A Model-Free Prediction-Based Flux Control Method for Switched Reluctance Motors

Abstract

This invention belongs to the field of motor control and discloses a flux linkage control method for switched reluctance motors based on model-free prediction. This method derives a model-free flux linkage prediction formula, utilizing only the flux linkage difference and voltage difference between adjacent sampling points to predict flux linkage, eliminating reliance on motor resistance, inductance, and other parameters, thus significantly reducing the impact of parameter variations on control accuracy under high-temperature environments. The steps include: generating a piecewise linear flux linkage distribution function, measuring the flux linkage and voltage difference in real time, predicting the flux linkage value under candidate switching states, and constructing a cost function to select the optimal switching state. This method effectively suppresses torque ripple and radial force ripple under both normal and high-temperature conditions, exhibiting high robustness and low computational burden, making it particularly suitable for high-temperature scenarios in aerospace applications. Simulation and experimental results show that the torque ripple suppression error is less than 5% in high-temperature environments, and the flux linkage tracking accuracy is significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of motor control and relates to a flux linkage control method for a switched reluctance motor based on model-free prediction. Background Technology

[0002] Switched reluctance motors (SRMs) have attracted widespread attention in recent years due to their robust and simple structure, ease of manufacture, high reliability, wide speed range, convenient mode switching, and adaptability to harsh conditions. They have become a strong candidate for drive systems in electric vehicles and multi-electric aircraft.

[0003] Switched reluctance motors (SRMs) are doubly salient pole motors with highly saturated magnetic circuits and nonlinear electromagnetic characteristics, whose output torque cannot be linearly represented by current. However, under traditional current control, the output torque of SRMs fluctuates significantly, and this inherent high torque ripple poses a potential threat to the safety of electric drive systems. To address this issue, researchers have chosen motor control as their research focus, proposing direct torque control (DTC), direct instantaneous torque control (DITC), torque distribution function (TSF), and model predictive control (MPC) to suppress SRM torque ripple. Among these, model predictive control's main characteristic is using a system model to predict the future behavior of the controlled variable. Based on predefined optimization criteria, the controller uses this information to obtain the optimal control signal. Therefore, model predictive control has received increasing attention in switched reluctance motor control due to its high efficiency and superior performance.

[0004] However, research has found that model predictive flux control can effectively reduce the dependence of vibration suppression strategies on motor parameters in switched reluctance motors and significantly improve the high-temperature adaptability of the control method. However, it has the following drawbacks:

[0005] 1) There is still a need for accurate motor phase resistance in the process of flux prediction. However, motor phase resistance is easily affected by temperature and changes. Existing methods for identifying high-temperature parameters of phase resistance require a balance between stability and identification performance and still rely on sensor data.

[0006] 2) It may impose a large computational burden on the control algorithm.

[0007] Further research in this invention reveals that by deriving the model-free flux prediction function formula and utilizing the flux difference and voltage difference of the phase winding at two consecutive sampling points, a general prediction model can be established, further enhancing the high-temperature robustness of the control method and fully leveraging the advantages of flux prediction control. This can further reduce the dependence of vibration suppression strategies on motor parameters. Summary of the Invention

[0008] To effectively reduce the dependence of vibration strategies of switched reluctance motors on motor parameters, this invention proposes a model-free predictive flux linkage control method to reduce the dependence of vibration strategies of switched reluctance motors on motor parameters.

[0009] The technical solution of the present invention:

[0010] A model-free prediction-based flux linkage control method for switched reluctance motors includes the following steps:

[0011] Step 1: At each control time k, acquire and store the phase voltage value u at the current time. ph (k), Phase current value i ph (k) and rotor position angle θ ph The current phase flux linkage value ψ is obtained through a pre-constructed lookup table ψ(i,θ). ph (k) and the phase flux linkage value at the previous time step Calculate the flux linkage difference:

[0012]

[0013] Step 2: Based on the rotor position angle θ ph Using a predefined piecewise linear flux linkage distribution function, determine the reference flux linkage value ψ at the current time. ref (θ ph );

[0014] Step 3: Construct a model-free flux linkage prediction function to predict the flux linkage in candidate switch state S. ph Phase flux linkage value at time k+1 under (k+1):

[0015]

[0016] In the formula, T s Indicates the control sampling period, Δu ph (k+1) represents the voltage difference between time k+1 and time k, calculated using the following formula:

[0017]

[0018] In the formula, u dc This represents the bus voltage of the power converter;

[0019] Step 4: Determine if the current rotor position is in the commutation zone. If so, traverse all candidate switch states in the commutation zone and predict the voltage difference Δu corresponding to each candidate switch state. ph (k+1) and phase flux linkage value ψ ph(k+1); if not, then traverse all candidate switch states in the single-phase conduction region and predict the voltage difference Δu corresponding to each candidate switch state. ph (k+1) and phase flux linkage value ψ ph (k+1);

[0020] Step 5: For each candidate switch state, calculate the cost function:

[0021]

[0022] In the formula, J represents the cost function value, and m represents the number of motor phases;

[0023] Step 6: Select the candidate switch state S that minimizes the cost function value. ph (k+1) is then applied to the control of the power converter.

[0024] Furthermore, the predefined piecewise linear flux linkage distribution function in step 2 is:

[0025]

[0026] In the formula, θ on θ represents the opening angle. ov θ represents the commutation overlap angle. off θ represents the shut-off angle. end ψ represents the angle of de-energization. ref1 This indicates the commutation overlap angle θ ov The reference flux linkage value at the location, ψ ref2 Indicates the angle of θ at the cutoff point off Reference flux linkage value at the location.

[0027] Furthermore, the model-free flux linkage prediction function in step 3 is obtained through the following process:

[0028] Step 3.1: Construct the voltage balance equation for the switched reluctance motor:

[0029]

[0030] In the formula, R ph i is the phase resistance of the motor. ph (k) represents the phase current at time k;

[0031] Step 3.2: Discretize the voltage balance equation to obtain the flux linkage prediction model:

[0032]

[0033]

[0034] Step 3.3: Utilize the relationship between phase flux linkage, phase inductance, and phase current:

[0035]

[0036] In the formula, L ph (k) represents the phase inductance value at time k;

[0037] Step 3.4: Substitute equation (9) into equation (7) to obtain:

[0038]

[0039] After simplification and derivation, we obtain:

[0040]

[0041] Similarly, rearranging equation (8) yields:

[0042]

[0043] Step 3.5: Due to the control of the sampling period T s Minimal, assuming the phase inductance values ​​at adjacent times are approximately equal:

[0044]

[0045] Step 3.6: Subtract equations (11) and (12) from step 3.4, and substitute the result into equation (13) to obtain:

[0046]

[0047] Step 3.7: Based on the control frequency being between 10kHz and 20kHz, control the sampling period T. s The phase resistance R of the motor is very small, thus eliminating the need for further investigation. ph ,get:

[0048]

[0049] Step 3.8: Substitute equations (1) and (3) into (15) to obtain the final model-free flux linkage prediction function:

[0050] .

[0051] Furthermore, the candidate switch state Sph (k+1) is represented as:

[0052] .

[0053] The beneficial effects of this invention are as follows: Addressing the issue of the flux linkage prediction formula's dependence on the motor's resistance parameters, this invention re-derives the flux linkage prediction formula, proposing a novel model-free flux linkage prediction formula. Based solely on the flux linkage and voltage difference between two consecutive sampling points, it can achieve accurate prediction of the flux linkage at the next moment, further improving the high-temperature adaptability of the switched reluctance motor vibration suppression method. Simultaneously, the new flux linkage prediction formula has a more reasonable sensitivity to phase voltage changes. Simulation results verify that it can achieve higher-precision flux linkage control, thereby reducing the motor's torque ripple and radial force ripple. Simulation results demonstrate that the model-free predictive flux linkage control method proposed in this chapter achieves better control performance at both room temperature and high temperature. Attached Figure Description

[0054] Figure 1 The principle of flux linkage prediction is given, where (a) is the phase flux linkage prediction process and (b) is the phase voltage prediction process.

[0055] Figure 2 This is a flowchart of model-free flux linkage prediction.

[0056] Figure 3 This is the overall block diagram of model-free predictive flux control based on flux distribution function.

[0057] Figure 4 This is a simulation diagram of the model-free predictive flux linkage control module.

[0058] Figure 5 The flux linkage distribution function and the simulated phase flux linkage are given under normal operating conditions with an SRM speed of 1000 rpm and a load of 10 Nm.

[0059] Figure 6 The flux linkage distribution function and the simulated phase flux linkage are given under the operating conditions of 350℃, SRM speed of 1000rpm, and load of 10Nm.

[0060] Figure 7 The experimental results of the traditional model predictive flux control method and the proposed model-free predictive flux control method are compared under normal operating conditions with an SRM speed of 2000 rpm and a load of 10 Nm. (a) shows the experimental results of the traditional model predictive flux control method, and (b) shows the experimental results of the proposed model-free predictive flux control method.

[0061] Figure 8The experimental results of the traditional model predictive flux control method and the proposed model-free predictive flux control method are compared under the conditions of 350℃, SRM speed of 2000rpm and load of 10Nm. (a) shows the experimental results of the traditional model predictive flux control method and (b) shows the experimental results of the proposed model-free predictive flux control method. Detailed Implementation

[0062] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings and technical solutions.

[0063] Example

[0064] A model-free prediction-based flux linkage control method for switched reluctance motors includes the following steps:

[0065] Using a 3kW three-phase 12 / 8-pole switched reluctance motor as a prototype, a model-free flux linkage control simulation model of the motor under normal and high temperature conditions was built in Matlab / Simulink. Torque pulsation and radial force pulsation were used as the criteria for evaluating vibration magnitude. The model-free predictive flux linkage control method was compared with the model-predictive flux linkage control method through simulation to verify the effectiveness of the derived model-free predictive control and its superior performance under high-temperature conditions.

[0066] Step 1: At each control time k, acquire and store the phase voltage value u at the current time. ph (k), Phase current value i ph (k) and rotor position angle θ ph The current phase flux linkage value ψ is obtained through a pre-constructed lookup table ψ(i,θ). ph (k) and the flux linkage value at the previous time step Calculate the flux linkage difference:

[0067]

[0068] Step 2: Based on the rotor position angle θ ph Using a predefined piecewise linear flux linkage distribution function, determine the reference flux linkage value ψ at the current time. ref (θ ph );

[0069] Step 3: Construct a model-free flux linkage prediction function to predict the flux linkage in candidate switch state S. ph Phase flux linkage value at time k+1 under (k+1):

[0070] (19)

[0071] In the formula, T s Indicates the control sampling period, Δu ph(k+1) represents the voltage difference between time k+1 and time k, calculated using the following formula:

[0072] (20)

[0073] In the formula, u dc This represents the bus voltage of the power converter;

[0074] Step 4: Determine if the current rotor position is in the commutation zone. If so, iterate through all candidate switch states in the commutation zone and predict the voltage difference Δu corresponding to each state. ph (k+1) and flux linkage value ψ ph (k+1); if not, then traverse all candidate switch states in the single-phase conduction region and predict the voltage difference Δu corresponding to each state. ph (k+1) and flux linkage value ψ ph (k+1);

[0075] Step 5: For each candidate switch state, calculate the cost function:

[0076] (twenty one)

[0077] In the formula, J represents the cost function value, and m represents the number of motor phases;

[0078] The candidate switch state S ph (k+1) is represented as

[0079] (twenty two)

[0080] Step 6: Select the switching state S that minimizes the cost function value. ph (k+1) is then applied to the control of the power converter.

[0081] Furthermore, the predefined piecewise linear flux linkage distribution function in step 2 is:

[0082] (twenty three)

[0083] Furthermore, the model-free flux linkage prediction function in step 3 is obtained through the following process:

[0084] Step 3.1: Construct the voltage balance equation for the switched reluctance motor:

[0085] (twenty four)

[0086] In the formula, R ph i is the phase resistance of the motor. ph (k) represents the phase current at time k;

[0087] Step 3.2: Discretize the voltage balance equation to obtain the traditional flux linkage prediction model:

[0088] (25)

[0089] (26)

[0090] Step 3.3: Utilizing the relationship between magnetic flux, inductance, and current:

[0091] (27)

[0092] In the formula, L ph (k) represents the phase inductance value at time k;

[0093] Step 3.4: Substitute equation (27) into equation (25) to obtain

[0094] (28)

[0095] After simplification and derivation, we obtain:

[0096] (29)

[0097] Similarly, rearranging equation (26) yields:

[0098] (30)

[0099] Step 3.5: Due to the sampling period T s Minimal, assuming the inductance values ​​at adjacent times are approximately equal:

[0100] (31)

[0101] Step 3.6: Subtract equations (29) and (30) from step 3.4, and substitute the result into equation (31) to obtain:

[0102] (32)

[0103] Step 3.7: Since the commonly used control frequency for switched reluctance motor control methods is between 10kHz and 20kHz, the sampling period T s It is very small, so the resistance term R can be eliminated. ph ,get:

[0104] (33)

[0105] Step 3.8: Substitute equations (18) and (20) into (33) and simplify to obtain the final model-free flux linkage prediction function:

[0106] (34)

[0107] In the formula, θ on θ represents the opening angle. ov θ represents the commutation overlap angle. off θ represents the shut-off angle. end ψ represents the angle of de-energization. ref1 This indicates the commutation overlap angle θ ov The reference flux linkage value at the location, ψ ref2 Indicates the angle of θ at the cutoff point off Reference flux linkage value at the location;

[0108] Existing model-free predictive control schemes for motors are mostly based on current model derivations. However, the predictive equations for current are more complex than those for flux linkage, as shown in equation (35). Even with formula transformations, it is impossible to perfectly separate the voltage and current differentials. Even if i is eliminated based on the principle of extremely short sampling periods... ph The parameters before (k) still cannot eliminate the characteristic parameters of the voltage difference. In actual control, it is still necessary to use methods such as increasing the number of sampling points and characteristic parameter estimation algorithms to obtain an accurate current prediction model. However, in the flux linkage prediction formula, the flux linkage difference and voltage difference can be perfectly separated. The voltage difference does not contain characteristic parameters related to motor parameters. This makes the flux linkage of switched reluctance motors very suitable for model-free predictive control, which can effectively simplify the prediction method and improve the reliability of the prediction.

[0109]

[0110] Based on equation (34), the flux linkage prediction formula for motor parameter characteristics is transformed into a model-free prediction formula. Only the voltage difference and flux linkage difference between two adjacent moments need to be measured to predict the phase flux linkage value at moment (k+1) under different switching vector control. The prediction principle in practical applications is as follows: Figure 1 As shown, the phase flux linkage values ​​at times (k-1), k, and (k+1) and the phase voltage value at time k are sampled. The phase flux linkage is predicted at time k. If S ph φ when (k+1)=1 ph (k+1)=1 compared to S ph (k+1)=0 and S ph φ when (k+1)=-1 ph (k+1) is closer to φ ref Then the switch state S ph (k+1)=1 is applied to the power converter at time (k+1).

[0111] Based on the flux linkage prediction formula, model prediction flux linkage control simulation modeling is performed, and simulation modules are built as follows: Figure 4 As shown.

[0112] Next, simulation verification of the control effect of the flux linkage distribution function was carried out, such as... Figure 5 and Figure 6 As shown. Under normal temperature conditions, with the motor at its rated speed of 1000 rpm and a load of 10 Nm, the flux linkage curve simulation using the model-free predictive flux linkage control method shows that the flux linkage can almost perfectly follow the given value at normal temperature. Comparison shows that compared to model predictive flux linkage control, the flux linkage curve controlled by the model-free predictive flux linkage control method is smoother, has less fluctuation, and can follow the given value more perfectly. Under 350℃ conditions, with the motor at its rated speed of 1000 rpm and a load of 10 N, the flux linkage curve simulation using the model-free predictive flux linkage control method shows that high-precision flux linkage control can still be achieved at high temperatures. Comparison of the two verifications reveals that there is almost no difference in flux linkage control performance at normal and high temperatures. Under the same speed and load, the reference flux linkage shapes are slightly different, which is the result of online optimization of the motor's control angle.

[0113] Experimental verification of model-free predictive flux linkage control at room temperature, as follows: Figure 7 As shown, the performance of model-free predictive flux control is analyzed by measuring the current, flux linkage, torque, and radial force of the motor under different control methods. Under normal temperature conditions, with the motor at its rated speed of 2000 rpm and a load of 10 Nm, the performance curves of the motor were measured using both the traditional model-predictive flux control method and the proposed model-free predictive flux control method. Under the proposed model-free predictive flux control, the motor's torque and radial force ripples are very small, and the motor characteristic curve is smoother. The model-free predictive flux formula has an advantage in control accuracy compared to the model-predictive formula.

[0114] Experimental verification of model-free predictive flux linkage control at high temperatures, as follows: Figure 8 As shown, under operating conditions of 350℃, with the motor at its rated speed of 2000 rpm and a load of 10 Nm, the performance curves of the motor were measured using both the traditional model predictive flux control method and the proposed model-free predictive flux control method. It can be seen that under the proposed model-free predictive flux control, the motor's torque ripple and radial force ripple are very small. Although the waveforms of phase torque, phase flux linkage, and phase radial force differ at different temperatures due to changes in motor characteristics, the torque ripple and radial force ripple remain very small and almost indistinguishable under different temperature conditions. The comparison reveals that the model-free predictive flux control method at high temperatures results in smaller torque ripple and radial force ripple, as well as a smoother performance waveform.

[0115] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention without departing from the principles and spirit of the present invention.

Claims

1. A flux linkage control method for a switched reluctance motor based on model-free prediction, characterized in that, Includes the following steps: Step 1: At each control time k, acquire and store the phase voltage value u at the current time. ph (k), Phase current value i ph (k) and rotor position angle θ ph The current phase flux linkage value ψ is obtained through a pre-constructed lookup table ψ(i,θ). ph (k) and the phase flux linkage value at the previous time step Calculate the flux linkage difference: Step 2: Based on the rotor position angle θ ph Using a predefined piecewise linear flux linkage distribution function, determine the reference flux linkage value ψ at the current time. ref (θ ph ); Step 3: Construct a model-free flux linkage prediction function to predict the flux linkage in candidate switch state S. ph Phase flux linkage value at time k+1 under (k+1): In the formula, T s Indicates the control sampling period, Δu ph (k+1) represents the voltage difference between time k+1 and time k, calculated using the following formula: In the formula, u dc This indicates the bus voltage of the power converter; Step 4: Determine if the current rotor position is in the commutation zone. If so, traverse all candidate switch states in the commutation zone and predict the voltage difference Δu corresponding to each candidate switch state. ph (k+1) and phase flux linkage value ψ ph (k+1); if not, then traverse all candidate switch states in the single-phase conduction region and predict the voltage difference Δu corresponding to each candidate switch state. ph (k+1) and phase flux linkage value ψ ph (k+1); Step 5: For each candidate switch state, calculate the cost function: In the formula, J represents the cost function value, and m represents the number of motor phases; Step 6: Select the candidate switch state S that minimizes the cost function value. ph (k+1) is then applied to the control of the power converter.

2. The flux linkage control method for a switched reluctance motor according to claim 1, characterized in that, The predefined piecewise linear flux linkage distribution function in step 2 is: In the formula, θ on θ represents the opening angle. ov θ represents the commutation overlap angle. off θ represents the shut-off angle. end ψ represents the angle of de-energization. ref1 This indicates the commutation overlap angle θ ov The reference flux linkage value at the location, ψ ref2 Indicates the angle of θ at the cutoff point off Reference flux linkage value at the location.

3. The flux linkage control method for a switched reluctance motor according to claim 1, characterized in that, The model-free flux prediction function in step 3 is obtained through the following process: Step 3.1: Construct the voltage balance equation for the switched reluctance motor: In the formula, R ph i is the phase resistance of the motor. ph (k) represents the phase current at time k; Step 3.2: Discretize the voltage balance equation to obtain the flux linkage prediction model: Step 3.3: Utilize the relationship between phase flux linkage, phase inductance, and phase current: In the formula, L ph (k) represents the phase inductance value at time k; Step 3.4: Substitute equation (9) into equation (7) to obtain: After simplification and derivation, we obtain: Similarly, rearranging equation (8) yields: Step 3.5: Due to the control of the sampling period T s Minimal, assuming the phase inductance values ​​at adjacent times are approximately equal: Step 3.6: Subtract equations (11) and (12) from step 3.4, and substitute the result into equation (13) to obtain: Step 3.7: Based on the control frequency being between 10kHz and 20kHz, control the sampling period T. s The phase resistance R of the motor is very small, thus eliminating the need for further investigation. ph ,get: Step 3.8: Substitute equations (1) and (3) into (15) to obtain the final model-free flux linkage prediction function: 。 4. The flux linkage control method for a switched reluctance motor according to claim 1, characterized in that, The candidate switch state S ph (k+1) is represented as: 。

Images