point Store a set of optimal stator currents .
[0044] In step S003, in the field weakening region, each excitation current is scanned with the goal of minimizing total copper loss. A fourth-order equation is established based on the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters, and the corresponding d-axis current and q-axis current are obtained by solving the equation. The total copper loss is calculated for each excitation current, and the excitation current that minimizes the total copper loss is taken as the excitation current reference value. The corresponding d-axis current reference value and q-axis current reference value are also obtained. Optionally, establishing a quartic equation includes: Substitute the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters into the voltage constraint equation and torque equation; By expressing the q-axis current as a function of the d-axis current using the torque equation, and substituting this into the voltage constraint equation, we obtain a fourth-order equation concerning the d-axis current.
[0045] Optionally, the corresponding d-axis current and q-axis current can be obtained by solving for: Solve the quartic equation to obtain the real roots; Based on the stator current limit equation and voltage limit equation, the root that satisfies the current limit and voltage limit is selected from the real roots as the physically feasible d-axis current. If there are multiple physically feasible d-axis currents, select the d-axis current that minimizes the total copper loss. Then calculate the q-axis current based on the torque equation.
[0046] Alternatively, the Ferrari method can be used to solve the quartic equation.
[0047] Regarding the d-axis current The quartic equation is: , Among them, A, B, C, and D are determined by the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, d-axis inductance, q-axis inductance, mutual inductance between the excitation winding and the stator d-axis, and the number of pole pairs.
[0048] The total copper loss is calculated for each excitation current, including: Total copper loss The calculation formula is: , in, For stator resistance, For d-axis current, For q-axis current, For excitation resistor, This represents the excitation current detected during scanning.
[0049] Optionally, the excitation current that minimizes total copper loss is used as the excitation current reference value, and the corresponding d-axis current reference values and q-axis current reference values are obtained, including: Select three scan points near the scan point corresponding to the minimum total copper loss; A quadratic function of total copper loss and excitation current is fitted based on three scanning points: Where a, b, and c are coefficients; Take the derivative of the quadratic function and set the derivative to zero:
[0050] Obtain the optimal excitation current : , The optimal excitation current is used as the excitation current reference value, and substituted into the voltage constraint equation and torque equation to obtain the corresponding d-axis current reference value and q-axis current reference value.
[0051] Total copper loss is the sum of stator copper loss and rotor excitation copper loss.
[0052] In step S004, current loop control and PWM control are performed based on the excitation current reference value, d-axis current reference value, and q-axis current reference value.
[0053] Optionally, a stator voltage command is output based on the d-axis current reference value and the q-axis current reference value, and an excitation voltage command is output based on the excitation current reference value; Generate PWM signals to drive the three-phase inverter based on stator voltage commands; The actual excitation current in the excitation winding is controlled according to the excitation voltage command.
[0054] This invention solves the technical problems of complex calculation, sensitivity to initial values, and potential convergence to local optima in the field weakening zone optimization algorithm of electrically excited synchronous motor control in the prior art; it also solves the technical problem that the existing control strategy only optimizes stator copper loss and does not include rotor excitation copper loss in the unified optimization objective; and it also solves the technical problems of the separation of control strategies in the constant torque zone and field weakening zone and the discontinuity of efficiency optimization in the full speed domain.
[0055] The full-speed-domain efficiency optimization method for electrically excited synchronous motors provided by this invention determines the operating region based on the actual speed and DC bus voltage. In the constant torque region, the reference values for the d-axis and q-axis currents are obtained based on the reference values for electromagnetic torque and excitation current. In the field weakening region, each excitation current is scanned. A fourth-order equation is established and solved with the goal of minimizing total copper loss. The total copper loss is calculated, and the excitation current corresponding to the minimum value is taken as the reference value for excitation current. The corresponding reference values for the d-axis and q-axis currents are obtained. Finally, current loop control and PWM control are performed uniformly, thereby avoiding initial value sensitivity and local optima, and realizing global collaborative optimization of stator and rotor copper losses and full-speed-domain efficiency optimization.
[0056] In one embodiment, determining the motor operating range includes: calculating the required stator voltage vector amplitude of the motor based on the electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters; If the required stator voltage vector amplitude is less than or equal to the voltage limit value, the motor is determined to be operating in the constant torque region; otherwise, the motor is determined to be operating in the weak field region.
[0057] Because the stator voltage required by a motor depends not only on the speed but also on the torque demand, motor parameters, and bus voltage. At high speeds and light loads, the required voltage may not reach its limit, allowing for constant torque control; however, at low speeds and heavy loads, the voltage may be insufficient due to high current. Therefore, this invention first calculates the required stator voltage vector amplitude under the current operating condition based on the electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters, and then compares it with the voltage limit value to determine whether the field weakening region has been entered. The advantages of this approach are: the judgment directly reflects the actual output capability of the inverter, avoiding premature field weakening leading to torque reduction, and avoiding late field weakening causing voltage saturation and current runaway; simultaneously, it can adapt to bus voltage fluctuations and load changes, achieving a smooth switch between the constant torque region and the field weakening region, ensuring optimized torque output capability and efficiency across the entire speed range.
[0058] The electromagnetic torque reference value is output by the speed loop PI controller.
[0059] The speed loop PI controller receives the reference speed. Compared with actual speed Calculate the rotational speed error: , The electromagnetic torque reference value is output after PI calculation. , Its discretization form can be: , in, This is the proportionality coefficient. The integral coefficient is... To control the cycle.
[0060] therefore, The calculations are performed internally by the controller, requiring no additional sensors.
[0061] The actual rotational speed can be acquired through speed sensors, such as encoders, rotary transformers, or Hall effect sensors; it can also be obtained through sensorless estimation algorithms, such as those based on back electromotive force, flux linkage observers, or model reference adaptive algorithms. In electrically excited synchronous motors, if a position sensor is installed, the actual rotational speed can be obtained from the differential position signal; if no position sensor is installed, the controller estimates the actual rotational speed based on stator voltage, stator current, and motor parameters.
[0062] The DC bus voltage is obtained through the bus voltage sampling circuit.
[0063] Motor parameters include d-axis inductance q-axis inductance Mutual inductance between the excitation winding and the stator d-axis Extreme logarithms Stator resistance Magnetizing resistor These parameters can be obtained from motor design values, nameplate parameters, factory tests, or offline identification, and are pre-stored in the controller's memory; they can also be corrected using online parameter identification algorithms, such as recursive least squares or model reference adaptive methods. In real-time control, the controller reads the motor parameters from the memory or uses the parameters obtained from online identification.
[0064] According to the electromagnetic torque reference value Actual speed Current excitation current Based on the motor parameters, calculate the required stator voltage vector amplitude for the current operating condition: .
[0065] in, It can be obtained from the current torque reference value and excitation current reference value through maximum torque-current ratio control or the current control strategy.
[0066] Required stator voltage vector magnitude With voltage limit value Compare: like The motor is determined to be operating in the constant torque region. like This determines that the motor is operating in the weak magnetic field region. The voltage limit value can be obtained based on the actual situation.
[0067] In one embodiment, in the constant torque region, the rated excitation current is read as the excitation current reference value; Based on the electromagnetic torque reference value and the excitation current reference value, the d-axis current reference value and the q-axis current reference value are calculated using the maximum torque-current ratio formula.
[0068] Read the rated excitation current: , Based on the mutual inductance between the excitation winding and the stator d-axis Calculate the excitation flux linkage: , Establish the maximum torque-to-current ratio relationship: For electrically excited synchronous motors, under the condition of maximum torque-to-current ratio, the reference values of d-axis current and q-axis current satisfy the following: , in: It is the d-axis inductance; It is the q-axis inductance; This is the reference value for the q-axis current.
[0069] This formula means that, under the condition of minimizing the total stator current, a given q-axis current corresponds to an optimal d-axis current.
[0070] Simultaneous torque equations: The reference value for electromagnetic torque is: , in: This is a reference value for electromagnetic torque. It is an extreme logarithm.
[0071] In step 2 Substituting the expression into the torque equation, we obtain the expression containing only... The equation: , Simplified to: , Solve for the q-axis current reference value: Numerical solutions were performed on the above equations to obtain electromagnetic torque reference values. of This equation is a nonlinear equation. In practice, offline calculation and online table lookup interpolation can be used to avoid solving the nonlinear equation in real time.
[0072] Calculate the reference value for d-axis current: The obtained Substituting back into the maximum torque-current ratio formula, the calculation is obtained. : .
[0073] The results obtained will prepare for subsequent steps.
[0074] In one embodiment, the rated excitation current is read in the constant torque region; By querying the pre-stored two-dimensional table based on the electromagnetic torque reference value and the rated excitation current, and performing interpolation, the d-axis current reference value and the q-axis current reference value are obtained.
[0075] In actual implementation, multiple electromagnetic torques are pre-set. and multiple excitation currents Under the combined conditions, the optimal result is calculated offline according to the above process. and Construct a two-dimensional table.
[0076] The two-dimensional table uses electromagnetic torque as the first dimension and excitation current as the second dimension, with each grid point storing a set of data. .
[0077] When running online: 1. Read the rated excitation current ; 2. Obtain the reference value of electromagnetic torque ; 3. According to and Perform bilinear interpolation in a two-dimensional table; 4. Output d-axis current reference value and q-axis current reference value .
[0078] The results obtained will prepare for subsequent steps.
[0079] In one embodiment, establishing the quartic equation includes: Substitute the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters into the voltage constraint equation and torque equation; By expressing the q-axis current as a function of the d-axis current using the torque equation, and substituting this into the voltage constraint equation, we obtain a fourth-order equation concerning the d-axis current.
[0080] 1. The voltage constraint equation, neglecting the voltage drop across resistance, is:
[0081] Torque equation:
[0082] 2. Define intermediate variables make:
[0083]
[0084]
[0085]
[0086]
[0087]
[0088]
[0089] The voltage constraint equation then becomes: , The torque equation is reduced to:
[0090] 3. Eliminate q-axis current: Solve from the torque equation :
[0091] Substituting into the voltage constraint equation:
[0092] Multiply both sides :
[0093] Rearrangement:
[0094] 4. Expand into a quartic equation make:
[0095] in:
[0096]
[0097] Further order:
[0098] in:
[0099]
[0100]
[0101] The equation then becomes:
[0102] After unfolding, we get:
[0103] Divide both sides by This yields the standard quartic equation:
[0104] 5. Expressions for coefficients A, B, C, and D:
[0105]
[0106]
[0107]
[0108] in: , , , , , , .
[0109] 6. Explanation: when hour, The above quartic equation holds true.
[0110] like ,but The quartic equation degenerates and needs to be handled separately.
[0111] Voltage limit This corresponds to the maximum phase voltage amplitude in the SVPWM linear modulation region.
[0112] This quartic equation is for a specific candidate excitation current. Established; during the excitation current scanning process, each time a new one is used... The coefficients A, B, C, and D will all change, requiring a new system to be established and solved.
[0113] By establishing quartic equations, the original multivariate optimization problem requiring simultaneous solution of d-axis and q-axis currents is transformed into a single-variable quartic equation problem, reducing the dimensionality of the solution. Subsequently, analytical methods can be used to directly find the roots, avoiding the dependence on initial values for numerical methods such as Newton's iteration, and preventing non-convergence or convergence to local optima. The equation coefficients are determined by real-time operating data, which can adapt to changes in speed, torque, and bus voltage. The calculation process is deterministic and the computational load is controllable, making it easy to execute in real time in a microprocessor. It provides a stable analytical solution for d-axis current for point-by-point scanning of excitation current in the field weakening region and comparison of total copper loss, thereby ensuring the real-time performance and reliability of efficiency optimization across the entire speed domain.
[0114] In one embodiment, the corresponding d-axis current and q-axis current are obtained by solving for: Solve the quartic equation to obtain the real roots; Based on the stator current limit equation and voltage limit equation, the root that satisfies the current limit and voltage limit is selected from the real roots as the physically feasible d-axis current. If there are multiple physically feasible d-axis currents, select the d-axis current that minimizes the total copper loss. Then calculate the q-axis current based on the torque equation.
[0115] 1. Solve the quartic equation to obtain four roots: The quartic equation is:
[0116] Using Ferrari's method to solve analytically, four roots were obtained: , These four roots could be real roots or conjugate complex roots.
[0117] 2. Filtering for real roots: For each root Determine its imaginary part:
[0118] If the following conditions are met: , Then the real part of the root is taken as the candidate d-axis current: , in, The preset real number discrimination threshold can be, for example, taken as... .
[0119] If symbolic analysis is used for computation, the root with a zero imaginary part is directly taken as the real root.
[0120] All candidate d-axis currents that meet the conditions are grouped into a candidate set:
[0121] in, The number of real roots, .
[0122] If the candidate set is empty, it is determined that there is no real solution under the current excitation current, and the next excitation current is scanned.
[0123] 3. Calculate the q-axis current for each candidate d-axis current: For each in the candidate set Calculate the corresponding q-axis current based on the torque equation:
[0124] Before calculating, determine if the denominator is zero or close to zero:
[0125] in, This is a preset small quantity. If the denominator does not meet this condition, the candidate d-axis current is discarded.
[0126] 4. Current limit verification: For each candidate solution Determine whether the stator current limit equation is satisfied:
[0127] in, This is the stator current limit value.
[0128] If the conditions are not met, the candidate solution is discarded.
[0129] 5. Voltage limit verification: Further determine whether the voltage limit equation is satisfied:
[0130] To account for numerical calculation errors, an allowable tolerance can be set.
[0131] If the conditions are not met, the candidate solution is discarded.
[0132] 6. Obtain the set of physically feasible solutions: After the above screening, candidate solutions that meet the following conditions are retained: It is a real root; The denominator in q-axis current calculations is not zero; Satisfies the stator current limit; Meets voltage limits; The excitation current is within a safe range.
[0133] These candidate solutions constitute the set of physically feasible solutions:
[0134] 7. If multiple physically feasible solutions exist, select the one with the minimum total copper loss: If the number of physically feasible solutions is greater than 1, then calculate the total copper loss for each candidate solution:
[0135] Select the candidate solution with the minimum total copper loss:
[0136] in, As a reference value for the d-axis current, Used as a reference value for the q-axis current.
[0137] 8. Output Results The final output is the current excitation current. The following corresponds to:
[0138] If there is only one physically feasible solution, then that solution is directly used as the d-axis current and q-axis current; if there is no physically feasible solution, then continue scanning the next excitation current.
[0139] By solving the quartic equation to obtain real roots, and selecting physically feasible d-axis currents based on the stator current limit equation and voltage limit equation, mathematical solutions that do not meet the actual operating constraints of the motor can be eliminated, avoiding voltage saturation, overcurrent, or control failure due to the selection of infeasible roots. When multiple physically feasible solutions exist, the d-axis current is further selected based on minimizing total copper loss. This allows for obtaining a solution that satisfies torque and voltage constraints while minimizing total copper loss in the field weakening region, avoiding the selection of non-optimal solutions. Finally, the q-axis current is calculated based on the torque equation to ensure that the output torque meets the electromagnetic torque reference value. Thus, this embodiment achieves a balance between physical feasibility and minimizing total copper loss in the field weakening region, improving the accuracy and reliability of efficiency optimization in the field weakening region. Furthermore, the solution process is deterministic, facilitating real-time control.
[0140] In one embodiment, the Ferrari method is used to solve the quartic equation.
[0141] The Ferrari solver subroutine is invoked to calculate the four roots of the quartic equation. The Ferrari method is a classic analytical algorithm for solving quartic equations. It obtains an exact analytical solution by completing the square and reducing the degree. Its solution process is stable and does not depend on the initial value.
[0142] In one embodiment, regarding the d-axis current The quartic equation is: , Among them, A, B, C, and D are determined by the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, d-axis inductance, q-axis inductance, mutual inductance between the excitation winding and the stator d-axis, and the number of pole pairs.
[0143] By substituting the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters into the voltage constraint equation and torque equation, and eliminating the q-axis current, a fourth-order equation concerning only the d-axis current is obtained. The equation coefficients A, B, C, and D are determined by the aforementioned real-time operating data and motor parameters. The advantages are: transforming the nonlinear optimization problem of multivariable coupling in the field weakening region into a single-variable quartic equation solution problem, reducing the solution dimension and complexity; determining the equation coefficients online according to real-time operating conditions and motor parameters, adapting to changes in speed, torque, bus voltage, and excitation current; and unifying the equation form, facilitating direct analytical root finding, avoiding dependence on initial values for numerical iteration, and preventing non-convergence or convergence to local optima.
[0144] In one embodiment, calculating the total copper loss for each excitation current includes: Total copper loss The calculation formula is: , in, For stator resistance, For d-axis current, For q-axis current, For excitation resistor, This represents the excitation current detected during scanning.
[0145] The total copper loss is calculated for each excitation current because when scanning the excitation current in the field weakening region, each Each will correspond to a different set If only stator copper loss is optimized, a larger value might be selected. This leads to excessive excitation copper loss; however, if only the excitation copper loss is optimized, an excessively small value may be selected. This leads to an increase in stator field weakening current and stator copper losses. Only by calculating stator copper losses and rotor excitation copper losses together can a fair comparison be made among candidate solutions corresponding to different excitation currents to find the solution that minimizes the total copper losses. .
[0146] This embodiment incorporates stator-side and rotor-side copper losses into the same evaluation index, avoiding localized optimization; it can fully utilize the advantage of adjustable EESM excitation current, and reasonably reduce copper losses in the weak magnetic region. This reduces the stator field weakening current while also taking into account the copper loss of the excitation winding; it provides a unified basis for selecting the optimal excitation current with the minimum total copper loss, thereby improving efficiency across the entire speed range, especially in the high-speed field weakening region.
[0147] In one embodiment, the excitation current that minimizes total copper loss is used as the excitation current reference value, and the corresponding d-axis current reference value and q-axis current reference value are obtained as follows: Select three scan points near the scan point corresponding to the minimum total copper loss; A quadratic function of total copper loss and excitation current is fitted based on three scanning points: Where a, b, and c are coefficients; Take the derivative of the quadratic function and set the derivative to zero:
[0148] Obtain the optimal excitation current : , The optimal excitation current is used as the excitation current reference value, and substituted into the voltage constraint equation and torque equation to obtain the corresponding d-axis current reference value and q-axis current reference value.
[0149] Three scanning points are selected near the scanning point corresponding to the minimum total copper loss. A quadratic function of total copper loss and excitation current is fitted based on these three scanning points, and the optimal excitation current is obtained by differentiating the quadratic function. This is because the scanning process uses a fixed step size, and the minimum total copper loss point among the discrete scanning points is usually only an approximate optimal point within the candidate range. The excitation current that truly minimizes total copper loss may lie between two scanning points. From the perspective of someone skilled in the art, the change of total copper loss with excitation current near the minimum value is approximately a quadratic function. The coefficients of this quadratic function can be determined using three scanning points. By differentiating and setting the derivative to zero, the optimal excitation current in the continuous domain can be analytically obtained. This avoids missing the true minimum value due to an excessively large step size, and also avoids increasing the computational load by infinitely reducing the step size in pursuit of accuracy. The obtained optimal excitation current is used as the reference value of the excitation current, and substituted back into the voltage constraint equation and torque equation to obtain the corresponding d-axis current reference value and q-axis current reference value, ensuring that the three reference values satisfy the torque and voltage constraints under the same operating conditions. This embodiment improves the accuracy of solving for the optimal excitation current without significantly increasing the computational load, making the total copper loss closer to the global minimum. In one embodiment, a stator voltage command is output based on the d-axis current reference value and the q-axis current reference value, and an excitation voltage command is output based on the excitation current reference value; Generate PWM signals to drive the three-phase inverter based on stator voltage commands; The actual excitation current in the excitation winding is controlled according to the excitation voltage command.
[0150] The stator voltage command is output based on the d-axis current reference value and the q-axis current reference value, and the excitation voltage command is output based on the excitation current reference value because the electrically excited synchronous motor has two independent control channels: the stator winding and the rotor excitation winding.
[0151] The field weakening optimization yields three current reference values, but these must ultimately be applied to the motor via a power converter: the d-axis and q-axis current reference values need to be adjusted by the stator current loop to generate stator voltage commands, which are then pulse-width modulated to generate switching signals to drive the three-phase inverter, ensuring that the actual stator current tracks the reference values; the excitation current reference value needs to be adjusted by the excitation current loop to generate excitation voltage commands, which are then adjusted by the excitation power converter to regulate the actual excitation current in the excitation winding, ensuring that the actual excitation current tracks the reference values. The effects of this approach are as follows: It enables independent closed-loop control of the stator and excitation sides, ensuring that the three optimized reference values are accurately applied to the motor; it improves DC bus voltage utilization and reduces stator current harmonics by controlling the three-phase inverter with pulse width modulation; it allows for independent adjustment of the excitation current via the excitation power converter, fully leveraging the adjustable magnetic field of the EESM, and weakening the air gap magnetic field in the field weakening region by reducing the excitation current, thus avoiding the increased stator copper losses and demagnetization risk caused by the need to inject a large negative d-axis current into the permanent magnet synchronous motor; and it outputs the same set of current reference values in both the constant torque and field weakening regions, and undergoes the same current loop and modulation control, achieving a unified control interface across the entire speed domain and smooth switching between regions, thereby improving system real-time performance and reliability.
[0152] The present invention also proposes a system employing the full-speed-domain efficiency optimization method for electrically excited synchronous motors as described in any of the preceding claims, comprising: Speed loop PI controller, excitation current optimization module, current loop PI controller, SVPWM module and excitation power converter; A speed loop PI controller is used to output an electromagnetic torque reference value based on the error between the reference speed and the actual speed. The excitation current optimization module is used to receive electromagnetic torque reference value, actual speed and DC bus voltage, determine the motor operating area, and calculate and output a set of reference values, including: excitation current reference value, d-axis current reference value and q-axis current reference value; A current loop PI controller is used to output stator voltage commands and excitation voltage commands based on a set of reference values. The SVPWM module is used to generate PWM signals to drive the three-phase inverter based on stator voltage commands. The excitation power converter is used to adjust the actual excitation current in the excitation winding according to the excitation voltage command.
[0153] Optionally, a preferred embodiment of the full-speed-domain efficiency optimization system for electrically excited synchronous motors (EESMs) of the present invention is described in detail, focusing on the microprocessor (MCU) as the core control unit. (Refer to...) Figure 2-12 Please provide an explanation.
[0154] Figure 2The overall software architecture of the system of the present invention is shown, demonstrating the complete data processing flow from signal input to PWM output and the logical relationship between each functional module.
[0155] Figure 3 The architecture of the motor control system was demonstrated, enabling precise control of the motor's three-phase current and torque, allowing the motor to drive the load as needed.
[0156] Figure 4 The typical torque-speed characteristic curves of an electrically excited synchronous motor are shown across the entire speed range, and the division between the constant torque region and the field weakening region is clearly defined.
[0157] Figure 5 Demonstrates the use of electrically excited synchronous motors The relationship between the current limiting circle and the voltage limiting ellipse in the coordinate system illustrates the constraints on motor operation and their variation with excitation current and speed.
[0158] Figure 6 The constant torque curves of motors with different salient pole characteristics are shown, illustrating the mapping relationship between stator current and electromagnetic torque.
[0159] Figure 7 By combining the constant torque curve and the MTPA curve, the optimization principle of the maximum torque-current ratio control strategy is intuitively demonstrated.
[0160] Figure 8 This paper comprehensively demonstrates the relationship between the current limiting circle, voltage limiting ellipse, constant torque curve, and MTPA curve, and defines the motor's... The complete operating area of the plane.
[0161] Figure 9 The schematic diagram of the topology of a two-level three-phase voltage source inverter is shown, which is the basis of the power circuit driven by the SVPWM module of this invention.
[0162] Figure 10 The voltage space vector diagram and its sector division are shown, illustrating the basic principles of the SVPWM algorithm.
[0163] Figure 11 A detailed flowchart of the weak magnetic region optimization algorithm, one of the core components of this invention, is shown, which specifically illustrates the hybrid optimization process based on Ferrari's method and iterative fitting.
[0164] Figure 12 A flowchart illustrating the Ferrari method's iterative fitting optimization process is presented.
[0165] 1. The core hardware of this system is a high-performance 32-bit microprocessor (MCU). This MCU needs to possess high-speed computing power, abundant on-chip peripherals and peripherals, and sufficient storage space to support complex floating-point mathematical operations and real-time control tasks. The MCU integrates multiple CPU cores, a high-precision ADC module, an enhanced PWM (ePWM) unit, and dedicated peripherals for motor control (such as an encoder interface), fully meeting the requirements of this system. The overall software architecture of the system is implemented by the program within the MCU, and its structure is as follows: Figure 2 As shown, the complete data processing flow from signal input to PWM output and the logical relationships between each functional module are clearly demonstrated. All algorithms are executed periodically in the MCU.
[0166] 2. Core Module Functions and Working Principles 2.1 Speed Loop PI Controller The MCU connects to external power circuits and sensors through its on-chip peripherals to form a complete closed-loop control system.
[0167] Reference speed Compared with actual speed error This error is controlled by a PI controller (proportional coefficient). Integral coefficient The calculation is performed to output a reference value for the electromagnetic torque. Its discretization calculation formula is: , in, To control the cycle, set it to 100μs. and The speed loop needs to be tuned according to the motor's moment of inertia and load characteristics to ensure a fast response and no overshoot.
[0168] 2.2 Excitation Current Optimization Module This module is key to achieving global optimization in this invention; its operational logic is implemented by a program in the MCU. This module receives... , and Its internal operating process is as follows Figure 11 As shown. The MCU first determines the operating region based on the current speed and voltage status. If it enters the weak magnetic field region, it activates the hybrid optimization algorithm: Step 1: Determine the running region. The module first determines which region to enter based on voltage utilization. The determination condition is: if... ≤ (Base speed) and the voltage vector magnitude required for the current calculation. much smaller If the torque is high, it enters the constant torque region; otherwise, it enters the weak magnetic region. It is determined by calculation based on motor parameters and rated voltage.
[0169] Step 2A: Constant Torque Region Implementation Method (MTPA Control). In this region, the goal is to maximize the torque-to-current ratio. For salient pole ratio... The MTPA curve of the EESM can be obtained by the following formula:
[0170] in To simplify real-time calculations, this implementation method employs offline calculation and online table lookup. Pre-calculation is performed at different torques... and excitation current Under the combination, calculate the optimal ( , (Combination). At runtime, based on... and current By performing a two-dimensional lookup table and interpolation, the optimal stator current command can be obtained. , Excitation current command In this area, it is usually maintained at the rated value. To provide maximum torque output capability.
[0171] Step 2B: Implementation method for the weak magnetic region. This implementation method is the core of the invention, and its goal is to solve for the total copper loss while satisfying voltage constraints and torque requirements. The smallest ( )combination.
[0172] Initialization: Set the initial value of the excitation current iteration. iteration step size Minimum copper loss Initialize to a maximum value.
[0173] Step 3: Iteration and Ferrari method for solution: a. Construct a quartic equation: for the current iteration value to Substituting into the voltage limit equation and the torque equation, and neglecting the voltage drop across the resistance, the voltage constraint can be simplified to: , The torque equation is: , b. Substitution and simplification: use and By expressing this and substituting it into the voltage constraint equation, we can obtain the following: quartic equation: , Coefficients A, B, C, and D are , , , and motor parameters , , , The function.
[0174] , c. Ferrari's method for analytical solution: This calls the Ferrari solver subroutine to calculate the four roots of the quartic equation. Ferrari's method is a classic analytical algorithm for solving quartic equations. It obtains an exact analytical solution by completing the square and reducing the degree. Its solution process is stable and does not depend on initial values. Figure 12 As shown.
[0175] d. Solution sifting: Select real roots from the four roots and determine the unique feasible one based on its physical meaning. Solution, and then calculate the corresponding .
[0176] Calculate and compare copper losses: based on feasible ( Calculate the total copper loss using the following formula: , like Then update And record the current current combination ( ).
[0177] Loop iteration: Let until Reduced to the preset safety lower limit To prevent demagnetization.
[0178] Second-order fitting optimization: After the iteration is completed, select three iteration points around the recorded minimum copper loss point. , ,( Use these three points to fit a quadratic function. By taking the derivative To obtain the optimal excitation current Finally, Substitute the values and fine-tune them to obtain the final optimal stator current command. , .
[0179] 2.3 SVPWM mode for modulation. make( , Through Clarke and Park inverse transformations, the voltage reference in the three-phase stationary coordinate system is obtained. , , ), and then the SVPWM algorithm was adopted. Figure 9 , to the voltage vector Combined into 8 basic voltage space vectors ~ The combination of vectors is used to calculate the duration of action of each vector. , , Ultimately, six PWM drive signals are generated to control the inverter bridge switching transistors.
[0180] 3. Technical Effects and Experimental Verification Through the above-described MCU-based hardware and software implementation, this invention has achieved significant technical effects: Efficiency Improvement: Thanks to the powerful computing capabilities of the MCU and its perfect support for the hybrid optimization algorithm, the system minimizes the total copper loss across all operating conditions, and the average operating efficiency is significantly improved, especially in the high-speed, weak magnetic field region, where the efficiency advantage is even more pronounced.
[0181] Reliability Guarantee: The Ferrari analytical method is stably executed by the MCU, fundamentally avoiding the risk of numerical iteration non-convergence and ensuring the robustness and reliability of the control system. Simultaneously, field weakening is achieved by adjusting the excitation current, avoiding the risk of demagnetization of the permanent magnet motor.
[0182] Real-time performance achieved: Despite the complexity of the algorithm, with the support of a high-performance MCU, all control tasks are completed within a short control cycle, fully meeting the real-time requirements of electric vehicle drive.
[0183] This specific implementation uses the MCU as the computing and control core, and provides a complete, efficient, reliable and implementable EESM full-speed domain optimization control scheme.
[0184] In one embodiment, the current loop PI controller includes a CSR current loop module, a voltage feedforward module, and a voltage limiting module; The SVPWM module includes a sector determination module, a time calculation module, a duty cycle calculation module, and an SVPWM waveform generation module; The system also includes a Res current sampling unit, a Clark transformation module, an angle acquisition module, a Park transformation module, an angle compensation module, a Park inverse transformation module, and an excitation regulator.
[0185] Reference Figure 2In one embodiment, the system's software architecture is shown, specifically including: Res (Res current sampling unit), Clark transformation module, angle acquisition module, Park transformation module, CSR current loop module, voltage feedforward module, voltage limiting module, angle compensation module, Park inverse transformation module, sector judgment module, time calculation module, duty cycle calculation module, SVPWM waveform generation module, and excitation regulator.
[0186] The Res current sampling unit acquires the three-phase stator current of the electrically excited synchronous motor and outputs the three-phase current signal to the Clark conversion module. The Clark conversion module converts the three-phase current in the three-phase stationary coordinate system into the current in the two-phase stationary αβ coordinate system and sends it to the Park conversion module. The angle acquisition module acquires the rotor position signal of the motor and outputs the rotor electrical angle. One path is sent to the Park conversion module to provide an angle reference for the positive Park coordinate transformation; the other path is sent to the angle compensation module for angle error compensation, and the compensated angle is input to the Park inverse conversion module.
[0187] The Park conversion module receives the two-phase stationary coordinate system current and rotor electrical angle, transforms them into d-axis and q-axis feedback currents in the dq rotating coordinate system, and outputs them to the CSR current loop module. The CSR current loop module calculates the deviation between the current setpoint and the feedback current and outputs the dq-axis voltage command signal. The voltage feedforward module receives the dq-axis voltage command and adds a voltage feedforward decoupling term to eliminate coupling interference between the dq axes and improve dynamic response performance. The output of the voltage feedforward module is sent to the voltage limiting module, which limits the amplitude of the dq-axis voltage command to prevent the inverter from entering voltage saturation.
[0188] The dq-axis voltage command output from the voltage limiting module is input to the Park inverse transformation module. The Park inverse transformation module, combined with the compensated electrical angle output from the angle compensation module, inversely transforms the dq rotating coordinate system voltage into a two-phase stationary αβ coordinate system voltage. This αβ voltage signal is sequentially sent to the sector determination module, time calculation module, duty cycle calculation module, and SVPWM generation module. The sector determination module identifies the sector where the reference voltage vector is located; the time calculation module calculates the duration of the basic voltage vector's action; the duty cycle calculation module converts the vector's action time into a PWM duty cycle; and the SVPWM generation module generates a PWM drive pulse signal based on the duty cycle to drive the electrically excited synchronous motor (EESM).
[0189] The excitation regulator independently outputs an excitation voltage (U_f) which is connected to the excitation winding of the electrically excited synchronous motor to achieve independent adjustment of the excitation magnetic field. The electrically excited synchronous motor EESM and the permanent magnet synchronous motor PMSM are mechanically coaxially connected and maintain the same mechanical speed, forming a coaxially coupled motor assembly.
[0190] The overall system forms a current closed-loop vector control architecture, which realizes torque and speed control of the electrically excited synchronous motor through stator current closed-loop feedback, coordinate transformation, feedforward decoupling, angle compensation and SVPWM modulation. The excitation circuit independently adjusts the excitation magnetic field to adapt to the coaxial dual motor coupled drive condition.
[0191] The system adopts CSR current loop closed-loop control, combined with voltage feedforward module to achieve dq axis decoupling, which can quickly track d and q axis current command, suppress current coupling disturbance, improve motor torque dynamic response speed, and achieve higher torque control accuracy.
[0192] The configuration angle compensation module corrects the electrical angle of the coordinate transformation, reduces the control deviation caused by rotor position detection error, improves the accuracy of coordinate transformation, and reduces current harmonics and torque pulsation.
[0193] A voltage limiting module is set up to constrain the amplitude of the output dq axis voltage command, preventing the inverter from entering the voltage saturation region, ensuring stable operation of SVPWM modulation, and avoiding control failure.
[0194] The excitation regulator independently applies excitation voltage to the excitation winding of the electrically excited synchronous motor (EESM), which can flexibly adjust the excitation magnetic field, optimize the motor power factor and operating efficiency over a wide speed range, and adapt to the joint drive conditions of coaxial coupling with PMSM.
[0195] The entire control chain relies on Clark / Park coordinate transformation and SVPWM space vector modulation, resulting in low harmonic content in the output voltage waveform, reducing motor stator losses and vibration noise, and improving the operational stability of the coaxial coupled motor assembly.
[0196] The present invention also proposes a vehicle that employs the full-speed-range efficiency optimization method for an electrically excited synchronous motor as described in any of the preceding claims.
[0197] Optionally, the electromagnetic torque reference value is output by the speed loop PI controller based on the error between the reference speed and the actual speed; the actual speed is acquired by the speed sensor; the DC bus voltage is acquired by the bus voltage sampling circuit; and the rated excitation current and motor parameters are pre-stored in the memory.
[0198] Optionally, decreasing the excitation current in fixed steps includes: Set the initial value of the excitation current iteration to the rated excitation current; Set the iteration step size to a negative value; Each time the current is decreased, it is determined whether the current excitation current is less than or equal to the preset excitation current safety lower limit; If the current excitation current is less than or equal to the safety lower limit, the scan will stop. If the current excitation current is greater than the safety lower limit, continue scanning.
[0199] Optionally, the stator voltage command is input to the SVPWM module to generate the PWM signal driving the three-phase inverter, including: The stator voltage command is subjected to Clarke transformation and inverse Park transformation to obtain the voltage reference in the three-phase stationary coordinate system. Based on the voltage reference in the three-phase stationary coordinate system, the SVPWM algorithm is used to generate 6 PWM drive signals; The six PWM drive signals are sent to the three-phase inverter to control the operation of the electrically excited synchronous motor.
[0200] Optionally, the electromagnetic torque reference value is output by the speed loop PI controller based on the error between the reference speed and the actual speed.
[0201] Optionally, in the field weakening region, the air gap magnetic field is weakened by reducing the excitation current, thus avoiding the injection of negative d-axis current into the stator for field weakening.
[0202] This invention solves the technical problems of complex calculation, sensitivity to initial values, and potential convergence to local optima in the field weakening zone optimization algorithm of electrically excited synchronous motor control in the prior art; it also solves the technical problem that the existing control strategy only optimizes stator copper loss and does not include rotor excitation copper loss in the unified optimization objective; and it also solves the technical problems of the separation of control strategies in the constant torque zone and field weakening zone and the discontinuity of efficiency optimization in the full speed domain.
[0203] The full-speed-domain efficiency optimization method for electrically excited synchronous motors provided by this invention determines the operating region based on the actual speed and DC bus voltage. In the constant torque region, the reference values for the d-axis and q-axis currents are obtained based on the reference values for electromagnetic torque and excitation current. In the field weakening region, each excitation current is scanned. A fourth-order equation is established and solved with the goal of minimizing total copper loss. The total copper loss is calculated, and the excitation current corresponding to the minimum value is taken as the reference value for excitation current. The corresponding reference values for the d-axis and q-axis currents are obtained. Finally, current loop control and PWM control are performed uniformly, thereby avoiding initial value sensitivity and local optima, and realizing global collaborative optimization of stator and rotor copper losses and full-speed-domain efficiency optimization.
[0204] As needed, the above technical solutions can be combined to achieve the best technical effect.
[0205] The above description is merely the principle and preferred embodiment of the present invention. It should be noted that, for those skilled in the art, several other modifications can be made based on the principle of the present invention, and these modifications should also be considered within the scope of protection of the present invention.
Claims
1. A method for optimizing the full-speed-domain efficiency of an electrically excited synchronous motor, characterized in that, include: Determine the motor's operating range based on the actual speed and DC bus voltage obtained; The motor operating range includes at least the constant torque region and the field weakening region; In the constant torque region, the d-axis current reference value and q-axis current reference value are obtained based on the acquired electromagnetic torque reference value and excitation current reference value. In the weak magnetic region, each excitation current is scanned with the goal of minimizing total copper loss. A fourth-order equation is established based on the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters, and the corresponding d-axis current and q-axis current are obtained by solving the equation. The total copper loss is calculated for each excitation current, and the excitation current that minimizes the total copper loss is taken as the excitation current reference value. The corresponding d-axis current reference value and q-axis current reference value are also obtained. Current loop control and PWM control are performed based on the excitation current reference value, d-axis current reference value, and q-axis current reference value.
2. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 1, characterized in that, Determining the motor's operating range includes: calculating the required stator voltage vector amplitude based on the electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters; If the required stator voltage vector amplitude is less than or equal to the voltage limit value, the motor is determined to be operating in the constant torque region; otherwise, the motor is determined to be operating in the weak field region.
3. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 1, characterized in that, In the constant torque region, the rated excitation current is read as the reference value of the excitation current; Based on the electromagnetic torque reference value and the excitation current reference value, the d-axis current reference value and the q-axis current reference value are calculated using the maximum torque-current ratio formula.
4. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 1, characterized in that, In the constant torque region, read the rated excitation current; By querying the pre-stored two-dimensional table based on the electromagnetic torque reference value and the rated excitation current, and performing interpolation, the d-axis current reference value and the q-axis current reference value are obtained.
5. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 1, characterized in that, Establishing a quartic equation includes: Substitute the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, and motor parameters into the voltage constraint equation and torque equation; By expressing the q-axis current as a function of the d-axis current using the torque equation, and substituting this into the voltage constraint equation, we obtain a fourth-order equation concerning the d-axis current.
6. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 5, characterized in that, The corresponding d-axis and q-axis currents obtained by solving include: Solve the quartic equation to obtain the real roots; Based on the stator current limit equation and voltage limit equation, the root that satisfies the current limit and voltage limit is selected from the real roots as the physically feasible d-axis current. If there are multiple physically feasible d-axis currents, select the d-axis current that minimizes the total copper loss. Then calculate the q-axis current based on the torque equation.
7. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 6, characterized in that, Ferrari's method is used to solve the quartic equation.
8. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 6, characterized in that, Regarding the d-axis current The quartic equation is: , Among them, A, B, C, and D are determined by the current excitation current, electromagnetic torque reference value, actual speed, DC bus voltage, d-axis inductance, q-axis inductance, mutual inductance between the excitation winding and the stator d-axis, and the number of pole pairs.
9. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 1, characterized in that, The total copper loss is calculated for each excitation current, including: Total copper loss The calculation formula is: , in, For stator resistance, For d-axis current, For q-axis current, For excitation resistor, This represents the excitation current detected during scanning.
10. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 9, characterized in that, The excitation current that minimizes total copper loss is used as the excitation current reference value, and the corresponding d-axis current reference values and q-axis current reference values are obtained, including: Select three scan points near the scan point corresponding to the minimum total copper loss; A quadratic function of total copper loss and excitation current is fitted based on three scanning points: Where a, b, and c are coefficients; Take the derivative of the quadratic function and set the derivative to zero: Obtain the optimal excitation current : , The optimal excitation current is used as the excitation current reference value, and substituted into the voltage constraint equation and torque equation to obtain the corresponding d-axis current reference value and q-axis current reference value.
11. The method for optimizing the full-speed-range efficiency of an electrically excited synchronous motor according to claim 1, characterized in that, The stator voltage command is output based on the d-axis current reference value and the q-axis current reference value, and the excitation voltage command is output based on the excitation current reference value. Generate PWM signals to drive the three-phase inverter based on stator voltage commands; The actual excitation current in the excitation winding is controlled according to the excitation voltage command.
12. A system employing the full-speed-domain efficiency optimization method for an electrically excited synchronous motor as described in any one of claims 1 to 11, characterized in that, include: Speed loop PI controller, excitation current optimization module, current loop PI controller, SVPWM module and excitation power converter; A speed loop PI controller is used to output an electromagnetic torque reference value based on the error between the reference speed and the actual speed. The excitation current optimization module is used to receive electromagnetic torque reference value, actual speed and DC bus voltage, determine the motor operating area, and calculate and output a set of reference values, including: excitation current reference value, d-axis current reference value and q-axis current reference value; A current loop PI controller is used to output stator voltage commands and excitation voltage commands based on a set of reference values. The SVPWM module is used to generate PWM signals to drive the three-phase inverter based on stator voltage commands. The excitation power converter is used to adjust the actual excitation current in the excitation winding according to the excitation voltage command.
13. A vehicle, characterized in that, The full-speed-range efficiency optimization method for electrically excited synchronous motors as described in any one of claims 1 to 11 is adopted.
