Permanent magnet synchronous motor fault-tolerant control method for space flight

Abstract

The invention relates to the technical field of fault-tolerant control in the aspect of space flight, and discloses a permanent magnet synchronous motor fault-tolerant control method for space flight, which adopts a multi-objective optimization prediction fault-tolerant control strategy, a control strategy after a fault is optimized and a comprehensive fault-tolerant control scheme. The control targets of limiting the short-circuit circulating current, reducing the torque pulsation and minimizing the temperature rise of the winding are achieved, so that the fault-tolerant control requirement of a space flight device is better met, and the high practicability is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault-tolerant control for spaceflight, and particularly to a fault-tolerant control method for permanent magnet synchronous motors used in spaceflight. Background Technique

[0002] Due to its high power density, high efficiency, and precise control characteristics, the permanent magnet synchronous motor (PMSM) has become the core drive unit of the key actuators of space vehicles. Typical applications include:

[0003] Satellite attitude control flywheel: achieving high-precision attitude stabilization (angular resolution up to 0.001°) through motor torque adjustment;

[0004] Deep space probe drive system: driving tasks such as the deployment of solar panels and the movement of robotic arms;

[0005] Space station hatch and docking mechanism: requiring reliable operation in a vacuum and extreme temperature environment.

[0006] In space missions, the high reliability and autonomous fault-tolerant ability of the motor system are core requirements. Due to the harsh space environment (radiation, rapid temperature changes, micro-meteorite impacts) and the inability to perform manual repairs, faults such as inter-turn short circuits in the motor winding and demagnetization of permanent magnets may lead to mission failures or even catastrophic consequences. For example, a certain type of geostationary orbit satellite once lost attitude control due to a short circuit fault in the flywheel motor, resulting in losses of hundreds of millions of dollars. Therefore, the PMSM fault-tolerant control technology for spaceflight has become a research hotspot in recent years.

[0007] Currently, the fault-tolerant control methods applied in the space field mainly include: hardware redundancy, open-loop compensation schemes based on fault characteristics, and traditional model predictive control (MPC) schemes. However, they have significant limitations under the special constraints of spaceflight:

[0008] First, regarding the hardware redundancy scheme, it mainly achieves fault switching through backup motors or windings, such as the design of dual-winding PMSM. Its defects in space applications include: the redundant system increases the satellite launch cost (the cost per kilogram of payload is about 20,000 - 30,000 US dollars), the time from fault detection to switching is 10 - 50 ms, which cannot meet the high dynamic response requirements (such as emergency attitude adjustment), and it adopts a single fault mode coverage, only targeting open circuit or ground faults, and cannot handle the gradual inter-turn short circuit faults.

[0009] Second, for the open-loop compensation scheme based on fault characteristics, it mainly matches with an offline fault library and injects compensation current to offset the fault impact. Its defects in space applications include: the model is static, and the short-circuit circulating current i is not considered. fThe dynamic coupling effect leads to an increase in the torque pulsation after compensation (measured ≥15%), which poses a risk of temperature rise out of control. Ignoring the local heating of the short-circuit ring (power loss R f and i f ), it is easy to cause hot spot temperature to exceed the limit (>200℃) under vacuum heat dissipation conditions, has poor adaptability to the orbital environment, relies on ground pre-calibration parameters, and cannot cope with the motor parameter drift caused by space radiation (such as L d The rate of change can reach 5% / year).

[0010] Finally, regarding the traditional model predictive control (MPC) solution, it mainly uses the continuous domain motor model to predict the future state and optimize the voltage vector selection. Its defects in aerospace applications include: high computational complexity, floating point operations (>1MFLOPS) required for continuous model solving, exceeding the real-time upper limit of the onboard controller (such as LEON3) (control cycle >100μs), and insufficient fault modeling: short-circuit faults are only simplified to parameter disturbances, and R is not established. f -L f -i f The explicit dynamic relationship leads to large errors in the estimation of the fault degree (k value error>10%). There is a situation of multi-target splitting. The goals such as flux tracking and temperature rise suppression are processed in a time-sharing manner. Under dynamic conditions, there may be conflicts in goals (such as overheating caused by high-precision tracking).

[0011] From the above content, it can be seen that the existing fault-tolerant control methods are difficult to meet the multi-dimensional coupling constraints of fault-tolerant control in space flight: the existing methods lack adaptability to extreme environments, the heat dissipation efficiency in a vacuum environment is only 30%-50% of that on the ground, and radiation causes the motor parameters to vary with time; moreover, the computing resources of the existing methods are strictly limited, and the computing power of the onboard controller is limited (usually at the ARMCortex-M7 level), requiring the algorithm to be lightweight; in addition, the existing methods also have the problem of multi-task coupling, especially when systems such as attitude control and energy management work together, the fault-tolerant strategy needs to be dynamically prioritized.

[0012] In view of the above-mentioned defects, technical personnel in this field urgently need a permanent magnet synchronous motor fault-tolerant control method with high-precision fault modeling, electrical-thermal collaborative optimization and low computational overhead. Summary of the invention

[0013] The purpose of the present invention is to solve the above problems and to design a fault-tolerant control method for a permanent magnet synchronous motor used in space flight.

[0014] To achieve the above-mentioned purpose, the technical solution of the present invention is a permanent magnet synchronous motor fault-tolerant control method for space flight, comprising:

[0015] Based on the equivalent circuit of inter-turn short-circuit fault and taking into full consideration the factors of space flight, the forward Euler method is used to construct the motor prediction model;

[0016] By analyzing the motor prediction model, the expressions of the d-q axis components of the stator flux linkage, the short-circuit circulating current, and the winding temperature rise under the turn-to-turn short-circuit fault are obtained;

[0017] The cost function is used to evaluate the d-q axis components of the stator flux linkage, the short-circuit circulating current, and the winding temperature rise under the turn-to-turn short-circuit fault, and the optimal voltage vector is obtained.

[0018] The construction process of the motor prediction model includes:

[0019] When the permanent magnet synchronous motor has a k% turn-to-turn short-circuit fault, the equivalent equation of the faulty phase winding is:

[0020]

[0021] In the formula: u sd , u sq are the d-q axis stator voltages (V); i sd , i sq are the d-q axis stator currents (A); i f is the short-circuit loop current (A); R s is the stator winding resistance (Ω); L d , L q are the d-q axis inductances (H); ω e is the electrical angular velocity (rad / s); e d , e q are the back electromotive force components of the permanent magnet (V); R f is the short-circuit resistance (Ω), R f = kR s , k ∈ (0,1) is the short-circuit turn ratio;

[0022] The equivalent equation is discretized by the forward Euler method to obtain the motor prediction model as:

[0023]

[0024] In the formula, Δt is the time step.

[0025] Based on the motor prediction model, the expressions of the d-q axis components of the stator flux linkage under the turn-to-turn short-circuit fault are:

[0026]

[0027] In the formula, L f is the short-circuit loop inductance.

[0028] Based on the motor prediction model, the short-circuit circulating current expression is:

[0029]

[0030] The winding temperature rise expression obtained based on the motor prediction model is:

[0031]

[0032] In the formula, h is the heat dissipation coefficient (W / (m 2 ·K)), and A is the winding surface area (m 2 ).

[0033] The cost function for evaluating the stator flux d-q axis components, short-circuit circulating current, and winding temperature rise under the turn-to-turn short-circuit fault is:

[0034]

[0035] In the formula, ψ d * , ψ q * are the expected fluxes (generated by the space flight conditions) respectively, while λ 1 , λ 2 , λ 3 and λ 4 are the weight coefficients (calibrated through the spacecraft thermal control and electromagnetic compatibility constraints);

[0036] Based on all candidate voltage vectors u s (n) The optimal voltage vector u s opt is solved as:

[0037]

[0038] Compared with the prior art, the present application has the following advantages:

[0039] 1. The technical solution of the present application has high-precision fault tolerance. Through the short-circuit circulating current dynamics model, online compensation for turn-to-turn short circuits with any ratio is realized, avoiding torque fluctuations caused by ignoring i f in the traditional method (experimental data: torque ripple reduced by 40%);

[0040] 2. The technical solution of the present application is based on electro-thermal co-control. The temperature rise prediction model is coupled with the electromagnetic equation, solving the risk of local overheating in the closed environment of the spacecraft (simulation shows that the peak hot spot temperature drops by 18%);

[0041] 3. The technical solution of the present application realizes lightweight real-time optimization. Through the forward Euler discretization + finite control set strategy, the measured single-cycle calculation time on the ARM Cortex-M7 is only 18 μs, meeting the kHz-level control requirements of the space motor;

[0042] 4. The technical solution of this application adopts dynamic multi-objective adaptation, and the weight coefficient λi can be automatically adjusted according to the space mission phases (such as orbit transfer and attitude adjustment) to ensure the optimal comprehensive performance under faults (Case: After a satellite motor fails, the output power remains at 98% of the rated value).

[0043] 5. The technical solution of this application is robust in extreme environments. The model parameters (such as R s , h) support on-orbit adaptive calibration to adapt to challenges such as space radiation and sudden temperature changes (Comparison test: The error of the traditional method > 10%, and the error of this model < 3%). Description of the Drawings

[0044] Figure 1 is the principle block diagram of a permanent magnet synchronous motor fault-tolerant control method for space flight according to the present invention;

[0045] Figure 2 is the equivalent circuit diagram of the stator inter-turn short circuit fault of the permanent magnet synchronous motor according to the present invention;

[0046] Figure 3 is the comparative analysis table of a permanent magnet synchronous motor fault-tolerant control method for space flight according to the present invention and the existing method;

[0047] Figure 4 is the parameter table of a permanent magnet synchronous motor fault-tolerant control method for space flight according to the present invention;

[0048] Figure 5 is the comparative analysis table of the simulation data of Embodiment 1 according to the present invention. Detailed Embodiments

[0049] The present invention will be specifically described below with reference to the drawings, as Figures 1-4 shown;

[0050] In the technical solution of this application, the space factors affecting the inter-turn short circuit equivalent circuit include temperature change and vibration, etc.; considering the influencing factors after the fault, it is necessary to establish the relationship between the short-circuit circulating current and the temperature rise; then it is necessary to comprehensively consider conditions such as flux linkage tracking, circulating current suppression, and temperature rise limitation and use the cost function to obtain the optimal voltage vector.

[0051] The specific content is as follows:

[0052] First, based on the equivalent circuit model of the inter-turn short circuit, and considering the influence of environmental temperature change on the resistance or the influence of vibration on the inductance in the space environment, modify the traditional state equation of the permanent magnet synchronous motor and add the current variable of the short-circuit loop; then discretize the continuous-time state equation of the permanent magnet synchronous motor by the forward Euler method and construct the motor prediction model.

[0053] Secondly, the flux linkage equation under fault conditions is derived and decomposed in the d-q coordinate system. Considering the influence of the short-circuit loop, the short-circuit circulating current expression and the temperature rise expression are established, such as the temperature rise caused by Joule heat.

[0054] Finally, using the cost function, the flux linkage error, the magnitude of the circulating current, and the temperature rise are taken as optimization objectives, and the optimal voltage vector that minimizes the cost is selected.

[0055] To achieve the control objectives of suppressing short-circuit circulating current, winding temperature rise, and reducing torque ripple of the faulty motor, this application adopts a predictive torque control strategy. Based on Figure 2 the equivalent circuit of the turn-to-turn short-circuit fault, the motor prediction model is constructed using the forward Euler method as shown in Equation (2):

[0056] A fault-tolerant control method for a permanent magnet synchronous motor used in spaceflight, comprising:

[0057] Based on the equivalent circuit of the turn-to-turn short-circuit fault and fully considering the factors of spaceflight, a motor prediction model is constructed using the forward Euler method;

[0058] By analyzing the motor prediction model, the expressions of the d-q axis components of the stator flux linkage, the short-circuit circulating current expression, and the winding temperature rise expression under the turn-to-turn short-circuit fault are obtained;

[0059] Using the cost function to evaluate the d-q axis components of the stator flux linkage, the short-circuit circulating current, and the winding temperature rise under the turn-to-turn short-circuit fault and obtaining the optimal voltage vector.

[0060] The construction process of the motor prediction model includes:

[0061] When the permanent magnet synchronous motor has a k% turn-to-turn short-circuit fault, the equivalent equation of the faulty phase winding is:

[0062]

[0063] Where: u sd , u sq are the d-q axis stator voltages (V); i sd , i sq are the d-q axis stator currents (A); i f is the short-circuit loop current (A); R s is the stator winding resistance (Ω); L d , L q are the d-q axis inductances (H); ω e is the electrical angular velocity (rad / s); e d , e q are the permanent magnet back electromotive force components (V); R f is the short-circuit resistance (Ω), R f = kR s, where \(k\in(0,1)\) is the short - circuit turn ratio;

[0064] The equivalent equation is discretized using the forward Euler method and the motor prediction model is obtained as:

[0065]

[0066] where \(\Delta t\) is the time step.

[0067] Based on the motor prediction model, the expressions for the d - q axis components of the stator flux under inter - turn short - circuit fault are:

[0068]

[0069] where \(L\) f is the short - circuit equivalent inductance.

[0070] Based on the motor prediction model, the expression for the short - circuit circulating current is:

[0071]

[0072] Based on the motor prediction model, the expression for the winding temperature rise is:

[0073]

[0074] where \(h\) is the heat dissipation coefficient \((W / (m\) 2 \(\cdot K))\), and \(A\) is the winding surface area \((m\) 2 \(^2)\).

[0075] The cost function for evaluating the d - q axis components of the stator flux, the short - circuit circulating current, and the winding temperature rise under inter - turn short - circuit fault is:

[0076]

[0077] where \(\psi\) d * , \(\psi\) q * are the desired fluxes (generated by the space flight conditions), and \(\lambda\) 1 , \(\lambda\) 2 , \(\lambda\) 3 and \(\lambda\) 4 are the weight coefficients (calibrated through spacecraft thermal control and electromagnetic compatibility constraints);

[0078] Based on all candidate voltage vectors \(u\) s (n) The optimal voltage vector \(u\) s opt is obtained as:

[0079]

[0080] Example 1

[0081] Based on the above method, as Figure 5 shown, the fault-tolerant control process of this application is as follows:

[0082] Step 1: Real-time identification of short-circuit faults;

[0083] Monitor the three-phase current i a , i b , i c using the Clarke / Park transformation to calculate the d-q axis currents i sd , i sq ;

[0084] If the detected current harmonic distortion rate exceeds the threshold (e.g., THD > 5%), or the amplitude of a certain phase current fluctuates abnormally (deviates from the nominal value by ±20%), trigger the short-circuit fault diagnosis;

[0085] Online identify the short-circuit resistance R f and inductance L f , and determine the short-circuit turn ratio k = R f / R s (with an accuracy of up to ±2%).

[0086] Step 2: Update the prediction model;

[0087] Substitute the identified R f , L f into the discretized motor equation:

[0088] i sd (k + 1) = f(i sd (k), i sq (k), i f (k), u sd (k))

[0089] i sq (k + 1) = g(i sq (k), i sd (k), u sq (k))

[0090] Synchronously update the temperature rise model:

[0091]

[0092] Step 3: Generate candidate voltage vectors;

[0093] According to the inverter topology (e.g., three-phase two-level), generate 8 groups of candidate voltage vectors u s (n) = [u sd(n) , u sq (n) (n = 0, 1, ..., 7).

[0094] For each u s (n) , substitute it into the prediction model to calculate the following moment's:

[0095] Stator magnetic flux ψ d (n) , ψ q (n) , short - circuit circulating current i f (n) , temperature rise increment ΔT (n) ;

[0096] Step 4: Cost function evaluation and optimization;

[0097] Calculate the cost function values of each candidate vector:

[0098]

[0099] Weight dynamic adjustment:

[0100] According to the spacecraft state (such as high torque accuracy is required during the orbit - changing phase, λ 1 , λ 2 increases; when heat dissipation is limited, λ 4 is enhanced).

[0101] Select the optimal vector:

[0102]

[0103] Step 5: Voltage vector application and state feedback;

[0104] Convert u s opt into inverter switching signals (SVPWM modulation) to drive the motor;

[0105] Collect the actual current i sd (k + 1), i sq (k + 1) and temperature T(k + 1).

[0106] Step 6: Online calibration of model parameters;

[0107] Electromagnetic parameter drift compensation:

[0108] For the changes in R s , L d , L q caused by space radiation, use the recursive least - squares method to update the model parameters;

[0109] Thermal model correction:

[0110] Adjust the heat dissipation coefficient h according to the temperature sensor data (h is reduced to 30%-50% of the ground value in a vacuum environment);

[0111] Adaptive adjustment for the space environment;

[0112] 1. Radiation hardening:

[0113] The model parameters (R s , L d , L q ) are calibrated using the radiation sensitivity coefficient to suppress the numerical jumps caused by single-event effects;

[0114] The control algorithm code is stored in a radiation-hardened FPGA through triple modular redundancy (TMR).

[0115] 2. Vacuum thermal management:

[0116] The temperature rise model incorporates the vacuum heat dissipation coefficient h vac = h air × 0.35 and is corrected in real time through thermocouple feedback;

[0117] If the predicted ΔT > T max (such as 150 °C), automatically reduce the torque command and trigger a heat dissipation warning.

[0118] 3. Long-term reliability:

[0119] Perform parameter self-check and model reset once every 24 hours to avoid cumulative errors;

[0120] In the event of a fault, prioritize ensuring the flux linkage tracking accuracy (λ 1 weight is increased to 70%) to ensure the stability of attitude control.

[0121] The scenario to which the technical solution of the present invention is applied is: on-orbit short-circuit fault of the satellite flywheel motor;

[0122] Fault conditions:

[0123] A 20% turn-to-turn short circuit occurs in phase B of the flywheel motor (k = 0.2), and the short-circuit circulating current i f = 4.2 A (exceeding the safety threshold).

[0124] Control response:

[0125] Detect the fault and update the model within 2 ms;

[0126] 2. Optimize the weight λ 3 (circulating current suppression) is increased to 0.4, and i f is reduced to 1.8 A;

[0127] 3. Predict ΔT = 22°C through the temperature rise model and trigger the intermittent start of the cooling fan;

[0128] 4. Keep the flux linkage tracking error within 3% and the satellite attitude angle deviation ≤ 0.05°.

[0129] Result:

[0130] No safety shutdown was triggered during the mission cycle, and there was no performance degradation after 12 months of continuous operation after the fault.

[0131] Through the closed-loop architecture of model-optimization-feedback, this control process realizes the highly reliable operation of aerospace motors under extreme environments and fault conditions, which is significantly better than the traditional methods relying on redundant hardware or fixed rules.

[0132] The above technical solutions only reflect the preferred technical solutions of the technical solutions of the present invention. Some changes that those skilled in the art may make to some parts thereof all reflect the principles of the present invention and fall within the protection scope of the present invention.

Claims

1. A permanent magnet synchronous motor fault-tolerant control method for space flight, characterized in that: include: Based on the equivalent circuit of inter-turn short-circuit fault and taking into full consideration the factors of space flight, the forward Euler method is used to construct the motor prediction model; By analyzing the motor prediction model, the expressions of stator flux dq axis components, short-circuit circulating current and winding temperature rise under turn-to-turn short-circuit fault are obtained. The cost function is used to evaluate the stator flux dq-axis components, short-circuit circulating current and winding temperature rise under turn-to-turn short-circuit fault and obtain the optimal voltage vector.

2. A permanent magnet synchronous motor fault-tolerant control method for space flight according to claim 1, characterized in that: The construction process of the motor prediction model includes: Assuming that a k% inter-turn short-circuit fault occurs in a permanent magnet synchronous motor, the equivalent equation of the faulty phase winding is: Where: u sd ,u sq are the dq axis stator voltages respectively; i sd ,i sq are the dq axis stator currents respectively; i f is the short-circuit current; R s is the stator winding resistance; L d ,L q are the dq axis inductance respectively; ω e is the electrical angular velocity; e d ,e q are the back electromotive force components of the permanent magnet respectively; R f is the short-circuit resistance, R f =kR s , k∈(0,1) is the short-circuit turns ratio; The forward Euler method is used to discretize the equivalent equation and obtain the motor prediction model: Where Δt is the time step.

3. A permanent magnet synchronous motor fault-tolerant control method for space flight according to claim 2, characterized in that: Based on the motor prediction model, the expression of the stator flux dq axis component under turn-to-turn short circuit fault is obtained as follows: Where, L f is the short-circuit ring inductance.

4. A permanent magnet synchronous motor fault-tolerant control method for space flight according to claim 3, characterized in that: Based on the motor prediction model, the short-circuit circulating current expression is:

5. A permanent magnet synchronous motor fault-tolerant control method for space flight according to claim 4, characterized in that: The winding temperature rise expression based on the motor prediction model is: Where h is the heat dissipation coefficient and A is the winding surface area.

6. A permanent magnet synchronous motor fault-tolerant control method for space flight according to claim 5, characterized in that: The cost function for evaluating the stator flux dq axis components, short-circuit circulating current and winding temperature rise under turn-to-turn short-circuit fault is: In the formula, ψ d * ,ψ q * are the expected flux, λ1, λ2, λ3 and λ4 are weight coefficients respectively; based on all candidate voltage vectors u s (n) Solve for the optimal voltage vector u s opt for:

Images